Engineered targeted enhancement of translation in mammalian cell

Abstract

Background

While the central dogma outlines DNA-to-protein information flow, existing gene regulators mainly target transcription. Here, we developed the λN-Guided RNA Targeting System (λGRTS), a CRISPR-independent platform enhancing mammalian mRNA translation via specific translation-guiding RNAs (tgRNAs).

Results

λGRTS integrates λN (high-affinity BoxB binder), mutated eIF4E1 (ablated non-specific 5′ cap binding, retains TIC recruitment), and auxiliary factors (HuR for dsRNA stabilization, PABP and RRM2-RRM3 for mRNA closed loops). Optimized 22-nt tgRNAs (targeting 58 bp upstream of mRNA ATG in 5′ UTR) and NES-tagged proteins maximized efficacy. λGRTS outperformed dCasRX (smaller ~ 50 kDa vs. ~ 150 kDa, single tgRNA vs. crRNA-tracrRNA), boosting functional proteins (e.g., GFP). It activated P53/PTEN, suppressing GBM cell proliferation, inducing G1 arrest, and reducing invasion in vitro. In vivo, lentiviral λGRTS inhibited orthotopic GBM in nude mice, extended survival by approximately 35–40 days, with IHC confirming P53/PTEN upregulation. Mass spectrometry showed no off-target effects. Cross-species tests (human, mouse, bovine cells) validated broad applicability via conserved eIF4E1.

Conclusions

λGRTS offers a specific, safe translation-centric tool for gene regulation and oncology research.

Graphical Abstract

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12915-026-02653-0.

Background

Gene expression regulatory tools, particularly those based on Cas proteins, have profoundly advanced biological research and therapeutic development by enabling precise and programmable modulation of gene expression, thereby offering innovative strategies to address genetic defects [1, 2]. The CRISPR-Cas9 system serves as a prominent representative, functioning through Cas-derived effector proteins guided by RNA molecules (gRNAs) that recognize target nucleic acids via sequence complementarity [3]. Through this mechanism, Cas proteins can be directed to specific genomic loci, where they modulate transcriptional processes and regulate gene expression [4]. Beyond transcriptional repression, CRISPR activation (CRISPRa) systems, such as those employing dCas9-VP64 fusion proteins, further expand the versatility of CRISPR technology by recruiting transcriptional activators to defined gene loci, thereby enhancing transcriptional output [5, 6].

Despite their widespread application in both prokaryotic and eukaryotic systems, CRISPR-Cas–based approaches remain constrained by several limitations. Notably, prolonged expression of exogenous Cas proteins can impose cellular stress and may lead to unintended cytotoxic effects, raising concerns regarding cellular homeostasis [7]. Additionally, RNA-targeting systems such as dCas13 have been developed for translational regulation and have demonstrated broad cross-species applicability [8]. Nevertheless, the increasing complexity and diversity of biological systems underscore an urgent need for alternative gene regulation strategies that offer greater flexibility, reduced cellular burden, and enhanced safety profiles.

In both basic research and clinical contexts, gene regulation has become an indispensable methodological cornerstone. Among existing strategies, transcriptional upregulation of target genes is widely employed to compensate for genetic deficiencies. However, a major drawback of such transcription-focused techniques often involves sustained or irreversible perturbations to gene expression programs, which may introduce unpredictable downstream effects and compromise cellular or organismal stability [9]. In contrast, RNA represents a transient and highly dynamic carrier of genetic information, rendering it an attractive substrate for reversible gene regulation. RNA-targeting approaches enable dynamic control of gene expression and allow regulatory effects to be readily attenuated or reversed in the event of adverse outcomes, thereby offering potential advantages in terms of safety and controllability for therapeutic applications [10]. The RNA epitranscriptome—encompassing RNA localization, stability, and translational efficiency—plays a central role in orchestrating RNA-mediated gene expression outcomes [11]. Although translational enhancement strategies have been extensively explored in prokaryotic systems, their implementation in mammalian cells, which possess with more intricate translational regulatory networks, remains comparatively underdeveloped, representing a critical gap in mammalian gene regulation.

To address the lack of targeted translation activation tools in mammalian cells, we propose a novel strategy designed to selectively enhance protein translation without directly altering transcriptional activity. Translation initiation is widely recognized as the rate-limiting step of protein synthesis; consequently, efficient recruitment of translation initiation complexes (TICs) is essential for boosting translational output [12]. Through systematic evaluation of multiple translation initiation factors, we identified eIF4E1 as particularly promising candidate for translational activation. Structural analyses indicate that eIF4E1 binds the 5′ cap of mRNA, a binding event that facilitates global translation initiation [13]. To achieve transcript-specific regulation, we engineered eIF4E1 via site-directed mutagenesis to disrupt its inherent 5′ cap-binding activity while preserving its capacity to recruit additional components of the translational machinery. To enable targeted recruitment of eIF4E1 to specific mRNAs, we compared three RNA–protein interaction modules—λN/BoxB, PP7cp/PP7, and MS2/MCP—and found that the λN/BoxB pair exhibited superior protein recruitment efficiency. A key advantage of the λN peptide is its small molecular size (3 kDa), which is substantially smaller than that of Cas13 proteins (63 kDa), thereby reducing the likelihood of steric hindrance and cellular perturbation. Moreover, λN displays high affinity for the BoxB RNA hairpin, ensuring stable and specific RNA–protein interactions. Based on this framework, we designed translation-guided RNAs (tgRNAs) containing both a BoxB motif and a sequence-specific targeting region. These tgRNAs function as molecular scaffolds that recruit the λN-eIF4E1 fusion protein to target mRNAs, thereby promoting localized translation initiation [14]. The resulting λN-Guided RNA Targeting System (λGRTS) was developed to overcome the limitations of existing translational regulation platforms by exploiting the high-affinity λN-BoxB interaction. Through complementary base pairing between tgRNAs and target mRNAs, λGRTS enables precise and programmable delivery of functional proteins such as λN-eIF4E1 to defined RNA substrates [15]. This targeted recruitment facilitates the assembly of translation initiation complexes, promotes ribosome loading, and ultimately enhances translation initiation at specific transcripts. Importantly, the optimized λGRTS v5.1 system achieves translational activation levels comparable to those of dCas13-based system, while offering distinct advantages, including reduced molecular size and enhanced regulatory flexibility. Collectively, these findings establish λGRTS as a simple, efficient, and highly specific platform for targeted translation enhancement, providing a valuable new strategy for RNA-based gene regulation in mammalian cells.

Results

Design and construction of an mRNA closed-loop–mediated translation enhancement system in mammalian cells

In Cas9 and Cas13-based systems, guide RNAs (gRNAs) are indispensable for biological functionality, as they direct effector proteins to specific target sequences [16, 17]. Accordingly, two critical prerequisites for effective RNA targeting are the accessibility of gRNA recognition sites on target transcripts and the stability of base pairing between gRNAs and their cognate RNA sequences. Inspired by the RNA-targeting capability of Cas13—which enables precise post-transcriptional regulation, we hypothesized that directly modulating the RNA translation process could offer enhanced responsiveness and improved safety compared with to transcription-focused strategies. Based on this rationale, we sought to develop a novel translation-targeting system analogous to Cas13, in which guide RNAs mediate the recruitment of functional proteins to predefined RNA loci through well-characterized protein-RNA interactions. Because the assembly of TICs is widely recognized as the rate-limiting step of mRNA translation [18], therefore, promoting TIC assembly at specific mRNA sites is essential for achieving targeted translation initiation.

To validate the feasibility and efficacy of this approach, we constructed an exogenous green fluorescent protein (GFP) reporter containing both a 5′ untranslated region (5′ UTR) and a 3′ untranslated region (3′ UTR) to closely mimic native mRNA architecture. This reporter was placed under the control of the low-activity L30 promoter (Fig. 1A) [19], thereby minimizing basal transcription and enhancing the sensitivity for detecting changes in translational efficiency. To identify an optimal RNA–protein interaction module for targeted recruitment, we evaluated three candidate pairs (Fig. 1B). Their binding affinities to the target transcript were assessed using RNA immunoprecipitation followed by PCR (RIP-PCR), revealing that the λN/BoxB interaction exhibited significantly stronger binding than PP7cp/PP7 [20] and MCP/MS2 [21] (Fig. 1C–F). Consequently, λN was selected as the “guide module” to direct functional proteins to target RNAs. To ensure structural flexibility and minimize steric hindrance, a flexible GGGGS linker was introduced at the C-terminus of λN to connect it with candidate translation initiation factors. In parallel, we designed translation-guided RNAs (tgRNAs) composed of two essential elements: a sequence complementary to the GFP mRNA (enabling target recognition) and a BoxB motif (facilitating recruitment of λN-conjugated functional proteins to the target site) (Fig. 1G). Using the Cas13 system as a conceptual framework, we thus established a translation enhancement platform based on the λN-BoxB interaction.

Fig. 1.

The basic component of λGRTS. A The upstream sequence of the GFP gene, including the L30 promoter and 5′ UTR; B The candidate protein component of λGRTS; C-F RNA immunoprecipitation PCR (RIP-PCR) in HEK293T cells with HA-antibody or isotype IgG (isotype immunoglobulin G) as the negative control to compare the binding strength of λN-BoxB, MCP-MS2, and PP7cp-PP7; G Schematic diagram of hairpin structure formed by tgRNA; n = 3; mean ± SD was used to characterize the statistics of three independent experiments

This system comprises four core components: (1) λN protein: a hairpin RNA-binding protein that functions as the central targeting module, enabling high-specificity recruitment of effector proteins to BoxB motif; (2) guided RNA (tgRNA), a chimeric RNA harboring both a BoxB hairpin—for λN binding and a sequence complementary to the target mRNA for site-specific localization; (3) functional protein, such as a translation initiation factor, which is recruited to the target mRNA in a proximity-dependent manner to promote translation; (4) translation auxiliary proteins, including double-stranded RNA-binding proteins (dsRBPs) and poly(A)-binding protein (PABP), which facilitate translational activation.

