In Vivo mRNA-Lipid Nanoparticle CAR-T Cell Engineering: Advances, Challenges, and Clinical Translation
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsOverall the article is well written and generally well referenced and has value, but there is potential to improve it by being less of a cheer leader for LNP mRNA in vivo CAR and be more of a neutral reporter on strengths and weaknesses of LNP in vivo vs other options. The section on challenges and limitations does highlight relevant issues but more realism in the early parts of the review would produce a more valuable review.
the main weakness of the manuscript is that it does not adequately note limitations of cited studies. While this is understandable for many studies, the key questions, such as what fraction of LNP-CAR nanoparticles are taken up by T cells as compared to other cells in various studies is not noted. Rather it is noted that the study technology improved that ratio, but numbers are not given. Context and quantitative information is needed. If as a made up example, one NP delivers 99% of mRNA to phagocytes and 1% to T cells, and an improved NP delivers 90/10, that is a strong improvement, but the implications of expression of most CARs in cells other than T cells is not mentioned. Examples are lines 332 notes that inclusion of antiCD3 targeting “resulted in efficient in vivo transfection of T lymphocytes”: what is “efficient”? 1%?10%, 50%? And how much was taken up by other cells. This would be much more useful for readers than trying to cheer lead about how this is working already.
One comparison being made is between ex vivo DNA modified (permanent) CAR T cells and in vivo LNP mRNA mediated CAR T cells. The obvious different is the DNA is permanent, as would a lentiviral vector delivered in LNP in vivo be permanent in the cells that integrate it. What is not considered is ex vivo LNP mRNA modified CAR T cells, which would also solve the permanence issue, but would still have other mostly cost limitations. Perhaps the authors think this is worth discussing.
The currently labeled table 3 attempts to compare ex vivo DNA based CAR T with in vivo mRNA LNP based car. The comment on ex vivo DNA CAR T for solid tumors is “limited by toxicity and persistence” and the comparative for LNP is “potential advantage due to transient exposure”. The reality is that toxicity and problems due to persistence are due to off target effects but the real failure so far of ex vivo CAR T for solid tumors is failure to impact the solid tumors. So the point while noting a key difference misleads readers into thinking that if the off target effects were controlled that ex vivo CAR systems would be usable.
Comment on line 71-73 that LNP can be designed to preferentially target T cells and cites reference 19 (Yin et al PNAS). That paper is about in vivo LNP for CAR transfection of cells, but my understanding of the paper is that there is no data showing preferential transduction of T cells. If that is not correct I would be happy to be told which figure shows that preferential T cell transduction in vivo. This is an important point for this paper and should use a references (multiple references would be an improvement) that is very clear on demonstrating preferential CAR expression in T cells after in vivo LNP mRNA. When available the actual quantitative changes by the targeting should be included.
Lines 293-8 first sentence discusses circular RNA but that sentence is not referenced and should be. Next sentence discusses self amplifying RNA and has reference 62, but that reference title is about circular RNA. Each sentence needs appropriate reference.
Table 2 (first one) typo “Kay finding”
Each version of table 2 has some in vivo preclinical studies, suggest a table with all preclinical studies and another table with human clinical studies
Table 4 notes run together so the specific study associated with the notes is not clear.
Reference 41 is important reference but lacks page number
Author Response
Dear Editors and Reviewers,
We are pleased to submit the revised version of our manuscript for your consideration. We sincerely thank both reviewers for their thorough and constructive critiques, which have substantially improved the quality of this work. In response to Reviewer 1, we have replaced qualitative descriptors with quantitative T-cell transfection figures, added explicit discussion of off-target LNP uptake by hepatocytes and phagocytes, expanded our treatment of circular and self-amplifying RNA with appropriate references, corrected a typographical error in Table 2, and split the original Table 2 into separate preclinical and clinical study tables for improved clarity. We have also revised the notes in Table 4 to clearly link each claim to its supporting study, added a missing page number to Reference 41, and updated the citation for in vivo preferential T-cell targeting with more appropriate references.
