Granulocyte-macrophage colony-stimulating factor (GM-CSF) deficiency drives autoimmune pulmonary alveolar proteinosis (aPAP), a disease characterized by impaired macrophage-mediated clearance of pulmonary surfactants. Clinical data suggest that inhaled recombinant GM-CSF reduces symptoms in aPAP patients, providing a rationale for mRNA-based GM-CSF replacement therapies. However, these require effective mRNA delivery after nebulization. Here, we report the iterative in vivo design of a lipid nanoparticle, named nebulized lung delivery 2 (NLD2), that efficiently delivers mRNA after nebulization. NLD2 carrying GM-CSF mRNA transfected alveolar macrophages in vivo, leading to interleukin-10 pathway activation and subsequent surfactant lipoprotein clearance. In a preclinical disease model of aPAP, GM-CSF mRNA delivery reduced surfactant protein thickness more than recombinant GM-CSF. These data support continued exploration of nebulized lipid nanoparticle therapies for aPAP.

Abstract Image-based spatial transcriptomics depends on cell segmentation to assign transcripts to individual cells, but how segmentation algorithms perform across tissues with distinct cellular architectures is poorly understood. This study presents the broadest independent benchmark to date of cell segmentation algorithms for spatial transcriptomics, comparing five approaches across ten mouse tissues using a 5,006-gene Xenium panel. To quantify segmentation errors, Co-expression Rejection in Segmentation Purity (CRISP) was developed, an open-source tool available in R and Python that measures cell purity through tissue-specific mutually exclusive marker co-expression without requiring ground truth annotations. This benchmark revealed that segmentation algorithms face a fundamental tradeoff between maximizing transcript capture and maintaining cell purity, and that the severity of this tradeoff is tissue-dependent. Proseg achieved the highest average performance across tissues, though the magnitude of its advantage varies with tissue architecture. Overall, CRISP provides per-tissue performance profiles as a practical resource for algorithm selection.

BA.3.2, a variant of SARS-CoV-2 containing ~40 mutations in its spike protein compared to its nearest ancestor, has spread globally since its first detection in South Africa in November 2024. Here, we report antigenic characterization of BA.3.2 viruses in three naive animal models, and visualize its antigenic phenotype in the context of SARS-CoV-2 evolution using antigenic cartography. We find that: (1) BA.3.2 is substantially antigenically divergent from existing SARS-CoV-2 variants; (2) infection with BA.3.2 in hamster and mouse animal models produces sera with lower homologous titer than infection with other variants. Both of these results may have implications for the selection of vaccine antigens.

Vaccination remains the most successful preventative measure against viral infection, but methods to stably deter rapidly-evolving pathogens have remained elusive. Vaccines capable of incorporating and anticipating viral evolution could address current challenges in seasonal vaccination efforts against SARS-CoV-2 and influenza where economic and disease burdens remain high despite decades of combined study. Rare epitope suppression (RES) is an underutilized concept within vaccine design, where humoral epitope targeting can be molded using complex antigen pools. Based in mRNA vaccine technology, 'wobble vaccines' represent the novel application of RES to human pathogens designed to anticipate and resist viral evolution. To establish this platform, public SARS-CoV-2 sequencing data was compiled from the first two years of the COVID-19 pandemic to identify high-diversity sites across the receptor binding domain (RBD) of the spike protein. Wobble RBD (WobbRBD) libraries reflecting that entropy were synthesized and incorporated into established self-amplifying (SA) vaccine constructs. Animals immunized with these complex antigen pools showed no obvious adverse effects. By three days-post vaccination, WobbRBD stimulated robust primary immune activation with distinctive characteristics compared to traditional single-strain vaccine modalities. By day 14, germinal centers, class switching, and antibody-secreting cells were induced, creating potent SARS-CoV-2 spike-binding IgG antibodies. Despite similar overall activation profiles, WobbRBD generated significantly increased breadth against SARS-CoV-2 variant spikes in comparison to single-strain controls — even against future-emerging strains. Taken together, wobble vaccines represent a novel method for anticipating and preventing viral escape with promising applications in SARS-CoV-2, influenza, HIV, and beyond.

