English

• 1. Introduction

  RNA interference (RNAi) is a natural process of post-transcriptional gene regulation presented in almost all eukaryotic organisms that affect protein synthesis by silencing specific gene expression [1]. In 1998, this biological mechanism was first discovered and harnessed in the nematode Caenorhabditis elegans via long double-stranded RNA (dsRNA) [2]. Building on this, the efficacy of RNAi in mammalian cells was then demonstrated via small interfering RNA (siRNA) by Tuschl's laboratory in 2001 [3]. siRNA, referred to short interfering RNA, is a dsRNA molecule consisting of 21–23 nucleotides [4]. It functions by specifically binding to the mRNA of target proteins through stringent base-pair complementarity, thereby inducing mRNA degradation. The mechanism of action is that the siRNA entering the cell is integrated into a complex containing the RNA-induced silencing complex and the Argonaute (Ago) protein. Within this complex, the sense strand of the siRNA is released and degraded, while the antisense strand binds to complementary sequence of target mRNA, resulting in its degradation (Fig. 1) [5]. This process effectively prevents mRNA from being translated into functional proteins at its source.

  Figure 1

  

  Figure 1.  Mechanism of siRNA mediated gene silencing (By Figdraw).

  DownLoad: Full-Size Img PowerPoint

  According to mechanism, theoretically, siRNAs can be designed to target any gene of interest, as long as the corresponding nucleotide sequence is designed properly based on the mRNA sequence. Thus, it not only expands the scope of therapeutically reachable targets but also significantly shorten the development timeline. Many clinically meaningful protein targets have been considered "undruggable" owing to the absence of proper ligand binding sites for conventional small molecule drugs and biopharmaceutics, whereas siRNAs enable those targets to become druggable via a simple base-pairing. And this "simple" mechanism would also lead to the shorter time-frame identification of siRNAs lead candidates, as well as a lower cost of siRNAs pipeline than conventional drug modalities [6]. Thus, siRNAs stand out from RNAi based drugs with unique advantages compared with conventional drugs, such as broad therapeutic scope, high clinical translatability, short development timeline and prolonged clinical durability. Following the Food and Drug Administration (FDA) landmark approval of Patisiran (trade name Onpattro) in 2018, the first siRNA therapeutic has been utilized, which is a lipid nanoparticle (LNP) designed to treat peripheral neuropathy associated with hereditary transthyretin-mediated amyloidosis (hATTR) [7]. To date, six siRNA therapeutics have already been commercially available, validating the RNAi technology for clinical use (Table 1) [8]. Meanwhile, in recent years, research-driven companies, led by Alnylam, Lonis and Quark, have continuously increased their investment in the development of siRNA therapeutics. Nowadays, hundreds of siRNA-related clinical trials are being conducted, demonstrating broad prospects.

  Table 1

  

  Table 1.  Clinically approved siRNA therapeutics.

  DownLoad: CSV

  Generic name	Dosage form	Launch date	Target	Clinical indication	Route of administration
  Patisiran	LNP	2018	Transthyretin (TTR)	Hereditary transthyretin amyloidosis	iv
  Givosiran	GalNAc-siRNA conjugate	2019	Aminolevulinate delta synthase 1 (ALAS1)	Acute hepatic porphyria	sc
  Inclisiran	GalNAc-siRNA conjugate	2020	Proprotein convertase subtilisin/kexin type 9 (PCSK9)	Primary hypercholesterolemia or mixed dyslipidemia	sc
  Lumasiran	GalNAc-siRNA conjugate	2021	Human hydroxyacid oxidase 1 (HAO1)	Primary hyperoxaluria type1	sc
  Vutrisiran	GalNAc-siRNA conjugate	2022	TTR	Hereditary transthyretin amyloidosis	sc
  Nedosiran	GalNAc-siRNA conjugate	2023	Lactate dehydrogenase (LDH)	Primary hyperoxaluria type1	sc

  Despite significant advances in siRNA therapeutics research, pivotal challenges, notably the delivery issue, remain to be overcome. These delivery challenges are closely linked to the physicochemical properties of the siRNAs. Upon systematic administration, siRNAs face their initial obstacle in delivery: Undergoing metabolic degradation and clearance during the processes of absorption and circulation. This challenge stems from the innate instability of siRNAs in the physiological environment. The phosphodiester bonds of unmodified naked siRNAs are susceptibility to RNases and phosphatases throughout the body that lead to the fragmentation of siRNAs [9,10]. Moreover, the molecular weight of siRNAs, approximately 14 kDa, fall below the renal clearance limit of 40–60 kDa, which leads to their swift degradation by the kidney and the reticuloendothelial system, consequently, results in reduced bioavailability and a brief half-life of merely a few minutes [11-13]. The second hurdle in siRNA delivery is accomplishing substantial enrichment and effective internalization within the specific tissues and target cells, respectively. As a result of their poor metabolic stability and targeting ability, siRNAs cannot be effectively disseminated across specific tissues or internalized by cells without the help of serum protein binding and targeting entities [6,14]. Besides, it is hard for siRNAs to diffuse across the lipid bilayer of target cells passively, due to large size, anionic charge as well as hydrophilic property. The third challenge of siRNA delivery is how to achieve efficient escape from endosomes and lysosomes. Even if siRNA successfully internalized into the target cells, siRNAs are too large, charged, and/or hydrophilic to diffuse across the lipid bilayers of endosomes and lysosomes, resulting in 99% of them to be retained, with only 1% or a smaller fraction entering the cytoplasm [15]. Hence, endo/lysosomal escape of siRNA therapeutics become the rate-limiting delivery problem.

  In the last few decades, significant advancements have been made in enhancing stability, reducing innate immune responses, overall charge and kidney clearance of siRNAs by introducing chemical modifications and nanocarriers, and these strategies play important roles in overcoming the extracellular barriers to the delivery of siRNAs in vivo [14,16,17]. However, overcoming the intracellular endo/lysosomal barriers remains the foremost problem to solve in siRNA delivery for wider therapeutic applications. This review highlights the latest advancements in reported strategies for overcoming endo/lysosomal barriers of siRNA therapeutics. All these strategies will be discussed from the viewpoints of cellular uptake and intracellular transport: (a) siRNA therapeutics endo/lysosomal escaping strategies for intracellular transporting via "endosomes-lysosomes" pathway, (b) siRNA therapeutics endo/lysosomal escaping strategies for intracellular transporting via "endosome-Golgi-endoplasmic reticulum (ER)" pathway, (c) siRNA therapeutics endo/lysosomal escaping strategies via nonendocytosis pathway. The discussion of the interface between biomaterials science and RNA biochemistry in this review highlights the profound influence of siRNA therapeutics on personalized medicine and complex diseases in the future.

  2. Cellular uptake and intracellular transport mechanisms of siRNA therapeutics

  As mentioned above, the challenges of siRNA therapeutic applications are closely related to the inherent physicochemical properties of unmodified naked siRNA molecules, and even with complete chemical stabilization, siRNAs would fundamentally remain ineffective without delivery systems due to low bioavailability and targeting efficiency [18,19]. Therefore, siRNAs are delivered by two common strategies, conjugate-mediated delivery and nanosystem-mediated delivery [20-22]. Based on these, the cellular uptake and intracellular transport mechanisms of siRNA therapeutics are primarily determined by the physicochemical properties of nanocarriers or the targeting entities. And endocytosis is the main route of entry into the cell harnessed by siRNA therapeutics.

  Endocytosis primarily involves phagocytosis and pinocytosis. Phagocytosis, executed by specialized phagocytes, is a key cellular process for engulfing and clearing pathogens and large particles (≥0.5 µm), performing a vital function within the body's immune defense system. Pinocytosis involves the uptake of liquid-phase and smaller particles, the most well-documented including clathrin-mediated endocytosis (CME), caveolae-mediated endocytosis (CvME), macropinocytosis, present in almost all eukaryotic cells (Fig. 2) [23,24].

  Figure 2

  

  Figure 2.  Endocytic pathways and intracellular transport process of siRNA therapeutics (By Figdraw).

  DownLoad: Full-Size Img PowerPoint

  CME stands out as a prominently investigated endocytic mechanisms and a pivotal pathway for the cellular uptake of siRNA therapeutics. Gilleron et al. used quantitative fluorescence imaging and electron microscopy to monitor the cellular uptake of siRNA-loaded LNPs, and found that siRNA-loaded LNPs entered cells mainly through CME and micropinocytosis [25]. Similarly, Xia et al. found that hyaluronic acid (HA)-modified siRNA-loaded selenium nanoparticles (HA-Se-PEI@siRNA NPs) are primarily internalized by HepG2 cells via CME, enhancing the treatment of liver cancer [26].

  CvME represents a cholesterol-dependent mechanism, which occurs at lipid raft domains of plasma membrane, related to the uptake of substances such as albumin, viruses and toxins [27]. Caveolae proteins can oligomerize on the cell membrane, forming structures known as caveolae, which are invaginations of the membrane. The intracellular route of caveolae is unknown and controversial. Whether the internalized caveolae can fuse with endosomes and follow the classical endocytotic degradative pathway or an alternative pathway with distinct cellular compartments to avoid lysosomal degradation is still unclear [27,28]. However, some studies revealed that the intracellular trafficking of polymeric NPs internalized via CvME might be related to the cell types and the composition of NPs. And this pathway had been utilized to enhance the intracellular delivery efficiency of the siRNAs [29]. This will be discussed in Section 3.2.

  Macropinocytosis, similarly to phagocytosis, involves the engulfment of fluid and micron-sized particles, where the plasma membrane invaginates to engulf a large volume of external fluid, particles and solutes, forming large vesicles known as macropinosomes (0.2–5 µm), thus mediating endocytosis [30]. Macropinocytosis allows non-phagocytic cells to internalize large macromolecules, and several lipid-based siRNA delivery systems have been observed to be internalized by cells via this mechanism [31].

  Accumulating evidence has indicated that siRNA loaded NPs are taken up by cells via various endocytic routes, which are dependent on the specific type of cell. For instance, Lu et al. used exosome-mimicking liposomes for vascular endothelial growth factor (VEGF) siRNA delivery, and found that the siRNA formulations were internalized into A549 cells through CvME, and membrane fusion manner, while entering human umbilical vein endothelial cells (HUVEC) through CvME alone [32]. Sun et al. demonstrated that siRNA-loaded nanoplex was taken up by melanoma cells via CvME as well as macropinocytosis [33].

