Functional chimeric mRNAs encode proteins in mammalian immunity

https://doi.org/10.1038/s41586-026-10982-x

Olivia Venezia, Harry Kane, Gang Du, Hannah D. Coughlan, Timothy M. Johanson, Hongshen Wang, Patricia Tilstam, Samuel W. Kazer, Georgia Gunner, Daisy A. Hoagland, Megha G. Basavappa, Kanchana Ravichandran, Eren Ada, Ademi Zhakyp, Sandhya Kumar, Coco Duizer, Owen Searle, Chris Geo Provido, Leonel Joannas, Shilong Yang, Liam B. Healy, Qiankun Wang, Joia Capocchi, Bahawar Sharif Dhillon, Sarah Slavoff, Liang Shan, Jorge Henao-Mejia, Ruth A. Franklin, Isaac M. Chiu, Jose Ordovas-Montanes, Kate L. Jeffrey, Judy Lieberman, Rhys S. Allan, Jonathan C. Kagan, Hao Wu & Ruaidhri Jackson

Department of Immunology, Harvard Medical School, Boston, MA, USA

Individual mammalian mRNAs and proteins are generally thought to originate from single genomic loci, with isoform diversity arising through cis-splicing of pre-mRNA. Whether mRNAs from distant genes can undergo trans-splicing to produce functionally relevant chimeric transcripts has remained unclear. Here, we developed a pipeline combining long-read direct RNA sequencing with non-targeted and targeted validation to identify chimeric transcripts in macrophages. Chromatin conformation capture studies revealed that inflammation induces interchromosomal DNA interactions, positioning parent genes in close proximity to facilitate chimeric mRNA formation. Notably, we identified a protein-coding chimeric mRNA representing a fusion between the pore-forming protein gasdermin D (GSDMD) and a C-terminal domain translated out of frame from Tmem106a (Gsdmd-Tmem106a) in mice. We show that inflammasome priming upregulates Gsdmd-Tmem106a, with the protein localizing to the plasma membrane. Upon inflammasome activation, GSDMD-TMEM106A directly interacts with canonical GSDMD N termini to accelerate and enhance pore formation and IL-1beta release. Finally, we demonstrate that GSDMD-TMEM106A balances host defense and immunopathology in vivo: its loss protects against lethal sepsis but compromises antibacterial defense, whereas overexpression enhances host protection while increasing sepsis lethality. We establish that protein-coding chimeric mRNAs generated by regulated transcript fusion events are functional during inflammation and immunity.

https://material-image.wanwang.xin/1353944556863376/public/7d68661e-52d5-4e14-9c98-de2ebc9fc5c4.pdf

Inflammation Drives chRNA Expression

The diversity of proteins produced by cells is typically defined by the repertoire of individual genes encoded in the genome. This diversity can be amplified through pre-mRNA cis-splicing, where varied pairings of exons within a given transcript are differentially fused. It remains unclear whether mechanisms beyond cis-splicing exist to diversify the protein-coding capacity of mammalian cells. In unicellular and invertebrate organisms, trans-splicing has been reported—a process where exons from distinct pre-mRNAs are fused to create hybrid proteins. In trypanosomes and nematodes, a common pre-mRNA is trans-spliced to various other mRNAs to promote transcript stability and translation; by contrast, in Drosophila, this process generates functionally diverse transcripts and proteins. In healthy mammals, few examples of trans-splicing exist, and functional chimeric mRNAs (chRNAs) are typically associated with oncogenic transformation of cells, where genomic translocations fuse disparate genes at the DNA level.

Despite advances and widespread adoption of RNA sequencing (RNA-seq), endogenously expressed chRNAs in mammals have largely remained undetected due to technical limitations. cDNA synthesis methods rely on viral reverse transcriptase enzymes to convert RNA into cDNA; however, these enzymes can create artificial fusion transcripts via template switching. Additionally, because mammalian chRNAs are not annotated in reference transcriptomes, candidate chRNAs often map to multiple genomic regions during alignment, leading to their routine exclusion from standard RNA-seq analyses.

To identify chRNAs while avoiding artifacts from cDNA synthesis, the team performed Oxford Nanopore PromethION direct RNA-seq on polyadenylated RNA from steady-state, tissue-repairing, and inflammatory mouse bone marrow-derived macrophages (BMDMs). Ten biological replicates yielded 52.9 million passing reads, with over 90% mapping to the mouse genome.

