A few years ago, mRNA technology in China experienced an unprecedented capital frenzy. Dozens of companies rushed into the field, billions of dollars poured in, and countless factories were built. At the time, people believed that mRNA was the master key to future drugs, and whoever held this key would open the door to the next era.

 

However, as the pandemic receded, this once-booming track quickly cooled down. Some companies were forced to shut down pipelines to survive, while many more names simply disappeared from industry news. People could not help but ask: what did those years of feverish investment in mRNA actually leave behind?

 

Lin Jinzhong is one of the few who can answer this question. He founded RNAimmune, which became the second company in China to receive emergency use authorization for a COVID-19 mRNA vaccine. In the early days of the venture, Lin and his core founding team, along with Fudan University students and key staff, independently completed the technical development of key steps, including antigen design, mRNA synthesis, LNP delivery, and industrial scale-up.

 

Recently, Lin gave an exclusive interview to TONACEA, reviewing RNAimmune's journey of building an mRNA technology platform from scratch, while also sharing the survival strategies and future outlook of an mRNA company in the post-pandemic industry ebb.

 

TONACEA 01: The Gamble of Antigen Design

 

Lin recalls that the day the Phase III clinical trial of RQ3013 was unblinded was the most stressful day of his career.

 

"My palms were sweating from morning till night. The regulatory authority required the use of a control vaccine rather than a placebo, with the lower bound of the 95% confidence interval for relative protective efficacy needing to exceed 30%. The WHO standard generally requires a lower bound greater than zero. This year, the lower bound for an mRNA influenza vaccine filed abroad was only 16.7%. This meant we needed to accumulate enough cases to achieve statistical significance. In the end, the results barely passed the threshold. When the results were announced, our clinical collaboration team was so moved that they cried."

 

Antigen design was the first technical challenge Lin faced. The success of RQ3013's Phase III clinical trial was largely due to an antigen design strategy that differed from mainstream approaches.

 

Lin explained that the most direct choice at the time would have been to use the antigen sequence of circulating strains, such as the Alpha, Beta variants, or the original strain. However, from antigen design to Phase III data readout, an mRNA vaccine typically takes over 18 months, while virus mutation cycles are measured in months. If a new immune-evasive variant emerged during the clinical trial, all previous data could become invalid.

 

Therefore, based on the limited information available at the time, he decided to design a chimeric antigen sequence that does not exist in nature.

 

"This decision was highly uncertain at the time," Lin said. He consulted independent experts, and the feedback was generally cautious, as a completely novel, self-designed antigen could face higher regulatory hurdles. But Lin ultimately stuck to his judgment.

 

The antigen design for RQ3013 was completed in December 2020, more than a year before the emergence and global pandemic of the Omicron variant.

 

The antigen uses the full-length S protein of the Alpha strain as its backbone, incorporating key escape mutation sites from the Beta strain, while modifying the furin cleavage site to stabilize the pre-fusion conformation. This chimeric antigen covers multiple hotspot positions in the two main neutralizing epitope regions, NTD and RBD. These positions have remained highly conserved in subsequent Omicron subvariants such as BA.1, BA.5, BQ.1.1, and XBB.1.5.

Although the actual circulating strains during the Phase III trial were BA.5.2 and BF.7, RQ3013 still demonstrated statistically significant protective efficacy.

 

This design approach was subsequently carried over to iterative products. In December 2023, RNAimmune's XBB.1.5 mRNA vaccine RQ3033 received emergency use authorization, becoming the first COVID-19 mRNA vaccine targeting XBB and other variants to be approved in China through immunogenicity bridging based on complete Phase III data.

 

RQ3033 was designed based on the full-length S protein of XBB.1.5, covering all key mutations of XBB.1.9 and most mutation sites of EG.5. Clinical data showed that 14 days after booster immunization, the neutralizing antibody GMT against XBB.1.9.1 was 622.73, a 35.21-fold increase from pre-boost levels; against EG.5, it was 673.48. Based on the platform data from RQ3013, RQ3033 was exempted from Phase I/II clinical trials and directly entered an immunogenicity bridging registrational clinical study, becoming the first vaccine product in China to be approved for EUA through this pathway.

 

RQ3013 completed the full-chain validation of a domestic mRNA platform, covering chimeric antigen design, nucleoside modification, and local clinical pathways. RQ3033, based on the same platform, achieved rapid iteration through immunogenicity bridging. On this basis, RQ3064, targeting the JN.1 variant, requires only an update of the antigen sequence, with all other technical elements remaining unchanged, and is being rapidly advanced under the platform technology evaluation framework. These three products, advancing sequentially, reflect the coordinated maturation of technology, regulation, and organization, marking the basic completion of the platform's construction.