  After internalization by cells, siRNA therapeutics is transported in a sequential manner via early endosomes (EE), late endosomes (LE), ultimately reaching lysosomes. The EE mature to form multivesicular LE, and the further maturation of endosomes involves structural changes and leads to the formation of multilamellar endosomes (a temporary heterotypic organelle) through fusion with lysosomes, ultimately transitioning into lysosomes. It is noteworthy that throughout the maturation process of endosomes, the pH level progressively declines when compared to physiological pH 7.4 (EE: pH < 6.5, LE: pH < 6.0, lysosomes: pH < 5.0). Most NPs carried siRNAs ultimately reach the lysosome, where contains a variety of degradative enzymes, including such as nucleases and phosphatases, once these siRNA-loaded NPs cannot rapidly escape from the lysosome, they are usually degraded, resulting in ineffective delivery of the siRNAs [34].

  Although there have been groundbreaking advancements in clinical application, the low delivery efficiency of non-viral vectors (including LNPs, polymer, NPs) due to lysosome capture greatly hinders siRNA therapeutics development. Studies have indicated that after being internalized by cells, only 1%–2% of siRNAs encapsulated within LNPs are proven to escape endo/lysosomal restrictions, whereas the escape efficiency of N-acetylgalactosamine (GalNAc)-siRNA is even lower (< 0.2%) [25]. Therefore, sufficient GalNAc-siRNA conjugates to achieve therapeutic levels rely on the extraordinarily high expression levels of sialoglycoprotein receptor (ASGPR), that enhance the cell uptake and storage of siRNAs in liver cell compartments, resulting in sustained therapeutic effects in vivo [35]. Although 1%–2% escaped siRNAs has been proved to be therapeutic in the liver, more siRNA are required to escape from the lysosome to produce therapeutic effect in other target organs/cells. It is urgent to develop new strategies to overcome endo/lysosomal barriers and improve siRNA delivery efficiency into target cells' cytosol.

  3. Advanced strategies for overcoming endosomal/lysosomal barriers in siRNA delivery

  With a deeper understanding of cellular uptake and intracellular transport mechanisms over recent years, various approaches for overcoming endo/lysosomal barriers in siRNA delivery have been developed by researchers, which are mainly classified into three main categories as outlined below:

  3.1 siRNA therapeutics endo/lysosomal escaping strategies for intracellular transporting via "endosome-lysosome" pathway

  Since the majority of siRNA therapeutics taken up by endocytosis would be trapped in the endosomal-lysosomal system, most endo/lysosomal escaping strategies of siRNA therapeutics take advantage of the physicochemical properties of materials to disturb the stability of late endosome and lysosome, leading to the siRNAs release into cytoplasm. pH and light responsive materials are commonly used materials applied in these strategies. This section highlights various strategies used to promote endo/lysosomal escape of siRNA therapeutics, and the strategies are categorized based on the underlying mechanisms (Fig. 3, Table 2 [36-60]).

  Figure 3

  

  Figure 3.  Schematic representation of endo/lysosomal escape mechanisms in the uptake of siRNA therapeutics via endosome-lysosome pathway (By Figdraw).

  DownLoad: Full-Size Img PowerPoint

  Table 2

  

  Table 2.  The summary of siRNA therapeutics endo/lysosomal escaping strategies for intracellular transporting via endosome-lysosome pathway.

  DownLoad: CSV

  Mechanism	Carrier	siRNA target site	Gene silencing efficiency	Indication	Ref.
  Proton sponge effect	PEG-b-PLA-PHis-ssPEI	Bcl-2	85.45%	Breast cancer	[36]
  	PAMAM nanocomplex	KRAS	85.3%	Lung cancer	[37]
  	HE25 peptide	SARS-CoV-2	85%	Coronavirus disease (COVID)-19	[38]
  	JBNTs with a lysine side chain	Glyceraldehyde-3-phosphate dehydrogenase (GAPDH)	90%	COVID-19	[39]
  	stEK peptide	Cyclophilin B Connective tissue growth factor (CTGF)	> 90% 49%–58%	Cervical cancer	[40]
  	Cyclic peptide modified liposomal delivery system (Cyc-LH)	PLK1	> 70%	Breast cancer	[41]
  Osmosis lysis effect	mPEG-P(DPAx-co-DMAEMAy)-PT	PLK1 PD-L1	86.9% ~45%	Colorectal cancer and liver cancer	[42]
  	CAP NPs	PD-L1 PLK1	~50% 12%–14%	Melanoma	[43]
  	NaGdF4	PD-L1	~70%	Colorectal cancer and breast cancer	[44]
  	Fe0-siRNA NPs	FHC	–	Cervical cancer	[45]
  Membrane destabilization	heat-stable iLAND	ANGTPL3	97%	Hyperlipidemia	[46]
  	CADs	EGFP PLK1	> 80%	Non-small cell lung cancer	[47]
  	HGP-PEI	GFP	50% (without HGP) vs. 80% (with HGP)	Cervical cancer	[48]
  	gH625-siRNA nanocarrier	GFP	40% (without gH625) vs. 70% (with gH625)	Triple-negative breast cancer	[49]
  Membrane fusion	RNA-LNPs with cubic and inverse hexagonal structure	–	–	Breast cancer and cervical cancer	[50]
  	HA-modified ferritin (Apn) nanocagnets (Apn/siRNA-HA)	TK1pro	–	Breast cancer	[51]
  Photochemical internalization	TPPS2a-lipo	Epidermal growth factor receptor (EGFR)	10% (PCI−) vs. 70% (PCI+)	Epidermoid carcinoma	[52]
  	TPPS2a-PLA TPPS2a-PLH TPPS2a-PLL	S100A4	15% (PCI−) vs. 90% (PCI+) 10% (PCI−) vs. 45% (PCI+) 10% (PCI−) vs. 80% (PCI+)	Osteosarcoma	[53]
  	ZnPc upconversion nanoparticles	Superoxide dismutase 1 (SOD1)	70% (PCI−) vs. 90% (PCI+)	Oral cell carcinoma	[54]
  	Porphyrin-periodic mesoporous Ionosilica nanoparticles	Luciferase	17% (PCI−) vs. 83% (PCI+)	–	[55]
  	Pyropheophorbide α-PEI	PLK1	40% (PCI+)	Breast cancer	[56]
  	NB-Br-siRNA NPs	PLK1	40% (PCI−) vs. 65% (PCI+)	Breast cancer and hepatocarcinogen	[57]
  	Porphyrin-patisiran	TTR	–	Hereditary transthyretin amyloidosis	[58]
  	TPPS2a-upconversion NPs	STAT3	25% (PCI−) vs. 50% (PCI+)	Melanoma	[59]
  	Fullerene (C60)	EGFP	17% (PCI−) vs. 53% (PCI+)	Breast cancer	[60]

  3.1.1   Proton sponge effect

  Proton sponge effect, a predominant strategy to induce endo/lysosomal escape, primarily relies on the use of the materials with high buffering characteristics, such as polyethyleneimine (PEI) [61] and poly(amidoamine) (PAMAM) [37,62]. In acidic conditions of the endo/lysosomes, the materials will adsorb hydrogen protons. In this case, Cl⁻ and inorganic ions will flow into endo/lysosomes to maintain the charge balance so that the osmotic pressure increases, resulting in vesicle swelling and rupture, thus releasing siRNAs into the cytoplasm to display gene silencing effect [63]. In general, polymers that are highly branched and rigid, with a positive charge, will facilely induce osmotic swelling and facilitate endosomal escape through the proton sponge effect [64]. Zhu et al. [36] developed a Bcl-2 siRNAs loaded copolymer (PEG-b-PLA-PHis-ssPEI) that is sensitive to redox potential and pH. In this system, siRNAs were self-assembled into PEG shielded polyplexes by electrostatic attractions. PEI not only can induce the proton sponge effects but also can cause membrane destabilization due to the electrostatic attractions between the PEI and the membranes. However, PEI/PAMAM-based or other cationic materials are considered high cytotoxicity while ensuring transfection efficiency [65]. Therefore, cation block strategies or surface modification were used to improve biosafety. Chen et al. [37] designed a glutathione (GSH)-sensitive anionic block copolymer as a shell for positive charge shielding to address the cytotoxicity issue associated with PAMAM dendrimers. The nanocomplex was mainly taken up by cells via CME pathway. The GSH-responsive shell would be disassembled and the kirsten rat sarcoma viral oncogene (KRAS) siRNA-loaded PAMAM re-exposed in tumor environment, which could trigger lysosome escape by the proton sponge effect. The expression of KRAS in A549 human lung carcinoma cells decreased to 85.3% after 48 h of treatment. The nanocomplex not only achieved effective intracellular siRNA delivery, but also significantly improved the biosafety of PAMAM.

  Moreover, amino acids with ionizable cationic side chains, including lysine, histidine and arginine, underwent protonation at acidic pH levels, resulting in a positive charge. Therefore, peptides or copolymers containing these amino acids can be readily synthesized for siRNA delivery. Tuttolomondo et al. [38] developed a HE25 peptide that formed a complex with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) siRNAs via electrostatic attraction. The complex was uptaken by clathrin-mediated endocytosis, once enter late endosomes, the histidine in the N-terminal domain of HE25 peptides became positively charged at pH 5.5, triggering the siRNA release through the histidine-proton sponge effect. These mechanisms resulted in the almost complete knock-down of SARS-CoV-2 expression, reaching up to 100% efficacy. Lee et al. [39] selected a Janus base nanotubes (JBNTs) with a lysine side chain for siRNA delivery. The results showed that siRNAs can successfully escape from late endosomes via the "proton sponge" effect, exerting gene silencing efficacy of ~90%, while the common LNPs only can silence approximately 60%–70%. Hyun et al. [40] developed an optimized LKH-stEK peptide which replaced several amino acids with histidine moieties. These changes could facilitate escape of the siRNAs entrapped in endosomes due to the changes of protonation states in endosomal compartments, and exerted > 90% gene knock-down in HeLa cells with 50 nmol/L siRNAs targeting cyclophilin B. Dang et al. [41] used stearic acid octahistidine (SA-H₈) compacted with polo-like kinase 1 (PLK1) siRNA, which could respond to the lysosome-to-cytoplasm pH variation and promote siRNA release.