According to the well-established closed-loop model of mRNA translation [22], modulation of mRNA cyclization is a key determinant of translational efficiency [23]. In this model, complexes assemble at the 5′ UTR to recruit ribosomal subunits, while interactions between the 5′ end and the 3′ poly(A) tail induce the formation of a closed-loop mRNA structure (as illustrated schematically). This cyclization enhances translational output by promoting efficient ribosomal scanning, facilitating ribosome recycling following translation termination, and protecting mRNAs from exonucleolytic degradation [24]. PABP, a canonical eukaryotic translation initiation factor, plays a central role in mediating this closed-loop formation by bridging the 3′ poly(A) tail and the 5′ cap-associated initiation complex [25, 26]. In addition to PABP, several dsRBPs such as NF90 [27], PKR [28], HuR [29], and HuD [30] were found to stabilize the interaction between tgRNA and their target mRNA (Fig. 1B), thereby enhancing the recruitment of translation initiation factors (TIFs) to target sites and further promoting translational activation.

Engineering targeted translation enhancement via mutated eIF4E1 and λN-BoxB RNA guidance

After establishing the basic framework of the translation enhancement system, we next sought to identify the most effective translation initiation factor for augmenting intracellular protein synthesis. To this end, we individually cloned and overexpressed core components of the translation initiation machinery, including eIF4A, eIF4E1, eIF4E3, and eIF4G, in mammalian cells and evaluated their capacity to enhance GFP expression using the L30-GFP reporter system. Among these factors, eIF4E1 overexpression produced the highest GFP fluorescence intensity, exhibiting an approximately 1.5-fold increase relative to cells expressing eIF4A, eIF4E3, or eIF4G (Fig. S1A and B). Notably, quantitative real-time PCR (qRT-PCR) analysis revealed no significant differences in GFP mRNA abundance among the groups (P > 0.05) (Fig. S1C), indicating that eIF4E1 enhances gene expression primarily at the translational level rather than through transcriptional regulation. Although this increase was statistically significant (N = 1.49; P = 0.0003), the overall magnitude of enhancement remained modest, limiting its practical utility for sensitive and targeted translational control.

To amplify this translational signal and improve detection sensitivity, we replaced the commonly used strong cytomegalovirus (CMV) promoter with the weaker L30 promoter (Fig. 2A). This substitution substantially reduced basal transcriptional activity, maintaining GFP mRNA at low and stable levels and thereby enabling more sensitive detection of translational changes (Fig. S1D–G). However, subsequent analysis of additional endogenous target genes revealed that eIF4E1 overexpression resulted in mild but consistent increases in protein expression across multiple transcripts (Fig. S1H), without corresponding changes in mRNA abundance (Fig. S1I). These findings suggest that wild-type eIF4E1 mediates global, non-specific translational enhancement rather than transcript-selective regulation. Consequently, we aimed to engineer a mutant form of eIF4E1 (eIF4E1^MUT) that retains translational activation capability while lacking intrinsic 5′ cap binding activity, thereby enabling targeted-regulation when coupled to the λN-BoxB targeting system (Fig. S1J).

Fig. 2.

Design of eIF4E1 mutations. A Regulation of green fluorescence expression by transfection of different translation initiation factors under the regulation of different promoters; B Molecular docking shows the 5′ cap binding site of eIF4E1 with the mRNA (dark colors represent hydrogen bonds); C Detection of the effect of eIF4E1 mutations on GFP expression; D Molecular docking reveals key sites for eIF4E1 binding to 4EBP; E The effect of eIF4E1^W73A mutation on GFP expression; F Effect of co-transfection of different fusion proteins with tgRNA on GFP fluorescence expression; G Effectiveness of different types and lengths of linkers for translation enhancement systems

eIF4E1 plays a central role in TIC assembly by recognizing and binding the 5′-m⁷GpppN cap structure of mRNAs [31]. Although eIF4E1 is widely regarded as one of the most potent translation-enhancing initiation factors, its affinity for the conserved 5′ cap inherently leads to broad, non-specific translational activation. To overcome this limitation, we performed site-directed mutagenesis to selectively disrupt eIF4E1’s cap-binding capacity while preserving its ability to recruit other TIC components such as eIF4G. Structural analysis of eIF4E1 (PDB: 1ipb) [32] identified several key residues—W102, E103, R157, K162, and T205—that form multiple hydrogen bonds with the mRNA 5′ cap (Fig. 2B). Individual mutation of these residues significantly reduced cap-binding capacity, as evidenced by a significant reduction in GFP fluorescence. Among these mutations, E103A and K162R exerted the most pronounced effects (N = 0.46, P < 0.05), and their combination nearly completely eliminated cap binding (Fig. 2C and Fig. S1K).

To further optimize translational activation, we introduced an additional W73A mutation to disrupt the interaction between eIF4E1 and the translational repressor 4EBP (PDB: 3u7x) [33], which inhibits TIC assembly by competitively blocking eIF4G recruitment. The W73A mutation relieved 4EBP-mediated inhibition, resulting in enhanced GFP expression relative to wild-type eIF4E1 (Fig. 2D and E), and substantially weakened the eIF4E1-4EBP interaction (Fig. S1L). Structural modeling confirmed that the resulting eIF4E1^MUT protein, harboring E103A, K162R, and W73A mutation substitutions, maintains an overall fold nearly identical to that of wild-type eIF4E1 (eIF4E1^WT), indicating preserved protein stability (Fig. S2A). Molecular docking analyses further demonstrated that while the W73A mutation selectively disrupts 4EBP binding, it does not impair eIF4G association (Fig. S2B–D), thereby preserving the capacity for TIC assembly. Consistently, interactions between eIF4E1^MUT and eIF4A were also unaffected (Fig. S2E and F), confirming that the engineered protein remains fully competent for incorporation into functional TICs.

For targeted RNA regulation, we paired the engineered eIF4E1^MUT with the λN-BoxB system. tgRNAs were designed to contain the BoxB motif and approximately 20-bp complementary sequences targeting distinct regions of GFP mRNA, including the 5′ UTR, coding sequence (CDS), and 3′ UTR (Fig. S1M). This design enabled systematic evaluation of positional effects on translational enhancement. Co-transfection experiments revealed that λN-eIF4E1^WT promoted GFP translation largely independent of tgRNA binding position, consistent with its residual cap binding activity. In contrast, λN-eIF4E1^MUT exhibited pronounced position-dependent effects, with tgRNA targeting the 5′ UTR producing the most robust translational enhancement (Fig. 2F).

Finally, to optimize the structural configuration of the fusion protein, we evaluated the effects of linker rigidity and length by incorporating either rigid (EAAAK) or flexible (GGGGS) linkers with varying repeats between λN and eIF4E1^MUT. Flexible GGGGS linkers consistently outperformed rigid EAAAK linkers, with the (GGGGS)₂ configuration producing the strongest translational enhancement (N = 2.85, P < 0.01) (Fig. 2G). This improvement is likely attributable to reduced steric constraints and enhanced conformational freedom, facilitating more effective interactions among the fusion protein, target mRNA, and translation initiation machinery.

Optimization of tgRNA constructs for enhanced translation in the λGRTS system

As described above, translation-guided RNAs (tgRNAs) within the λGRTS platform are modular molecules composed of multiple functional elements. To determine the optimal tgRNA architecture for maximizing translational enhancement, we designed a series of tgRNA variants with distinct structural features and constructed their corresponding expression plasmids (Fig. S3A). These plasmids were co-transfected with vectors encoding the λGRTS effector protein (λN-m-eIF4E1), and the translational regulatory effects of each tgRNA variant were systematically evaluated using GFP as a reporter.

To assess the requirement for tgRNA structural integrity in λGRTS-mediated translational activation, we generated truncated tgRNA variants lacking individual functional components such as the BoxB motif or the target-complementary sequence. These variants were then tested for their ability to promote translation. The results showed that robust translational activation was observed only when all λGRTS components were present. Under this intact configuration, recruitment of λN-m-eIF4E1 resulted in a pronounced increase in translational output, with GFP expression elevated by approximately 4.7-fold compared with controls (P < 0.001) (Fig. S3B). These findings demonstrate that the full structural integrity of tgRNAs is essential for λGRTS functionality.

We next examined two key structural parameters of tgRNAs: the number of BoxB motifs responsible for λN binding and the length of the linker connecting distinct tgRNA domains. By systematically testing tgRNA variants containing one, two, or three BoxB motifs in combination with linkers of varying lengths (10–30 bp), we found that translational enhancement was maximal when tgRNAs contained a single BoxB motif and a 10-bp linker (Fig. S3C–E). To gain mechanistic insight, three-dimensional structural simulations were generated to model interactions between tgRNA variants and the λN-m-eIF4E1 fusion protein. These simulations supported the conclusion that the 1 × BoxB plus 10-bp linker configuration minimizes steric hindrance and facilitates more efficient protein-RNA interactions (Fig. S3D–F).

In addition, we investigated whether the relative orientation of tgRNA functional elements influences translational activation. To this end, we constructed “reverse tgRNA” variants in which the order of the BoxB motif and the target-complementary sequence was inverted relative to the standard design (Fig. S3G). Fluorescence intensity measurements revealed significantly reduced GFP expression in cells transfected with reverse tgRNAs compared with those expressing standard tgRNAs (Fig. S3H). Consistently, flow cytometry showed an approximately 1.6-fold decrease in the proportion of GFP-positive cells in the reverse tgRNA group (N = 1.6, P < 0.01) (Fig. S3I and J). Collectively, these results define an optimized tgRNA architecture for the λGRTS: a cis-oriented tgRNA containing a single BoxB motif, a 10-bp linker, and a target-complementary sequence.