 

TONACEA 02: Building an Autonomous Process Chain from Scratch

 

In biopharmaceuticals, molecular design is often only one-tenth of the work. The remaining nine-tenths is turning that molecule into a stable, safe, manufacturable, and regulatory-compliant drug product. This is especially true for mRNA drugs. From DNA template preparation to in vitro transcription, from capping and purification to LNP encapsulation and filling, every step along the way can become a breakpoint. The smoothness of the entire process chain ultimately determines whether the product remains in a paper or reaches the market.

 

"From a technological standpoint, an mRNA platform can be divided into chemical and biological components. Chemistry further breaks down into nucleotide chemistry and lipid chemistry. We've been accumulating our biological side. On the chemical side, the team focused early on nucleotide chemistry to solve problems related to mRNA synthesis, capping, modification, and short- and long-chain processes," Lin said.

 

In 2020, RNAimmune turned to lipid chemistry research, aiming to solve the delivery problem.

 

For delivery, LNP remains the mainstream approach. Lin categorizes mRNA delivery systems into five generations: 1.0 is validated standard LNP; 2.0 improves upon 1.0 in safety, stability, and accessibility; 3.0 achieves organ or cell-type targeting; 4.0 pursues immune silencing for long-term repeat dosing; 5.0 targets precise whole-tissue delivery and crossing physiological barriers. He believes the industry is currently at the 3.0 stage.

 

Building on previous research, RNAimmune's R&D team made multiple improvements to the LNP delivery system. For example, during scale-up production, they identified a previously unreported impurity with a large molecular weight that could not be effectively separated by conventional detection methods. After identifying the issue, the team modified the lipid molecule to improve stability and safety.

 

Additionally, the team made numerous subtle adjustments to the lipid molecules at the chemical structure level, such as adding or removing carbon atoms, changing substitution positions, introducing branching, adjusting pKa, and altering linker types. Cumulatively, these adjustments led to significant improvements in safety and efficacy.

 

In terms of formulation, RNAimmune did not adopt microfluidic technology. Lin's rationale was that microfluidic devices are primarily designed for small-scale research applications, with limited single-batch capacity, posing challenges for scale-up to commercial production. Therefore, he chose the most basic T-junction mixer, which, at sufficiently high flow rates, can also achieve good mixing results.

 

He had purchased T-junction mixers of different apertures from the market and tested them one by one, even customizing chips by opening his own molds. The final process he settled on was both simple and stable, producing LNPs with excellent uniformity under electron microscopy.

 

Lin also purchased a high-end imported formulation device, the first of its model to be imported into China. After research and comparison, he found that its core mixing principle was consistent with that of the standard T-junction mixer, further validating the feasibility of the team's self-developed process route.

 

Industrial scale-up presented another dimension of challenge. Increasing single-batch yield from one or two grams to tens of grams is a significant leap for mRNA. mRNA molecules are relatively unstable, and a single batch can cost millions of RMB. "It would often just disappear during processing," Lin said.

 

But RNAimmune ultimately succeeded in validating this process route, acquiring a complete process solution for mRNA synthesis to final drug product, along with proprietary equipment, achieving core technology accumulation.

 

Through repeated validation and optimization from upstream to downstream, the accumulation of domestic mRNA technology platforms has significantly advanced.

 

"Now the domestic industry chain has developed, and many CDMOs are performing excellently," Lin said. In the early days of the venture, there were no ready-made raw material solutions in China; many key raw materials were co-developed by RNAimmune and its partners. Today, most raw materials can be sourced from the domestic supply chain, eliminating the need for RNAimmune to produce them itself. The only exceptions are capping technology and cap analogs, which RNAimmune continues to develop and supply independently due to the need for continuous iteration. At this point, China's mRNA supply chain has completed its journey from nothing to something, from zero to one – the most valuable systemic legacy of that period for the industry.

 

TONACEA 03: Next Stops – Cancer Vaccines vs. In Vivo CAR-T

 

COVID-19 vaccines were just the starting point for mRNA technology. Beyond delivery systems, Lin also categorizes mRNA molecules themselves into five generations: Stage 1.0 is primarily used for vaccines. With continued molecular engineering, by the 5.0 era, mRNA could achieve long-term expression lasting over a month. Combined with iterative improvements in delivery systems, the ultimate goal is systemic, whole-tissue targeting.

 

"I hope that in the 5.0 era, mRNA can keep pace with AAV," Lin said. He believes the main applications of mRNA in the CGT field are personalized cancer vaccines and in vivo CAR-T.

 

The dose safety of mRNA cancer vaccines has already been validated, and multiple leading pipelines have shown progress. Lin predicts that we may see products approved for marketing within a few years.

 

At ASCO 2026, Merck and Moderna presented five-year follow-up data from a Phase IIb clinical study of their co-developed tumor neoantigen mRNA therapy, intismeran autogene, in patients with high-risk melanoma. The data showed that combination with Keytruda reduced the risk of recurrence or death by 49%, with a trend toward improved overall survival. This therapy has now entered multiple Phase II/III clinical trials, covering indications including melanoma, non-small cell lung cancer, bladder cancer, and renal cell carcinoma.