The computational tools LongGF, JAFFAL, and Genion were benchmarked to detect BCR-ABL1 in K562 cells and used to identify chRNAs containing annotated splice donor and acceptor sites from two distinct genes. This analysis identified 30,390 candidates across activation states. These chRNAs had a median length of 1,681 nucleotides, compared with 2,190 nucleotides for their parent transcripts. Parent genes were distributed throughout the genome, with approximately 88% of chRNA species arising from interchromosomal pairings and about 4.5% and 7.5% originating from proximal and distal intrachromosomal loci, respectively. These findings suggest contributions from both readthrough transcription via cis-splicing and trans-splicing involving distal loci.

Limited sequencing depth and modest agreement among long-read detection tools precluded robust quantification, prompting orthogonal validation. The team generated high-depth Illumina RNA-seq data and analyzed them using six fusion-detection algorithms, yielding an implausibly large candidate set consistent with false positives arising from reverse transcription and PCR template switching. Nevertheless, comparison with the direct RNA catalog identified over 250 chRNAs with matching junctions, including candidates associated with differential macrophage polarization.

Since short reads rarely span sufficient sequence on both sides of a chRNA junction, this approach may yield false negatives for specific isoforms. To address this, the team designed tandem NanoString nCounter probes that require simultaneous hybridization across each RNA junction. More than 500 probe sets were created to target chRNAs detected by all three long-read tools, along with scrambled controls expected to produce false negatives. Over 100 high-quality probes identified constitutive or polarization-regulated chRNAs.

PCR and Sanger sequencing independently validated chRNA junctions, including a Gsdmd-Tmem106a amplicon spanning Gsdmd exons 1-2 fused to Tmem106a exons 6-9. Gsdmd-Tmem106a was among the top inflammation-induced candidates identified by NanoString, and both parent genes are known to play established roles in lipopolysaccharide (LPS)-induced inflammation. The same exon-exon junction was detected in C57BL/6J, BALB/cJ and wild-derived mice. Quantitative RT-PCR analysis further confirmed LPS-inducible expression, which peaked at 6 hours alongside the parent genes. In vivo, Gsdmd-Tmem106a was detected and induced in lung macrophages following influenza A infection, brain macrophages/microglia during Escherichia coli meningitis, and peritoneal macrophages after LPS administration.

Applying this pipeline to steady-state and inflammatory human monocyte-derived macrophages, the team identified over 900 chRNAs, including inter- and intrachromosomal species with properties similar to those of mouse chRNAs. qPCR further confirmed their detection and differential regulation by LPS.

Despite the lower sequencing depth in the human dataset, cross-species analysis identified 33 chRNAs sharing parent genes between mice and humans. Although the team did not identify a GSDMD-TMEM106A chRNA in these cells, approximately 8 displayed at least 50% mRNA sequence homology, and 28 encoded predicted proteins with at least 50% conservation, featuring closely matched domain architectures. These included HDAC8-CITED1 and TBC1D22B-RNF8,, which retained functional domains from both parent genes and were validated by Sanger sequencing in mouse and human macrophages.

Therefore, integrating direct RNA-seq with short-read RNA-seq, NanoString, PCR, and expression analyses provides a generalizable framework for identifying, validating, and prioritizing endogenous chRNAs for functional studies in immunity and inflammation.

https://material-image.wanwang.xin/1353944556863376/public/b1e5a8d9-186c-4ab9-a1e4-7fb3acd4bb82.pdf

Inflammation Regulates chRNA Gene Contacts

To differentiate chromosomal translocation from RNA-level fusion, the team conducted long-range genomic PCR on DNA from untreated and LPS-stimulated BMDMs using primers targeting junction exons, including Gsdmd exon 2 and Tmem106a exon 6. Neither Gsdmd-Tmem106a nor Cd274-Lacc1 yielded the expected kilobase-scale amplicon under either condition. This finding argues against detectable chromosomal translocations and supports an RNA-level fusion event mediated by trans-splicing.

Since the chRNA catalogue was defined using annotated splice sites, the team tested whether canonical splicing machinery contributes to chRNA formation. Inhibiting the SF3B spliceosome complex with pladienolide B or knocking down U1 snRNA abolished LPS-induced Gsdmd-Tmem106a expression. Similarly, inhibiting RNA polymerase II with actinomycin D eliminated Gsdmd-Tmem106a and Cd274-Lacc1, as well as their parent transcripts, supporting co-transcriptional joining of newly synthesized RNAs.

The team next investigated whether chRNA formation is driven by spatial proximity between parent genes. Genome-wide in situ Hi-C analysis showed that interchromosomal parent-gene pairs detected by direct RNA-seq exhibited significantly more DNA-DNA interactions than randomized exon pairs. Since CTCF regulates genome organization, the team depleted Ctcf using small interfering RNA (siRNA) and performed Hi-C after 6 hours of LPS stimulation. While LPS increased contacts between chRNA parent genes, Ctcf depletion eliminated this effect, demonstrating that these interactions are dynamically regulated and depend on CTCF.