Inhalation of Rhodococcus equi causes severe pneumonia in humans and animals worldwide, most commonly affecting horse foals. The standard for preventing R. equi pneumonia in foals is transfusion of hyperimmune plasma, which is expensive and carries the risk of adverse effects. Our goal was to passively immunize foals against R. equi by nebulizing mRNA encoding an equine monoclonal antibody (mAb) against the virulence-associated protein A (VapA) directly into the lungs. VapA-specific memory B cells from an immunized horse were used to identify and select the sequence for an equine immunoglobulin (Ig)G1 mAb. In vitro-transcribed mRNA encoding this sequence expressed full-length, VapA-specific mAbs in vitro and safely and effectively produced intrapulmonary mAb in foals for at least 5 days following nebulization. These findings establish a platform to generate mRNA-encoded mAbs for immunotherapeutic and immunoprophylactic applications in horses and demonstrate the feasibility of delivering nebulized mRNA-mAb for intrapulmonary mAb expression in neonates.

Bovine trichomonosis is caused by the urogenital parasite Tritrichomonas foetus (T. foetus). In the United States, approved therapies are lacking, and management is limited to culling infected bulls. Preputial therapy with synthetic mRNA could lead to effective new treatments. We developed synthetic mRNA encoding bovine IgG1 against two epitopes of the T. foetus cell surface antigen TF1.17 and used the mRNA to transfect bovine cells in vitro. Transfected cells expressed membrane anchored or secreted versions of the antibodies with a NanoLuciferase (NanoLuc) reporter molecule fused to each light chain. Luminescence in cells and supernatants collected 24 and 48 h post-transfection confirmed the production of anti-TF1.17 and was significantly higher than in non-transfected controls (p < 0.05). Anti-TF1.17 bound to live parasites as indicated by significantly higher luminescence following treatment with 24 and 48 h post-transfection supernatants compared to transfection controls (p = 0.001). Treatment of T. foetus with concentrated anti-TF1.17 antibody decreased parasite viability. When T. foetus were added to mRNA transfected kidney cells 48 h post transfection, cytopathic effects of the parasites were reduced following 24 h of co-culture with cells producing anti-TF1.17 as compared to controls (p < 0.05). To our knowledge, this is the first use of mRNA transfection of bovine cells to induce the expression of antibodies that can bind to T. foetus, decrease their viability and their cytopathic effects on host cells. This work forms the basis for the development of novel mRNA-mediated approaches to treat or prevent bovine trichomonosis.

Broadly neutralizing antibodies (bnAbs) prevent HIV infection but face administration and cost challenges. We present single-chain mRNA-encoded bnAbs that improve heavy-light chain assembly and enable three enhancement approaches: co-expression, isotype selection, and engineered nanobodies. We observe enhanced in vitro neutralization through co-expression of PGT121 and VRC07 (targeting V3-glycan and CD4-binding site, respectively) and by replacing IgG constant heavy chain (IgG-C-H) with IgA-C-H or incorporating an IgM tailpiece into IgG-C-H. While IgG versions fail to neutralize several SHIV/HIV strains, co-expressing PGT121 and VRC07 as IgM-like multimers restores neutralizing capability (IC50 < 100 ng/mL) and substantially improves breadth. Vaginal explants from rhesus macaques, treated with intravaginal aerosolized mRNAs, show robust protection against ex vivo challenges with multiple SHIV strains when PGT121 and VRC07 are co-expressed as IgM-like multimers. Our data present novel drug compositions for intravaginal mRNA delivery to prevent HIV acquisition and advance antibody-design strategies that enhance breadth and potency of existing bnAbs.

Chimeric antigen receptor (CAR) T cell immunotherapy relies on CAR targeting of tumor-associated antigens; however, heterogenous antigen expression, interpatient variation and off-tumor expression by healthy cells remain barriers. Here we develop synthetic antigens to sensitize solid tumors for recognition and elimination by CAR T cells. Unlike tumor-associated antigens, we design synthetic antigens that are orthogonal to endogenous proteins to eliminate off-tumor targeting and that have a small genetic footprint to facilitate efficient tumor delivery to tumors by lipid nanoparticles. Using a camelid single-domain antibody (VHH) as a synthetic antigen, we show that adoptive transfer of anti-VHH CAR T cells to female mice bearing VHH-expressing tumors reduced tumor burden in multiple syngeneic and xenograft models of cancer, improved survival, induced epitope spread, protected against tumor rechallenge and mitigated antigen escape in heterogenous tumors. Our work supports the in situ delivery of synthetic antigens to treat antigen-low or antigen-negative tumors with CAR T cells.

