2025 Volume 48 Issue 5 Pages 706-712
mRNA vaccines have emerged as promising platforms for the prevention of infectious diseases and cancer treatment. The antigenic protein has a signal peptide added to the N-terminus for extracellular secretion. However, it remains unclear whether the optimization of signal peptides has been sufficiently compared and examined for antigen protein secretion and immunogenicity. This study investigated the effects of various signal peptides on the extracellular secretion of a model protein, NanoLuc luciferase (Nluc), in different cell lines. We compared the secretion efficiency of Nluc fused to artificial (#38 and #34) and natural signal peptides (cystatin S, lactotransferrin, and tissue plasminogen activator) in human embryonic kidney 293, C2C12, and HepG2 cells. Luciferase assays and Western blot analysis revealed that the cystatin S signal peptide consistently induced the highest secretion of Nluc among all cell types tested. Notably, the cystatin S signal peptide outperformed previously reported tissue plasminogen activator signal peptides in terms of secretion efficiency. Furthermore, we observed no correlation between Nluc secretion and mRNA expression levels for each signal peptide, suggesting that enhanced secretion was not attributable to increased transcription. Our findings highlight the potential of the cystatin S signal peptide in enhancing the extracellular secretion of antigenic proteins in mRNA vaccines by improving the efficiency of protein translation.
mRNA vaccines have been the subject of research and development for clinical application since the 1990s, primarily focusing on cancer therapy.1) However, significant challenges, including the rapid degradation of mRNA within the body and unintended activation of innate immune responses, impede their practical application.2) Recent advances in drug delivery systems and modified nucleic acids have facilitated substantial progress in mRNA vaccine research.3) Notably, mRNA vaccines were first implemented during the coronavirus disease 2019 pandemic, demonstrating their potential efficacy in preventing infectious diseases.4,5)
mRNA vaccines can be designed within a short timeframe. Unlike DNA-based gene therapies, mRNA vaccines do not require nuclear delivery, thereby eliminating the risk of genomic integration and ensuring a favorable safety profile.4,5) Theoretically, mRNA vaccines can target any protein, expanding their therapeutic applications beyond infectious diseases to include cancer and genetic disorders.6) Despite ongoing research, these applications have yet to achieve widespread clinical implementation. For instance, regarding cancer therapy, the immune evasion capabilities of tumor cells may necessitate combination treatments such as the use of immune checkpoint inhibitors in conjunction with mRNA vaccines.
A critical aspect of mRNA vaccine efficacy is the efficient secretion of the encoded antigen. In vivo, mRNA is susceptible to degradation and can elicit inflammatory responses. To mitigate these effects, mRNA vaccines encapsulate antigen-encoding mRNA within lipid nanoparticles (LNPs), thereby protecting the mRNA and attenuating inflammation.7) Upon intramuscular administration, the mRNA is translated into antigens within the cells. The secreted antigen is subsequently recognized by antigen-presenting cells, initiating both cellular and humoral immune responses.3) Enhancing the extracellular secretion of antigens could potentially improve their recognition by antigen-presenting cells, leading to more efficacious immune responses at lower mRNA doses. This strategy focuses on optimizing the levels of signal peptides that govern the localization and transport of secreted proteins.
Signal peptides are N-terminal sequences in secreted proteins comprising an initial methionine, a positively charged n-region, a hydrophobic core, and a c-region terminating at the cleavage site. The hydrophobic core exhibits high variability and is abundant in amino acids, such as leucine, valine, and alanine. During protein synthesis, the mRNA of secreted proteins is initially translated by free ribosomes in the cytoplasm, generating a signal peptide. Signal recognition particles (SRPs) recognize this peptide and direct the ribosome–mRNA complex to the endoplasmic reticulum (ER).8–10) Consequently, the complex binds to a translocator on the ER membrane, and the signal peptide is cleaved by a signal peptidase, producing a mature protein. This protein undergoes folding and post-translational modifications in the ER and Golgi apparatus before extracellular secretion.11)
Previous studies have identified both artificial (#38 and #34) and natural (cystatin S and lactotransferrin) signal peptides that enhance protein secretion by replacing the native signal peptide of alkaline phosphatase with a model secretory protein.12) Furthermore, replacing the signal peptide with the tissue plasminogen activator (tPA) has been reported to increase the protein secretion levels of antigens in vaccines.5,13) However, comparisons of protein secretion levels between artificial and natural signal peptides, as well as evaluation of cell types beyond human embryonic kidney 293 (HEK293) cells, remain limited in the literature. Considering that LNPs tend to accumulate in the liver after intramuscular injection,14) it is important to assess the efficacy of various signal peptides in different cell types, including muscles and hepatocytes. A recent study demonstrated that modifying the signal sequence in mRNA vaccines could significantly enhance antigen expression, thereby strengthening immune protection against viral infections.7,15) These findings underscore the potential of signal sequence optimization to improve the efficacy of mRNA vaccines.