Functionally, Ctcf depletion reduced LPS-induced Gsdmd-Tmem106a without affecting Gsdmd or Tmem106a expression, and similarly reduced Cd274-Lacc1 without altering either parent gene. Finally, junction-specific primers detected only the Gsdmd exon 2 and Tmem106a exon 6 fusion isoform, demonstrating that the fusion is highly selective rather than a random consequence of proximity. Together, these findings show that inflammation promotes CTCF-dependent interactions between chRNA parent genes, enabling specific trans-splicing and chRNA formation in macrophages.

https://material-image.wanwang.xin/1353944556863376/public/c48a67d2-649d-4709-893f-67a74d3e1e67.pdf

Gsdmd-Tmem106a Encodes a Fusion Protein

To test the protein-coding potential of chRNA, the team focused on Gsdmd-Tmem106a, which retains the Gsdmd 5' untranslated region and canonical start codon. The Tmem106a-derived sequence was predicted to be translated out-of-frame, producing a novel C-terminus absent from the annotated mouse proteome.

The team generated Gsdmd-Tmem106a-MYC knock-in mice (G-TMYC mice) with a C-terminal MYC tag inserted into the chimeric open reading frame within exon 6 of Tmem106a, upstream of the predicted chimeric stop codon. Since the Tmem106a-derived sequence is translated out-of-frame, the tag is specific to GSDMD-TMEM106A. MYC immunoblotting revealed an LPS-inducible protein in G-TMYC bone marrow-derived macrophages (BMDMs), but not in wild-type BMDMs. A junction-specific siRNA selectively depleted Gsdmd-Tmem106a, while an independent siRNA targeting Gsdmd exon 2 (encoded on chromosome 15) was used to determine if this Gsdmd exon is required for producing the chromosome-11-encoded MYC tag. Both treatments reduced the MYC signal, confirming its chimeric origin.

A custom antibody targeting the novel C-terminal peptide also detected an LPS-inducible protein reduced by both siRNAs. The endogenous GSDMD-TMEM106A migrated above its predicted molecular weight, while overexpressed GSDMD-TMEM106A with a C-terminal HA tag yielded both the expected ~15 kDa protein and higher-molecular-weight species, indicating multiple modified states.

Finally, the team generated Gsdmd-Tmem106aStop (G-TStop) mice by introducing a single G-to-T substitution into Tmem106a exon 6. This introduced a stop codon after the third Tmem106a-derived amino acid without affecting canonical GSDMD or TMEM106A translation. GSDMD-TMEM106A was absent in G-TStop BMDMs but robustly induced by LPS in wild-type cells, confirming that Gsdmd-Tmem106a encodes a stable, LPS-inducible protein.

https://material-image.wanwang.xin/1353944556863376/public/5e4e4448-5a30-4163-9e8e-c99739ed54f8.pdf

GSDMD-TMEM106A Drives IL-1beta Immunity

To define GSDMD-TMEM106A function, the team assessed its role in inflammasome activity. Canonical GSDMD is cleaved by inflammatory caspases, releasing GSDMD-NT to oligomerize and form membrane pores that enable secretion of caspase-1 (CASP-1)-processed IL-1beta and IL-18. Following siRNA-mediated knockdown of Gsdmd-Tmem106a, LPS- and nigericin-induced IL-1beta and IL-18 release were significantly reduced. LDH release after prolonged nigericin treatment was also reduced, though less markedly than early cytokine secretion, while non-inflammasome-dependent IL-6 secretion remained unaffected.

BMDMs from G-TStop mice similarly showed reduced early IL-1beta release but normal IL-6 secretion, with a modest decrease in late-stage cell lysis. Thus, both acute depletion and genetic loss of GSDMD-TMEM106A selectively impair inflammasome-dependent cytokine release and pyroptosis.

Since IL-1beta promotes antibacterial immunity and GSDMD protects against Salmonella enterica subspecies enterica serovar Typhimurium (hereafter, S. Typhimurium), the team challenged G-TStop and wild-type mice with S. Typhimurium. G-TStop mice showed significantly higher bacterial burdens in the spleen and liver compared to their co-housed littermates. Conversely, in lethal LPS-induced sepsis—where GSDMD worsens disease severity—G-TStop mice were protected from early hypothermia; while most G-TStop mice survived for 96 hours, all wild-type controls died within 48 hours.

