Nanoparticle delivery of a prodrug-activating bacterial enzyme leads to anti-tumor responses

Huayamares, Sebastian G.; Lian, Liming; Rab, Regina; Hou, Yuning; Radmand, Afsane; Kim, Hyejin; Zenhausern, Ryan; Achyut, Bhagelu R.; Gilbert Ross, Melissa; Lokugamage, Melissa P.; Loughrey, David; Peck, Hannah E.; Echeverri, Elisa Schrader; Da Silva Sanchez, Alejandro J.; Shajii, Aram; Li, Andrea; Tiegreen, Karen E.; Santangelo, Philip J.; Sorscher, Eric J.; Dahlman, James E.

doi:10.1038/s41467-025-58548-1

Download PDF

Article
Open access
Published: 12 April 2025

Nanoparticle delivery of a prodrug-activating bacterial enzyme leads to anti-tumor responses

Nature Communications volume 16, Article number: 3490 (2025) Cite this article

7704 Accesses
2 Citations
Metrics details

Subjects

Abstract

Most cancer patients diagnosed with late-stage head and neck squamous cell carcinoma are treated with chemoradiotherapy, which can lead to toxicity. One potential alternative is tumor-limited conversion of a prodrug into its cytotoxic form. We reason this could be achieved by transient and tumor-specific expression of purine nucleoside phosphorylase (PNP), an Escherichia coli enzyme that converts fludarabine into 2-fluoroadenine, a potent cytotoxic drug. To efficiently express bacterial PNP in tumors, we evaluate 44 chemically distinct lipid nanoparticles (LNPs) using species-agnostic DNA barcoding in tumor-bearing mice. Our lead LNP, designated LNP intratumoral (LNP^IT), delivers mRNA that leads to PNP expression in vivo. Additionally, in tumor cells transfected with LNP^IT, we observe upregulated pathways related to RNA and protein metabolism, providing insight into the tumor cell response to LNPs in vivo. When mice are treated with LNP^IT-PNP, then subsequently given fludarabine phosphate, we observe anti-tumor responses. These data are consistent with an approach in which LNP-mRNA expression of a bacterial enzyme activates a prodrug in solid tumors.

Prodrug-conjugated tumor-seeking commensals for targeted cancer therapy

Article Open access 21 May 2024

Long non-coding RNA LINC-PINT as a novel prognostic biomarker in human cancer: a meta-analysis and machine learning

Article Open access 29 March 2024

Acid-degradable lipid nanoparticles enhance the delivery of mRNA

Article 23 August 2024

Introduction

The ability to manufacture mRNA vaccines^1,2 suggests lipid nanoparticle (LNP)-mRNA drugs could be cancer nanomedicines. This concept is substantiated by LNP-mRNA drugs encoding immunomodulatory mRNA delivered intratumorally into patients^3,4,5 as well as preclinical results demonstrating that intratumoral LNP-mRNA drugs can lead to anti-tumor responses. In one example, tumor growth was inhibited in a B16F10 murine melanoma model following LNP-mRNA drugs encoding immunostimulatory cytokines⁶. Other examples have demonstrated that intratumoral injection of self-replicating mRNAs, circular mRNAs, or mRNAs encoding gene editing constructs could drive anti-tumor responses^7,8,9. These data support the exploration of LNPs that efficiently deliver mRNA after intratumoral administration.

One tumor type in which intratumoral delivery could be impactful is head and neck squamous cell carcinoma (HNSCC). Head and neck cancer is the sixth most common malignancy worldwide¹⁰, accounts for 3-4% of cancers in the United States, and is expected to increase 30% by 2030¹¹. Late-stage HNSCC afflicts otherwise healthy patients and can confer chronic pain, uncontrolled bleeding, debility, and death¹⁰. Patients are given surgery and/or chemoradiotherapy as a first-line treatment approach; however, this often leads to insufficient anti-tumor activity and significant toxicity¹².