Nuclear export of λGRTS working proteins enhances translation efficiency in mammalian cells

Following the establishment of optimal tgRNA parameters, we next focused on engineering the λGRTS effector proteins to further enhance translational activity. Subcellular localization is a critical determinant of protein function [34], and this principle is particularly relevant for translational regulators, as eukaryotic protein synthesis predominantly occurs in the cytoplasm. Although λGRTS effector proteins like λN-m-eIF4E1 are expressed in the cytoplasm, inherent variability in protein sorting can lead to their mislocalization to other compartments (e.g., nucleus), thereby reducing their availability for cytoplasmic translation. We therefore hypothesized that enforcing cytoplasmic localization would increase translational efficiency. To test this hypothesis, nuclear export signals (NES) or nuclear localization signals (NLS), well-characterized peptide motifs that direct proteins to specific subcellular compartments [35, 36], were fused to the λGRTS effector proteins. NES or NLS sequences were introduced at three distinct positions within the baseline v1.1 construct: the N-terminus, the C-terminus, or the junction between the λN and m-eIF4E1 domains (Fig. 3A). The resulting variants were expressed in mammalian cells and evaluated for their ability to promote GFP translation.

Fig. 3.

Improving λGRTS to harvest higher translation enhancements. A Protein pattern map with NLS and NES sequences added; B Detection of fluorescence intensity after co-transfection of different tgRNA and effector proteins; C Protein patterns map with various dsRBPs added; D Fluorescence detection of the effect of adding different dsRBPs on fusion proteins on the intensity of fluorescence expression; E Detection of fluorescence changes after addition of different RRM domains on fusion proteins; F Comparison of the effects of the λGRTS system and dCasRX on eukaryotic gene expression; G The fluorescence intensity in different groups was measured by a plate reader; H Flow cytometry was used to detect the intensity of green fluorescence in the different groups

Constructs containing NLS tags exhibited GFP expression levels comparable to those of control cells, indicating minimal translational enhancement. In contrast, NES-tagged variants produced significantly stronger translational activation (Fig. 3B). Notably, the most pronounced effect was observed when the NES motif was fused to the N-terminus of the effector protein, resulting in a 4.8-fold increase in GFP expression relative to controls (N = 4.8, P < 0.05) (Fig. S4A–D). These results demonstrate that enforced cytoplasmic localization of λGRTS effector proteins substantially enhances translational output particularly via N-terminal NES fusion, likely by increasing the effective cytoplasmic concentration TICs.

Enhancement of the λGRTS through dsRBP-mediated stabilization of tgRNA-mRNA interactions

In the λGRTS system, sequence-specific base pairing between tgRNAs and their target mRNAs constitutes the molecular basis for targeted translational enhancement. However, a well-recognized limitation of RNA-based targeting strategies is that double-stranded RNA (dsRNA) duplexes formed through RNA-RNA interactions are inherently less stable than double-stranded DNA (dsDNA) counterparts [37]. In the initial design of λGRTS, the tgRNA-mRNA complementary region was set to 20 bp, a length selected to balance targeting specificity with binding affinity. Nevertheless, we reasoned that the intrinsic instability of such dsRNA duplexes could lead to premature dissociation of the tgRNA-effector complex during recruitment of TICs, thereby impairing ribosomal assembly and limiting translational enhancement.

To address this limitation, NF90 was selected as a representative dsRBP and inserted at three distinct positions within the effector protein: the N-terminus, the junction between λN and m-eIF4E1, or the C-terminus. Analysis of GFP expression revealed that insertion of NF90 between λN and m-eIF4E1 produced the strongest translational enhancement (Fig. S4E–G), indicating that this configuration optimally supports both dsRNA stabilization and effector protein functionality. Based on this result, four effector protein variants were subsequently constructed, each incorporating one of the candidate dsRBPs at this optimized position (Fig. 3C).

Comparative functional analysis demonstrated that the variant containing HuR (designated v3.5) exhibited the highest translational enhancement, producing an approximately 1.25-fold increase in GFP expression relative to the dsRBP-lacking control (N = 1.25, P < 0.05) (Fig. 3D). This result identifies HuR as the most effective dsRBP for the λGRTS platform, potentially reflecting its strong affinity for dsRNA and its structural compatibility with the λGRTS effector architecture.

Closed-loop mRNA formation enhances translation in the λGRTS

The closed-loop model of mRNA translation supported by extensive experimental evidence and is widely recognized as a key mechanism for enhancing translational efficiency [38, 39]. This model relies on two essential structural features of eukaryotic mRNAs: the 5′-m⁷GpppN cap at the 5′ end and the poly(A) tail at the 3′ end. PABP, a highly conserved cytoplasmic factor, binds the poly(A) tail with strong sequence-independent affinity in mammalian cells and serves as a central molecular scaffold for mRNA closed-loop formation.

In the λGRTS, we exploited this fundamental principle to promote mRNA cyclization and thereby enhance translation. Formation of a closed-loop confers multiple translational advantages, including accelerated scanning of the mRNA by the 40S ribosomal subunit, which reduces the time required to locate the AUG start codon, and enhanced ribosome recycling following translation termination. Together, these effects increase the overall rate of protein synthesis. In addition, PABP binding to the poly(A) tail protects mRNA from exonucleolytic degradation, thereby stabilizing transcripts and supporting sustained translational enhancement [40, 41]. Structurally, cytoplasmic PABP contains four evolutionarily conserved RNA recognition motifs (RRMs), connected by flexible carboxy-terminal helical regions that mediate both RNA binding and protein–protein interactions essential for closed-loop assembly.

To optimize PABP incorporation into the λGRTS effector protein, we first examined how PABP positioning influences translational activity. Effector protein variants were generated by fusing full-length PABP to the N-terminus, C-terminus, or the junction between λN and m-eIF4E1, using the v4.1 construct as a baseline. Fluorescence analysis revealed that N-terminal fusion of PABP produced the strongest enhancement of GFP expression (Fig. S4H–J), identifying this configuration as optimal for PABP-mediated promotion of closed-loop formation.

To further reduce effector protein size and minimize potential steric constraints, we next sought to identify the minimal PABP domain capable of supporting closed-loop formation. Each of the four individual RRM domains was separately cloned into the effector protein in place of full-length PABP and evaluated for translational activity. Single RRM domains failed to enhance GFP expression, indicating that individual RRMs are insufficient to promote closed-loop formation. In contrast, tandem combinations of RRMs were functional, with the RRM2-RRM3 pair exhibiting the strongest effect. Incorporation of this tandem domain resulted in an approximately 1.32-fold increase in GFP expression (N = 1.32, P < 0.05) (Fig. 3E). These results indicate that the RRM2-RRM3 module represents a minimal functional unit capable of mediating PABP-dependent mRNA closed-loop formation within the λGRTS framework, thereby contributing to enhanced translational efficiency.

λGRTS exhibits translational activation comparable to dCasRX

The dCasRX system is a well-established and widely used platform for RNA-targeted translational enhancement, with broad applications in both basic research and preclinical studies [42]. To position λGRTS within the current landscape of RNA-based translation regulation tools, we performed a direct comparative analysis of λGRTS and dCasRX to evaluate their targeting specificity and translational activation efficiency (Fig. 3F).

In these experiments, both systems were engineered to target the same GFP reporter gene driven by the L30 promoter, and translational output was quantified using fluorescence intensity measurements (microplate reader) and flow cytometry. The results demonstrated that λGRTS achieved target protein expression levels comparable to those obtained with dCasRX, indicating that λGRTS functions as an effective translational activator in mammalian cells.

Notably, λGRTS exhibited several practical advantages over dCasRX. Protein (RRM2&3-NES-λN-HuR-m-eIF4E1) has a substantially smaller molecular weight (~ 50 kDa) than dCasRX (~ 150 kDa), which reduces the likelihood of vector packaging constraints—particularly in viral delivery systems and may lessen unintended perturbation of cellular processes. Second, λGRTS requires only a single tgRNA for targeting, whereas dCasRX relies on a crRNA-tracrRNA duplex, thereby simplifying vector design and experimental implementation.

When translational activation efficiency was further analyzed after normalization for vector size and system complexity, λGRTS exhibited an approximately 3.47-fold higher effective translational output compared with dCasRX (P < 0.05) (Fig. 3G and H). Collectively, these results establish λGRTS as a competitive and streamlined alternative to dCasRX, offering comparable translational efficacy with improved modularity and practicality for RNA-based gene regulation in mammalian cells.

Determining the optimal tgRNA length and targeting location in the λGRTS

Although the λGRTS system effectively enhances the translation of proteins encoded by complementary RNA strands, an important question remained as to whether additional design parameters could further improve translational output. Drawing inspiration from CRISPR-based transcriptional activation systems—where guide RNA (gRNA) is strongly influenced by target genomic position [43], we hypothesized that the translational enhancement efficiency of λGRTS might likewise depend on both the binding position and complementary length of translation-guided RNAs (tgRNAs). To test this hypothesis and establish general principles for tgRNA optimization, we systematically constructed a tgRNA library comprising variants with distinct target positions on mRNA and varying lengths of sequence complementarity.

This investigation was motivated by the central role of translation initiation complex (TIC)-mediated ribosome assembly as the rate-limiting step in mRNA translation. The efficiency of translation initiation is closely linked to the site at which ribosomes assemble on the mRNA, with certain positions favoring rapid initiation while minimizing unnecessary energetic costs associated with excessive distance from the start codon. Optimal positioning facilitates efficient ribosome loading and saturation of the mRNA template, thereby maximizing translational output [44]. Based on this relationship, we reasoned that the tgRNA binding site by dictating the location of TIC recruitment would be a major determinant of λGRTS performance.

Following successful incorporation of the closed-loop mechanism to enhance translation, we next optimized tgRNA design by systematically varying two core parameters: the length of the tgRNA-mRNA complementary region and the binding position of tgRNAs along the target mRNA. To this end, we constructed a comprehensive tgRNA library featuring complementary lengths ranging from 15 to 25 nucleotides (nt) and binding sites distributed across the 5′ UTR, coding sequence (CDS), and 3′ UTR of the GFP reporter mRNA. Each tgRNA variant was co-transfected into mammalian cells, and translational output was quantified by measuring GFP fluorescence intensity.