Stress granules (SGs) are dynamic, cytoplasmic foci that form in response to environmental stresses, including viral infections, and function to restore cellular homeostasis by regulating mRNA translation, storage, and decay. To inhibit SG formation and subvert their antiviral effects, viruses from diverse families sequester or cleave G3BP1, the key SG nucleating protein. We found that an infection with simian hemorrhagic fever virus (SHFV), a member of the family Arteriviridae, does not induce the formation of bona fide SGs despite inducing phosphorylation of PKR and eIF2α. The SG proteins, G3BP1, G3BP2, TIA-1, Caprin-1, and USP10, but not the translation initiation proteins, eIF3A, eIF4G, and small ribosomal protein S6 (rpS6), were redistributed into foci located in the same intracellular region as the viral dsRNA foci. However, SGs could be induced in infected cells by exogenous inducers. LC-MS/MS analysis of the proteins co-immunoprecipitating with endogenous G3BP1 from SHFV-infected cell lysates detected multiple viral replication/transcription complex proteins. Interaction between G3BP1 and the nsp2 and N proteins of SHFV was observed in reciprocal co-immunoprecipitation assays, and colocalization was detected by IFA. A conserved FGAP motif in nsp2 and a FAEP motif in the N protein were shown to be required for interaction with G3BP1. We also detected G3BP cleavage products in the SHFV-infected cell lysates and hypothesize that cleavage is mediated by a viral protease. These findings suggest that SG formation is not induced by an SHFV infection due to recruitment of G3BP to sites of viral replication and cleavage of G3BP by viral proteins.IMPORTANCEEukaryotic cells shut down translation by assembling stress granules (SGs) in response to environmental stresses, including viral infections. Viruses require cellular translation machinery for protein synthesis and have developed mechanisms to subvert SG assembly. Simian hemorrhagic fever virus (SHFV), a simian arterivirus, causes asymptomatic infections in African cercopithecoid monkeys but fatal hemorrhagic fever disease in Asian macaques. Even though intracellular production of SHFV RNA activates the stress sensor, PKR, SGs are not induced. G3BP1, the main nucleating protein of SGs, is recruited to foci located near viral replication complexes through interaction with the viral proteins nsp2 and N. An FGAP motif in nsp2 and an FAEP motif in the N protein are required for interaction with G3BP1. Cleavage of G3BP1 was identified as an additional mechanism of viral counteraction of SG formation.

Dengue is a major global health threat, and there are no approved antiviral agents. Prior research using Cas13 only demonstrated dengue mitigation in vitro. Here we demonstrate that systemic delivery of mRNA-encoded Cas13a and guide RNAs formulated in lipid nanoparticles can be used to treat dengue virus (DENV) 2 and 3 in mice. First, we identified guides against DENV 2 and 3 that demonstrated in vitro efficacy. Next, we confirmed that Cas13 enzymatic activity is necessary for DENV 2 or DENV 3 mitigation in vitro. Last, we show that a single dose of lipid-nanoparticle-formulated mRNA-encoded Cas13a and guide RNA, administered 1 day post-infection, promotes survival of all infected animals and serum viral titre decreases on days 2 and 3 post-infection after lethal challenge in mice. Off-target analysis in mice using RNA sequencing showed no collateral cleavage. Overall, these data demonstrate the potential of mRNA-encoded Cas13 as a pan-DENV drug. Potent CRISPR guides targeting conserved dengue virus regions can treat dengue-2 and -3 infection in vivo in mice when co-delivered with Cas13 by lipid nanoparticles.

The CRISPR-Cas13 system has been proposed as an alternative treatment of viral infections. However, for this approach to be adopted as an antiviral, it must be optimized until levels of efficacy rival or exceed the performance of conventional approaches. To take steps toward this goal, we evaluated the influenza viral RNA degradation patterns resulting from the binding and enzymatic activity of mRNA-encoded LbuCas13a and two crRNAs from a prior study, targeting PB2 genomic and messenger RNA. We found that the genome targeting guide has the potential for significantly higher potency than originally detected, because degradation of the genomic RNA is not uniform across the PB2 segment, but it is augmented in proximity to the Cas13 binding site. The PB2 genome targeting guide exhibited high levels (>1 log) of RNA degradation when delivered 24 hours post-infection in vitro and maintained that level of degradation over time, with increasing multiplicity of infection (MOI), and across modern influenza H1N1 and H3N2 strains. Chemical modifications to guides with potent LbuCas13a function, resulted in nebulizer delivered efficacy (>1-2 log reduction in viral titer) in a hamster model of influenza (Influenza A/H1N1/California/04/09) infection given prophylactically or as a treatment (post-infection). Maximum efficacy was achieved with two doses, when administered both pre- and post-infection. This work provides evidence that mRNA-encoded Cas13a can effectively mitigate Influenza A infections opening the door to the development of a programmable approach to treating multiple respiratory infections.