Given that HNSCC is often characterized by tumors that cannot be fully resected due to their location yet can be accessed via an intratumoral injection, we reasoned that we could design a two-step anti-tumor therapy (Fig. 1a). In the first step, an LNP would be formulated to carry mRNA and injected intratumorally. The mRNA would encode Escherichia coli (E. coli) purine nucleoside phosphorylase (PNP), an enzyme that cleaves fludarabine to 2-fluoroadenine (F-Ade), which is cytotoxic. In the second step, we would administer fludarabine phosphate; this would be converted to fludarabine by endogenous enzymes¹³, then subsequently cleaved to F-Ade in tumor cells expressing PNP¹⁴. Since E. coli PNP is not expressed in human cells, and because fludarabine is a poor substrate for human PNP analogs¹⁵, we expected that resulting cell death would not extend beyond the tumor (i.e., where LNP-mediated PNP expression occurs). This is supported by existing clinical data. In a phase 1 clinical trial, 12 patients were treated with an adenoviral vector expressing PNP, then given fludarabine phosphate¹⁶. All 12 completed the study without dose-limiting toxicity, and a dose-dependent tumor response, including tumor regressions in some patients, was observed¹⁶. While such data are promising, adenoviral vectors can be limited by pre-existing antibodies in humans¹⁷ that make them difficult to redose¹⁸. By contrast, LNP-RNA drugs are regularly readministered to patients. An LNP-mRNA-based approach would require efficient mRNA delivery after intratumoral injection.

**Fig. 1: High-throughput DNA barcoding can be used to optimize the intratumoral (IT) delivery of mRNA to tumor cells in vivo using LNPs.**

Here, we use species-agnostic nanoparticle delivery screening (SANDS)¹⁹ to evaluate 44 chemically distinct LNPs for intratumoral delivery in mice carrying FaDu-based tumors, a model of HNSCC^20,21. This high-throughput in vivo LNP selection contrasts previous work optimizing LNP tumor delivery in cell culture, which can poorly predict delivery in adult animals²². The in vivo studies identify a lead LNP, termed LNP intratumoral (LNP^IT), which subsequently delivers PNP-encoding mRNA to tumor models in vivo. Consistent with our hypothesis, intratumoral administration of LNP^IT-PNP mRNA followed by fludarabine phosphate leads to anti-tumor responses in vivo without overt systemic toxicity.

Results

In vivo identification of an LNP for intratumoral delivery

We first evaluated whether four chemically distinct ionizable lipids that have different in vivo tropism would deliver mRNA to FaDu tumors in NU/J immunocompromised mice (Fig. 1b). We tested (1) DLin-MC3-DMA, which delivers siRNA to the liver in humans after an intravenous administration^23,24, (2) cKK-E12, which delivers siRNA to the liver in non-human primates (NHPs) after an intravenous administration²⁵, (3) SM-102, which is used in the Moderna COVID vaccine²⁶, and (4) C12-200, which delivers siRNA to the liver in NHPs²⁷ (Supplementary Table 1a). We formulated the LNPs to carry mRNA encoding anchored nanoluciferase (NanoLuc), injected these intratumorally at a dose of 5 µg/tumor, and measured luminescence 48 hours later (Fig. 1c). Since C12-200 had the highest bioluminescence, we performed a screen¹⁹ using C12-200-like lipids to optimize intratumoral mRNA delivery (Fig. 1d, e). We designed 64 LNPs using stereopure isoforms of the C12-200 lipid, which were recently shown to deliver mRNA more efficiently than racemic C12-200²⁸, as well as the previously studied²⁹ polyethylene glycol (PEG)-lipid C₁₈PEG₂₀₀₀. We synthesized eight lipids in total: four lipid lengths, with each chiral form (S and R) (Fig. 1f). Since cholesterol^30,31,32 and helper lipid structure^33,34,35,36 can affect delivery, we varied these constituents as well (Fig. 1g). Finally, to control for molar ratio-dependent effects, we formulated each potential combination with two molar ratios (Supplementary Table 1b).