To first determine the optimal mRNA region for tgRNA targeting, we designed tgRNAs with a fixed 20-nt complementary region which are selected to balance binding specificity and minimize potential off-target interactions that targeted the 5′ UTR (GFP-1), CDS (GFP-2), or 3′ UTR (GFP-3) of GFP mRNA. Fluorescence analysis revealed a strong positional dependence: tgRNAs targeting the 5′ UTR induced robust translational enhancement, whereas tgRNAs targeting the CDS or 3′ UTR produced minimal or reduced GFP expression relative to the non-targeting control. These results identify the 5′ UTR as the most favorable region for λGRTS mediated translational activation and prompted us to focus subsequent optimization efforts on this region.

To further refine tgRNA targeting within the 5′ UTR, we constructed additional tgRNA variants targeting three subregions: upstream (proximal to the promoter), midstream, and downstream (proximal to the ATG start codon) (Fig. 4A). Co-transfection and fluorescence analysis demonstrated that tgRNAs targeting the midstream subregion of the 5′ UTR consistently produced higher GFP expression than those targeting either upstream or downstream segments (Fig. 4B–E). This finding suggests that the midstream portion of the 5′ UTR represents a particularly favorable binding site for tgRNA-mediated recruitment of translation initiation machinery, and it was therefore selected for further optimization.

Fig. 4.

The 22-nt tgRNA (58 bp upstream of ATG) showed the strongest translation promotion efficiency. A Three tgRNAs targeted to different position were designed in the 5′ UTR region; B The effect of the formation of translation initiation complexes at different positions on the GFP translation process was detected by a plate reader; C Fluorescence photography shows the fluorescence intensity of cells forming translation initiation complexes at different locations; D Flow cytometry was used to detect the differences in fluorescence intensity of cells forming translation initiation complexes at different locations; E The MFI of each group of cells was calculated from the flow cytometry results; F The tgRNA sequences of different lengths were designed according to 5′ UTR sequences and corresponding fluorescence intensity was detected with microplate reader; G The tgRNA sequences with length of 22 bp for different positions were designed according to the 5′-UTR sequence, and corresponding fluorescence intensity was detected with microplate reader

Using the midstream 5′ UTR region as the target site, we next examined the effect of complementary sequence length on translational enhancement. A baseline tgRNA containing a 20-nt complementary region and targeting a site 60 bp upstream of the ATG start codon was used as a reference. We then generated a series of tgRNA variants with complementary lengths ranging from 15 to 25 nt; red fluorescent protein (RFP)-targeting tgRNAs were included as negative controls exclude non-specific translational effects.

Fluorescence quantification revealed a clear length-dependent trend: GFP expression increased progressively as the complementary region lengthened from 15 to 22 nt, with maximal enhancement observed at 22 nt. When the complementary region exceeded 22 nt (23–25 nt), translational enhancement plateaued or declined modestly. Comparative analysis confirmed that a 22-nt complementary region produced the highest translational output, corresponding to an approximately 1.22-fold increase relative to controls (P < 0.05) (Fig. 4F). This length likely represents an optimal balance between duplex stability and minimal steric interference with TIC assembly.

Finally, using the optimized 22-nt complementary length, we systematically mapped the precise tgRNA binding position within the 5′ UTR. A panel of tgRNAs targeting positions spanning − 65 to − 50 bp upstream of the ATG start codon was constructed. Co-transfection assays revealed a positional gradient in translational efficiency: GFP expression increased as the tgRNA binding site moved toward the site approximately 58 bp upstream of the ATG. Targeting this position resulted in an approximately 1.14-fold increase in GFP expression relative to controls (P < 0.05). Notably, targeting sites either further upstream or closer to the ATG led to reduced enhancement (Fig. 4G). These results indicate that the − 58 bp position within the 5′ UTR represents an optimal tgRNA binding site for λGRTS-mediated translational activation, likely reflecting favorable spatial alignment with translation initiation machinery while avoiding interference with ribosome scanning.

Construction and functional validation of λGRTS fusion plasmids

To enhance the versatility of the λGRTS platform for both in vitro and in vivo applications, we engineered an all-in-one vector incorporating several essential functional elements. These include an ampicillin resistance gene for bacterial selection during plasmid amplification, a puromycin resistance cassette to enable stable cell line generation, and lentiviral packaging elements (e.g., the ψ packaging signal) to facilitate viral delivery. This integrated vector is designed to promote translation initiation by directing the λGRTS effector complex (RRM2&3-NES-λN-HuR-m-eIF4E1) to target mRNA loci via an embedded tgRNA (22-nt complementary region targeting a site 58 bp upstream of the ATG start codon), thereby reducing experimental variability and simplifying implementation. The complete plasmid map and sequence information has been deposited in the Addgene public repository to facilitate reproducibility and broad adoption of the λGRTS (Fig. S5A).

To evaluate the intrinsic cytotoxicity of key λGRTS components—including λN, eIF4E1^MUT, HuR, PABP, and the RRM2-RRM3 domain, we constructed individual overexpression plasmids and transfected them into U251 cells. Cell viability was subsequently assessed using the CCK-8 assay. Overexpression of each component did not result in significant changes in cell viability compared with control cells, indicating that none of the components exhibited obvious toxicity to cellular function (P > 0.05) (Fig. S5B).

We next assessed the targeting specificity and potential off-target effects of λGRTS. HEK293T cells stably expressing an L30-driven GFP reporter were transfected with λGRTS fusion constructs, using GFP-21 as the tgRNA target sequence. At 6 h post-transfection, total cellular proteins were harvested and analyzed by quantitative mass spectrometry. Comparative analysis between λGRTS-treated and control cells revealed a pronounced upregulation of GFP protein in the experimental group (Fig. S5C). In contrast, no substantial aberrant expression patterns were detected among other proteins, suggesting a high degree of target specificity.

To determine whether λGRTS can enhance the translation of exogenously supplied mRNA, we transfected U251 cells with in vitro-transcribed (IVT) GFP mRNA together with the λGRTS. In the presence of a GFP-targeting tgRNA, λGRTS significantly increased translation of the exogenous mRNA, yielding an approximately 4.8-fold enhancement relative to controls (P < 0.001) (Fig. S5D). These results indicate that λGRTS is capable of activating translation independently of endogenous transcription and support its potential applicability in mRNA-based therapeutic strategies.

Because promoter strength can markedly influence transcript abundance, we further examined whether λGRTS-mediated translational activation is dependent on promoter activity. Stable GFP-expressing cell lines driven by distinct promoters were subjected to λGRTS treatment. In all cases, GFP protein levels increased by approximately fourfold (Fig. S5E), demonstrating that λGRTS-mediated translation enhancement is largely independent of promoter-driven transcriptional strength. In addition, PCR analysis revealed increased GFP mRNA abundance following λGRTS treatment (Fig. S5F), indicating enhanced mRNA stability.

Collectively, these findings demonstrate that λGRTS enhances target gene expression through a dual mechanism: stabilization of target mRNAs and activation of their translation. This combined mode of action underscores the robustness and versatility of λGRTS as a post-transcriptional gene regulation platform.

λGRTS enables translational activation of endogenous mammalian genes

To further assess the practical utility of λGRTS for regulating endogenous genes in mammalian cells which is a critical step toward applications in biological research and therapeutic development, we selected two well-characterized tumor suppressor genes, P53 and PTEN, as representative target for in vitro validation. Based on our prior optimization strategy, we identified optimal tgRNA target sites corresponding to 22-nt sequences located approximately 58 bp upstream of the ATG start codon within the 5′ UTRs of the respective mRNAs (Fig. 5A). Species-specific tgRNA sequences were designed for human transcripts, chemically synthesized, and cloned into the previously described all-in-one λGRTS plasmid vector for expression in the U251 glioblastoma cell line.

Fig. 5.

Effects of targeted activation of P53 and PTEN translation with λGRTS on GBM cell function. A RNA sequences targeting P53 and PTEN genes 58 bp upstream of the ATG at a length of 22 bp were identified; Detection of the translational enhancement effect of λGRTS on P53 (B) and PTEN (C); The mRNA level of PCR of P53 (D) and PTEN (E) were detected by RT-qPCR after transfected with λGRTS plasmid in U87 and U251; F, G RT-qPCR assay revealed that λGRTS improved the mRNA stability of target gene. Actinomycin D (ActD) was used to inhibit the mRNA synthesis process in the GBM cells after transfected with λGRTS for 0, 6, 12 h; H–K The CCK-8 experiment revealed the effect of λGRTS-activated P53 and PTEN expression on proliferation of U87 and U251 GBM cells; L–O The cell cycle experiments were performed to detect the cell cycle distribution with λGRTS-activated P53 and PTEN in U87 and U251; P, Q Transwell assays were applied to detect the invasion ability after targeted activation of P53 and PTEN by λGRTS in U87 and U251 GBM cells

Following transfection of the λGRTS constructs into U251 cells, endogenous protein expression of P53 and PTEN was evaluated by western blot analysis. Compared with cells transfected with a non-targeting tgRNA control (tgRNA-NT), cells expressing λGRTS targeting P53 or PTEN exhibited pronounced upregulation of the corresponding proteins (Fig. 5B and C), demonstrating effective translational activation of endogenous targets.

To elucidate the mechanism underlying the observed increases mRNA abundance detected by RT-qPCR (Fig. 5D and E), we performed an actinomycin D (ActD) chase assay. ActD inhibits de novo transcription, thereby enabling assessment of mRNA stability independently of transcriptional activity. Notably, even in the presence of ActD, λGRTS-treated cells maintained significantly higher levels of P53 and PTEN mRNA compared with tgRNA-NT controls (Fig. 5F and G). These findings indicate that the elevated mRNA levels are not attributable to transcriptional activation but instead result from λGRTS-mediated stabilization of target transcripts.

Collectively, these results demonstrate that λGRTS enhances endogenous gene expression through a post-transcriptional mechanism that combines mRNA stabilization with targeted translational activation, thereby reinforcing its potential as a versatile platform for endogenous gene regulation in mammalian cells.