In response to Reviewer 2, we have substantially expanded Section 3.3 on SORT technology, added comparative analyses and discussion of conflicting results throughout Section 5, and introduced a new Table 6 comparing in vivo mRNA–LNP CAR-T with viral in vivo delivery, CRISPR-based engineering, allogeneic CAR-T, CAR-NK, and CAR-macrophage approaches. We have also added a discussion of ex vivo mRNA-transfected CAR-T cells as a distinct intermediate modality.
We believe these revisions have fully addressed the reviewers’ concerns and have significantly strengthened the manuscript. A detailed point-by-point response is enclosed. We hope the revised manuscript is now suitable for publication.
Reviewer-1.
Comments-1: The main weakness of the manuscript is that it does not adequately note the limitations of cited studies. While this is understandable for many studies, the key questions, such as what fraction of LNP-CAR nanoparticles are taken up by T cells as compared to other cells in various studies, are not noted. Rather, it is noted that the study technology improved that ratio, but numbers are not given. Context and quantitative information are needed. If, as a made-up example, one NP delivers 99% of mRNA to phagocytes and 1% to T cells, and an improved NP delivers 90/10, that is a strong improvement, but the implications of expression of most CARs in cells other than T cells are not mentioned. Examples are lines 332 notes that inclusion of anti-CD3 targeting “resulted in efficient in vivo transfection of T lymphocytes”: what is “efficient”? 1%?10%, 50%? And how much was taken up by other cells. This would be much more useful for readers than trying to cheerlead on how this is working already.
Response: We are grateful to the reviewer for this important and well-taken criticism. We have revised the section to (i) replace qualitative descriptors such as "efficient" with the quantitative T-cell transfection figures actually reported in the cited studies, (ii) state explicitly that even optimized targeted LNPs deliver only a minority of the administered dose to T cells, with the remainder taken up by hepatocytes and phagocytes, and (iii) add a new paragraph discussing the methodological heterogeneity of reported uptake metrics, the largely uncharacterized biological consequences of CAR expression in off-target cells, and the need for standardized quantitative reporting. We agree that providing this context offers more value to readers than uncritical endorsement of the platform's progress.
Comments-2: One comparison being made is between ex vivo DNA-modified (permanent) CAR T cells and in vivo LNP mRNA-mediated CAR T cells. The obvious difference is that the DNA is permanent, as would a lentiviral vector delivered in LNP in vivo be permanent in the cells that integrate it. What is not considered is ex vivo LNP mRNA modified CAR T cells, which would also solve the permanence issue, but would still have other mostly cost limitations. Perhaps the authors think this is worth discussing.
Response: We thank the reviewer for this insightful observation. We have added a brief comparison clarifying that ex vivo mRNA-transfected CAR-T cells represent a distinct intermediate modality, sharing the transient, non-integrating safety profile of in-vivo mRNA–LNP CAR-T while retaining the manufacturing-related cost and logistical limitations of conventional autologous platforms. This addition makes explicit the three-way comparison between permanent (DNA/lentiviral) modification, ex-vivo mRNA modification, and in vivo mRNA–LNP modification.
Comments-3: The currently labeled Table 3 attempts to compare ex vivo DNA-based CAR T with vivo mRNA LNP-based CAR. The comment on ex vivo DNA CAR T for solid tumors is “limited by toxicity and persistence,” and the comparative for LNP is “potential advantage due to transient exposure”. The reality is that toxicity and problems due to persistence are due to off-target effects, but the real failure so far of ex vivo CAR T for solid tumors is the failure to impact the solid tumors. So, the point while noting a key difference misleads readers into thinking that if the off-target effects were controlled, ex vivo CAR systems would be usable.
Response: We thank the reviewer for this helpful observation. Table 4 has been revised so that each claim is now placed in a discrete cell directly linked to its supporting study. A clarifying caption has also been added. We believe these changes make the specific study underlying each statement immediately identifiable.