Supplementary Figure 6: PMTRIP FISH-PLA can quantify COX-2-TIAR interactions in lung tissue. A) 20 nM of V5 tagged PMTRIPs targeting COX-2 mRNA were delivered as FISH probes to colon cancer samples. PLA (green) was performed between the V5 tag and TIAR. 20x cropped image is indicated by the white box. Extended focus images are shown, with a 1 mm scale bar for full tissue. B) Extended focus images of FISH-PLA (green) controls are shown. Scale bar is 1 mm. C) Extended focus cropped images of full tissues in (A) are indicated by white boxes. Scale bars are 100 μm. D) Extended focus image of cancer tissue FISH-PLA at 63x is also shown. Scale bars for (D) are 10 μm. E) Extended focus cropped images of full tissues in (B) are indicated by white boxes. Extended focus images of tissue FISH-PLA at 63x are also shown. COX-2 FISH shown in red. Scale bars are 100 μm for 20x and 10 μm for 63x images. F) Quantification of PLA signal in (A,B). Statistics were performed with a one-way ANOVA with a Tukey's multiple comparisons test, where n=3 and * p < 0.015 and ** p < 0.0077. Standard deviations are shown in red.

Supplementary Figure 4: PMTRIP FISH-PLA can quantify COX-2-TIAR interactions in colon tissue. A) 20 nM of V5 tagged PMTRIPs targeting COX-2 mRNA were delivered as FISH probes to colon cancer samples. PLA (green) was performed between the V5 tag and TIAR. Extended focus images are shown, with a 1 mm scale bar. B) Extended focus images of FISH-PLA (green) controls are shown. Scale bar is 1 mm. C) 20x cropped image for (A) are shown, as indicated by the white boxes. Extended focus images are shown, with a 100 μm scale bar. D) Extended focus image of tissue FISH-PLA at 63x is shown, with a 10 μm scale bar. COX-2 FISH shown in red. E) Extended focus cropped images of full tissues in (B) are indicated by white boxes. Extended focus images of tissue FISH-PLA at 63x are also shown. COX-2 FISH shown in red. Scale bars are 100 μm for 20x and 10 μm for 63x images. F) Quantification of PLA signal in (a-b). Statistics were performed with a one-way ANOVA with a Tukey's multiple comparisons test, where n=3 and * p < 0.046. Standard deviations are shown in red. G) Immuohistochemistry for TIAR was performed on colon cancer and healthy colon tissue. Single plane representative images are shown. Scale bars are 50 μm. Quantification of the area of HuR expression normalized by the area of nuceli is shown. Statistics were a Mann-Whitney t-test, where n= 5 and * p < 0.032.

Supplementary Figure 3: PMTRIP FISH-PLA can quantify PolyA-TIAR interactions in colon tissue. A-C) Cells were treated with 5 μ g/ml of ActD 1 h prior to fixation. 5 nM of V5 tagged PMTRIPs targeting polyadenylated mRNA were delivered as FISH probes. PLA was performed between the V5 tag and HuR. PLA signal is quantified by A) the number of PLA puncta, B) volume of PLA signal, or C) the sum of intensities of PLA signal per cell. Statistics were performed with a one-way ANOVA with a Dunn's multiple comparisons test, where n=30, ** p < 0.002, and **** p < 0.0001. 95% confidence intervals are shown in red. D) 5 nM of V5 tagged PMTRIPs targeting polyadenylated mRNA were delivered as FISH probes to colon cancer samples. PLA (green) was performed between the V5 tag and TIAR. Extended focus images are shown, with a 1 mm scale bar for full tissue. E) Extended focus images of FISH-PLA (green) controls are shown. Scale bar is 1 mm. F) Extended focus cropped images of full tissues in (D) are indicated by white boxes. Scale bars are 100 μm. G) Extended focus image of cancer tissue FISH-PLA at 63x is shown. Scale bar is 10 μm. h) Extended focus cropped images of full tissues in (E) are indicated by white boxes. Extended focus images of control tissue FISH-PLA at 63x are also shown. Scale bars are 100 μm for 20x and 10 μm for 63x images. I) Quantification of PLA signal in (D-E). Statistics were performed with a one-way ANOVA with a Tukey's multiple comparisons test, where n=3 and ** p < 0.0075. Standard deviations are shown in red.