Gain-of-function experiments yielded reciprocal effects. Lentiviral overexpression of GSDMD-TMEM106A increased IL-1beta release following NLRP3 activation. Similarly, transfection with Gsdmd-Tmem106a mRNA enhanced IL-1beta secretion and accelerated LDH release, whereas mRNA encoding only the GSDMD-derived region (Gsdmd-Tmem106AdeltaCT, or G-TdCT-encoding mRNA) did not, demonstrating that the TMEM106A-derived sequence is required.

The team next administered LNP-encapsulated Gsdmd-Tmem106a, Gsdmd-Tmem106aΔCT, or non-translating control mRNA 18 hours before S. Typhimurium challenge. Gsdmd-Tmem106a administration significantly reduced liver bacterial burden at 24 hours compared with both controls, demonstrating that exogenous GSDMD-TMEM106A enhances antibacterial defense. However, in LPS-induced sepsis, Gsdmd-Tmem106a administration accelerated hypothermia within 3 hours compared with Gsdmd-Tmem106aΔCT and control mRNAs. Since inflammasome cytokines and GSDMD activity promote sepsis lethality, the team measured serum cytokines 6 hours after LPS exposure. Gsdmd-Tmem106a specifically increased IL-1beta without altering IL-6, IFN-γ, TNF, CCL2, or IL-12p70. Nearly all mice treated with Gsdmd-Tmem106a reached humane endpoints within 72 hours, whereas those treated with Gsdmd-Tmem106aΔCT or control mRNA were protected.

Finally, because Gsdmd-/- mice resist lethal LPS sepsis, the team tested whether GSDMD-TMEM106A requires endogenous GSDMD. GSDMD-TMEM106A-induced sepsis lethality was completely abolished in Gsdmd-/- mice, and GSDMD-TMEM106A overexpression did not induce IL-1beta release in Gsdmd-/- immortalized BMDMs (iBMDMs). Thus, GSDMD-TMEM106A functions through canonical GSDMD to amplify IL-1beta release, enhancing antibacterial defense while exacerbating sepsis immunopathology.

https://material-image.wanwang.xin/1353944556863376/public/0286b8ca-e247-4654-a858-eb67ccbbda80.pdf

GSDMD-TMEM106A Enhances Pore Formation

To position GSDMD-TMEM106A within the inflammasome pathway, we first examined priming and proteolytic activation. Neither siRNA-mediated depletion nor genetic loss in G-TStop macrophages altered LPS-induced Il1b, Il6, Nlrp3, or Gsdmd expression. Loss of Gsdmd-Tmem106a also did not affect GSDMD or CASP-1 cleavage following LPS and nigericin treatment, indicating its activity occurs downstream of inflammasome assembly and GSDMD cleavage.

Because GSDMD-TMEM106A requires canonical GSDMD, the team tested whether these proteins function cooperatively. In Gsdmd-/- iBMDMs expressing a fixed amount of GSDMD, increasing the dose of GSDMD-TMEM106A-encoding mRNA enhanced IL-1beta release in a dose-dependent manner. Although Gsdmd-Tmem106a accounts for less than 10% of parental Gsdmd expression, supplementing GSDMD-encoding mRNA with 10% GSDMD-TMEM106A-encoding mRNA nearly doubled IL-1beta release, whereas an equivalent increase in GSDMD had little effect. GSDMD-TMEM106A-encoding mRNA did not enhance secretion, demonstrating that the TMEM106A-derived C-terminus is required for synergy with GSDMD-NT.

Since GSDMD-TMEM106A lacks the inhibitory GSDMD C-terminus, the team investigated whether it associates with cellular membranes prior to inflammasome activation. Recombinant FLAG-tagged GSDMD-TMEM106A showed modest binding to inner-leaflet lipids—phosphatidic acid and phosphatidylserine—as well as the mitochondrial lipid cardiolipin. Epitope-tagged GSDMD-TMEM106A localized to the plasma membrane before inflammasome activation and accumulated at membrane protrusions during pyroptosis, a stage characterized by cell rounding and swelling. Subcellular fractionation further revealed that endogenous GSDMD-TMEM106A translocates to membranes during LPS priming.

Mutation of the GSDMD-derived membrane-anchor residues Phe50 and Trp51 in GSDMD-TMEM106A-encoding mRNA resulted in cytoplasmic retention of the GSDMD-TMEM106AF50G/W51G-HA protein. In contrast, the non-functional GSDMD-TMEM106AdeltaCT-HA protein remained localized to membranes, demonstrating that membrane targeting depends exclusively on the GSDMD-derived region. Only full-length, membrane-localized GSDMD-TMEM106A enhanced IL-1beta release and cell lysis, indicating that membrane localization is required for activity.