This study compared the protein secretion efficiencies of artificial and natural signal peptides in various cell lines and identified the most effective signal peptide to enhance extracellular protein secretion, thereby improving the immunogenicity and efficiency of mRNA vaccines.
For preparation of insert DNA fragments, 5 μL of 100 μM each oligonucleotide containing #38, #34, cystatin S, lactotransferrin, and tPA signal peptide sequences (Supplementary Table S1), 2 μL of 10× oligonucleotide annealing buffer (Invitrogen, Waltham, MA, U.S.A.), and 8 μL of nuclease-free water were mixed and incubated at 95°C for 4 min. After incubation, the mixtures were allowed to equilibrate at room temperature (r.t.) for 10 min.
Annealed oligonucleotides of the signal peptides were used as insert DNAs. The pNL1.3.CMV[secNluc/CMV] vector (Promega Corporation, Madison, WI, U.S.A.), which expresses NanoLuc® luciferase (Nluc; naturally occurring secretion signals are removed) N-terminally fused with the native interleukin-6 (IL-6) signal peptide by the cytomegalovirus (CMV) enhancer/promoter, was employed as a DNA template for the vector. The pNL1.3.CMV vector without the IL-6 sequence was linearized by inverse PCR using the primers listed in Supplementary Table S2 and KOD FX Neo (TOYOBO, Osaka, Japan).
The In-Fusion HD Cloning Kit was obtained from TaKaRa Bio (Shiga, Japan), and the standard solution comprised the following components: 25–100 ng linearized vector, insert DNA (2 : 1 M ratio of insert to vector), 2 μL of 5× In-Fusion HD enzyme premix, and nuclease-free water to a total volume of 10 μL. The In-Fusion reaction mixture was incubated at 50°C for 15 min in accordance with the manufacturer’s instructions. All plasmids constructed in this study were confirmed through DNA sequencing conducted by Eurofins Genomics, Tokyo, Japan.
Cell CulturesHEK293 cells were maintained in high-glucose Dulbecco’s modified Eagle’s medium (DMEM; FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan) supplemented with 10% (v/v) fetal bovine serum (FBS) and 1% penicillin–streptomycin solution (×100; FUJIFILM Wako Pure Chemical Corporation). HepG2 cells were maintained in low-glucose DMEM (FUJIFILM Wako Pure Chemical Corporation) supplemented with 10% (v/v) FBS and 1% penicillin–streptomycin solution (×100). C2C12 cells were provided by Dr. Katsuya Hirasaka (Nagasaki University, Nagasaki, Japan) and maintained in high-glucose DMEM supplemented with 10% (v/v) FBS and 1% penicillin–streptomycin solution (×100). These cells were cultured at 37°C in a humidified atmosphere containing 5% CO2.
TransfectionThe transfection procedure was previously described,16) and the protocols provided by each manufacturer were followed. HEK293 and HepG2 cells in 24-well plates were transfected with the Nluc vectors and Firefly (Ffly) luciferase vector pGL4.54[luc2/TK] (Promega Corporation) using Screenfect™ A Transfection Reagent (FUJIFILM Wako Pure Chemical Corporation). Ffly vectors served as the internal standards. After 24 h, the medium was replaced with serum-free Opti-MEM, and the culture supernatant and cells were collected after an additional 24 h.
C2C12 cells in 24-well plates were transfected with the Nluc vectors and Ffly vector pGL4.53[luc2/PGK] (Promega Corporation) using GenJet™ Reagent (SignaGen Laboratories, Ijamsville, MD, U.S.A.). After 48 h, the medium was replaced with DMEM (high glucose) containing 2% (v/v) horse serum to induce differentiation. After 72 h, the culture supernatant and the cells were collected.