We constructed all 64 LNPs to carry a DNA barcode as well as mRNA encoding a glycosylphosphatidylinositol (GPI)-anchored camelid single-variable domain on a heavy chain (VHH) antibody (anchored-VHH, aVHH)^19,37. LNP-1, with chemical composition 1, carried aVHH mRNA and DNA barcode 1, whereas LNP-N, with chemical composition N, carried aVHH mRNA and DNA barcode N. By using a sensitive DNA barcode³⁸, we were able to formulate the mRNA:barcode at a ratio of 10:1. After mixing lipid and nucleic acid phases together using microfluidics³⁹, we characterized hydrodynamic diameter and stability of each LNP using dynamic light scattering (DLS). LNPs with a hydrodynamic diameter between 50 and 200 nm were deemed acceptable for injection. Of the 64 LNPs initially formulated, 44 met inclusion criteria and were pooled, dialyzed into PBS, and sterile filtered. As a control, we measured the hydrodynamic diameter of the pooled LNPs and found it to be within the range of diameters of LNPs constituting the pool, suggesting that the LNPs did not aggregate (Fig. 1h).

We then intratumorally administered the pooled LNP library to NU/J mice carrying FaDu HNSCC xenografts at a total nucleic acid dose of 6 µg/tumor (i.e., 0.13 µg/LNP, for all 44 LNPs on average). As a negative control, we added an unencapsulated DNA barcode, which is endocytosed and delivered into the cancer cells less efficiently than barcodes within LNPs²². Sixteen hours later, which is sufficient for cells to express the aVHH protein, we digested tumors, isolated human (i.e., human CD47⁺) cells that were functionally transfected (i.e., aVHH⁺) via fluorescence-activated cell sorting (FACS), and sequenced the CD47⁺ aVHH⁺ cells to identify the barcodes within them (Supplementary Fig. 1). As a final control, we measured the normalized delivery⁴⁰ of barcodes carried in LNPs and unencapsulated barcodes. As expected, unencapsulated barcodes were found less frequently than barcodes carried by LNPs (Fig. 1i).

We subsequently analyzed the barcodes to identify LNP characteristics found in both top and bottom performing LNPs (Fig. 2a, Supplementary Fig. 2). We found that cationic DC-cholesterol⁴¹ was negatively enriched and neutral cholesterol was positively enriched; that is, the best-performing LNPs were formulated with cholesterol whereas the worst-performing LNPs were formulated with DC-cholesterol. We then calculated the normalized delivery for each LNP in the screen (Supplementary Table 2), which led us to select two lead LNPs: LNP²⁸ and LNP^IT. LNP²⁸ and LNP^IT were formulated with neutral cholesterol, contained the same molar ratios of four components, and formed stable LNPs with hydrodynamic diameters less than 200 nm (Fig. 2b, Supplementary Fig. 3a). When compared with the LNP composed of racemic C12-200, both LNP²⁸ and LNP^IT yielded higher transfection in FaDu cells (Supplementary Fig. 3b).

**Fig. 2: IT LNPs that functionally deliver mRNA to human tumor cells in vivo.**

We next formulated lead LNPs with anchored NanoLuc-encoding mRNA and injected these at a dose of 3 µg/tumor into bilateral FaDu tumors in NU/J mice. Two days later, we isolated tumors and off-target tissues and quantified NanoLuc protein expression using an in vivo imaging system (IVIS). We observed luciferase expression in the FaDu tumors (Fig. 2c, Supplementary Fig. 3c). As a control for mRNA-specific effects and to compare delivery in mouse parenchymal and human tumor cells, we formulated the lead LNPs with mRNA encoding aVHH and injected at a dose of 6 µg/tumor. Sixteen hours later, we observed LNP-mediated transfection in parenchymal and tumor (human CD47⁺) cells (Fig. 2d, Supplementary Fig. 3d, e). Consistent with the fact that NU/J mice lack T cells and show partial defects in B cell development due to the Foxn1^nu homozygous mutation, we observed no T cells. Of the few immune cells we did observe, most were macrophages, which is consistent with previous characterizations of NU/J mice⁴².