  3.1.2   Osmosis lysis effect

  As previously mentioned, the proton sponge effect leads to increased osmotic pressure caused by H⁺ and Cl⁻. Similarly, many inorganic metal ion or disassemble polymer subunits at low pH also can cause osmosis lysis effect which is benefit to the disruption of endo/lysosomes. Li et al. [42] prepared a series of quaternary ammonium-based polymer [mPEG-P(DPA_x-co-DMAEMA_y)-PT] nanoparticles featuring a pH-sensitive hydrophobic core. Through adjusting the proportion of the blocks to accurately respond to different pH, the polymer micelle could quickly disassembled in the endosome pH range (pH 6.5–6.8). The results showed that the polyplexes escaped from the endosomes and lysosomes after transfection for 3–5 h, only 50 nmol/L siPLK1 resulted in > 80% gene silencing in HepG2 cells (Fig. 4).

  Figure 4

  

  Figure 4.  (A–C) Schematic of siRNA-loaded amphiphilic triblock polymers for endo/lysosomal escape. PDDT-Ms/siRNA can quickly respond to acidic pH and disassembled around pH 6.5–6.8, leading to effective endosomal escape of siRNA. Copied with permission [42]. Copyright 2021, American Chemical Society.

  DownLoad: Full-Size Img PowerPoint

  A number of studies showed that inorganic nanoparticles, such as Ca²⁺ [66], Zn²⁺ [67], Gd³⁺ [44], Fe²⁺ [45], can release inorganic ions in the acidified lysosomal compartments which cause a significant rise in the lysosomal internal osmotic pressure, disturb the osmotic balance and release siRNAs into the cytoplasm. The most representative one is calcium phosphate (CAP) nanoparticle. Wu et al. [43] designed new lipid-decorated calcium carbonate/phosphate (LCCP) nanoparticles for programmed cell death ligand 1 (PD-L1) siRNAs and PLK1 siRNAs delivery respectively, the study confirmed that the nanoparticles (NPs) can achieve the early endosomal escape and accelerate mRNA inhibition. Yu et al. [44] investigated the interaction mechanism between siRNAs and ligand-free NaGdF4 nanoparticles. In a neutral environment, the PD-L1 siRNA phosphate skeleton coordinates with Gd on the surface of NaGdF4 nanoparticles to form siPD-L1/NaGdF4 spherical nanoparticles. In the acidic environment, the phosphate groups could be protonated, reducing the affinity of siRNAs to NaGdF4 nanoparticles, and siRNAs were dissociated from NaGdF4. The intracellular escape process of siRNAs was observed by fluorescence confocal microscopy. PD-L1 expression was silenced by 70% in 4T1 in situ breast cancer models. Wang et al. [45] reported a Fe⁰-siRNA nanoparticles by embedding ferritin heavy chain (FHC) siRNAs into Fe⁰ nanoparticles, ferrous iron possessed a superior ability to generate hydroxyl radicals (⋅OH), which facilitated endo/lysosomal escape of siRNAs by disrupting the membranes.

  In addition, it is proved that inhibition of autophagy to promote lysosome escape is another promising strategy used for cytosolic siRNA delivery. As a widely used autophagy flux inhibitor, chloroquine (CQ) can impair the fusion of autophagosomes with lysosomes and thus promotes lysosomal escape [68,69]. Studies revealed that CQ could be pronated in lysosomes, causing water influx and osmotic swelling of lysosomes and thus causing lysosomal membrane permeabilization, and potentially increase the amount of siRNAs that reached the cytosol. Du Rietz et al. [70] probed membrane damage properties of endosomal escape of cholesterol-siRNA conjugates induced by CQ, and chloroquine could lower chol-siRNA half maximal inhibitory concentration (IC₅₀) from 289 nmol/L to 17 nmol/L. Bhattarai et al. [71] also verified that loading the mesoporous silica nanoparticles with chloroquine enhanced silencing activity of siRNA.

  3.1.3   Membrane destabilization

  Under acidic conditions, positively charged or ionizable materials acting as siRNA delivery carriers will electrostatically interact with negatively charged lysosomal membranes, thereby disrupting the membrane to allow siRNA to enter the cytoplasm [72-74]. The application of cationic lipids in siRNA delivery is based on this principle. And the most typical ionizable lipid Dlin-MC3-DMA have been used in commercially available siRNA therapeutic Patisiran [73]. In addition, various ionizable lipids have been developed and widely used in many studies. Hu et al. [46] designed and screened a new heat-stable ionizable lipid-assisted nucleic acid delivery system (iLAND), for delivering apolipoprotein C3 (ApoC3) siRNAs or angiopoietin like protein 3 (ANGTPL3) siRNAs. siRNA@iLAND would be protonated after accumulation in the endo/lysosomes due to the amine groups on the lipid surface. The increasing charges resulted in the creation of ion pairs between ionizable lipids and endo/lysosomal membrane, the membranes changed from a columnar shape to a conical structure. Both membrane destabilization and proton sponge effects worked together and triggered siRNAs to release into the cytoplasm, and the study effectively delivered antihyperlipidemia siRNAs with an effective dose of ~0.18 mg/kg.

  There were also studies using exogenous cationic low molecular weight adjuvants to induced lysosomal membrane permeabilization [75]. Joris et al. [47] showed that cationic amphiphilic drugs (CADs) were passively diffuse into the acidic lysosomes and protonated. Protonated CADs inserted into lysosomal membranes and further caused the disassociation of ASM proteins that are bound to the membrane, inducing a temporary destabilization of the lysosomal membrane and facilitated siRNAs release without cytotoxicity.

  Another mechanism for inducing destabilization of the lysosomal membranes is pore formation. In theory, pore formation is determined by the balance between a line tension that closes the pore and a membrane tension that enlarges the pore [76]. Certain agents can reduce the line tension of the pores and maintains a constant pore radius. In nature, bacterial toxins showed an irregular coil structure at neutral pH, some amino acids in peptide protonated to form amphiphilic α-helices with high affinity to the endosomal membrane under the acidic environment and adhered to or inserted into the endosomal membrane to form barrel stave or toroidal pores [77-79]. For example, Listeriolysin O (LLO) toxin has been developed as a means to facilitate escape from lysosomes [80]. LLO rapidly aggregate in neutral environment but remain stable under acidic pH conditions, which allows to exert its full activity within late endosomes or lysosomes [81]. Upon attachment to the membrane, up to 50 monomers will self-assemble into a pre-pore, eventually destabilizes the lysosomal membrances [80,82]. HIV gp41-derived peptide have also been reported to be capable of inducing pore formation in membranes because of the amphipathic α-helical structures [48,83]. Kwon et al. [48] modified green fluorescent protein (GFP) siRNAs delivery vehicles by incorporating a peptide that is based on the endodomain of HIV gp41 (HGP). Compared with other peptide from adenovirus, HGP mediated efficient escape of polycation vectors at low concentrations, and gene silencing efficiency increased by 30% compared with the group without HGP.

  In addition, there are also studies suggesting that cell-penetrating peptides (CPPs), short peptides with positive-charged, can mediate the siRNAs escape from endo/lysosomes via pore formation mechanism. Ben et al. [49] designed a gH625-functionalized siRNA nanocarrier (CS-MSN) to investigate its cellular uptake in a breast cancer model. The results indicated that gH625 did not alter the uptake pathway of CS-MSN, and the presence of the peptide gH625 enhanced the siRNAs escape from the lysosomal compartment. The CS-MSN demonstrated a 1.7-fold increase in efficacy in inhibiting GFP compared to nanovectors without gH625. The authors proposed that it may be due to the membrane-interacting domain of the 1^st type herpes simplex virus carried by the gH625, which formed transient pores in the endosomal membrane.

  3.1.4   Membrane fusion

  Membrane fusion is the process by which two closely apposed lipid bilayers merge to form a single bilayer. It was proposed that the liposomes, extracellular vesicles or other nanocarriers with lipid bilayer can be fused with biological membranes that lead to the partial release of encapsulated contents from the endosomes into the cytoplasm [30,50,84,85]. It was shown that RNA-LNPs with cubic and inverse hexagonal structured are facilitated to fuse with endosomal membranes, which will lead to a greater degree of endosomal escape compared to lamellar structured RNA-LNPs [50]. Pei et al. [85] also confirmed that siRNA-loaded exosomes escaped from the lysosomes, as evidenced by Mander's correlation coefficient of siRNAs with lysosomes decreasing from 0.72 to 0.32.

  However, most studies indicated that the role of the lipid bilayers itself in mediating escape is negligible [86]. Thus, for nanocarriers that have been endocytosed into endo/lysosomes, more potent mediators of endo/lysosomal escape are also necessary to promote membrane fusion triggered at certain acidic pH conditions. Most viruses contain single integral membrane peptides that undergo conformational changes under external stimuli such as pH changes, which will trigger the fusion in the lipid bilayer. Enveloped viruses thus show superior endosomal escape ability [87]. Based on this, simulating the fusion of viral envelope with membranes is another strategy to facilitate endosome escape. Hemagglutinin-2 (HA2), a typical peptide from influenza virus shell, can mediate membrane fusion by promoting irreversible conformational changes in the glycoprotein at acidic pH [88,89]. Xu et al. [90] reported a kind of HA2-modified solid lipid nanoparticles (HA2-SLNs) with a notable feature of endosomal escape. Cellular uptake assay indicated that CME was involved. After endocytosis, the HA2-SLNs first was transported to EE and showed less colocalization with LE and lysosomes in the presence of HA2. Cao et al. [51] constructed siRNA-loaded ferritin (Apn) nanocages and modified with HA peptide (Apn/siRNA-HA). Apn nanocages were cleaved and siRNAs and HA peptides were released in the acidic pH of lysosomes. The released HA peptides could break lysosome membranes, resulting in the release of siRNAs to exert efficiently gene silencing effect.

  Similarly, GALA peptides and KALA peptides also exhibits highly water-soluble characteristics at neutral pH, but can insert into endo/lysosomal membranes and promote membrane fusion under acidic conditions, which is also widely used in studies of lysosomal escape [91].