To test whether membrane-localized GSDMD-TMEM106A interacts with GSDMD-NT, the team performed co-immunoprecipitation following NLRP3 activation. Endogenous GSDMD-NT co-precipitated with GSDMD-TMEM106A-HA but not with cytoplasmic GSDMD-TMEM106AF50G/W51G-HA or the non-translating control, indicating a physical interaction restricted to the plasma membrane.

AlphaFold 3 predicted co-folding of GSDMD-TMEM106A with the GSDMD-NT pore, positioning its TMEM106A-derived region within the pore. This region contains four residues conserved in GSDMD-NT acidic patches that are required for IL-1beta release. Substitution of these residues (E74A, D89A, and D92A) altered the predicted co-folding and abolished GSDMD-TMEM106A-mediated enhancement of IL-1beta release, establishing their functional importance.

Endogenous GSDMD-TMEM106A loss slowed GSDMD-NT pore formation, as measured by propidium iodide (PI) uptake, while overexpression accelerated it. In cells with doxycycline-inducible GSDMD-NT, expression of Gsdmd-Tmem106a—but not Gsdmd-Tmem106adeltaCT or control mRNA—increased the rate of PI uptake, further demonstrating activity downstream of inflammasome activation. Finally, in cell-free liposome permeabilization assays using purified proteins, recombinant GSDMD-TMEM106A-FLAG, but not recombinant GSDMD-TMEM106AdeltaCT-FLAG, directly enhanced GSDMD-NT-mediated pore formation.

Together, these findings establish GSDMD-TMEM106A as a membrane-localized cofactor that physically and functionally cooperates with GSDMD-NT to accelerate pore formation, amplify early cytokine release, and enhance subsequent pyroptosis.

https://material-image.wanwang.xin/1353944556863376/public/77ce1063-6cd3-4b1e-ae49-f41e96fa2619.pdf

Discussion

Mammalian cells harbor substantial functional genomic dark matter. Over the past several decades, as technology advanced, the strict criteria defining eukaryotic protein-coding genes have gradually expanded, enabling the discovery of novel protein-coding entities. Non-AUG start codons driving translation initiation, evidence of bicistronic translation, and small ORF-encoded microproteins have all been identified in mammalian cells across diverse processes ranging from immunity to muscle function. Furthermore, beyond classical linear exon splicing, backsplicing events generating circular RNAs with coding potential have been discovered. Here, we demonstrate that distinct genes located far apart in the genome can produce chRNAs during normal physiological responses.

A key finding of this study is that chRNA biogenesis is a rapid, stimulus-responsive process, not a byproduct of genomic instability or splicing noise. Pro-inflammatory signals strongly induce chRNAs in macrophages; these form at annotated exon boundaries and require the canonical spliceosome. For some chRNAs, parent genes are brought together via interchromosomal DNA contacts that increase within 6 hours of inflammation and depend on the genome-organizing protein CTCF. chRNA formation is highly precise: for example, Gsdmd and Tmem106a consistently join at Gsdmd exon 2 and Tmem106a exon 6. The mechanisms driving this specificity remain unclear but may involve repetitive elements similar to those in circular RNA biogenesis, specialized trans-splicing, RNA-binding factors, nuclear condensates, or chromatin and RNA modifications. Notably, tumor-associated fusion transcripts like JAZF1-JJAZ1 and PAX3-FOXO1 can also arise through physiological trans-splicing without chromosomal rearrangement. Combined with the proposed RNA-before-DNA model and evidence that 3D genome organization influences translocation partner selection, these findings suggest that chRNAs mark loci that frequently interact in nuclear space and may become prone to translocation during genotoxic DNA repair.

The findings support a model in which LPS-induced GSDMD-TMEM106A localizes to the plasma membrane prior to inflammasome-mediated cleavage of GSDMD, creating a pre-positioned docking site that accelerates recruitment and assembly of released GSDMD-NT. Consistent with this model, GSDMD-TMEM106A physically interacts with cleaved GSDMD-NT at the membrane, and AlphaFold 3 predicts that its TMEM106A-derived region cooperatively folds with the membrane-inserting region of the GSDMD-NT oligomer. Since GSDMD pores are thought to assemble from membrane-associated monomers, this interaction may stabilize only after GSDMD-NT oligomerization, potentially explaining its absence in the cytoplasmic GSDMD-TMEM106AF50G/W51G-HA mutant. Analogous to Drosophila Myd88 and mammalian PIP2-binding adaptors that concentrate immune effectors at specific membrane domains, binding of GSDMD-TMEM106A to phosphatidic acid and phosphatidylserine may enhance the spatial precision and kinetics of pore formation. Accordingly, its loss attenuates but does not abolish cytokine release and pyroptosis, indicating that it amplifies rather than enables canonical pore formation. Its additional binding to cardiolipin suggests a potential role in GSDMD-mediated mitochondrial damage.