 

Intismeran autogene is an individualized mRNA-based neoantigen therapy encoding up to 34 neoantigens. Designed and manufactured based on the DNA mutation profile of a patient's tumor, it aims to activate a specific anti-tumor immune response.

 

Lin revealed that RNAimmune has also laid out a cancer vaccine pipeline and received ethics approval, but the process cost issue remains unresolved, with current production costs around RMB 1 million per dose. To address this, RNAimmune is working to reduce costs to below RMB 100,000 per dose.

 

A second important application direction is in vivo CAR-T. mRNA technology is highly suited to this scenario because in vivo CAR-T does not require long-term, sustained drug effect. mRNA expression is transient; CAR expression naturally decays after discontinuation. This "reversible" feature actually provides a safety advantage, effectively equipping the therapy with a built-in "brake."

 

According to Insight statistics, as of recently, over 20 in vivo CAR-T R&D projects using the mRNA approach (non-viral vector) have been disclosed globally, with most concentrated in the targeted lipid nanoparticle (tLNP) direction.

 

Lin points out that the real challenge for mRNA remains the delivery system – specifically, achieving systemic administration and precise targeting of T cells. "Current options carry significant safety risks. Simply taking the LNP used in traditional infectious disease vaccines and coupling it with a targeting antibody is a dangerous approach."

 

Therefore, RNAimmune has invested considerable time in re-screening its lipid library. The principle guiding the筛选 is: not to pursue the molecule with the highest expression efficiency, but to prioritize the molecule with the best safety window.

 

TONACEA 04: Navigating the Cycle

 

Around 2021, investment interest in mRNA in China reached its peak. Numerous startups completed large financing rounds during this period. But with the end of the pandemic, the fervor quickly subsided.

 

As of October 2025, venture capital investment in mRNA vaccine startups fell 82% compared to the same period in 2024, indicating a marked decline in investor enthusiasm. Revenues of leading companies such as BioNTech and Moderna dropped significantly, while many once-prominent mRNA companies in China either disbanded or fell into stagnation. RNAimmune is one of the few companies still steadily advancing its business.

 

In this wave of mRNA, very few companies have truly completed the entire closed loop from laboratory to industrialization and continued to operate to this day. Compared with the billions or even tens of billions raised during the industry's peak, RNAimmune was always more restrained in its fundraising, reflecting the shift of capital from frenzy back to rationality in the mRNA track.

 

In January 2026, Lin and his scientific team took full operational control of RNAimmune. The company remains a considerable distance from profitability – that is the objective reality. After taking over, Lin's first priority was to figure out how to keep the operation running.

 

His strategy is pragmatic and clear. The first step is to generate revenue. RNAimmune possesses a complete set of solutions covering delivery systems, plasmid templates, mRNA synthesis, purification, formulation, and instrumentation. By collaborating with other companies, it can generate revenue by providing technology, personnel, and equipment services. "Exporting China's mRNA platform to other countries is also an achievement."

 

Second, while undertaking some CDMO and technical service projects is not a transformation into a contract manufacturer, it allows external collaborations to support platform iteration, process validation, and team operations. Particularly in the production of CAR-T-related samples, with the technical foundation of an approved vaccine and compliant manufacturing facilities, the CMC quality and compliance of RNAimmune-produced samples are more persuasive, making it easier to gain the recognition of partners and research institutions.

 

To "cut costs," RNAimmune is also considering streamlining its pipeline. Infectious disease vaccines will primarily be handled through BD collaborations, allowing the company to focus on therapeutic products such as CAR-T and cancer vaccines.

 

Lin said that in recent years, RNAimmune has invested the most R&D effort in delivery solutions and has built a family of patents around this area. The patents cover not a single molecule, but an entire class of structures. After the data from the CAR-T project are read out, RNAimmune may consider splitting it off for licensing. "The overall goal is to keep the team and platform operating healthily, navigating the technology cycle, and waiting for the real application scenarios to mature."

 

Lin does not often talk about personal insights. When asked about his biggest takeaway from these years of entrepreneurship, he replied: "I don't have time for insights. I don't look at the future, and I don't look at the past. I just focus on doing the present well." This does not mean he fails to assess direction. Regarding the evolution of the technology platform, he remains full of imagination and anticipation. "The current generation of technology platforms can fully meet current market demands, but it is certainly not the best; there is always room for improvement."

 

RNAimmune's path shows that building a technology platform cannot bypass every specific problem in the industrialization process. There are no shortcuts to solving these problems, but with each one solved, the platform's foundation grows stronger. As for the next step, Lin believes the direction is already set; the remaining task is to do the work at hand well.

 

Reference:

• Formation and Evolution of China's mRNA Technology Platform: From COVID-19 Vaccines to the Next Generation of Drug Systems; Lin Jinzhong