  3.1.5   Photochemical internalization (PCI)

  PCI is based on the reactive oxygen species (ROS) and energy produced by hydrophilic photosensitizers after irradiated at a specific wavelength in endosomes or lysosomes, which causes the endo/lysosomal membrane ruptured. It is different from photodynamic therapy (PDT) although similar in components that the PDT induces cell death through an overabundance of ROS, whereas PCI results in the localized breakdown of endo/lysosomal membranes with low cytotoxicity [92]. Commonly used photosensitizers in siRNA delivery systems include TPPS_2a [52,53], ZnPc [54], porphyrin [55], pyropheophorbide-α [56], brominated sulfur-substituted Nile Blue (NB-Br) [57]. Mo et al. [58] reported a light-activated siRNA endosomal release (LASER) strategy, in which porphyrin was incorporated into clinically-approved Patisiran formulations, ROS was generated under near-infrared (NIR) irradiation and enabled LASER due to endosomal membranes disruption, allowing siRNAs to be released into the cytoplasm. Yang et al. [57] proposed a kind of amphiphilic conjugate consisting of siRNA and a photosensitizer (NB-Br) that can self-assemble into nanoparticles through electrostatic interactions (siRNA-NB NPs) (Fig. 5). When cells were treated with siRNA-NB NPs and irradiation allowed more siRNAs to escape from the lysosomes compared with the group without irradiation. This was evidenced by a decrease in the Pearson's coefficient of siRNA and lysosomes from 0.812 to 0.470. Jayakumar et al. [59] developed a novel NIR-to-UV–vis upconversion nanoparticles loaded with signal transducer and activator of transcription 3 (STAT3) siRNAs and TPPS_2a. The efficiency of gene silencing was increased by > 30% with the effect of PCI compared to the group without endosomal escape facilitation. However, most of the above-mentioned photosensitizers required the use of high-power lasers to achieve excitation. Contrastingly, even under weak visible light, fullerene (C60) can induce constant ROS. Wang et al. [60] prepared an amphiphilic fullerene derivative (C60-Dex-NH₂) for delivering enhanced green fluorescent protein (EGFP) siRNAs. The results showed that upon exposure to visible light, C60 could trigger ROS which could destruct the lysosome structure, and the gene silencing efficiency enhanced. And at the same time the micelle-like aggregate structures of C60-Dex-NH2 prevent siRNAs from destroying by ROS. In MDA-MB-231 and 4T1 tumor-bearing mice, gene silencing efficiency reached 53% and 69%, respectively.

  Figure 5

  

  Figure 5.  Schematic illustration of the intracellular trafficking and endo/lysosomal escaping of siPLK1-NB NPs via PCI. (a) NB-Br were grafted onto siPLK1 and self-assembled into NPs by electrostatic interaction. (b) The ROS generated by NB-Br could rupture the lysosome membrane under light irradiation and facilitate siRNA release. Copied with permission [57]. Copyright 2023, John Wiley & Sons, Inc.

  DownLoad: Full-Size Img PowerPoint

  In this section, we have summarized techniques for promoting lysosomal escape, most of which involve pH-responsive carriers. Although these materials provide a promising strategy for endo/lysosomal escaping, pH difference between normal and tumor tissues, the extracellular microenvironment and endo/lysosomes may not be significant enough achieve precise siRNA delivery. To overcome the challenge, it is necessary to design new carriers with higher pH sensitivity that respond to the subtle pH variations, thereby enhancing cytosolic siRNA delivery efficiency and specificity.

  3.2 siRNA therapeutics endo/lysosomal escaping strategies for intracellular transporting via "endosome-Golgi-ER" pathway

  In order to avoid the siRNA degradation in lysosomes, researchers have been committed to improving the lysosomal escape efficiency of siRNAs through various delivery strategies. However, modulating the intracellular fate of siRNAs is a more effective way to avoid degradation and exert gene silencing effect, as the bottleneck of lysosome entrapment degradation can be solved from the source by circumventing the lysosomes.

  CvME represents an alternative pathway that bypass lysosomes for efficient cytoplasmic delivery of siRNAs. Song et al. [29] constructed a mannose-modified trimethylchitosan cysteine/tripolyphosphoryl nanoparticles (MTC/TPP NPs) to deliver TNF-α siRNAs, and thoroughly studied their intracellular transport process to uncover the underlying molecular mechanisms. The results revealed that siRNA-loaded NPs were mainly internalized via CvME, then were shipped into the Golgi-ER, bypassing the lysosomal degradation pathway, which contributed to the enhanced gene silencing. And the study also proved that syntaxin6 and Niemann-Pick type C1 (NPC1) were the key regulators for this cytosolic siRNA delivery of MTC/TPP NPs. Wang et al. [93] reported a polypeptide-polysaccharide conjugate, KDEL-grafted chondroitin sulfate (CK), with a function of altering the intracellular delivery path of siRNAs. CK composed of endoplasmic reticulum retention signaling peptide KDEL and natural polysaccharide chondroitin sulfate, was adsorbed on the surface of liposomes for delivering autophagy-related gene 7 (ATG7) siRNAs. Lip/siATG7/CK provided excellent RNAi efficiency to hepatic stellate cells (HSCs) by acting as a boat that followed the CD44-Golgi ER flow rather than entering the lysosomes.

  In addition, ER membrane is a promising tool for controlling the intracellular destiny of siRNAs via the endosome-Golgi-ER pathway. Being the largest intracellular membrane system, the ER membrane (EM) has the capability to facilitate directed transport between the Golgi complex and the ER, utilizing coat protein complex I (COPI) and COPII vesicles. Thus biomimetic delivery systems fused this natural cell membrane or organelle membrane into delivery vectors can be used for cytosolic delivery of siRNAs. Qiu et al. [94] developed an EM-decorated hybrid siEGFR-loaded liposomes to enhance siRNA transfection by modulating the cellular fate of siRNAs. The results showed that the internalization of EM-decorated NPs primarily relied on the CvME pathway via soluble N-ethylmaleimide-sensitive factor attachment protein receptors (SNARE) proteins expression in EM, and had a markedly increased retention within the ER. The gene silencing effect of the EM-decorated NPs was 35% superior to that of the unmodified nanoparticles (Fig. 6).

  Figure 6

  

  Figure 6.  Schematic illustration of EhCv/siRNA that was transported intracellularly via "endosome-Golgi-ER" pathway. (a) EM was fused with siRNA-loaded cationic vesicles. (b) EhCv/siRNA would transport siRNA via Golgi-ER pathway, thus avoiding lysosomal degradation and enhancing siRNA silencing effects. Copied with permission [94]. Copyright 2019, The Author(s).

  DownLoad: Full-Size Img PowerPoint

  3.3 siRNA therapeutics endo/lysosomal escaping strategies via nonendocytosis pathway

  While the bottleneck of endo/lysosomal escape in siRNA delivery can also be solved from the source by circumventing the endocytic pathway, the following section will describe strategies for direct cytosolic delivery of siRNAs. We will review the benefits and drawbacks of these systems, as well as tackle the challenges of cytosolic delivery in siRNA delivery (Fig. 7).

  Figure 7

  

  Figure 7.  Schematic representation of nonendocytosis pathways that can bypass endo/lysosomes (By Figdraw).

  DownLoad: Full-Size Img PowerPoint

  3.3.1   Cytosolic delivery of siRNA delivery via cell-penetrating peptides

  CPPs are extensively utilized to facilitate the intracellular transport of diverse biomolecules, encompassing peptides, proteins, nucleic acids, and nanoparticles [95,96]. Studies have shown that CPPs can deliver siRNAs directly into the cytosol in a manner independent of endocytosis [97]. Pan et al. [98] developed a novel series of CPPs by incorporating stearic acid and histidine, valine, and lysine residues. Among these, a specific CPP termed STR-KV (stearoylated-HHHKKKVVVVV) was selected for siRNA delivery and exhibited the strongest capability in inducing gene silencing. To elucidate the internalization pathway of the STR-KV/siRNA complex, the researchers employed various inhibitors of endocytosis and low-temperature treatment strategies. The results indicated that internalization was insensitive to these treatments. Subsequent intracellular trafficking and internalization dynamics experiments conducted that the STR-KV/siRNA complex delivered siRNAs through a non-endocytic pathway involving direct membrane translocation.

  CADY is another amphiphilic CPP capable of forming a robust complex with siRNAs, and the CADY/siRNA complexes significantly trigger temporary membrane permeability, which is promptly recovered through membrane fluidity. The researchers verified that the primary mechanism by which CADY/siRNA complexes enter the cells is direct transmembrane transport. CADY is able to interact with phospholipids on cell membranes, helping to disrupt cell membranes and increasing uptake. Subsequently, the complex quickly localizes to the cytoplasm through a process that operates independently of endocytic pathways, ensuring that cell integrity and viability remain unimpaired [99,100].

  Although CPPs have been widely investigated for macromolecules delivery, their cell uptake mechanisms are not fully understood, and in some cases controversial. According to the generally accepted statement, the cellular uptake pathways of CPPs include both non-endocytic and endocytic pathways, each with its own unique characteristics [101,102]. We have previously investigated cell uptake mechanisms of the CPP-siRNA conjugates, and revealed that these conjugates were internalized by cells through various routes, which included direct traversal of the plasma membrane as well as caveolae- and clathrin-independent endocytic processes [101]. And a growing body of research revealed that the cell uptake mechanism of CPPs depended on their physicochemical characteristics, the chosen experimental settings, as well as cargos. Therefore, application of CPPs to overcome endo/lysosomal barriers in siRNA delivery should be designed based on the intrinsic property of siRNAs as well as the physicochemical characteristics of CPPs and carriers [103,104].

  3.3.2   Cytosolic delivery of siRNA delivery via pH low insertion peptides (pHLIP)

  pHLIP is a pH-induced transmembrane structure derived from bacteriorhodopsin helix C, which can reversibly form transmembrane α-helices and insert into the cell membrane in acidic environments, facilitating the transport of hydrophobic molecules across the membrane through a non-endocytic transport mechanism [105,106]. Zhao et al. [107] constructed a peptide-siRNA conjugates wherein pHLIP were covalently attached to CDCA1 siRNAs through a disulfide linkage, enabling specific delivery to prostate cancer (PCa) cells. The localized acidic environment of prostate cancer triggered the allosteric formation of a transmembrane helix by pHLIP, resulting in the formation of a transmembrane helix that enabled insertion into the lipid bilayer of the cell membrane. Consequently, the CDCA1 siRNA, linked to the pHLIP complex, successfully achieved cytosolic transport. In the intracellular translocation experiments, delivery of CDCA1 siRNAs to the cytosol was realized at pH 6.2, whereas it did not happen at the physiological pH 7.4. Subsequently, the disulfide bond linking pHLIP to CDCA1 siRNA was cleaved within the cytosolic reducing environment, releasing the free siRNA to exert its gene-silencing effect. Similarly, Son et al. [108] constructed a siRNA delivery system by conjugating the peptide nucleic acid (PNA) analog of siRNAs that targeting CEACAM6 to a peptide possessing pHLIP, resulting in the pHLIP-siCEACAM6 conjugates for the transmembrane transport of siRNAs. The PNA and peptide were connected by a disulfide bond that was reducible within the cytosol; hence, the linkage of PNA to pHLIP facilitated the delivery of PNA into the cell. Studies revealed that exposed to pHLIP-siCEACAM6 under acidic conditions showed reduced expression of the native CEACAM6 protein and exhibited significant tumor suppressive effects.