Luciferase AssayCells were lysed by the addition of 300 μL/well of Passive Lysis Buffer (Promega Corporation) and subjected to agitation at r.t. for 10 min. The lysate and culture medium were centrifuged and the resulting supernatant was transferred into new tubes.
The Nano-Glo® Live Cell Assay System was used to analyze the culture supernatants, while the Nano-Glo® Dual-Luciferase® Reporter Assay System was utilized for the cell lysates to measure the luminescence of Nluc and Ffly, following their respective manufacturer’s protocols. Luminescence measurements were obtained using a LuMate 4400 plate reader (Awareness Technology, Palm, FL, U.S.A.).
Western Blotting (WB)This methodology was based on previously established protocols,17,18) and the membranes were incubated overnight with primary antibodies at 4°C, followed by incubation with secondary antibodies at r.t. for 1 h. The primary antibodies used were anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1 : 1000; sc-32233, Santa Cruz Biotechnology Inc., Dallas, TX, U.S.A.) and anti-NanoLuc (1 : 1000; N700A, Promega Corporation). Horseradish peroxidase-conjugated goat anti-mouse immunoglobulin G (1 : 5000; 1031-05, Southern Biotech, Birmingham, AL, U.S.A.) was used as the secondary antibody. GAPDH served as the loading control for WB, and the intensity of the Western blot bands was quantified using ImageJ software.
RT-PCRAll cells were maintained in a 6-well plate and subjected to total RNA extraction using TRI Reagent (Molecular Research Center Inc., Cincinnati, OH, U.S.A.) according to the manufacturer’s instructions. Template cDNA was synthesized using ReverTra Ace (TOYOBO) in a reaction conducted at 30°C for 1 min and 42°C for 60 min.
PCR amplification was conducted using 0.1 μL of Paq5000 DNA polymerase (5.0 U/μL; Agilent Technologies Inc., Santa Clara, CA, U.S.A.), 1 μL of 10 × Paq5000™ Reaction Buffer, 1 μL of dNTP mixture (2.5 mM each), 0.1 μL of forward primer (20 μM), 0.1 μL of reverse primer (20 μM; sequences of oligonucleotides were listed in Supplementary Table S3), 6.7 μL of nuclease free water, and 1 μL of template cDNA (described in this section) were added in PCR tubes. PCR amplification was performed under the following cycling conditions: initial denaturation at 95°C for 2 min, followed by 25 cycles of denaturation at 95°C for 20 s, annealing at a temperature 5°C for 20 s, and extension at 70°C for 30 s. The final extension step was performed at 72°C for 5 min. GAPDH and β-actin were used as reference genes for normalization.
Statistical AnalysisAll data are expressed as mean ± standard error of the mean. Statistical comparisons were performed using Dunnett’s test, and significance was set at p < 0.05. All analyses were performed using JMP Pro 17 software (version 17.2.0, SAS Institute Inc., Cary, NC, U.S.A.).
Previous studies have shown that both artificial (#38 and #34) and natural (cystatin S and lactotransferrin) signal peptides enhance protein secretion.12) We compared the protein secretion levels mediated by each signal peptide (IL-6, #38, #34, cystatin S, and lactotransferrin; Table 1) with the native IL-6 signal peptide, using Nluc luciferase as a model for secretory proteins. Vectors encoding Nluc fused with various signal peptides (IL-6, #38, #34, cystatin S, and lactotransferrin) at the N-terminus, or vectors encoding Nluc without a signal peptide (deletion), were transfected into HEK293 cells. The fly luciferase vector was co-transfected into all cells as an internal standard, and luciferase assays were performed using the collected cells and culture supernatants. In HEK293 cells, the luminescence of intracellular Nluc was significantly higher after deletion (Fig. 1A). Conversely, extracellular Nluc demonstrated that cystatin S was the most effective signal peptide for protein secretion (Fig. 1B).