Notably, endogenous GSDMD-TMEM106A migrates primarily above its predicted molecular mass, while HA-tagged overexpression yields both higher- and expected-molecular-mass species, indicating distinct modified, oligomeric, and monomeric pools. Since the retained GSDMD region contains known sites for phosphorylation, oxidation, succination, and ubiquitination, characterizing these biochemical states and their effects on GSDMD-TMEM106A activity represents a key future direction.

The study uncovers a hidden layer of mammalian gene regulation where spatial genome organization and regulated splicing link distant genes to produce protein-coding chRNAs. Their conservation in mouse and human macrophages suggests that transcript fusion is an underappreciated mechanism for expanding transcriptomic and proteomic diversity. By encoding proteins with novel domain architectures and functions, chRNAs may also broaden the set of pharmacologically targetable proteins. However, their physiological prevalence—including potential fusions involving multiple parent genes and cell type- or context-specific networks—remains largely unclear. This study provides a generalizable framework for discovering, validating, and manipulating chRNAs, demonstrating that chRNA-specific interference, mRNA delivery, and gene editing can modulate their activity in cells and in vivo. Together, these findings establish chRNAs as a therapeutically actionable layer of mammalian gene regulation that expands the functional proteome and opens new avenues for biological discovery and therapeutic intervention.

https://material-image.wanwang.xin/1353944556863376/public/815019ae-348a-4c13-b715-403c20e1a314.pdf

Research Team and Affiliations

This study was conducted by a collaborative team from:

- Department of Immunology, Harvard Medical School, Boston, MA, USA

- Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical School, Boston, MA, USA

- Program in Cellular and Molecular Medicine, Boston Children's Hospital, Boston, MA, USA

- Walter and Eliza Hall Institute of Medical Research, Parkville, Victoria, Australia

- Department of Medical Biology, The University of Melbourne, Parkville, Victoria, Australia

- Broad Institute of MIT and Harvard, Cambridge, MA, USA

- Division of Gastroenterology, Boston Children's Hospital and Harvard Medical School, Boston, MA, USA

- Ragon Institute of MGH, MIT, and Harvard, Cambridge, MA, USA

- Department of Stem Cell and Regenerative Biology, Harvard University, Cambridge, MA, USA

- Moderna, Cambridge, MA, USA

- Department of Pathology and Laboratory Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA

- Institute for Immunology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA

- Division of Infectious Diseases, Department of Medicine, Washington University School of Medicine, St. Louis, MO, USA

- Department of Chemistry, Yale University, New Haven, CT, USA

- Division of Protective Immunity, Department of Pathology and Laboratory Medicine, Children's Hospital of Philadelphia, University of Pennsylvania, Philadelphia, PA, USA

- Harvard Stem Cell Institute, Harvard University, Cambridge, MA, USA

- Gene Lay Institute of Immunology and Inflammation, Brigham and Women's Hospital, Mass General Hospital, Harvard Medical School, Boston, MA, USA

- Division of Gastroenterology, Hepatology, and Nutrition, Boston Children's Hospital, Boston, MA, USA

- Program in Immunology, Harvard Medical School, Boston, MA, USA

- Department of Pediatrics, Harvard Medical School, Boston, MA, USA

Corresponding author: Ruaidhri Jackson (ruaidhri_jackson@hms.harvard.edu), Department of Immunology, Harvard Medical School.

https://material-image.wanwang.xin/1353944556863376/public/3f7aded6-12af-4ed2-8b6b-fabce9b28166.pdf

References

1. Liu, X. et al. Inflammasome-activated gasdermin D induces pyroptosis by forming membrane pores. Nature 535, 153-158 (2016).

2. Shi, J. et al. Cleavage of GSDMD by inflammatory caspases determines pyroptotic cell death. Nature 526, 660-665 (2015).

3. Michaeli, S. Trans-splicing in trypanosomes: machinery and its impact on the parasite transcriptome. Fut. Microbiol. 6, 459-474 (2011).

4. Dorn, R., Reuter, G. & Loewendorf, A. Transgene analysis confirms mRNA trans-splicing at the complex mod(mdg4) locus in Drosophila. Proc. Natl Acad. Sci. USA 98, 9724-9729 (2001).