  3.3.3   Cytosolic delivery of siRNA delivery via scavenger receptors Bi (SR-BI)-mediated nonendocytosis

  The high-density lipoprotein (HDL) constitutes a naturally occurring nanoparticle composed of a complex array of biomacromolecules and is involved in cholesterol transport. Apolipoprotein A-1 (APOA-1) is the most significant and predominant protein within HDL [109], providing an innate interaction with the SR-BⅠ receptor located on the membrane surface. Due to this interaction, hydrophobic high-density lipoprotein cholesterol esters (HDL-CE) can diffuse into the cytosol through a non-aqueous "channel" provided by SR-BI, without the requirement for endocytosis, a mechanism that facilitates direct cytosolic delivery of siRNAs and avoids siRNA degradation after endocytosis. Studies found that SR-B1 receptors were overexpressed on the membranes of malignant tumor cells. Therefore, HDL was a promising delivery vector of siRNAs for tumor therapy [110]. This siRNA delivery strategy bypasses the endo/lysosomal pathway, enhancing the amplification of siRNA gene-silencing effects [111,112]. Yang et al. [113] described a HDL mimetic nanocarrier, termed peptide-phospholipid scaffold (HPPS), which facilitated direct delivery of siRNA into the cytosol via the SR-BI pathway. The HPPS was designed to emulate ApoA-1, the primary apolipoprotein in HDL, and was primarily constructed from a combination of cholesterol oleate, phospholipidsand amphipathic α-helical peptides. Compared to Lipofectamine 2000, the majority of HPPS nanoparticles were found to localize within the cytosol of target cells. Moreover, the HPPS nanoparticles retained the ability to target cancer cells expressing SR-BI. Following treatment with HPPS-chol-siBcl2, a dose-dependent decrease in Bcl2 protein level was observed. This strategy effectively delivered siRNAs directly to the cytoplasm, circumventing the risk of degradation that siRNAs faced upon being captured by endo/lysosomes following cellular internalization. Ding et al. [114] engineered a recombinant form of HDL (rHDL) designed for the delivery of cholesterol-conjugated siRNA (Chol-siRNA) into the cytosol of tumor cells, with a focus on targeting breast cancer angiogenesis for therapeutic purposes. Initially, they successfully prepared a liposome complex attached with apolipoprotein APOA-I carrying Chol-siRNA (Lipo/Chol-siRNA complex). Cell uptake experiments demonstrated that FAM-Chol-siRNA bypassed endosomal capture and was predominantly localized in the cytosol. Concurrently, the marker for rHDL, Cy5.5-APOA-I, was observed to remain associated with the cell membrane after 6 h incubation period. These findings suggested that Chol-siRNA can directly permeate into the cytosol via a non-endocytic "portal" provided by SR-BI, thereby evading lysosomal degradation of the siRNAs.

  3.3.4   Cytosolic delivery of siRNA delivery via membrane fusion

  Fusogenic liposomes have been proposed for siRNA delivery, as they can directly fuse with the cell membrane, bypassing endocytosis [115]. However, liposomes have a low capacity for nucleic acid payload, and effective cargo leakage may occur during storage. Therefore, Kim et al. [116] developed a siRNA delivery platform that combined fusogenic lipids with solid porous silicon nanoparticles (pSiNPs) as the core. This system efficiently loaded siRNAs through a calcium precipitation strategy. When conjugated with tumor-homing peptides, fusion nanoparticles (FNPs) exhibited strong gene silencing effects on macrophages. DiI labeled FNPs displayed fusion and transferred to the plasma membrane. And the cells showed minimal colocalization between FNPs and lysosomes, indicating that FNPs were not internalized via endocytosis. They proposed that the fusogenic lipids contained in the nanoparticles promote direct merging with the cell membrane, thereby circumventing typical receptor-mediated endocytotic pathways. Zhao et al. [117] developed a polymer-locking fusogenic liposome that can cross BBB and deliver siRNA into the cytoplasm. The researchers point out that this fusogenic liposome, equipped with a reversible ROS-responsive "lock", only fuses after penetrating the BBB and entering the elevated ROS environment typical of GBM tissues. FRET experimental results demonstrated that 4-arm PEG-ODP modulated the structural stability of Plofsome by regulating the lateral fluidity of lipid molecules, thus directly releasing siMDK from cytoplasm. Wang et al. [118] designed cell-derived nanovesicles, termed eFT-CNVs, which co-expressed GPC3-targeting scFv and fusogenic proteins to achieve targeted drug delivery to the cytosol. The bioengineered fusogenic protein is a glycoprotein with a binding defect but retaining fusion capabilities, derived from the Sendai virus. After incubating eFT-CNVs with HepG2 cells for 20 min, the authors found that eFT-CNVs fused with the HepG2 plasma membrane. Cell experiments confirmed that CNVs labeled with PKH67 were barely absorbed without anti-GPC3 scFv, indicating that eFT-CNV could evade lysosomal degradation. Subsequently, Sox2 siRNAs were loaded into the nanovesicles using electroporation for cytoplasmic delivery. The results demonstrated that eFT-CNV could effectively bind to cancer cells overexpressing GPC3, induced membrane fusion, and achieved endo/lysosomal escape, thereby accomplishing cytoplasmic drug delivery. Zhuo et al. [119] demonstrated that exosomes enriched with cholesterol have a greater tendency to enter cancer cells through membrane fusion, facilitating direct siRNA cytoplasmic delivery. Using molecular dynamics (MD) simulations, they found that exosomes with membrane-enriched cholesterol undergo deformation, which expands their surface area in contact with the target cell membrane, thereby leading to membrane fusion. siPLK1-loaded exosomes modified with 30% cholesterol inhibited the expression of PLK1. Subsequently, in vivo studies confirmed the efficacy of 30% Chol MEs/siPLK1 through both oral and intravenous administration, showing that they successfully inhibited the growth of both orthotopic and ectopic colorectal tumors.

  Furthermore, to overcome the obstacles of siRNA cytoplasmic delivery, Tai et al. reported a bifunctional chemical tag designed for direct transport of siRNAs into the cytoplasm. This tag comprised a siRNA-binding component that associated non-covalently with siRNAs and a steroid segment capable of merging with the cell membrane. Unlike traditional covalently conjugated siRNA-steroid complexes, which primarily entered cells through endocytosis, the non-covalently labeled siRNAs were cell membrane-permeable, circumventing the endocytic pathway [120]. Following fusion with the cell membrane, siRNAs have the potential to release the hydrophobic components, allowing it to slide into the cytoplasm rather than remaining attached to the membrane [121].

  4. Conclusions and perspectives

  siRNA has become the focus of global drug discovery and development, and its application in disease treatment is increasingly widespread. Researchers have recently devoted themselves to overcoming the extracellular and intracellular barriers of siRNA delivery using nanotechnology, coupling technology, chemical modification, etc. Among which, the dilemma of lysosomal entrapment has not received enough attention in current research. Studies showed that ~99% siRNAs will be trapped in endo/lysosomes after endocytosis by clathrin or macropinocytosis and unable to enter the cytoplasm to display gene silencing effect, which remains as the most daunting step for intracellular delivery of siRNAs.

  In this case, various pH-sensitive nanocarriers, peptides, toxins, and photosensitizers were used to promote endo/lysosomal escape, however there are still many challenges in future scientific research or clinical transformation. (1) The endosomal escape problem should be solved in a nontoxic manner. It has to be admitted that the methods mentioned above can effectively promote the endo/lysosomal escape efficiency of siRNAs, but few studies have conducted a comprehensive evaluation of physiological toxicity caused by endo/lysosomal destruction, which is precisely the necessary condition for clinical acceptance. Lysosomes are important degradation and recycling stations within cells, containing a variety of hydrolases. The methods of causing membrane rupture, such as proton sponge effect, membrane destabilization, osmosis lysis effect, may lead to the leakage of lysosomal contents into the cytoplasm, not only causing disruption of the intracellular environment, but also activating innate immunity or other toxic pathways, ultimately resulting in cytotoxicity to varying degrees [122]. Certain molecules that promote lysosomal escape may be immunostimulatory, such as certain cationic lipids or polymers, thereby increasing the risk of immune-related side effects. Moreover, some methods, such as the proton sponge effect, may cause DNA damage and increase the risk of gene mutations. Contrastingly, PCI typically shows low cytotoxicity, and local immune reactions are usually mild, only high doses of photosensitizers can cause cell damage. In addition, methods that bypass the lysosomes also have shown lower toxicity, as the integrity of the lysosomes is not compromised. Therefore, improving endo/lysosomal escape efficiency while maintaining low toxicity is a key balance. In particular, the evaluation in vivo models is of greater significance, the research should not only be limited to short-term biosafety, but also should pay enough attention to the impact of long-term exposure to nanomaterials [123]. (2) Better techniques are needed to quantify endosomal escape of siRNA therapeutics. Current endosomal escape assays rely on indirect and qualitative measurements which is not sufficient to accurately visualize and quantify the escape of siRNAs. In general, most of lysosomal escape assays are conducted by colocalization studies via multicolor fluorescence microscopy. In these experiments, siRNA and endo/lysosomes are simultaneously labeled with fluorescent molecules or probes, their location within the cell were examined and characterized by the Pearson's coefficient. The assay is subjective, because it relies on the researcher's threshold for determining whether or not escape has occurred based on intracellular distribution of fluorescent signals from siRNA and endo/lysosomes. In addition, experimental results are influenced by multiple factors, including the property of siRNA carriers, the interaction between siRNA and carriers (covalent or noncovalent), etc. Therefore, new techniques with high sensitivity to yield quantitative information of lysosomal escape efficiency, and methods that enable high throughput analysis and screening of large numbers of samples are urgently needed. These techniques are critical for guiding the development and rational design of siRNA delivery systems [25,124]. (3) Verified the exact escape site will provide vital insights for additional enhancements in siRNA delivery. The endosomal-lysosomal system is composed of a series of intracellular membranous compartments, including EE, LE, and lysosome. These intracellular membranous compartments serve different roles within the cell and differ in contents. Once endocytosed by cells, siRNA loaded nanoparticles are constantly interacting with the surrounding environments, thus the nanoparticles escaping from different sites (EE, LE, lysosomes) may be difference due to differences in physichemical properties, which in turn affects the siRNA delivery efficiency. Therefore, endo/lysosomal escape sites of different nanoparticles and the relationship between carrier properties and escape sites needs thoroughly investigation to provide information that guiding the development of siRNA delivery system. (4) Modulating the intracellular fate of siRNAs would benefits from a better understanding of intracellular transport pathway. From the perspective of intracellular transport mechanism, many studies have proved that CvME can bypass the lysosome and carry out intracellular transport through the endosome-Golgi-ER pathway, a method regarded as more efficient and safe. However, which key signaling molecules that participate in the regulation of intracellular transport? In addition, are there other pathways that can mediate the efficient intracellular transport of siRNAs, and what is the relationship between these pathways? Unraveling these questions will provide vital insights for advancing siRNA delivery and enhance the efficacy of siRNA-based therapeutic.