| Artificial signal peptide | |
| #38 | MWWRLWWLLLLLLLWPMVWA |
| #34 | MRPTWAWWLFLVLLLALWAPARG |
| Natural signal peptide | |
| Cystatin S | MAYLLHAQLFLLTTFILVLNMRLCPVL |
| Lactotransferrin | MKLVFLVLLFLGALGLCLA |
| tPA | MDAMKRGLCCVLLLCGAVFVSP |
Intracellular (A) and extracellular (B) Nluc luminescence levels in HEK293 cells transfected with vectors encoding Nluc fused to different signal peptides (IL-6, #38, #34, cystatin S, and lactotransferrin) or without a signal peptide (deletion). Comparison of intracellular (C) and extracellular (D) Nluc luminescence levels of the cystatin S and tPA signal peptides. Nluc protein levels were quantified using a luciferase assay and normalized to firefly luciferase (Ffly) as the internal standard. Data are expressed as fold increase relative to the IL-6 signal peptide group (filled in gray; mean ± standard error of the mean [S.E.M.], n = 4 [Figs. 1A, 1B] or n = 6 [Figs. 1C, 1D]; Dunnett’s test, vs. IL-6, *p < 0.05, **p < 0.01).
The tPA signal peptide also enhanced protein secretion into the extracellular space.13) Therefore, we compared the protein expression levels of Nluc by cystatin S and tPA using a luciferase assay (Figs. 1C, 1D). The results showed that cystatin S (Fig. 1C) exhibited higher Nluc luminescence levels than tPA in the extracellular space (Fig. 1D).
In HEK293 cells, the protein secretion levels directed by each signal peptide were compared using WB (Figs. 2A, 2B). Consistent with the results of the luciferase assay, cystatin S was the most effective signal peptide for protein secretion (Fig. 2B).
Western blotting of (A) intracellular and (B) extracellular Nluc expression in HEK293 cells transfected with different signal peptides. GAPDH was used as a loading control. The quantification of Nluc levels was normalized to that of GAPDH. Data are shown as the fold increase relative to the IL-6 signal peptide group (filled in gray; mean ± S.E.M., n = 3; Dunnett’s test vs. IL-6, *p < 0.05).
Given that mRNA vaccines are administered via intramuscular injection, we used C2C12 cells, a myoblast cell line derived from the mouse skeletal muscle. As mRNA vaccines encapsulated in LNPs tend to accumulate in the liver, we used HepG2 cells, a human hepatocellular carcinoma-derived cell line. Protein secretion levels induced by each signal peptide or deletion of the signal peptide were compared in C2C12 and HepG2 cells.
Cystatin S was the most potent signaling peptide for protein secretion in C2C12 (Fig. 3B) and HepG2 cells (Fig. 3D), although the levels of lactotransferrin and #38 differed from those in HEK293 cells. These findings suggest that the signal peptide of cystatin S most effectively enhanced the extracellular secretion of Nluc in all cells examined.
Intracellular (A) and extracellular (B) Nluc expression in C2C12 cells transfected with different signal peptides. (C) Intracellular and (D) extracellular Nluc expression levels in HepG2 cells transfected with different signal peptides. Luciferase assays were performed to quantify secretion levels normalized to firefly luciferase as an internal standard. Data are expressed as fold increase relative to the IL-6 signal peptide group (filled in gray; mean ± S.E.M., n = 3 [Figs. 3A, 3B] or n = 4 [Figs. 3C, 3D]; Dunnett’s test, vs. IL-6, *p < 0.05, ***p < 0.005).
Luciferase assays and WB demonstrated that cystatin S enhanced the extracellular secretion of Nluc in HEK293, C2C12, and HepG2 cells. However, this result could be attributed to increased Nluc mRNA expression. To investigate this possibility, vectors expressing Nluc fused with or deleted of various signaling peptides (IL-6, #38, #34, cystatin S, and lactotransferrin) at the N-terminus were transfected into HEK293, C2C12, and HepG2 cells. Subsequently, Nluc mRNA expression levels in these cells remained unchanged across all cell types (Figs. 4A–4C). These findings suggest that the signal peptide cystatin S consistently promoted extracellular secretion of the Nluc protein.
Nluc mRNA expression in (A) HEK293, (B) C2C12, and (C) HepG2 cells. Nluc mRNA expression was normalized to GAPDH or β-actin as internal standards. No significant differences were observed among the groups, indicating that the variations in protein secretion levels were not due to differences in mRNA expression. Data are presented as fold increase relative to the IL-6 signal peptide group (filled in gray; mean ± S.E.M., n = 3).