Comments-4: Comment on lines 71-73 that LNP can be designed to preferentially target T cells and cites reference 19 (Yin et al PNAS). That paper is about in vivo LNP for CAR transfection of cells, but my understanding of the paper is that no data shows preferential transduction of T cells. If that is not correct, I would be happy to be told which figure shows that preferential T cell transduction in vivo. This is an important point for this paper, and it should use a reference (multiple references would be an improvement) that is very clear on demonstrating preferential CAR expression in T cells after in vivo LNP mRNA. When available, the actual quantitative changes to the targeting should be included.
Response: Thank you for this important comment. We agree that the previously cited reference did not sufficiently support the statement regarding preferential T-cell targeting in vivo. Accordingly, we revised the text and added multiple appropriate references that more clearly demonstrate preferential CAR expression/transfection in T cells following in vivo LNP-mediated mRNA delivery.
Comments-5: Lines 293-8 first sentence discusses circular RNA, but that sentence is not referenced and should be. The next sentence discusses self-amplifying RNA and has reference 62, but that reference title is about circular RNA. Each sentence needs an appropriate reference.
Response: We thank the reviewer for this comment. We have revised the text in Lines 293–298 and added the appropriate references for each statement.
Comments-6: Table 2 (first one) typo “Kay finding.”
Response: Addressed
Comments-7: Each version of Table 2 has some in vivo preclinical studies, suggesting a table with all preclinical studies and another table with human clinical studies
Response: We thank the reviewer for this excellent suggestion. We agree that separating preclinical and clinical studies would provide much better clarity and organization. We have now split the original Table 2 into two distinct tables:
- Table 2: Preclinical Studies of mRNA-LNP In-Vivo CAR-T Cell Generation
- Table 3: Clinical Studies of In Vivo CAR-T Using LNPs
Comments-8: Table 4 notes run together, so the specific study associated with the notes is not clear.
Response: Thank you for this observation. The notes in Table 4 have been revised and reformatted to clearly distinguish the specific study associated with each note, thereby improving clarity and readability.
Comments-9: Reference 41 is important, but lacks a page number
Response: Thank you for highlighting this issue. The page number for Reference 41 has now been added, and the citation has been updated accordingly.
Reviewer-2.
Comments-1: Section 3.3 on SORT technology is very brief compared to active targeting (Section 3.2). Given the importance of SORT for achieving spleen/lymphoid tropism, this section deserves a more detailed discussion of the underlying principles and key studies.
Response: We thank the reviewer for this helpful comment. Section 3.3 has been substantially expanded to provide a more detailed discussion of the underlying principles of SORT, recent mechanistic insights into protein corona-mediated targeting, and key studies applying anionic SORT to splenic T-cell and other immune cell engineering, together with a brief discussion of translational advantages and current limitations.
Comments-2: Section 5 summarizes studies but lacks a comparative evaluation and discussion of conflicting results.
Response: We thank the reviewer for highlighting this gap. To address it, we have added concise comparative analyses at the end of each subsection in Section 5, contrasting key platforms (anti-CD3 vs anti-CD5 targeting), discussing inconsistencies in CAR-T persistence and transfection efficiency between murine, humanized, and NHP models, and noting where conflicting or preliminary data warrant cautious interpretation. We believe these additions strengthen the critical balance of the section while preserving its overall structure.
Comments-3: The manuscript does not sufficiently compare with viral in vivo CAR delivery, CRISPR-based in vivo engineering, allogeneic/off-the-shelf CAR-T, CAR-NK, or CAR-macrophage therapies.
Response: We thank the reviewer for this important point. To address it, we have added a new comparative Table 6 contrasting in-vivo mRNA–LNP CAR-T with viral in vivo CAR delivery, in vivo CRISPR engineering, allogeneic/off-the-shelf CAR-T, CAR-NK, and CAR-macrophage therapies across engineering mechanisms, advantages, limitations, and clinical maturity. This contextualizes the mRNA–LNP approach within the broader landscape of next-generation engineered cell therapies and clarifies its distinct positioning.