5. Daley, G. Q., Van Etten, R. A. & Baltimore, D. Induction of chronic myelogenous leukemia in mice by the P210bcr/abl gene of the Philadelphia chromosome. Science 247, 824-830 (1990).

6. Nowell, P. C. & Hungerford, D. A. Chromosome studies on normal and leukemic human leukocytes. J. Natl Cancer Inst. 25, 85-109 (1960).

7. Li, H., Wang, J., Mor, G. & Sklar, J. A neoplastic gene fusion mimics trans-splicing of RNAs in normal human cells. Science 321, 1357-1362 (2008).

8. Bracht, J. R. et al. Chromosome fusions triggered by noncoding RNA. RNA Biol. 14, 620-631 (2017).

9. Yuan, H. et al. A chimeric RNA characteristic of rhabdomyosarcoma in normal myogenesis process. Cancer Discov. 3, 1394-1403 (2013).

10. Cocquet, J., Chong, A., Zhang, G. & Veitia, R. A. Reverse transcriptase template switching and false alternative transcripts. Genomics 88, 127-131 (2006).

11. Heyer, E. E. et al. Diagnosis of fusion genes using targeted RNA sequencing. Nat. Commun. (2019).

12. Latysheva, N. S. & Babu, M. M. Discovering and understanding oncogenic gene fusions through data-intensive computational approaches. Nucleic Acids Res. 44, 4487-4503 (2016).

13. Almeida da Paz, M., Warger, S. & Taher, L. Ignoring multimappers introduces bias in the functional assessment of NGS data. BMC Genom. (2024).

14. Chen, C., Haddox, S., Tang, Y., Qin, F. & Li, H. Landscape of chimeric RNAs in non-cancerous cells. Genes (2021).

15. Singh, S. et al. The landscape of chimeric RNAs in non-diseased tissues and cells. Nucleic Acids Res. 48, 1764-1778 (2020).

16. Liu, Q. et al. LongGF: a computational algorithm and software tool for rapid and accurate detection of gene fusions via long-read transcriptome sequencing. BMC Genom. 21, 793 (2020).

17. Davidson, N. M. et al. JAFFAL: detecting fusion genes with long-read transcriptome sequencing. Genome Biol. 23, 10 (2022).

18. Karaoglanoglu, F., Chauve, C. & Hach, F. Genion: an accurate tool for detecting gene fusions from long transcriptomics reads. BMC Genom. 23, 129 (2022).

19. Drexler, H. L. et al. Revealing nascent RNA processing dynamics with nano-COP. Nat. Protoc. 16, 1343-1375 (2021).

20. Uhrig, S. et al. Accurate and efficient detection of gene fusions from RNA sequencing data. Genome Res. (2021).

21. Nicorici, D. et al. FusionCatcher: a tool for detecting somatic fusion genes in paired-end RNA-sequencing data. bioRxiv (2014).

22. Haas, B. J. et al. Accuracy assessment of fusion transcript detection via read-mapping and de novo fusion transcript assembly-based methods. Genome Biol. 20, 213 (2019).

23. Melsted, P. et al. Fusion detection and quantification via pseudoalignment. bioRxiv (2017).

24. Davidson, N. M., Majewski, I. J. & Oshlack, A. JAFFA: high-sensitivity transcriptome-focused fusion gene detection. Genome Med. (2015).

25. Caldas, P. et al. Transcription readthrough is prevalent in healthy human tissues and associated with inherent genomic features. Commun. Biol. 7, 100 (2024).

26. Ribeiro, D. M., Ziyani, C. & Delaneau, O. Shared regulation and functional relevance of local gene co-expression revealed by single-cell analysis. Commun. Biol. 5, 876 (2022).

27. Zhang, X. et al. Inactivation of TMEM106A promotes lipopolysaccharide-induced inflammation via the MAPK and NF-κB signaling pathways in macrophages. Clin. Exp. Immunol. 203, 125-136 (2021).

28. Kotake, Y. et al. Splicing factor SF3b as a target of the antitumor natural product pladienolide. Nat. Chem. Biol. 3, 570-575 (2007).

29. Cretu, C. et al. Structural basis of splicing modulation by antitumor macrolide compounds. Mol. Cell 70, 265-273 (2018).

30. Perry, R. P. & Kelley, D. E. Inhibition of RNA synthesis by actinomycin D: characteristic dose-response of different RNA species. J. Cell. Physiol. 76, 127-139 (1970).