We then quantified delivery in a patient-derived xenograft (PDX) model. We inoculated NOD scid gamma (NSG) immunocompromised mice with patient-derived mixed/crude (non-clonal) HNSCC tumor cells extracted from a human lateral neck soft mass (328373-195-R-J1-PDC). We again compared LNP²⁸ and LNP^IT, first with mRNA encoding NanoLuc. PDX tumors injected with LNP^IT had high bioluminescence (Fig. 3a, Supplementary Fig. 4a). LNP²⁸ transfected PDX tumors less effectively than LNP^IT, highlighting the utility of comparing LNPs in multiple tumor models. We then repeated the experiment with aVHH; once again, LNP²⁸ delivered mRNA less efficiently than LNP^IT, which transfected human PDX cancer cells as well as murine cells (Fig. 3c, Supplementary Fig. 4b). The extracellular matrix and tumor microenvironment (TME) can contribute to intratumoral retention of therapeutic agents^43,44 and may interact differently with constituents of LNPs. Differences in microenvironment may provide one potential mechanism that helps explain the gene transfer performance of LNP²⁸ versus LNP^IT. Notably, NSG mice lack B and T cells; thus, we found fewer immune cells than in the FaDu model. To confirm the activity of LNP^IT, we increased the dose to 40 µg/tumor; this led to the expected dose-dependent increase in delivery to PDX and mouse parenchymal cells (Fig. 3d). We did not observe off-target delivery at the 40 µg/tumor dose (Supplementary Fig. 4d).

**Fig. 3: IT LNPs can functionally deliver mRNA to tumor cell types in patient-derived xenografts in vivo.**

Single-cell LNP delivery and cell response in PDX tumors

One limitation of many drug delivery studies, including those above, is their inability to (i) relate LNP delivery to cell heterogeneity and (ii) quantify the detailed response to LNPs. As a result, the extent to which heterogeneity and intracellular response to LNPs affect tumor delivery in vivo remains understudied. We therefore used a bespoke single-cell RNA-sequencing (scRNA-seq)-based approach^36,45 to overlay LNP^IT-mediated aVHH delivery. After intratumorally injecting mice with 40 µg/tumor of LNP^IT-mRNA or a PBS control, we mapped aVHH protein expression onto the t-distributed stochastic neighbor embedding (t-SNE) of 2,573 PDX tumor cells mapped to the human genome (Fig. 4a). We then examined the expression levels of canonical marker genes for aggressive human HNSCC (KRT14, KRT17, KRT6A, KRT5, KRT19, KRT8, KRT16, KRT18, KRT6B, KRT15, KRT6C, KRTCAP3, EPCAM, SFN)⁴⁶ and found that this 12-gene signature was expressed primarily by cells in clusters 2 and 10 (Fig. 4b). Interestingly, we found different levels of LNP^IT aVHH delivery quantified via cellular indexing of transcriptomes and epitopes by sequencing⁴⁷, with the highest amount in clusters 9 and 10 (Fig. 4c, d, Supplementary Fig. 5a). The high delivery and malignancy in cluster 10 provide one early line of evidence relating malignant gene expression to LNP-based transfection. Although significant future work is required to validate this potential relationship, it is important to note this finding would have been difficult to observe using traditional delivery readouts.

**Fig. 4: LNP^IT transfected malignant cell types of HNSCC tumors and upregulated RNA and protein metabolic pathways.**

We noted a number of variably up- or down-regulated genes in each cluster (Supplementary Fig. 5b). For example, we found clusters 2, 5, 8, and 10 had the highest gene expression of HNSCC stromal cell gene markers (ALDH1A1, BCL11B, BMI1, CD44)⁴⁸ (Supplementary Fig. 5c). We also found CD47 expressed ubiquitously across all clusters (Supplementary Fig. 5d). When evaluating SLA2, a prognostic marker in HNSCC that correlates with immune cell infiltration of the TME⁴⁹, we observed no expression (Supplementary Fig. 5e). This is consistent with low immune cell infiltration in an immunocompromised murine model.