λGRTS-mediated activation of P53 and PTEN suppresses malignant phenotypes in GBM cell

To determine whether λGRTS-mediated translational activation of the tumor suppressors P53 and PTEN influences malignant phenotypes of GBM cells, we conducted a series of functional assays using two human GBM cell lines, U87 and U251, which are widely employed as in vitro models of glioma progression. Given the central roles of P53 and PTEN in regulating cell cycle progression and cellular proliferation, we first examined the effect of λGRTS on GBM cell growth using the Cell Counting Kit-8 (CCK-8) assay. Targeted activation of P53 or PTEN via λGRTS resulted in a pronounced reduction in proliferative capacity in both U87 and U251 cells over a 96-h period, with proliferation rates decreased by approximately 40–50% relative to cells transfected with a non-targeting tgRNA (tgRNA-NT) (Fig. 5H–K).

We next assessed the impact of λGRTS on cell cycle distribution, as both P53 and PTEN function as key regulators of cell cycle checkpoints. Cell cycle profiles were analyzed by flow cytometry following staining with Red Nucleus I, a DNA-specific dye. In both U87 and U251 cells, λGRTS-mediated activation of P53 led to a marked reduction in the proportion of cells in the S and G2/M phases, accompanied by a corresponding increase in the G1 population, indicative of G1 phase arrest (Fig. 5L–O). Similarly, activation of PTEN resulted in a significant decrease in S-phase cells, consistent with its established role in suppressing cell cycle progression through inhibition of the PI3K/Akt signaling pathway.

Because invasive behavior is a defining feature of GBM malignancy, we further evaluated the effect of λGRTS on cellular invasiveness using Transwell invasion assays. Compared with tgRNA-NT controls, U87 and U251 cells with λGRTS-mediated activation of P53 or PTEN exhibited substantially reduced invasive capacity. Specifically, the number of invading cells decreased by approximately 60% following P53 activation and by approximately 55% following PTEN activation (Fig. 5P and Q).

Collectively, these findings demonstrate that λGRTS-mediated translational activation of endogenous tumor suppressor genes can significantly attenuate proliferative, cell cycle, and invasive phenotypes of GBM cells. These results highlight the potential of λGRTS as a post-transcriptional strategy for modulating functionally relevant genes and suppressing malignant cellular behaviors in cancer models.

Broad applicability of λGRTS for translational enhancement across mammalian species

To evaluate the broader applicability of the λGRTS platform across eukaryotic cells, with particular emphasis on mammalian species, we first examined the evolutionary conservation of eIF4E1, a core component of the λGRTS effector protein. Amino acid sequences of eIF4E1 from multiple mammalian species were retrieved from the UniProt database, and multiple sequence alignment was performed to assess sequence homology. eIF4E1 orthologs from Homo sapiens (human), Bos taurus (cow), Rattus norvegicus (rat), and Mus musculus (mouse), which were additionally cross validated using NCBI database entries, displayed a high degree of conservation, with pairwise sequence identities exceeding 95% (Fig. S6A).

To further assess structural conservation (a critical determinant of functional compatibility), we generated comparative three-dimensional structural models of these eIF4E1 orthologs and visualized them using PyMOL. The resulting models exhibited highly similar overall conformations, particularly within domains essential for TIC recruitment, including the eIF4G interaction interface and the engineered cap-binding region (Fig. S6B). These observations support the premise that λGRTS, which relies on eIF4E1-mediated TIC assembly, is likely to function across diverse mammalian species.

To experimentally validate cross-species functionality, we examined λGRTS performance in cell lines derived from three mammals: human HEK293T cells, mouse GL261 cells, and bovine MDBK cells. Based on the optimized tgRNA design principles established earlier, species-specific tgRNAs were designed to target a conserved site within the 5′ UTR of P53 mRNA, corresponding to a 22-nucleotide complementary sequence located approximately 58 bp upstream of the AUG start codon (Fig. S6C). Species-matched λGRTS plasmids were constructed and transfected into the corresponding cell lines.

Endogenous P53 expression was subsequently quantified at both the mRNA level by RT-qPCR and the protein level by western blot analysis. In all three cell lines, λGRTS treatment resulted in robust upregulation of P53 protein expression, corresponding to an approximately 4.1-fold increase relative to controls (P < 0.05), accompanied by elevated P53 mRNA levels. In contrast, expression of PTEN, which served as a non-targeted control, remained unchanged at both the mRNA and protein levels across all tested cell lines (Fig. S6D and E).

Collectively, these results demonstrate that λGRTS-mediated translational enhancement is not restricted to human cells but is conserved across a phylogenetically diverse set of mammalian species, underscoring the broad applicability and versatility of the λGRTS platform for cross-species gene regulation.

Targeted activation of tumor suppressors via λGRTS inhibits GBM progression in vivo

To evaluate the translational potential of λGRTS in a preclinical cancer model, we engineered a lentiviral vector encoding the core components of the λGRTS, including the optimized effector protein (NES-PABP-λN-HuR-m-eIF4E1) and tgRNAs targeting either P53 or PTEN. U251 cells were transduced with the corresponding lentiviruses to establish stable cell lines, which were subsequently used to generate an orthotopic glioblastoma model in nude mice. The overall experimental workflow is illustrated in Fig. 6A. Briefly, male BALB/c nude mice (4–5 weeks old) were intracranially implanted with U251 cells, and tumor progression and body weight were monitored at predefined time points.

Fig. 6.

λGRTS inhibits the GBM development and malignant ability through targeted activation of P53 and PTEN. A The work schedule of xenograft GBM mice, the xenograft GBM mice were injected with λGRTS using a lentivirus vector system; B Representative images show bioluminescent signals of tumor formation in xenograft model. Fluorescent signal intensity was used to characterize tumor size. The statistics are expressed as mean ± SD, n = 5; C The total flux of each nude mice group with different cell injection; D Kaplan–Meier survival analysis showed the prognosis of each mice group with different cell injection; E The body weight change of mice in different groups; F Representative P53 and PTEN IHC images of each xenograft GBM mice group with different cell injection; G The heatmap shows the IHC scores of Ki67, p53, and PTEN in different mice

Tumor growth was assessed weekly by bioluminescence imaging (BLI). Mice treated with λGRTS-P53 or λGRTS-PTEN exhibited markedly reduced bioluminescent signals compared with control mice, indicating suppressed tumor progression (Fig. 6B and C). Quantitative analysis at the experimental endpoint (4 weeks post-tumor implantation) confirmed that tumor volumes were significantly reduced in λGRTS-treated groups, with decreases of approximately 65% for P53-targeted tumors and 60% for PTEN-targeted tumors relative to controls. Survival analysis further demonstrated a significant extension of lifespan in mice receiving λGRTS-P53 and λGRTS-PTEN treatment, with median survival prolonged by approximately 40 days and 35 days, respectively, compared with the control group (Fig. 6D). In addition, mice in the λGRTS treated groups exhibited reduced body weight loss, over the course of the experiment, indicative of improved overall health status (Fig. 6E).

To validate the in vivo efficacy of λGRTS at the histological and molecular levels, excised tumor tissues were subjected to hematoxylin and eosin (HE) staining and immunohistochemistry (IHC) analysis. HE staining revealed that tumors from λGRTS-treated mice displayed disrupted tissue architecture and reduced cellular density compared with control tumors. Consistently, IHC analysis demonstrated significant upregulation of P53 and PTEN protein expression in the corresponding treatment groups, with staining intensities increased by approximately 2.5-fold for P53 and 2.3-fold for PTEN (Fig. 6F). Heatmap visualization of IHC scores across individual mice further confirmed the robust and consistent targeting capability of λGRTS (Fig. 6G).

Taken together, these in vivo findings are consistent with our in vitro results and demonstrate that λGRTS can effectively activate the translational expression of tumor suppressor genes in an orthotopic glioblastoma model. Collectively, the data indicate that λGRTS-mediated, RNA-guided translational activation of tumor suppressors significantly inhibits GBM progression, providing strong preclinical evidence supporting the potential of λGRTS as a novel therapeutic strategy for cancer treatment.

Discussion

In the present study, we developed a novel translational activation strategy based on the λN-Guided RNA Targeting System (λGRTS), addressing a critical unmet need in mammalian gene regulation—namely, the lack of precise, translation-centric tools that circumvent the limitations of transcriptional manipulation and CRISPR-based platforms. As a research tool, λGRTS exhibits several distinct advantages over traditional gene overexpression approaches, including its compact molecular architecture (utilizing the 3-kDa λN protein rather than large Cas proteins), a streamlined operational workflow (requiring only tgRNA design instead of complex fusion protein engineering), inherent structural stability (derived from the high-affinity λN-BoxB interaction), and high targeting specificity (mediated by sequence-complementary tgRNAs). Unlike conventional overexpression plasmids, which depend on exogenous promoters (e.g., CMV) to drive non-specific transcriptional upregulation that often results in ectopic protein accumulation and off-target cellular effects, λGRTS directly modulates the translation of endogenous mRNAs, thereby enabling more physiologically relevant control of gene expression. Importantly, λGRTS also demonstrates notable potential as a candidate clinical tool. Its functional components are largely humanized, incorporating mutated human eIF4E1 and human-derived PABP domains, which represents a significant advantage over archaea-derived CRISPR systems that may elicit autoimmune responses due to their foreign origin [45]. This high degree of human compatibility enhances the translational feasibility of λGRTS for therapeutic applications.

Taken together, λGRTS holds substantial promise for both basic biological research and translational biomedicine. In basic research settings, it provides a versatile platform for precise and reversible control of protein expression, allowing functional interrogation of essential genes, toxic proteins, or key signaling nodes without permanently altering the genome. Moreover, the modular nature of λGRTS enables its integration with inducible or tissue-specific promoters, permitting spatiotemporal regulation of gene expression in diverse contexts such as developmental biology, neurobiology, and cellular dynamics. The system can also be adapted for multiplexed regulation, enabling simultaneous targeting of multiple mRNAs to model polygenic diseases or investigate complex regulatory networks.