Reviewer 2 Report
Comments and Suggestions for AuthorsThis manuscript presents a comprehensive review of the in vivo generation of CAR-T cells using mRNA-loaded lipid nanoparticles (LNPs). The topic is highly timely and clinically relevant, given the limitations of conventional ex vivo CAR-T manufacturing and the rapid progress in LNP delivery platforms. The authors cover a broad scope, from mRNA biology and ionizable lipid chemistry through targeting strategies, preclinical evidence, clinical translation, and future directions. The review is well-structured and scientifically grounded. However, the manuscript requires revision to address some issues, including;
1) Section 3.3 on SORT technology is very brief compared to active targeting (Section 3.2). Given the importance of SORT for achieving spleen/lymphoid tropism, this section deserves a more detailed discussion of the underlying principles and key studies.
2) Section 5 summarizes studies but lacks a comparative evaluation and discussion of conflicting results.
3) The manuscript does not sufficiently compare with viral in vivo CAR delivery, CRISPR-based in vivo engineering, allogeneic/off-the-shelf CAR-T, CAR-NK, or CAR-macrophage therapies.
Author Response
Dear Editors and Reviewers, We are pleased to submit the revised version of our manuscript for your consideration. We sincerely thank both reviewers for their thorough and constructive critiques, which have substantially improved the quality of this work.
In response to Reviewer 2, we have substantially expanded Section 3.3 on SORT technology, added comparative analyses and discussion of conflicting results throughout Section 5, and introduced a new Table 6 comparing in vivo mRNA–LNP CAR-T with viral in vivo delivery, CRISPR-based engineering, allogeneic CAR-T, CAR-NK, and CAR-macrophage approaches. We have also added a discussion of ex vivo mRNA-transfected CAR-T cells as a distinct intermediate modality.
We believe these revisions have fully addressed the reviewers’ concerns and have significantly strengthened the manuscript. A detailed point-by-point response is enclosed. We hope the revised manuscript is now suitable for publication.
Reviewer-2.
Comments-1: Section 3.3 on SORT technology is very brief compared to active targeting (Section 3.2). Given the importance of SORT for achieving spleen/lymphoid tropism, this section deserves a more detailed discussion of the underlying principles and key studies.
Response: We thank the reviewer for this helpful comment. Section 3.3 has been substantially expanded to provide a more detailed discussion of the underlying principles of SORT, recent mechanistic insights into protein corona-mediated targeting, and key studies applying anionic SORT to splenic T-cell and other immune cell engineering, together with a brief discussion of translational advantages and current limitations.
Comments-2: Section 5 summarizes studies but lacks a comparative evaluation and discussion of conflicting results.
Response: We thank the reviewer for highlighting this gap. To address it, we have added concise comparative analyses at the end of each subsection in Section 5, contrasting key platforms (anti-CD3 vs anti-CD5 targeting), discussing inconsistencies in CAR-T persistence and transfection efficiency between murine, humanized, and NHP models, and noting where conflicting or preliminary data warrant cautious interpretation. We believe these additions strengthen the critical balance of the section while preserving its overall structure.
Comments-3: The manuscript does not sufficiently compare with viral in vivo CAR delivery, CRISPR-based in vivo engineering, allogeneic/off-the-shelf CAR-T, CAR-NK, or CAR-macrophage therapies.
Response: We thank the reviewer for this important point. To address it, we have added a new comparative Table 6 contrasting in-vivo mRNA–LNP CAR-T with viral in vivo CAR delivery, in vivo CRISPR engineering, allogeneic/off-the-shelf CAR-T, CAR-NK, and CAR-macrophage therapies across engineering mechanisms, advantages, limitations, and clinical maturity. This contextualizes the mRNA–LNP approach within the broader landscape of next-generation engineered cell therapies and clarifies its distinct positioning.