We then studied the transcriptomic changes in transfected cancer cells by analyzing differential expression. To control for potential bystander effects in the microenvironment (compared to PBS-treated mice) (Supplementary Fig. 6a), we compared cells that were targeted by LNPs (i.e., aVHH⁺ cells) to cells that were not (aVHH^- cells) from the same tumors. We found that 928 genes were significantly (p < 0.05) upregulated, including human HNSCC genes such as MYH9⁵⁰ and PRDX5⁵¹, and 17 were downregulated in aVHH⁺ cells compared to aVHH^-. When the top significantly upregulated genes (p < 0.001) from aVHH⁺ cells transfected by LNP^IT (Fig. 4e) were analyzed using the ReactomeGSA database⁵², we found 28 significantly enriched protein metabolic pathways (R-HSA-392499.10). Interestingly, all 28 of these pathways were related to cellular RNA or protein management. Of these, 14 were directly associated with mRNA translation into protein (R-HSA-72766) and stemmed directly from the core of the metabolic pathway tree (Biological Process GO:0019538), consistent with previously reported transcriptional responses to mRNA-carrying LNPs³⁶ (Fig. 4f, g, Supplementary Fig. 6b). Some of the most upregulated genes found in the PDX model in vivo were also explored in a different PDX model (PDX2) transfected with LNP^IT (Supplementary Fig. 6c). These data provide an early line of evidence that HNSCC cancer cells respond to LNPs carrying mRNA in part by altering genes related to the manufacture and processing of RNA and protein.

LNP^IT-PNP followed by fludarabine yield anti-tumor responses

The screening data, confirmation studies, and scRNA-seq results led us to conclude that LNP^IT transfected HNSCC cells in mice. We therefore used LNP^IT to test the hypothesis that LNP-PNP followed by fludarabine phosphate could have anti-tumor effects. We synthesized a chemically modified mRNA encoding E. coli PNP, formulated it within LNP^IT, and injected intratumorally at a dose of 6 µg/tumor. Since there is not an extensively validated antibody for PNP, we quantified tumor PNP enzymatic activity using high-performance liquid chromatography (HPLC)^15,53. PNP activity was highest six to 24 hours after injection and decreased at 48 hours (Fig. 5a, b), with no off-target PNP activity in the liver or spleen (Supplementary Fig. 7a, b). When we increased the PNP mRNA dose to 40 µg mRNA/tumor, we observed a dose-dependent increase in PNP activity (Fig. 5c).

**Fig. 5: LNP^IT carrying PNP mRNA, combined with fludarabine phosphate, induces tumor suppression of HNSCC patient-derived xenografts.**

We then performed a therapeutic study in NSG mice. After inoculating animals with 10⁷ PDX cells, we monitored tumor growth until the malignant masses reached volumes between 150 and 250 mm³. In group one, mice were then injected with LNP^IT-mRNA, then 24 hours later with fludarabine phosphate; in group two, mice were injected with LNP^IT-mRNA, then a DMSO control; in group three, a PBS control followed by fludarabine phosphate; and in group four, PBS then DMSO (Fig. 5d). Mice from group one had significantly smaller tumor volumes than each of the three control groups after treatment (Fig. 5e, Supplementary Fig. 8a), and anti-tumor responses were observed as early as one day after completing treatment (Supplementary Fig. 8b). We did not observe significant weight loss in group one relative to other groups, which provides an early line of evidence suggesting the intervention was tolerated (Supplementary Fig. 8c). We then repeated the experiment with groups one and four, this time using a clinical scoring system incorporating tumor size, tumor ulceration, body condition, and mobility (Supplementary Fig. 9). Treated mice survived longer than mice treated with controls (Fig. 5f). These data are consistent with the hypothesis that LNP^IT-mRNA treatment followed by fludarabine phosphate can lead to anti-tumor responses.