In biomedical research, λGRTS opens new avenues for therapeutic development. Its ability to selectively enhance the translation of endogenous tumor suppressors, such as P53 and PTEN, without inducing DNA damage responses positions it as an attractive candidate for cancer therapy, particularly in settings where CRISPR-based approaches raise safety concerns. Beyond oncology, λGRTS may be applicable to monogenic disorders caused by haploinsufficiency, where modest upregulation of the remaining functional allele could restore physiological protein levels. The system also shows potential in regenerative medicine, for example, by enhancing the expression of pro-regenerative or anti-apoptotic factors in stem or progenitor cells. Furthermore, its compatibility with viral delivery platforms, including lentivirus and AAV, supports its use in preclinical disease models and eventual clinical translation.

A central design principle of λGRTS is the targeted recruitment of ribosomal components and translation initiation machinery to the 5′ UTR of mRNAs, the primary rate-limiting site for translation initiation in eukaryotes [46, 47]. By fusing λN, a phage-derived BoxB-binding module, to a rationally engineered mutant of eIF4E1 (m-eIF4E1), we generated a chimeric protein that can be precisely guided to the 5′ UTR of target mRNAs via tgRNA-BoxB recognition. This targeted localization enhances translational efficiency by facilitating the assembly of TICs, resulting in robust upregulation of protein expression.

To validate this mechanism, we constructed a reporter system consisting of a GFP coding sequence fused to a native 5′ UTR and driven by the low-activity L30 promoter [48]. This design ensured low basal GFP mRNA expression, thereby increasing sensitivity to changes in translational efficiency and enabling accurate assessment of λGRTS activity.

Through systematic functional validation, we confirmed the pivotal role of eIF4E1 in TIC assembly. Among several candidate translation initiation factors (eIF4A, eIF4E1, eIF4E3, and eIF4G), eIF4E1 exhibited the strongest capacity to promote TIC formation, consistent with its established role as a key regulator of cap-dependent translation [49]. Sequential transfection of the three core λGRTS components (tgRNA, λN-m-eIF4E1, and auxiliary factors) revealed a subtle yet reproducible increase in GFP expression following tgRNA transfection alone. This observation suggests that tgRNA may contribute to target mRNA stabilization, potentially by reducing exonucleolytic degradation, thereby complementing the system’s primary translational activation mechanism. Notably, the translational efficacy of the λN-eIF4E1 fusion protein was highly dependent on tgRNA binding position. GFP expression increased as the targeting site was moved from 65 to 58 bp upstream of the ATG start codon but declined sharply when positioned beyond this optimal window. This positional sensitivity highlights the importance of balancing proximity to the 5′ cap (where TIC assembly initiates) with avoidance of steric interference during ribosomal scanning [50, 51].

Mechanistically, wild-type eIF4E1 binds the 5′ m⁷GpppN cap to drive global translation, a process that lacks transcript specificity. To confer target selectivity, we introduced site-directed mutations (E103A/K162R) into the cap-binding domain of eIF4E1, abolishing non-specific cap recognition while preserving its ability to recruit other TIC components. Upon tgRNA-mediated localization to the 5′ UTR, m-eIF4E1 recruits eIF4A, eIF4G, and eIF4B, thereby promoting assembly of the 43S pre-initiation complex (PIC) through interactions with eIF1, eIF1A, eIF3, and the ternary complex (eIF2-GTP-Met-tRNAi) [52–56]. Subsequent interactions between eIF4G and PABP facilitate mRNA circularization, accelerating ribosome scanning and enhancing translation efficiency in accordance with the closed-loop model of eukaryotic translation [57, 58]. Additionally, incorporation of the RNA-binding protein HuR further stabilizes RNA molecules by maintaining double-stranded RNA (dsRNA) interactions, increasing transcript abundance and further enhancing protein output.

Following reporter validation, we extended λGRTS to the activation of endogenous full-length genes. Using transcriptomic data from the NCBI database, we identified conserved 22-bp targetable sequences approximately 58 bp upstream of the ATG start codon and designed corresponding tgRNAs. Application of gene-specific λGRTS constructs to a panel of low abundance endogenous genes, including tumor suppressors such as P53 and PTEN, resulted in robust upregulation of protein expression, demonstrating the system’s capacity to modulate biologically relevant targets.

Notably, λGRTS overcomes key limitations of existing translational regulation systems. Although prokaryotic translational enhancement has been achieved using systems such as dCasRX-IF3 [59], the complexity of eukaryotic translation which requires coordinated interactions among multiple initiation factors has limited the development of analogous tools for mammalian cells [60, 61]. While dCasRX enables RNA targeting in eukaryotes, its large molecular size (~ 150 kDa) complicates viral packaging and may disrupt mRNA localization through steric hindrance [62]. In contrast, λGRTS employs the 3-kDa λN protein as its targeting module, whose high affinity interaction with BoxB ensures specificity while minimizing interference with mRNA trafficking and ribosomal dynamics [63, 64].

Compared with direct delivery strategies involving exogenous mRNA or DNA, λGRTS offers a more physiologically relevant mode of gene regulation. Direct delivery approaches often result in non-physiological protein overexpression, lack endogenous regulatory control, and may trigger immune responses due to foreign nucleic acids or excessive protein levels [65, 66]. Additionally, delivered mRNA typically exhibits a short half-life, limiting its utility in applications requiring sustained protein expression [67, 68]. By contrast, λGRTS regulates translation of endogenous mRNAs, preserving native transcriptional context, post-translational modifications, and subcellular localization. Its effects can integrate with intrinsic cellular signaling pathways and feedback mechanisms, enabling context-dependent amplification of protein expression rather than constitutive overproduction. Moreover, stable tgRNA expression via lentiviral delivery may enable long-term and potentially reversible translational enhancement, offering promise for chronic disease management [69].

Finally, we emphasize that λGRTS represents a flexible and customizable addition to the gene regulation toolkit, overcoming the rigidity of single-mode systems such as constitutive overexpression vectors or CRISPRa. Nonetheless, certain limitations warrant future optimization. In particular, reliance on 5′ UTR targeting restricts application to genes with sufficiently long or accessible 5′ UTRs. Future efforts may explore alternative targeting regions, such as internal ribosomal entry sites (IRESs), or engineered tgRNAs capable of recognizing structured RNA motifs. Additionally, refinement of viral delivery strategies could improve tissue specificity and reduce off-target effects in vivo.

Conclusions

In conclusion, λGRTS establishes a novel, CRISPR-independent paradigm for targeted translational activation in mammalian cells. By integrating the specificity of λN-BoxB binding, engineered translation factors, and closed-loop mRNA architecture, λGRTS achieves superior compactness, specificity, and safety relative to existing tools. Its demonstrated ability to activate endogenous tumor suppressors and inhibit cancer progression underscores its translational potential for both basic research and therapeutic development. As gene regulation strategies increasingly shift toward precise, transient, and RNA-centric modalities, λGRTS represents a significant advance toward physiologically relevant control of protein expression.

Methods

Cell lines and cell culture

Mammalian cell lines, including HEK293T, GL261, and MDBK, were obtained from the American Type Culture Collection (ATCC, USA). All mammalian cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, Gibco, USA) supplemented with 10% fetal bovine serum (FBS, Gibco, USA) and 1% penicillin–streptomycin (pan-strep, Gibco, USA). Incubation was carried out at 37 °C in a humidified atmosphere containing 5% carbon dioxide (CO₂). All cellular manipulations were performed during the logarithmic growth phase to ensure consistent cell viability and physiological state.

Plasmid cloning and lentivirus production

All plasmids were constructed via the Gibson assembly method and verified by Sanger sequencing conducted by Genewis Company (Suzhou, China). Plasmid vectors utilized in this study were procured from Addgene, including pCDH-myr-Akt (Addgene: Plasmid #46,969) [70], pLKO (Addgene: Plasmid #10,878) [71], and CRISPR-v2 (Addgene: Plasmid #52,961) [72]. Target gene fragments were amplified by polymerase chain reaction (PCR) using synthetic primers. Following agarose gel electrophoresis (incorporating 0.01% ethidium bromide), target bands were visualized under ultraviolet illumination, and the corresponding DNA fragments were purified using the FastPure Gel DNA Extraction Mini Kit (Vazyme: DC301). Each translation-guided RNA (tgRNA) was chemically synthesized and cloned into the CRISPR-v2 plasmid, which was engineered to co-express both the fusion protein and tgRNA. For lentivirus production, HEK293T cells were co-transfected with the lentiviral λGRTS plasmid and packaging plasmids using polyethylenimine (PEI) at a concentration of 50 μg/mg. Approximately 3 days post-transfection, lentiviruses were harvested, purified, and concentrated using a commercially available lentivirus purification kit, following the manufacturer’s protocols.

Strains and culture conditions

DH-5α competent Escherichia coli (E. coli, NEB Company) were employed for plasmid transformation and amplification throughout the study. Bacterial cultures were grown in Luria–Bertani (LB) broth at 37 °C with constant shaking at 200 rpm. For the selection of plasmid-transformed strains, bacterial cells were plated on LB agar plates supplemented with 20 μg/ml ampicillin, and single colonies were picked for subsequent expansion and plasmid extraction.

Fluorescence measurements

Cells transfected with different plasmids were seeded into 96-well plates at a standardized density. Optical density (OD) and fluorescence intensity were measured using a PERKLIN Microplate Reader. The endpoint fluorescence of green fluorescent protein (GFP) was detected with an excitation wavelength of 450 nm and an emission wavelength of 532 nm; fluorescence values were normalized to the OD at 488 nm to account for variations in cell number. For kinetic experiments, OD values were recorded every 30 min over the experimental period. All statistical data derived from these measurements were used for image generation and quantitative analysis. Notably, to ensure the accuracy of fluorescence readings, the culture medium in each well was carefully aspirated and replaced with phosphate-buffered saline (PBS) prior to measurement, eliminating potential interference from medium components.