  Despite all the challenges, we are witnessing the thriving development of siRNA therapeutics. The smart designs of multifunctional carriers integrated with pH, photochemical-sensitive modules or peptides further enhance the efficacy of siRNA delivery. Nonetheless, many gaps still exist between preclinical and clinical researches for siRNA development, which currently hinder the clinical translation of siRNA. Improving the preclinical characterization of siRNA preparations, gaining a deeper consideration of the clinical approval criteria, reducing the discrepancies between in vitro and in vivo studies are likely to enhance the success rate of siRNAs in clinical applications. With the development, it is believed that more and more siRNA therapeutics will be proved for clinical use.

  Declaration of competing interest

  The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

  CRediT authorship contribution statement

  Rui Li: Writing – original draft. Mengxi Zhu: Writing – original draft. Xiwen Hu: Writing – original draft. Jiaxuan Chen: Investigation. Fei Yu: Investigation. Stefan Barth: Supervision. Lu Sun: Writing – review & editing. Huining He: Writing – review & editing.

  Acknowledgments

  This work was supported by National Natural Science Foundation of China (No. 82173769), the National Key R & D Program of China (No. 2021YFE0106900), Applied Basic Research Multi-investment Foundation of Tianjin (No. 21JCYBJC01540) and the Science & Technology Development Fund of Tianjin Education Commission for Higher Education (No. 2023ZD019).

  
  
• 1. [1]

     R.L. Setten, J.J. Rossi, S.P. Han, Nat. Rev. Drug. Discov. 18 (2019) 421–446. doi: 10.1038/s41573-019-0017-4

  2. [2]

     A. Fire, S. Xu, M.K. Montgomery, et al., Nature 391 (1998) 806–811. doi: 10.1038/35888

  3. [3]

     S.M. Elbashir, J. Harborth, W. Lendeckel, et al., Nature 411 (2001) 494–498. doi: 10.1038/35078107

  4. [4]

     M.A. Ashfaq, V.D. Kumar, P.S.S. Reddy, et al., J. Biosci. 45 (2020) 128. doi: 10.1007/s12038-020-00098-3

  5. [5]

     M. Friedrich, A. Aigner, BioDrugs 36 (2022) 549–571. doi: 10.1007/s40259-022-00549-3

  6. [6]

     Q. Tang, A. Khvorova, Nat. Rev. Drug Discov. 23 (2024) 341–364. doi: 10.1038/s41573-024-00912-9

  7. [7]

     D. Adams, A. Gonzalez-Duarte, W.D. O'Riordan, et al., N. Engl. J. Med. 379 (2018) 11–21. doi: 10.1056/nejmoa1716153

  8. [8]

     J.R. Androsavich, Nat. Rev. Drug Discov. 23 (2024) 421–444. doi: 10.1038/s41573-024-00943-2

  9. [9]

     B. Hu, L. Zhong, Y. Weng, et al., Signal Transduct. Target. Ther. 5 (2020) 101. doi: 10.1038/s41392-020-0207-x

  10. [10]

      L. Zhang, Z. Dong, S. Yu, et al., Chin. Chem. Lett. 35 (2024) 109101. doi: 10.1016/j.cclet.2023.109101

  11. [11]

      S. Schenk, G.J. Schoenhals, G. de Souza, M. Mann, BMC Med. Genom. 1 (2008) 41. doi: 10.1186/1755-8794-1-41

  12. [12]

      B. Godinho, E.G. Knox, S. Hildebrand, et al., Mol. Ther. Nucleic. Acids 29 (2022) 116–132. doi: 10.1016/j.omtn.2022.06.005

  13. [13]

      A. Biscans, A. Coles, R. Haraszti, et al., Nucleic Acids Res. 47 (2019) 1082–1096. doi: 10.1093/nar/gky1239

  14. [14]

      J. Zhang, B. Chen, C. Gan, et al., Int. J. Nanomed. 18 (2023) 7605–7635. doi: 10.2147/ijn.s436038

  15. [15]

      S.F. Dowdy, RNA 29 (2023) 396–401. doi: 10.1261/rna.079507.122

  16. [16]

      S. Guo, M. Zhang, Y. Huang, Trends Mol. Med. 30 (2024) 13–24.

  17. [17]

      K. Paunovska, D. Loughrey, J.E. Dahlman, Nat. Rev. Genet. 23 (2022) 265–280. doi: 10.1038/s41576-021-00439-4

  18. [18]

      Y. Jia, X. Wang, L. Li, et al., Adv. Mater. 36 (2024) e2305300. doi: 10.1002/adma.202305300

  19. [19]

      K. Alamoudi, P. Martins, J.G. Croissant, et al., Nanomedicine 12 (2017) 1421–1433. doi: 10.2217/nnm-2017-0021

  20. [20]

      J.P. Bost, H. Barriga, M.N. Holme, et al., ACS Nano 15 (2021) 13993–14021. doi: 10.1021/acsnano.1c05099

  21. [21]

      B. Hu, S. Kong, Y. Weng, et al., Chin. Chem. Lett. 34 (2023) 108210. doi: 10.1016/j.cclet.2023.108210

  22. [22]

      F. Hu, J. Qi, Y. Lu, H. He, W. Wu, Chin. Chem. Lett. 34 (2023) 108250. doi: 10.1016/j.cclet.2023.108250

  23. [23]

      M. Sousa de Almeida, E. Susnik, B. Drasler, et al., Chem. Soc. Rev. 50 (2021) 5397–5434. doi: 10.1039/d0cs01127d

  24. [24]

      X. Wang, Y. Yin, H. Zhou, et al., Exploration 5 (2025) 20240036. doi: 10.1002/EXP.20240036

  25. [25]

      J. Gilleron, W. Querbes, A. Zeigerer, et al., Nat. Biotechnol. 31 (2013) 638–646. doi: 10.1038/nbt.2612

  26. [26]

      Y. Xia, M. Guo, T. Xu, et al., Int. J. Nanomed. 13 (2018) 1539–1552. doi: 10.2147/ijn.s157519

  27. [27]

      E. Linnane, S. Haddad, F. Melle, Z. Mei, D. Fairen-Jimenez, Chem. Soc. Rev. 51 (2022) 6065–6086. doi: 10.1039/d0cs01414a

  28. [28]

      A.L. Kiss, E. Botos, J. Cell. Mol. Med. 13 (2009) 1228–1237. doi: 10.1111/j.1582-4934.2009.00754.x

  29. [29]

      Y. Song, Y. Wu, L. Xu, et al., ACS Nano 15 (2021) 8267–8282. doi: 10.1021/acsnano.0c08596

  30. [30]

      S. Patel, J. Kim, M. Herrera, et al., Adv. Drug Deliv. Rev. 144 (2019) 90–111. doi: 10.1016/j.addr.2019.08.004

  31. [31]

      M. Sako, F. Song, A. Okamoto, et al., Int. J. Pharm. 569 (2019) 118606. doi: 10.1016/j.ijpharm.2019.118606

  32. [32]

      M. Lu, X. Zhao, H. Xing, et al., Int. J. Pharm. 550 (2018) 100–113. doi: 10.1016/j.ijpharm.2018.08.040

  33. [33]

      J. Sun, C. Qiu, Y. Diao, et al., Mol. Ther. Nucleic. Acids 10 (2018) 75–90. doi: 10.1016/j.omtn.2017.11.002

  34. [34]

      C. Qiu, F. Xia, J. Zhang, et al., Research 6 (2023) 0148. doi: 10.34133/research.0148

  35. [35]

      E. Fakhr, F. Zare, L. Teimoori-Toolabi, Cancer Gene Ther. 23 (2016) 73–82. doi: 10.1038/cgt.2016.4

  36. [36]

      J. Zhu, M. Qiao, Q. Wang, et al., Biomaterials 162 (2018) 47–59. doi: 10.1016/j.biomaterials.2018.01.042

  37. [37]

      D. Chen, P. Zhang, M. Li, et al., Drug Deliv. 29 (2022) 574–587. doi: 10.1080/10717544.2022.2032874

  38. [38]

      M. Tuttolomondo, S.T.D. Pham, M.G. Terp, et al., Mol. Ther. Nucleic. Acids 35 (2024) 102227. doi: 10.1016/j.omtn.2024.102227

  39. [39]

      J. Lee, I. Sands, W. Zhang, L. Zhou, Y. Chen, Proc. Natl. Acad. Sci. U. S. A. 118 (2021) e2104511118. doi: 10.1073/pnas.2104511118

  40. [40]

      S. Hyun, Y. Choi, H.N. Lee, et al., Chem. Sci. 9 (2018) 3820–3827. doi: 10.1039/c8sc00074c

  41. [41]

      Y. Dang, Y. Feng, X. Chen, et al., Chin. Chem. Lett. 35 (2024) 109660. doi: 10.1016/j.cclet.2024.109660

  42. [42]

      C. Li, J. Zhou, Y. Wu, et al., Nano Lett. 21 (2021) 3680–3689. doi: 10.1021/acs.nanolett.0c04468

  43. [43]