Since HNSCC tumors are heterogenous, we tested therapeutic effect in two additional in vivo models: FaDu tumors in immunocompromised NU/J mice (Fig. 6a) and MOC1 murine oral cancers (syngeneic HNSCC) in immunocompetent C57BL/6 mice (Fig. 6b). Tumors treated with the combination therapy were significantly smaller than those in the control groups. Given that MOC1 tumors are characterized by increased MHC-I expression and CD8+ T cell infiltration into the tumor microenvironment⁵⁴ while FaDu tumors were grown in NU/J mice with T cell deficiency⁵⁵, these data provide additional evidence that the approach is active in a number of different host immunologic contexts.

**Fig. 6: LNP^IT-PNP and fludarabine phosphate combination therapy has anti-tumor effects validated across various HNSCC preclinical models.**

To evaluate therapeutic response in distinct human cancer cells, we also performed an in vitro cell killing assay (Fig. 6c). We plated six different HNSCC cell lines: human FaDu, murine MOC1, and four HNSCC patient-derived xenograft lines, termed PDX1, PDX2, PDX3, and PDX4 (Supplementary Fig. 10a). We treated cells with LNP^IT-PNP and MeP-dR (a nucleoside analogue of fludarabine that has served as a prototype for in vitro testing). We showed the expected PNP and nucleoside-dependent cell killing and also observed time-dependent conversion from prodrug to oncolytic drug (Supplementary Fig. 10b). The negative controls, LNP-PNP^IT alone, MeP-dR alone, and no treatment, behaved as expected. In contrast, cells treated with the positive control MeP (the toxic base released following MeP-dR hydrolysis by PNP) led to the expected ablation of cells in culture.

Discussion

LNPs delivering RNA drugs to the liver lead to clinical responses in patients^56,57,58. These data suggest the potential impact of LNP-RNA drugs that target non-hepatic tissues⁵⁹. Yet, evidence also indicates that an LNP optimized for one route of administration is not optimized for another⁶⁰, highlighting the utility of studying how LNPs behave after an intratumoral injection. Here, we used species-agnostic DNA barcoding to test 44 LNPs in tumor-bearing mice. Our results provide early evidence that LNPs formulated with neutral cholesterol may outperform LNPs formulated with cationic cholesterol for intratumoral delivery. This finding is consistent with the hypothesis that negatively charged microenvironments⁶¹, including certain tumor microenvironments^62,63, can sequester positively charged drugs. Our analysis also suggests a second lesson: future screens should take place in multiple tumor models. Here, we evaluated all 44 LNPs in FaDu tumors, leading us to discovery of LNP²⁸ and LNP^IT. When the two lead LNPs were tested individually, they both delivered mRNA in FaDu tumors as predicted. Yet, when the LNPs were tested in PDX tumors, only LNP^IT efficiently delivered mRNA. This second lesson is timely; in the past, testing several dozen LNPs in multiple in vivo tumor models would have required hundreds of animals. DNA barcoding makes this feasible with far fewer. When stereopure C12-200 isoforms were previously screened intravenously, the S-isomers yielded higher expression²⁸, while the most effective intratumoral LNP candidate here was composed of an R-isomer. This highlights the importance of screening via the intended route of administration when selecting LNP candidates for RNA therapeutics⁶⁴.

Complementing the lessons learned as a result of barcode sequencing were the observations made by analyzing LNP-transfected tumors with scRNA-seq. We found it interesting that an unbiased approach evaluating cellular processes identified 28 significant pathways—all of which were related to RNA or protein. These data are consistent with the hypothesis that cancer cells respond to LNPs carrying mRNA by changing the ways in which mRNA and protein are processed. We foresee future work focused on understanding how these processes may be exploited to further improve LNP delivery.