In this study, we used Nluc as a model of secreted proteins and evaluated the efficiency of various signal peptides in extracellular secretion. Luciferase assays and WB demonstrated that Nluc, carrying the cystatin S signal peptide, exhibited the highest secretion levels in HEK293, C2C12, and HepG2 cells. These findings suggest that the cystatin S signal peptide was the most effective of all studied peptides in enhancing the extracellular secretion of Nluc. Lactotransferrin and #38 also showed relatively high secretion levels; however, their rankings varied depending on the cell line. These results suggest that differences in mRNA translation activity and protein production capacity of the cells led to differences in Nluc secretion. Future studies should be similarly compared using other cell types.
Previous studies reported that the tPA signal peptide facilitates highly efficient secretion.13) However, in this study, the cystatin S signal peptide demonstrated a superior secretion efficiency, which may be due to variations in the model proteins used. The effectiveness of a signal peptide may be influenced by the characteristics of the protein that follows it. A recent study by Zhang et al. examined the impact of signal peptides using the receptor-binding domain of SARS-CoV-2 and revealed that the IL-6 signal peptide exhibited higher secretion efficiency than the tPA signal peptide.19) These findings underscore the necessity of evaluating signal peptides using antigenic proteins in vaccine development.
Furthermore, although previous studies have not directly compared the secretion efficiencies of artificial and natural signal peptides,12) our findings indicate that proteins containing natural signal peptides exhibit higher extracellular secretion than those containing artificial signal peptides. The central hydrophobic region of the signal peptide is instrumental in SRP recognition. Herein, artificial signal peptides #38 and #34 consisted entirely of hydrophobic amino acids, such as leucine and tryptophan, resulting in highly hydrophobic sequences (Table 1). Previous studies employing hidden Markov models to generate high bit-score domain sequences have demonstrated that artificial signal peptides #34 and #38 enhance protein secretion.12) However, our results suggest that factors other than increased hydrophobicity improve SRP recognition. Structural differences in signal peptides have been implicated in enhancing SRP recognition.20) The cystatin S signal peptide, which contains a high proportion of hydrophobic amino acids in its core region, also contains non-hydrophobic amino acids. Additionally, the lactotransferrin signal peptide comprises a relatively shorter amino acid sequence than artificial signal peptides. These structural differences may contribute to superior SRP recognition by, leading to more efficient protein secretion.
Furthermore, no correlation was observed between Nluc secretion and mRNA expression levels for each signal peptide, as determined by luciferase assays, suggesting that signal peptides do not influence transcription. Considering that signal peptides are cleaved in the ER, they could not be directly involved in post-translational secretion pathways. Instead, enhanced recognition by the SRP increases translation efficiency, leading to increased protein secretion. A computational simulation study by Zhang et al. demonstrated a positive correlation between the binding affinity of SRP54, a key component of SRP,21) and translation efficiency of antigenic proteins.19) However, our results demonstrated that the amount of Nluc protein in the intracellular space did not consistently correlate with that secreted into the extracellular space. Protein insertion into the ER is highly dependent on the cleavage of signal peptides by signal peptidase.22) Furthermore, proteins entering the ER are selectively degraded by a quality control system termed the ER stress-induced preemptive quality control (ERpQC) prior to translocation.23,24) Although the mechanism by which ERpQC discriminates between signal peptides remains to be elucidated, these processes may also contribute to the observed differences in secretion efficiency.
Although this investigation focused exclusively on Nluc, subsequent studies should incorporate these signal peptides into mRNA vaccine constructs encapsulated in LNPs and evaluate their in vivo extracellular secretion and immunogenicity. Such investigations are essential for optimizing mRNA vaccine design to enhance protein expression and immune response induction. If more efficient signal peptides are identified through this research, it will minimize the side effects and costs by reducing the RNA content of RNA vaccines.
This work was partly supported by JSPS KAKENHI (Grant Number: 23K17479; to SK). We are grateful to Dr. Katsuya Hirasaka at Nagasaki University for generously providing the C2C12 cells.
The authors declare no conflict of interest.
This article contains supplementary materials.