Flow cytometry

Flow cytometry analyses were performed using a CytoFLEX flow cytometry system. The efficiency of the translation enhancement system was evaluated by quantifying the population of green fluorescent cells. Each experiment was independently replicated four times to ensure statistical robustness. For sample preparation, cells were trypsinized, followed by three rounds of rinsing with PBS to remove residual culture medium and cell debris. Samples were then analyzed directly on the CytoFLEX system, with a minimum of 20,000 cells collected per experimental group to ensure representative data. Flow cytometry data were processed, analyzed, and visualized using FlowJo_V10 software.

RNA extraction and real-time quantitative PCR

Total RNA was extracted from cultured cell lines using the Simply P Total RNA Extraction Kit (Bioflux, China), following the manufacturer’s instructions to minimize RNA degradation. Isolated total RNA was reverse-transcribed into complementary DNA (cDNA) using the Polymerus Reverse Transcription Kit (Cat. No. #R2020L, UElandy, China). RT-qPCR reactions were performed using Mon Amp Rapid Start Universal SYBR Green qPCR Mix (Monad, China) on a real-time PCR instrument, with cycling conditions optimized according to the kit specifications. The sequences of all primers used for RT-qPCR are provided in Table 1.

Table 1.

The primer of RT-qPCR

Gene (Homo sapiens)	Forward (5′−3′)	Reverse (5′−3′)
GAPDH	GGAGCGAGATCCCTCCAAAAT	GGCTGTTGTCATACTTCTCATGG
GFP	CGACAAGCAGAAGAACGGCATC	GACTTGAAGAAGTCGTGCTGCT
P53	GGAGCACTGCCCAACAACAC	GTGGTACAGTCAGAGCCAACC
PTEN	GCCAGATCATGTTGCAGCAAT	CAGAGGCGTAATGATCCCAG

Western blotting

Cells were seeded into 6-well plates, and plasmid transfection was performed when the cell confluency reached 60% to ensure optimal transfection efficiency. For protein extraction, cells were lysed using radioimmunoprecipitation assay (RIPA) buffer (Solarbio, China) supplemented with 1% phenylmethylsulfonyl fluoride (PMSF, Cat. No. #36,978, ThermoFisher, USA) to inhibit protease activity. The resulting cell lysates were centrifuged at 12,000 rpm for 15 min at 4 °C to remove cellular debris, and the supernatants (containing soluble proteins) were collected. The protein concentration in each supernatant was quantified using a bicinchoninic acid (BCA) protein assay kit to ensure equal loading of samples. Prior to electrophoresis, protein samples were denatured by boiling at 100 °C for 10 min in sodium dodecyl sulfate (SDS) loading buffer.

Denatured proteins were separated via 10% SDS–polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred onto a polyvinylidene difluoride (PVDF) membrane (Millipore). The PVDF membrane was pre-activated with methanol to enhance protein binding, and the transfer was conducted at a constant current of 400 mA for 1 h. After transfer, the membrane was blocked with 5% bovine serum albumin (BSA) in Tris-buffered saline with Tween-20 (TBST) at room temperature for 2 h to block non-specific binding sites. Following blocking, the membrane was incubated overnight at 4 °C with the appropriate primary antibodies. After three washes with TBST (each wash lasting 10 min), the membrane was incubated with horseradish peroxidase (HRP)-conjugated secondary antibodies (Proteintech: SA00001-1 and SA00001-2) at room temperature for 1 h. Finally, protein bands were visualized by incubating the membrane with ECL hypersensitive luminescent solution (Cat. No. #32,106, ThermoFisher, USA) and imaged using a GV6000M developer (GelView 6000pro). The primary antibodies used in this study included GAPDH (Proteintech, 10,494–1-AP), HA tag (Proteintech, 51,064–2-AP), GFP (Proteintech, 50,403–2-AP), eIF4E1 (Proteintech, 11,149–1-AP), 4EBP (Proteintech, 60,246–1-Ig), P53 (Proteintech, 10,442–1-AP), PTEN (Proteintech, 60,300–1-Ig), MMP-2 (Proteintech, 10,373–2-AP), MMP-9 (Proteintech, 10,375–2-AP), E-Cadherin (Proteintech, 20,847–1-AP), N-Cadherin (Proteintech, 22,018–1-AP), PCNA (Proteintech, 10,205–2-AP), and Ki67 (Proteintech, 28,074–1-AP).

Co-immunoprecipitation

Cells were lysed in ice-cold lysis buffer (25 mM Tris–HCl, pH 7.4; 150 mM NaCl; 1 mM EDTA; 1% NP-40; 5% glycerol) supplemented with 1% PMSF (Solarbio) to prevent protein degradation. After lysis on ice for 15 min, the cell lysates were scraped into microcentrifuge tubes and centrifuged at 12,000 rpm for 15 min at 4 °C to clarify the lysates. A 1% aliquot of the clarified supernatant was retained as the “input” sample to verify protein expression. The remaining supernatant was incubated overnight at 4 °C with the target primary antibody (normal rabbit IgG was used as a non-specific control to rule out background binding). The following day, protein A/G beads (Selleck) were added to the antibody-lysate mixture, and the samples were incubated for an additional 3 h at 4 °C with gentle rotation to allow the formation of antibody-bead complexes.

The beads were then collected by centrifugation and washed three times with ice-cold lysis buffer to remove unbound proteins and impurities. Bound proteins were eluted by resuspending the beads in SDS loading buffer and boiling at 100 for 10 min. The eluted proteins, along with the input samples, were separated by SDS-PAGE and analyzed via western blotting as described above. The primary antibodies used for Co-IP were 4EBP and HA tag.

RNA immunoprecipitation

HEK293T cells were seeded into 10-cm culture dishes and transfected with the indicated plasmids following the previously described protocol. To crosslink protein-RNA complexes, the cells were irradiated with 254 nm ultraviolet (UV) light at a dose of 300 mJ/cm². The next day, the cells were harvested and fractionated into nuclear and cytoplasmic fractions using a subcellular fractionation kit (not specified, standard protocol) to analyze protein-RNA interactions in distinct cellular compartments. The nuclear pellets were sonicated using a Bioruptor Pico device (5 cycles of 30 s ON/30 s OFF) to shear genomic DNA and disrupt nuclear structures, generating soluble nuclear lysates.

For immunoprecipitation, the nuclear and cytoplasmic lysates were incubated overnight at 4 °C with the HA tag antibody to capture HA-λN-associated RNA. The following day, protein A/G magnetic beads (Invitrogen) were added to the lysate-antibody mixture and incubated for 3 h at 4 °C with gentle rotation. After incubation, the magnetic beads were collected using a magnetic stand and washed sequentially with low-salt, high-salt, and LiCl wash buffers (standard RIP wash buffers) to remove non-specifically bound RNA. To isolate the immunoprecipitated RNA, the beads were treated with protease K at 55 °C overnight to digest associated proteins. The RNA was then extracted using a combination of chloroform (WAKO) and Trizol reagent (Thermo Fisher Scientific, Germany) according to the manufacturer’s instructions. The relative abundance of target RNA in the immunoprecipitated samples was quantified using quantitative real-time PCR (qRT-PCR).

Cell cycle detection

U87 and U251 cells were infected with λGRTS-P53, λGRTS-PTEN lentiviruses, or empty lentivirus (normal control group). At 48–72 h post-infection, the cells were trypsinized, washed twice with phosphate-buffered saline (PBS), and fixed in 80% ethanol at − 20 °C overnight to preserve cell morphology and DNA integrity. Prior to staining, the fixed cells were centrifuged to remove ethanol, washed twice with PBS, and resuspended in PBS containing Red Nucleus I (a fluorescent DNA dye) according to the manufacturer’s recommended concentration. The cells were incubated in the staining solution at 37 °C for 30 min in the dark to allow uniform dye binding to double-stranded DNA.

Cell cycle distribution was analyzed using a flow cytometer, with fluorescence intensity detected in the FL2 channel (excitation wavelength: 638 nm). For each sample, at least 10,000 events were collected to ensure statistical reliability. The flow cytometry data were processed using FlowJo software, and the proportions of cells in the G1, S, and G2/M phases of the cell cycle were calculated based on the DNA content histogram.

Cell proliferation assay

Cell proliferation was evaluated using the Cell Counting Kit-8 (CCK-8) assay (Yeason, China), which measures the activity of mitochondrial dehydrogenases in viable cells. Briefly, cells were seeded into 96-well plates at a density of 3 × 10³ cells per well and cultured under standard conditions (37 °C, 5% CO₂). At 24, 48, 72, and 96 h post-seeding, 10 μL of CCK-8 reagent was added to each well, and the plates were incubated for an additional 3 h at 37 °C to allow the formation of the colored formazan product. The absorbance of each well was measured at 450 nm using a microplate reader, with higher absorbance values indicating greater cell viability and proliferation. All assays were performed in triplicate, and the results were expressed as the mean ± standard deviation (SD) of three independent experiments.

Transwell assay

Transwell assays were conducted to assess cell migration and invasion capabilities using 24-well Transwell chambers (Corning, USA). For both assays, cells were harvested 48 h post-transfection, washed twice with PBS, and resuspended in serum-free DMEM to eliminate the influence of serum-derived growth factors.

Migration assay

A total of 5 × 10⁴ cells in 200 μL of serum-free DMEM were added to the upper chamber of uncoated Transwell inserts. The lower chamber was filled with 500 μL of DMEM supplemented with 25% FBS to serve as a chemoattractant.

Invasion assay

Prior to seeding cells, the upper chamber membranes were pre-coated with Matrigel (Yeasen, China) diluted 1:8 in serum-free DMEM and incubated at 37 °C for 30 min to form a gel matrix that mimics the extracellular matrix (ECM). A total of 5 × 10⁴ cells in 200 μL of serum-free DMEM were then added to the upper chamber, and the lower chamber was filled with 500 μL of 25% FBS-supplemented DMEM.

After incubation at 37 °C for 24–48 h (migration: 24 h; invasion: 48 h), the cells remaining on the upper surface of the membrane were gently removed using a cotton swab. The cells that had migrated or invaded to the lower surface of the membrane were fixed with 4% paraformaldehyde for 15 min, stained with 0.1% crystal violet for 20 min, and rinsed with PBS to remove excess dye. The stained cells were observed and imaged under an inverted microscope, and the number of migrated or invaded cells was counted in five randomly selected fields per insert to quantify migration/invasion efficiency.