It is also important to acknowledge the limitations of the study. Our applications of human-derived tumor models required testing mice with deficient immune systems. One question that will be answered in future work is whether LNP delivery occurs within other intratumoral immune cell subtypes in mice with intact immune systems and spontaneous tumors. If LNP^IT does transfect immune cells, the delivery system could be considered for immunostimulatory mRNA approaches⁷, such as those that have been tested in patients^3,4,65. A second limitation of the work involves delivery readouts; it is unclear how well delivery in mice predicts delivery in NHPs¹⁹. NHP studies will be particularly important to understand on- and off-target delivery and subsequent toxicity. In our experiments, we did not observe substantial off-target delivery, which likely reduced the chance of systemic toxicity. An additional source of toxicity is F-Ade released from tumor parenchyma. However, previous data suggest that escaped F-Ade would be diluted and metabolized via systemic xanthine oxidase^15,16,66 to non-toxic concentrations. Finally, the expression of PNP, which is a bacterial protein, could elicit an unexpected immune response. While this will require further study, we did not find any evidence suggesting a broad, undesired immune response in our current experiments.

Despite these limitations and need for future work, our results are consistent with the hypothesis that LNPs can deliver a bacterial enzyme that activates a prodrug and leads to anti-tumor responsiveness. We believe the data support additional preclinical work focused on evaluating this anti-cancer approach.

Methods

All experiments and research performed comply with all relevant ethical regulations in accordance with Emory University’s IACUC.

C12-200 synthesis

Stereopure C12-200 ionizable lipids²⁸ were formulated by adding amine (0.2 mmol) and chiral epoxide (1.4 mmol, 7equiv) to a 10 mL reaction vial with magnetic stirring. The vial was sealed, heated to 80 °C, and stirred for 48 hours. The crude mixture was then purified by column chromatography using a silica gel (300–400 mesh) and eluting with a DCM:MeOH:NH₄OH (90:9:1) mixture, affording desired compounds as colorless oils in 53–62 yields. Thin layer chromatography (TLC) was carried out using precoated silica Gel GF plates and visualized using KMnO4 stains. ¹H and ¹³C NMR spectra were recorded on a Bruker AVANCE 400 (400 MHz), and a Bruker AVANCE 600 (600 MHz) spectrometer at 25 °C. All ¹H Chemical shifts (in ppm) were assigned according to CDCl₃ (δ = 7.24 ppm), and all ¹³C NMR was calibrated with CDCl₃ (δ = 77.00 ppm). Coupling constants (J) are reported in hertz (Hz). High‐resolution mass spectra (HRMS) were recorded on LC/MS (Agilent Technologies 1260 Infinity II/6120 Quadrupole) and a time‐of‐flight mass spectrometer by electrospray ionization. A general reaction mechanism can be found in Supplementary Fig. 11. NMR/mass spec data for the ionizable lipids can be found in Supplementary Fig. 12.

Messenger RNA (mRNA) synthesis

E. coli PNP was designed based on ORF sequences. UTRs consisting of a 5’ untranslated region (UTR) from Kozak sequence and a 3’ UTR from mouse alpha-globin (GenBank accession no. NM_001083955) similar to the constructs previously described^67,68 were used. The plasmids were linearized with Not-I HF (New England Biolabs) overnight at 37 °C. Linearized templates were purified by ammonium acetate (Thermo Fisher Scientific) precipitation and resuspended with nuclease-free water. In vitro transcription was performed overnight at 37 °C using the HiScribe T7 Kit (NEB) following the manufacturer’s instructions (N1-methyl-pseudouridine modified). The resulting RNA was treated with DNase I (Aldevron) for 30 min to remove the template and purified using lithium chloride precipitation (Thermo Fisher Scientific). The RNA was heat denatured at 65 °C for 10 min before capping (via Cap-1 structure) using guanylyl transferase and 2′-O-methyltransferase (Aldevron). mRNA was then purified by lithium chloride precipitation, treated with alkaline phosphatase (NEB), and purified again. mRNA concentration was measured using a NanoDrop protocol. mRNA stock concentrations were ~4 mg/ml. Purified mRNA products were analyzed by gel electrophoresis to ensure purity. Messenger RNA encoding anchored nanoluciferase or anchored VHH (aVHH)^19,69 was synthesized such that GPI-anchored VHH and NanoLuc sequences were ordered as a DNA gBlock from Integrated DNA Technologies (IDT) with the same 5’ and 3’ UTRs used for PNP mRNA. Purification of linearized templates, in vitro transcription, DNase I treatment, heat denaturation, and downstream purification with lithium chloride precipitation and NEB were all performed exactly as detailed above for PNP mRNA.