      Y. Wu, W. Gu, J. Tang, Z.P. Xu, J. Mater. Chem. B 5 (2017) 7194–7203. doi: 10.1039/C7TB01635B

  44. [44]

      C. Yu, K. Li, L. Xu, et al., Nano Res. 15 (2022) 9160–9168. doi: 10.1007/s12274-022-4573-2

  45. [45]

      J. Wang, H. Ding, Y. Zhu, et al., Angew Chem. Int. Ed. 62 (2023) e202302255. doi: 10.1002/anie.202302255

  46. [46]

      B. Hu, B. Li, K. Li, et al., Sci. Adv. 8 (2022) eabm1418. doi: 10.1126/sciadv.abm1418

  47. [47]

      F. Joris, L. De Backer, T. Van de Vyver, et al., J. Control. Release 269 (2018) 266–276. doi: 10.1016/j.jconrel.2017.11.019

  48. [48]

      E.J. Kwon, J.M. Bergen, S.H. Pun, Bioconjug. Chem. 19 (2008) 920–927. doi: 10.1021/bc700448h

  49. [49]

      S. Ben Djemaa, K. Hervé-Aubert, L. Lajoie, et al., Biomacromolecules 20 (2019) 3076–3086. doi: 10.1021/acs.biomac.9b00637

  50. [50]

      L. Zheng, S.R. Bandara, Z. Tan, C. Leal, Proc. Natl. Acad. Sci. U. S. A. 120 (2023) e2301067120. doi: 10.1073/pnas.2301067120

  51. [51]

      X. Cao, X. Shang, Y. Guo, et al., Anal. Bioanal. Chem. 413 (2021) 3493–3499. doi: 10.1007/s00216-021-03297-5

  52. [52]

      S. Oliveira, M.M. Fretz, A. Høgset, G. Storm, R.M. Schiffelers, Biochim. Biophys. Acta 1768 (2007) 1211–1217. doi: 10.1016/j.bbamem.2007.01.013

  53. [53]

      J.A. Jørgensen, A.S. Longva, E. Hovig, S.L. Bøe, Nucleic Acid Ther. 23 (2013) 131–139. doi: 10.1089/nat.2012.0403

  54. [54]

      Z. Zhang, M.K.G. Jayakumar, X. Zheng, et al., Nat. Commun. 10 (2019) 4586. doi: 10.1109/cvpr.2019.00472

  55. [55]

      B. Mezghrani, L.M.A. Ali, S. Richeter, et al., ACS Appl. Mater. Interfaces 13 (2021) 29325–29339. doi: 10.1021/acsami.1c05848

  56. [56]

      A. Puri, M. Viard, P. Zakrevsky, et al., Nanomedicine 26 (2020) 102176.

  57. [57]

      Y. Yang, H. Ning, T. Xia, et al., Adv. Mater. 35 (2023) e2301409. doi: 10.1002/adma.202301409

  58. [58]

      Y. Mo, M.H.Y. Cheng, A. D'Elia, K. Doran, et al., ACS Nano 17 (2023) 4688–4703. doi: 10.1021/acsnano.2c10936

  59. [59]

      M.K. Jayakumar, A. Bansal, K. Huang, et al., ACS Nano 8 (2014) 4848–4858. doi: 10.1021/nn500777n

  60. [60]

      J. Wang, L. Xie, T. Wang, et al., Acta Biomater. 59 (2017) 158–169. doi: 10.3901/JME.2017.08.158

  61. [61]

      S. Roy, D. Zhu, W.J. Parak, N. Feliu, ACS Nano 14 (2020) 8012–8023. doi: 10.1021/acsnano.9b10219

  62. [62]

      Y. Ma, M. Sha, S. Cheng, et al., Nanomaterials 8 (2018) 433. doi: 10.3390/nano8060433

  63. [63]

      C. Wang, T. Zhao, Y. Li, et al., Adv. Drug Deliv. Rev. 113 (2017) 87–96.

  64. [64]

      M. Wojnilowicz, A. Glab, A. Bertucci, F. Caruso, F. Cavalieri, ACS Nano 13 (2019) 187–202. doi: 10.1021/acsnano.8b05151

  65. [65]

      Y.J. Choi, S.J. Kang, Y.J. Kim, Y.B. Lim, H.W. Chung, Drug Chem. Toxicol. 33 (2010) 357–366. doi: 10.3109/01480540903493507

  66. [66]

      X. Xu, Z. Li, X. Zhao, L. Keen, X. Kong, Regen. Biomater. 3 (2016) 187–195. doi: 10.1093/rb/rbw010

  67. [67]

      K. Dong, Z. Wang, Y. Zhang, J. Ren, X. Qu, ACS Appl. Mater. Interfaces 10 (2018) 31998–32005. doi: 10.1021/acsami.8b11972

  68. [68]

      Y. Sun, Y. Sha, G. Cui, F. Meng, Z. Zhong, Adv. Drug Deliv. Rev. 192 (2023) 114624.

  69. [69]

      E. Pandey, E.N. Harris, Mol. Ther. Nucleic Acids 33 (2023) 430–443.

  70. [70]

      H. Du Rietz, H. Hedlund, S. Wilhelmson, P. Nordenfelt, A. Wittrup, Nat. Commun. 11 (2020) 1809.

  71. [71]

      S.R. Bhattarai, E. Muthuswamy, A. Wani, et al., Pharm. Res. 27 (2010) 2556–2568. doi: 10.1007/s11095-010-0245-0

  72. [72]

      M. Lu, H. Xing, A. Zheng, Y. Huang, X.J. Liang, Acc. Chem. Res. 56 (2023) 224–236. doi: 10.1021/acs.accounts.2c00464

  73. [73]

      S. Liu, Q. Cheng, T. Wei, et al., Nat. Mater. 20 (2021) 701–710. doi: 10.1038/s41563-020-00886-0

  74. [74]

      Q. Wang, Q. Jiang, D. Li, et al., Chin. Chem. Lett. 35 (2024) 108683.

  75. [75]

      A. Debisschop, B. Bogaert, C. Muntean, S.C. De Smedt, K. Raemdonck, Curr. Opin. Chem. Biol. 83 (2024) 102531.

  76. [76]

      A.K. Varkouhi, M. Scholte, G. Storm, H.J. Haisma, J. Control. Release 151 (2011) 220–228.

  77. [77]

      E.E. Haywood, N.B. Handy, J.W. Lopez, M. Ho, B.A. Wilson, J. Biol. Chem. 297 (2021) 101347.

  78. [78]

      A.K. Mondal, K. Lata, M. Singh, et al., Biochim. Biophys. Acta Biomembr. 1864 (2022) 184013.

  79. [79]

      F.R. Ulhuq, G. Mariano, Microbiology 168 (2022) 001154.

  80. [80]

      I. Plaza-Ga, V. Manzaneda-González, M. Kisovec, et al., J. Nanobiotechnology 17 (2019) 108.

  81. [81]

      D.W. Schuerch, E.M. Wilson-Kubalek, R.K. Tweten, Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 12537–12542. doi: 10.1073/pnas.0500558102

  82. [82]

      R.K. Tweten, E.M. Hotze, K.R. Wade, Annu. Rev. Microbiol. 69 (2015) 323–340. doi: 10.1146/annurev-micro-091014-104233

  83. [83]

      A. Sáez-Cirión, S. Nir, M. Lorizate, et al., J. Biol. Chem. 277 (2002) 21776–21785.

  84. [84]

      B.S. Joshi, M.A. de Beer, B.N.G. Giepmans, I.S. Zuhorn, ACS Nano 14 (2020) 4444–4455. doi: 10.1021/acsnano.9b10033

  85. [85]

      X. Pei, X. Zhang, L. Zhang, et al., Chem. Eng. J. 421 (2021) 129774.

  86. [86]

      A. Wittrup, A. Ai, X. Liu, et al., Nat. Biotechnol. 33 (2015) 870–876. doi: 10.1038/nbt.3298

  87. [87]

      Z. Zhao, P. Qin, Y.W. Huang, Cell Calcium 94 (2021) 102360.

  88. [88]

      Y. Wang, B. Jin, B. Li, et al., Int. J. Pharm. 623 (2022) 121890.

  89. [89]

      D.J. Benton, S.J. Gamblin, P.B. Rosenthal, J.J. Skehel, Nature 583 (2020) 150–153. doi: 10.1038/s41586-020-2333-6

  90. [90]

      Y. Xu, Y. Zheng, L. Wu, et al., ACS Appl. Mater. Interfaces 10 (2018) 9315–9324. doi: 10.1021/acsami.8b00507

  91. [91]

      Y. Wan, P.M. Moyle, M.P. Christie, I. Toth, Nanomedicine 11 (2016) 907–919. doi: 10.2217/nnm.16.27

  92. [92]

      L.M.A. Ali, M. Gary-Bobo, Cancers 14 (2022) 3597. doi: 10.3390/cancers14153597

  93. [93]

      Y. Wang, L. Xiong, Z. Dong, et al., ACS Appl. Mater. Interfaces 15 (2023) 28941–28953. doi: 10.1021/acsami.3c03416

  94. [94]

      C. Qiu, H.H. Han, J. Sun, et al., Nat. Commun. 10 (2019) 2702.

  95. [95]

      J. Zhou, Z. Shao, J. Liu, et al., ACS Appl. Bio. Mater. 3 (2020) 2686–2701. doi: 10.1021/acsabm.9b01131

  96. [96]

      M.M. Khan, N. Filipczak, V.P. Torchilin, J. Control. Release 330 (2021) 1220–1228.

  97. [97]

      S.S. Ali Zaidi, F. Fatima, S.A. Ali Zaidi, et al., Nanobiotechnology 21 (2023) 381.