Supplementary Figure 7: PMTRIP FISH-PLA can quantify PolyA-HuR interactions in frozen colon tissue. A) 5 nM of V5 tagged PMTRIPs targeting polyadenylated mRNA were delivered as FISH probes to tissue samples. PLA (green) was performed between the V5 tag and HuR. Extended focus images are shown, with a 1 mm scale bar. B) Extended focus cropped images of full tissues in (A) are indicated by white boxes. Extended focus images of tissue FISH-PLA at 63x are also shown. Scale bars are 100 μm for 20x and 10 μm for 63x images. C) Quantification of PLA signal in (A) is shown. Standard deviations are shown in red.

Supplementary Figure 2: PMTRIPs FISH-PLA is specific, effective, and sensitive without dextran sulfate. A). 5 nM of V5-tagged PMTRIPs targeting polyadenylated mRNA were delivered as FISH probes, with FISH hybridization buffer containing either 0%, 5% or 10% dextran sulfate. PLA (white) was performed between the V5 tag and HuR. Extended focus images are shown. Scale bars indicate 10 µm. B) Quantification of PLA is shown. Statistics were performed with a one-way ANOVA with a Dunn's multiple comparisons test, where n=30 and **** p < 0.0001. 95% confidence intervals are shown in red. C) 5 nM V5 tagged MTRIPs targeting polyadenylated mRNA were assembled using either PNA or 2'OM oligos and were delivered as FISH probes. PLA (white) was performed between the V5 tag and HuR. Extended focus images are shown. Scale bars indicate 10 µm. Quantification of PLA is shown. Statistics were performed with a one-way ANOVA with a Dunn's multiple comparisons test, where n=30 and ** p < 0.002 and **** p < 0.0001. 95% confidence intervals are shown in black. D) Representative extended focus control images for figure 2A are shown. PLA is in white. Scale bars indicate 10 µm. E) Representative extended focus control images for figure 2B are shown. PLA is in white. Scale bars indicate 10 µm. F) Representative extended focus control images for figure 2C are shown. PLA is in white. Scale bars indicate 10 µm. G) Representative extended focus control images for figure 2D are shown. PLA is in white. Scale bars indicate 10 µm.

Supplementary Figure 1: PNA increases mRNA-protein PLA specificity upon live delivery and reduces the need for formamide and dextran sulfate for FISH. A) 5 nM of of V5-tagged MTRIPs with either PNA or 2OM oligonucleotides targeting polyadenylated mRNA were delivered to live cells via SLO. PLA (white) was performed between V5 and HuR. Extended focus images are shown. Scale bars indicate 10 μm. Quantification of PLA is shown. Statistics were performed with a two-way ANOVA with a Tukey's multiple comparisons test, where n=30 and **** p < 0.0001. 95% confidence intervals are shown in black. B) Cells were infected with RSV, they were exposed to FISH hybridization buffer with either 0% or 10% formamide overnight, and stained for F (red) and M (green). Extended focus images are shown. Scale bars indicate 10 μm. C) 10 μg of Dylight 650 pre-labeled luciferase mRNA (red), either naked or complexed with Viromer Red, was delivered to mouse muscle. FISH (green) was performed either with cy3B-PMTRIPs or Quazar 570 DNA oliognucleotides. D) 200 ng of cy3B pre-labeled luciferase mRNA (red) was delivered to cells via L2K transfection. FISH (green) was performed either with PMTRIPs or DNA oliognucleotides. E) Representative images of sham controls for (C) and (D). F) PMTRIP FISH was performed in A549s for polyadenylated mRNA or for a scrambled sequence. G) PMTRIP FISH was performed in HT-29s for COX-2, Actin-β or for a scrambled sequence. H) PMTRIP FISH was performed in DLD-1s for COX-2. In all images scale bars represent 10 μm.

Supplementary Figure 5: PMTRIP FISH-PLA can quantify PolyA-TIAR interactions in lung tissue. A) 5 nM of V5 tagged PMTRIPs targeting polyadenylated mRNA were delivered as FISH probes to colon cancer samples. PLA (green) was performed between the V5 tag and TIAR. Extended focus images are shown, with a 1 mm scale bar for full tissue. B) Extended focus images of FISH-PLA (green) controls are shown. Scale bar is 1 mm. C) Extended focus cropped image of full tissues in (A) are indicated by white boxes. Scale bars are 100 μm. D) Extended focus image of tissue FISH-PLA at 63x is shown. Scale bar is 10 μm. E) Extended focus cropped images of full tissues in (B) are indicated by white boxes. Extended focus images of tissue FISH-PLA at 63x are also shown. Scale bars are 100 μm for 20x images and are 10 μm for 63x images. F) Quantification of PLA signal in (A, B). Statistics were performed with a one-way ANOVA with a Tukey's multiple comparisons test, where n=3 and ** p < 0.0087. Standard deviations are shown in red.