“This was unexpected. It tells us that other cells are producing the vaccine antigen and handing it off to the immune system.”
That is what Dr. Brian D. Brown, lead author of the study “mRNA vaccine immunity is enhanced by hepatocyte detargeting and not dependent on dendritic cell expression”, published in April 2026 in Nature Biotechnology, said.
The unexpected finding: immune response following mRNA vaccination does not depend on whether dendritic cells directly express the mRNA sequence. Dendritic cells were the stated design target of the technology — formulated by its developers themselves.
What Sahin Described#
The mRNA-LNP technology used in BNT162b2 did not emerge from nowhere. Ugur Sahin describes in “Projekt Lightspeed” (Hoffmann und Campe, 2021) the purpose of the ionizable lipids BioNTech used for its vaccines. Dr. Sabine C. Stebel quotes page 158 verbatim as:
“The task of these new lipids was the same as the cancer vaccine: to transport the mRNA to the dendritic cells.”
— Joe Miller / Ugur Sahin / Özlem Türeci: Projekt Lightspeed, p. 158 (quote via Stebel/DrBine, not verified against physical copy)
Dendritic cells are professional antigen-presenting cells (pAPCs). They link innate and adaptive immune response — precisely what a vaccine is meant to trigger. That they were to be the primary target of the LNP-mRNA was not a marginal note. It was the stated mechanism behind the technology, formulated in the same design context as the cancer vaccine strategy.
Katalin Karikó describes in “Breaking Through” (Crown, 2023), in Stebel’s reading, the early DC transfection experiments with Drew Weissman as the foundation of LNP platform development — there too, DC targeting was the explicit objective.
What Nature Biotechnology 2026 Shows#
Marks et al. (2026) used synthetic microRNA targeting sequences (miRT) to selectively silence mRNA expression in defined cell types — without affecting other cells. The goal: to determine which cells actually need to express the mRNA for an immune response to occur. The tested models are murine cancer models and SARS-CoV-2 spike constructs — the same LNP system, the same distribution biology, the same cell types as Comirnaty.
The result in three points:
Dendritic cells: dispensable. When mRNA expression in pAPCs (professional antigen-presenting cells, which include dendritic cells) is completely silenced, antigen-specific T cell response remains intact. DC expression is not required for T cell priming.
Hepatocytes: suppressive. When liver cells express the mRNA, immune response is actively dampened — via the PD1/PDL1 signaling pathway. Completely silencing hepatocyte expression tripled T cell response in mouse models.
Myocytes (muscle cells): enhancing. mRNA expression in muscle cells produced equal or stronger immune responses — including against SARS-CoV-2. The mechanism: muscle cells produce the antigen and hand it over to dendritic cells via cross-presentation, which then present it to the immune system.
Brown describes the result in the EurekAlert press statement of April 29, 2026: “This was unexpected. It tells us that other cells are producing the vaccine antigen and handing it off to the immune system.”
The paper itself states: “These results were unexpected, as pAPCs are required for priming antigen-specific T cells and the ability of RNA-LNPs to transfect DCs and produce antigens for MHC-I presentation autonomously has been considered an important feature for RNA-LNP efficacy.”
The Gap#
The design target was DC transport. The actual mechanism is muscle cell production and cross-presentation. The Brown paper demonstrates that immune response remains the same even when DCs are never directly transfected.
This gap is documented — and it is not academically trivial. The EMA describes Comirnaty’s mechanism of action in its public documents as: “When a person is given the vaccine, some of their cells will read the mRNA instructions and temporarily produce the spike proteins.” No cell type specified. The regulatory mechanism description deliberately remained cell-type agnostic. What Sahin and Karikó formulated in their books as a design narrative — DC transport as the goal — thus occupies a different register: it was the scientific rationale behind the technology, not the regulatory mechanism description on which authorization rested. The Marks 2026 paper shows that the regulatory description was the biologically correct one.
Counterposition#
The strongest form of the counterargument: Sahin and Karikó described what their research goal was. Preclinical design targets and regulatory mechanism descriptions are different categories. The efficacy of the vaccines against COVID-19 was clinically established — exactly how the mechanism looked at the cellular level was an open question at the time of authorization.
This counterposition is correct as far as it goes. It does not explain why the DC-targeting narrative in public science communication — book publications, interviews, press releases — was presented as the mechanism of action, while EMA documentation remained cell-type agnostic. That was not forced simplification: Sahin described the transport mandate explicitly, not as one possibility among several.
Credit#
Dr. Sabine C. Stebel (DrBines verbales Vitriol) was the first to explicitly name the discrepancy between the stated DC-targeting goal (Sahin book, p. 158) and the findings of Marks et al. 2026, and to support it with book citations. This article goes directly to primary sources rather than through her newsletter — the discovery of the connection is hers.
Primary sources: Marks et al., Nature Biotechnology 2026 — EurekAlert Brown statement, April 29, 2026 — EMA Comirnaty mechanism — Joe Miller / Ugur Sahin / Özlem Türeci: “Projekt Lightspeed” (2021, quote via Stebel) — Katalin Karikó: “Breaking Through” (2023, paraphrase via Stebel)