  98. [98]

      R. Pan, W. Xu, Y. Ding, S. Lu, P. Chen, Mol. Pharm. 13 (2016) 1366–1374. doi: 10.1021/acs.molpharmaceut.6b00030

  99. [99]

      A. Rydström, S. Deshayes, K. Konate, et al., PLoS One 6 (2011) e25924. doi: 10.1371/journal.pone.0025924

  100. [100]

       K. Konate, A. Rydstrom, G. Divita, S. Deshayes, Curr. Pharm. Des. 19 (2013) 2869–2877. doi: 10.2174/1381612811319160004

  101. [101]

       J. Ye, X. Pei, H. Cui, et al., J. Ind. Eng. Chem. 63 (2018) 103–111.

  102. [102]

       Y. Xu, X. Zhang, N. Wang, et al., Int. J. Pharm. 584 (2020) 119469.

  103. [103]

       M. Zorko, S. Jones, Ü. Langel, Adv. Drug Deliv. Rev. 180 (2022) 114044.

  104. [104]

       A. Gori, G. Lodigiani, S.G. Colombarolli, G. Bergamaschi, A. Vitali, ChemMedChem 18 (2023) e202300236.

  105. [105]

       C. Gupta, Y. Ren, B. Mertz, Biophys. J. 115 (2018) 2403–2412.

  106. [106]

       S.Z. Hanz, N.S. Shu, J. Qian, et al., Angew Chem. Int. Ed. 55 (2016) 12376–12381. doi: 10.1002/anie.201605203

  107. [107]

       Z. Zhao, C. Li, B. Song, et al., Biochem. Biophys. Res. Commun. 503 (2018) 1761–1767.

  108. [108]

       S.M. Son, J. Yun, S.H. Lee, et al., Sci. Rep. 9 (2019) 11607.

  109. [109]

       M.G. Damiano, R.K. Mutharasan, S. Tripathy, K.M. McMahon, C.S. Thaxton, Adv. Drug. Deliv. Rev. 65 (2013) 649–662.

  110. [110]

       K.M. McMahon, C.S. Thaxton, Expert. Opin. Drug Deliv. 11 (2014) 231–247. doi: 10.1517/17425247.2014.866089

  111. [111]

       A. Kontush, M. Lindahl, M. Lhomme, et al., Handb. Exp. Pharmacol. 224 (2015) 3–51. doi: 10.1007/978-3-319-09665-0_1

  112. [112]

       R. Masimov, G. Büyükköroglu, Comb. Chem. High Throughput Screen. 26 ˘ (2023) 2541–2553. doi: 10.2174/1386207326666230406124524

  113. [113]

       M. Yang, H. Jin, J. Chen, et al., Small 7 (2011) 568–573. doi: 10.1002/smll.201001589

  114. [114]

       Y. Ding, Y. Wang, J. Zhou, et al., Biomaterials 35 (2014) 7214–7227.

  115. [115]

       V.P. Torchilin, Nat. Rev. Drug Discov. 4 (2005) 145–160. doi: 10.1038/nrd1632

  116. [116]

       B. Kim, S. Sun, J.A. Varner, et al., Adv. Mater. 31 (2019) e1902952.

  117. [117]

       Y. Zhao, J. Qin, D. Yu, et al., Nat. Nanotechnol. 19 (2024) 1869–1879. doi: 10.1038/s41565-024-01769-0

  118. [118]

       L. Wang, G. Wang, W. Mao, et al., Nat. Commun. 14 (2023) 3366.

  119. [119]

       Y. Zhuo, Z. Luo, Z. Zhu, et al., Nat. Nanotechnol. 19 (2024) 1858–1868. doi: 10.1038/s41565-024-01785-0

  120. [120]

       W. Tai, J. Control. Release 307 (2019) 98–107.

  121. [121]

       W. Tai, X. Gao, Biomaterials 178 (2018) 720–727.

  122. [122]

       A. Serrano-Puebla, P. Boya, Ann. N Y Acad. Sci. 1371 (2016) 30–44. doi: 10.1111/nyas.12966

  123. [123]

       S.F. Dowdy, R.L. Setten, X.S. Cui, S.G. Jadhav, Nucleic Acid Ther. 32 (2022) 361–368. doi: 10.1089/nat.2022.0004

  124. [124]

       M.A. Beach, S.L.Y. Teo, M.Z. Chen, et al., ACS Appl. Mater. Interfaces 14 (2022) 3653–3661. doi: 10.1021/acsami.1c18359

• 

  Figure 1  Mechanism of siRNA mediated gene silencing (By Figdraw).

  下载: 全尺寸图片 幻灯片
  

  Figure 2  Endocytic pathways and intracellular transport process of siRNA therapeutics (By Figdraw).

  下载: 全尺寸图片 幻灯片
  

  Figure 3  Schematic representation of endo/lysosomal escape mechanisms in the uptake of siRNA therapeutics via endosome-lysosome pathway (By Figdraw).

  下载: 全尺寸图片 幻灯片
  

  Figure 4  (A–C) Schematic of siRNA-loaded amphiphilic triblock polymers for endo/lysosomal escape. PDDT-Ms/siRNA can quickly respond to acidic pH and disassembled around pH 6.5–6.8, leading to effective endosomal escape of siRNA. Copied with permission [42]. Copyright 2021, American Chemical Society.

  下载: 全尺寸图片 幻灯片
  

  Figure 5  Schematic illustration of the intracellular trafficking and endo/lysosomal escaping of siPLK1-NB NPs via PCI. (a) NB-Br were grafted onto siPLK1 and self-assembled into NPs by electrostatic interaction. (b) The ROS generated by NB-Br could rupture the lysosome membrane under light irradiation and facilitate siRNA release. Copied with permission [57]. Copyright 2023, John Wiley & Sons, Inc.

  下载: 全尺寸图片 幻灯片
  

  Figure 6  Schematic illustration of EhCv/siRNA that was transported intracellularly via "endosome-Golgi-ER" pathway. (a) EM was fused with siRNA-loaded cationic vesicles. (b) EhCv/siRNA would transport siRNA via Golgi-ER pathway, thus avoiding lysosomal degradation and enhancing siRNA silencing effects. Copied with permission [94]. Copyright 2019, The Author(s).

  下载: 全尺寸图片 幻灯片
  

  Figure 7  Schematic representation of nonendocytosis pathways that can bypass endo/lysosomes (By Figdraw).

  下载: 全尺寸图片 幻灯片

  Table 1.  Clinically approved siRNA therapeutics.

  Generic name	Dosage form	Launch date	Target	Clinical indication	Route of administration
  Patisiran	LNP	2018	Transthyretin (TTR)	Hereditary transthyretin amyloidosis	iv
  Givosiran	GalNAc-siRNA conjugate	2019	Aminolevulinate delta synthase 1 (ALAS1)	Acute hepatic porphyria	sc
  Inclisiran	GalNAc-siRNA conjugate	2020	Proprotein convertase subtilisin/kexin type 9 (PCSK9)	Primary hypercholesterolemia or mixed dyslipidemia	sc
  Lumasiran	GalNAc-siRNA conjugate	2021	Human hydroxyacid oxidase 1 (HAO1)	Primary hyperoxaluria type1	sc
  Vutrisiran	GalNAc-siRNA conjugate	2022	TTR	Hereditary transthyretin amyloidosis	sc
  Nedosiran	GalNAc-siRNA conjugate	2023	Lactate dehydrogenase (LDH)	Primary hyperoxaluria type1	sc

  下载: 导出CSV

  Table 2.  The summary of siRNA therapeutics endo/lysosomal escaping strategies for intracellular transporting via endosome-lysosome pathway.

  Mechanism	Carrier	siRNA target site	Gene silencing efficiency	Indication	Ref.
  Proton sponge effect	PEG-b-PLA-PHis-ssPEI	Bcl-2	85.45%	Breast cancer	[36]
  	PAMAM nanocomplex	KRAS	85.3%	Lung cancer	[37]
  	HE25 peptide	SARS-CoV-2	85%	Coronavirus disease (COVID)-19	[38]
  	JBNTs with a lysine side chain	Glyceraldehyde-3-phosphate dehydrogenase (GAPDH)	90%	COVID-19	[39]
  	stEK peptide	Cyclophilin B Connective tissue growth factor (CTGF)	> 90% 49%–58%	Cervical cancer	[40]
  	Cyclic peptide modified liposomal delivery system (Cyc-LH)	PLK1	> 70%	Breast cancer	[41]
  Osmosis lysis effect	mPEG-P(DPAx-co-DMAEMAy)-PT	PLK1 PD-L1	86.9% ~45%	Colorectal cancer and liver cancer	[42]
  	CAP NPs	PD-L1 PLK1	~50% 12%–14%	Melanoma	[43]
  	NaGdF4	PD-L1	~70%	Colorectal cancer and breast cancer	[44]
  	Fe0-siRNA NPs	FHC	–	Cervical cancer	[45]
  Membrane destabilization	heat-stable iLAND	ANGTPL3	97%	Hyperlipidemia	[46]
  	CADs	EGFP PLK1	> 80%	Non-small cell lung cancer	[47]
  	HGP-PEI	GFP	50% (without HGP) vs. 80% (with HGP)	Cervical cancer	[48]
  	gH625-siRNA nanocarrier	GFP	40% (without gH625) vs. 70% (with gH625)	Triple-negative breast cancer	[49]
  Membrane fusion	RNA-LNPs with cubic and inverse hexagonal structure	–	–	Breast cancer and cervical cancer	[50]
  	HA-modified ferritin (Apn) nanocagnets (Apn/siRNA-HA)	TK1pro	–	Breast cancer	[51]
  Photochemical internalization	TPPS2a-lipo	Epidermal growth factor receptor (EGFR)	10% (PCI−) vs. 70% (PCI+)	Epidermoid carcinoma	[52]
  	TPPS2a-PLA TPPS2a-PLH TPPS2a-PLL	S100A4	15% (PCI−) vs. 90% (PCI+) 10% (PCI−) vs. 45% (PCI+) 10% (PCI−) vs. 80% (PCI+)	Osteosarcoma	[53]
  	ZnPc upconversion nanoparticles	Superoxide dismutase 1 (SOD1)	70% (PCI−) vs. 90% (PCI+)	Oral cell carcinoma	[54]
  	Porphyrin-periodic mesoporous Ionosilica nanoparticles	Luciferase	17% (PCI−) vs. 83% (PCI+)	–	[55]
  	Pyropheophorbide α-PEI	PLK1	40% (PCI+)	Breast cancer	[56]
  	NB-Br-siRNA NPs	PLK1	40% (PCI−) vs. 65% (PCI+)	Breast cancer and hepatocarcinogen	[57]
  	Porphyrin-patisiran	TTR	–	Hereditary transthyretin amyloidosis	[58]
  	TPPS2a-upconversion NPs	STAT3	25% (PCI−) vs. 50% (PCI+)	Melanoma	[59]
  	Fullerene (C60)	EGFP	17% (PCI−) vs. 53% (PCI+)	Breast cancer	[60]

  下载: 导出CSV
•