Binding site screening and molecular docking

To identify key amino acid residues involved in protein–protein interactions, a functional analysis of known protein complexes was first performed to narrow down potential interaction interfaces. Structural files containing detailed protein binding information were retrieved from the Protein Data Bank and UniProt database. Automated protein docking simulations were carried out using a standard docking protocol to predict the three-dimensional (3D) structure of the protein complex, and the interaction sites were visualized to identify critical amino acid residues that mediate binding.

PyMOL software was used to generate high-resolution visualizations of the protein complex, with a focus on highlighting covalent bonds, hydrogen bonds, and hydrophobic interactions between the interacting proteins. This approach allowed for the precise mapping of key binding sites, which was essential for validating and optimizing the design of the λGRTS.

Construction of xenograft model

Male BALB/c nude mice (4–5 weeks old, GemPharmatech, Nanjing, China) were randomly divided into three groups (n = 10 per group) and housed under specific-pathogen-free (SPF) conditions with a 12-h light/dark cycle, ad libitum access to food and water, and controlled temperature (22–25 °C) and humidity (50–60%). To establish subcutaneous xenograft tumors, 5 × 10⁶ transfected GBM cells were resuspended in 100 μL of PBS and injected subcutaneously into the right flank of each mouse.

Tumor volume was measured weekly using a digital caliper, and the volume was calculated using the formula: Volume = (length × width²)/2. Four weeks after the lentivirus carrying the λGRTS, the mice were humanely euthanized under isoflurane anesthesia. The subcutaneous tumor tissues were dissected, and the tumor volume and mass were measured to assess tumor growth inhibition. All animal experiments were approved by the Animal Ethics Committee of the Second Affiliated Hospital of Nanchang University, and all procedures strictly complied with the institutional guidelines for the care and use of laboratory animals.

HE and IHC staining

Mouse brain tumor tissues were surgically excised and immediately fixed in 4% paraformaldehyde for 24 h at room temperature to preserve tissue morphology. After fixation, the tissues were dehydrated using a gradient of sucrose solutions (10%, 20%, 30%) to prevent ice crystal formation during freezing, embedded in paraffin, and sectioned into 3-μm-thick slices using a microtome.

HE staining

The paraffin sections were dewaxed in xylene (twice for 10 min each), rehydrated through a gradient of ethanol (100%, 95%, 80%, 70%), and stained with hematoxylin for 5 min to visualize cell nuclei. The sections were then differentiated in 1% hydrochloric acid–ethanol for 30 s, blued in running tap water for 10 min, and counterstained with eosin for 3 min. After dehydration and clearing, the sections were mounted with neutral balsam and imaged under a light microscope.

IHC staining

For IHC, dewaxed and rehydrated sections were treated with 3% hydrogen peroxide (H₂O₂) for 10 min at room temperature to block endogenous peroxidase activity. Antigen retrieval was performed by boiling the sections in 0.01 M sodium citrate buffer (pH 6.0, item #G1202, Servicebio, Wuhan, China) for 15 min using a microwave oven, followed by cooling to room temperature. The sections were blocked with 5% BSA for 1 h at room temperature to prevent non-specific antibody binding, then incubated overnight at 4 °C with the appropriate primary antibody (e.g., P53, PTEN) at the recommended dilution.

The next day, the sections were washed three times with PBS and incubated with HRP-conjugated secondary antibody for 2 h at 37 °C. After three additional washes with PBS, the sections were stained with 3,3′-diaminobenzidine (DAB) substrate solution until the desired brown color developed, and the reaction was stopped by immersing the sections in distilled water. The sections were counterstained with hematoxylin for 30 s, dehydrated, cleared, and mounted. IHC images were captured using a Leica microscope (Wetzlar, Germany), and the staining intensity was quantified using ImageJ software to assess the expression levels of target proteins in tumor tissues.

Mass spectrometry analysis

To evaluate the specificity and potential off-target effects of the λGRTS, quantitative mass spectrometry was employed to analyze global protein expression changes in HEK293T cells stably expressing the L30-GFP reporter. Cells were transfected with the λGRTS fusion plasmid (targeting GFP-21) or a non-targeting control plasmid. At 6 h post-transfection, cells were harvested and lysed in RIPA buffer supplemented with protease inhibitors. Protein concentration was determined using a BCA assay.

Liquid chromatography-tandem mass spectrometry (LC–MS/MS) was performed on an Orbitrap Fusion Lumos mass spectrometer (Thermo Fisher Scientific) coupled to an EASY-nLC 1200 system. Peptides were separated on a 75 μm × 25 cm C18 column with a 120-min gradient. MS1 spectra were acquired at 120,000 resolution, and MS2 spectra were generated using higher-energy collisional dissociation (HCD) at 38% collision energy.

Raw data were processed using MaxQuant (version 2.0.3.0) with the built-in Andromeda search engine against the human UniProt database. TMTpro quantification was performed with reporter ion mass tolerance set to 0.003 Da. Protein abundance ratios (λGRTS/Control) were log2-transformed, and significance was assessed using a two-sided t-test with Benjamini–Hochberg correction. Proteins with a fold change > 1.5 and adjusted P value < 0.05 were considered significantly altered.

GFP mRNA synthesis and delivery

GFP mRNA was synthesized in vitro using a T7 promoter-driven linearized DNA template containing the GFP coding sequence and a poly(A) tail (≥ 70 nucleotides). The transcription reaction was performed at 37 °C for 2 h using a commercial kit (e.g., NEB HiScribe T7 ARCA mRNA Kit), incorporating the anti-reverse cap analog (ARCA) and a poly(A) tail (≥ 70 nucleotides). Following synthesis, the DNA template was digested with DNase I, and the resultant mRNA was purified via lithium chloride precipitation, resuspended in RNase-free water, and quantified spectrophotometrically. Integrity was confirmed by agarose gel electrophoresis. For cell delivery, HEK293T stably expressing the λGRTS was seeded in a 24-well plate and transfected with Lipofectamine MessengerMAX at a fusion rate of 70–90%. Briefly, mRNA (0.2–1.0 µg per well) and the transfection reagent were diluted separately in Opti-MEM, combined, incubated for 5–15 min to form complexes, and added to the cells. GFP expression was assessed by fluorescent microscopy and/or flow cytometry 24–48 h post-transfection.

Statistical analysis

All statistical analyses were performed using GraphPad Prism 9 software. For comparisons between two groups, an unpaired Student’s t-test was used to determine significant differences in protein expression, cell proliferation, or tumor volume. For comparisons involving three or more groups, one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test was employed to account for multiple comparisons.

All experimental data are presented as the mean ± standard deviation (SD) or mean ± standard error of the mean (SEM), as specified in the figure legends. A P value less than 0.05 (P < 0.05) was considered statistically significant, with the following notation used to indicate significance levels: n.s. (not significant, P > 0.05), * (P < 0.05), ** (P < 0.01), and *** (P < 0.001). Each experiment was independently replicated at least three times to ensure the reliability and reproducibility of the results.

Abbreviations

λGRTS

λN-Guided RNA Targeting System

CRISPR

Clustered Regularly Interspaced Short Palindromic Repeats

tgRNAs

Translation-guided RNAs

eIF4E1

Eukaryotic translation initiation factor 4E1

TICs

Translation initiation complexes

PABP

Poly(A)-binding protein

GFP

Green fluorescent protein

GBM

Glioblastoma

HE

Hematoxylin and eosin

IHC

Immunohistochemistry

4EBP

EIF4E binding protein

dsRBPs

Double-stranded RNA-binding proteins

TIFs

Translation initiation factors

CMV

Cytomegalovirus

NES

Nuclear export signals

NLS

Nuclear localization signals

IRESs

Internal ribosomal entry sites

PBS

Phosphate-buffered saline

CCK-8

Cell Counting Kit-8

Authors’ contributions

Subject design, project conception and manuscript writing were carried out by Kai Huang; Junzhe Liu, Yonghui Lv and Zhihong Zhou performed the experimental operation and data collation; Linzhen Huang, Kunjian Lei, Zewei Tu and Chuandong Gong provided help for statistical analysis. Chong Wang and Zhansheng Fang was responsible for plasmid construction; Juan Wang gave guidance and suggestions for many experiments; Xingen Zhu, Minhua Ye and Shigang Lv provided key resources and necessary help for the experiment. All authors read and approved the final manuscript.

Funding

This study was supported by the National Natural Science Foundation of China (grant no. 82472827, 82273068, 82460504, and 82172989); Key Research and Development Projects in Jiangxi (grant no. 20212BBG73021); Jiangxi Training Program for Academic and Technical Leaders of Major Disciplines—Young Talents Program (grant no. 20212BCJ23023); Jiangxi Provincial Natural Science Foundation (grant no. 20232BAB206101); Project Fund of Jiangxi Provincial Health Commission (20204343, 202130873, 202210044); Jiangxi Clinical Medical Research Center of Nervous System Diseases (20223BCG74009); and Jiangxi Provincial Postgraduate Innovation Fund Project (No. YC2025-B001).

Data availability

All data generated or analysed during this study are included in this published article and its supplementary information files. The protein sequence data with accession numbers P06730 [73], Q8BMB3 [74] P63074 [75] and P63073 [76] were obtained from the UniProt database [77]. The gene sequences were provided by the National Center for Biotechnology Information Sequence Read Archive with accession number PRJNA270632 [78]. The three-dimensional structure of the protein was obtained from the PDB database with accession number 1IPD [32] and 3U7X [33]. The raw data of quantitative mass spectrometry are publicly available via figshare (https://figshare.com/articles/dataset/_/31742590) [79].

Declarations

Ethics approval and consent to participate

The procedures were conducted in accordance with the principles of Animal Care of Nanchang University (Nanchang, China) and approved by the Laboratory Animal Committee (NCULAE-20250717001).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.