31. Ling, J. Q. et al. CTCF mediates interchromosomal colocalization between Igf2/H19 and Wsb1/Nf1.. Science 312, 269-272 (2006).

32. Kayagaki, N. et al. Caspase-11 cleaves gasdermin D for non-canonical inflammasome signaling. Nature 526, 666-671 (2015).

33. Netea, M. G. et al. IL-1beta processing in host defense: beyond the inflammasomes. PLoS Pathog. 6, e1000661 (2010).

34. Schultz, M. J. et al. Role of interleukin-1 in the pulmonary immune response during Pseudomonas aeruginosa pneumonia. Am. J. Physiol. Lung Cell. Mol. Physiol. 282, L285-L290 (2002).

35. Chen, K. W. et al. Noncanonical inflammasome signaling triggers gasdermin D-dependent neutrophil extracellular traps. Sci. Immunol. 3, eaar6676 (2018).

36. Leventis, P. A. & Grinstein, S. Distribution and function of phosphatidylserine in cellular membranes. Annu. Rev. Biophys. 39, 407-427 (2010).

37. Van Meer, G., Voelker, D. R. & Feigenson, G. W. Membrane lipids: where they are and how they behave. Nat. Rev. Mol. Cell Biol. 9, 112-124 (2008).

38. Yu, P. et al. Pyroptosis: mechanisms and diseases. Signal Transduct. Target. Ther. 6, 128 (2021).

39. Xia, S. et al. The Gasdermin D pore structure reveals preferential release of mature interleukin-1. Nature 593, 607-611 (2021).

40. Evavold, C. L. et al. Regulation of gasdermin D oligomerization and pyroptosis by the Ragulator-Rag-mTORC1 pathway. Cell 184, 4495-4511 (2021).

41. Jackson, R. et al. The translation of non-canonical open reading frames controls mucosal immunity. Nature 564, 434-438 (2018).

42. Brubaker, S. W., Gauthier, A. E., Mills, E. W., Ingolia, N. T., & Kagan, J. C. A bicistronic MAVS transcript reveals a class of truncated variants in antiviral immunity. Cell (2014).

43. Anderson, D. M. et al. A micropeptide encoded by a putative long noncoding RNA regulates muscle performance. Cell 160, 595-606 (2015).

44. Wang, Y. & Wang, Z. Efficient backsplicing produces translatable circular mRNAs. RNA (2015).

45. Memczak, S. et al. Circular RNAs are a large class of animal RNAs with regulatory potency. Nature (2013).

46. Chen, C.-K. et al. Structured elements drive extensive circular RNA translation. Mol. Cell 81, 4300-4318 (2021).

47. Jeck, W. R. et al. Circular RNAs are abundant, conserved, and associated with ALU repeats. RNA 19, 141-157 (2013).

48. Rowley, J. D. & Blumenthal, T. The cart before the horse. Science 321, 1302-1304 (2008).

49. Engreitz, J. M., Agarwala, V., & Mirny, L. A. Three-dimensional genome architecture influences partner selection for chromosomal translocations in human disease. PLoS ONE 7, e44196 (2012).

50. Lin, C. et al. Nuclear receptor-induced chromosomal proximity and DNA breaks underlie specific translocations in cancer. Cell 139, 1069-1083 (2009).

51. Marek, L. R. & Kagan, J. C. Phosphoinositide Binding by the Toll Adaptor dMyD88 controls antibacterial responses in Drosophila. Immunity 36, 612-622 (2012).

52. Miao, R. et al. Gasdermin D permeabilization of mitochondrial inner and outer membranes accelerates and enhances pyroptosis. Immunity 56, 2523-2541 (2023).

53. Chu, X. et al. Gasdermin D-mediated pyroptosis is regulated by AMPK-mediated phosphorylation in tumor cells. Cell Death Dis. 14, 469 (2023).

54. Devant, P. et al. Gasdermin D pore-forming activity is redox-sensitive. Cell Rep. (2023).

55. Humphries, F. et al. Succination inactivates gasdermin D and blocks pyroptosis. Science (2020).

56. Shi, Y. et al. E3 ubiquitin ligase SYVN1 is a key positive regulator of GSDMD-mediated pyroptosis. Cell Death Dis. (2022).

57. Chu, X. et al. Ubiquitination of the gasdermin D N-terminal domain directs its membrane translocation and pore formation during pyroptosis. Cell Death Dis. 16, 181 (2025).

This article is licensed under a Creative Commons Attribution 4.0 International License.

The Author(s) 2026

https://material-image.wanwang.xin/1353944556863376/public/f7ad22fb-38f1-4ec6-a905-aa9322c13ac4.pdf