Nanoparticle formulation

Nanoparticles for in vivo screening were formulated using a microfluidic device³⁹ at a flow rate ratio of 3:1 of nucleic acid: lipid phases. Larger batches of LNP^IT with PNP mRNA for preclinical tumor regression studies were formulated using the NanoAssemblr Ignite (Precision Nanosystems). DNA barcodes and/or mRNA were diluted in 10 mM citrate buffer (Teknova). DNA barcodes were purchased from IDT. PEGs, cholesterols, and helper lipids were diluted in 100% ethanol and purchased from Avanti Lipids. Citrate and ethanol phases were combined in a microfluidic device or the NanoAssemblr Ignite using glass (Hamilton Company) or plastic (BD) syringes, respectively, at a flow rate ratio of 3:1.

Nanoparticle characterization

The diameter and polydispersity of all LNPs were assessed via dynamic light scattering (DLS) (DynaPro Plate Reader III, Wyatt). Four microliters of LNPs were diluted in 96 microliters of sterile 1X PBS. LNPs were sterile purified using a 0.22 μm filter and injected only if three criteria were met: 20 nm <diameter <200 nm, and correlation function with 1 inflection point (monodisperse distribution). For screens, particles that met these cut-offs were pooled. Particles were dialyzed in 20 kD dialysis cassettes (Thermo Scientific). NanoDrop (Thermo Scientific) was used to assess nucleic acid concentration.

Encapsulation assay

Encapsulation was measured according to the Precision NanoSystems RiboGreen assay protocol. In duplicates, 50 μL of 6 ng/μL LNP (diluted in TE) was added to 50 μL of 1X TE (Thermo Fisher) or 50 μL of a solution containing a 1:50 dilution of Triton X-100 (Sigma Aldrich). After 10 minutes of incubation at 37 °C, 100 μL of 1:100 of RiboGreen reagent (Thermo Fisher) was added to each well. The fluorescence was quantified using a plate reader (BioTek Synergy H4 Hybrid) at an excitation wavelength of 485 nm and an emission wavelength of 528 nm.

Zeta potential

A Malvern Zetasizer Nano Z was used to measure the zeta potential of LNPs. Eight hundred microliters of the nanoparticles were loaded into a Malvern disposable folded capillary cell and software run under the following settings: material refractive index of 1.4, absorbance of 0.01, dispersant viscosity of 0.882cp, refractive index of 1.33, and dielectric constant of 79.

TNS assay

A stock solution of 10 mM HEPES (Sigma Aldrich), 10 mM MES (Sigma Aldrich), 10 mM sodium acetate (Sigma Aldrich), and 140 mM sodium chloride (Sigma Aldrich) was prepared and adjusted with hydrogen chloride and sodium hydroxide to the following pH values: 3, 4, 5, 6, 7, 8, 9, 10. Using four replicates for each pH, 140 μL pH-adjusted buffer was added to a 96-well plate, followed by treatment with 5 μL of 2-(p-toluidino)-naphthalene-6-sulfonic acid (60 μg/mL). Five microliters of LNP were next added to each well, followed by incubation for 5 minutes with constant shaking at 300 rpm. Fluorescence absorbance was quantified with excitation wavelength of 325 nm and an emission wavelength of 435 nm using a plate reader (BioTek Synergy H4 Hybrid).

Animal experiments

All animal experiments were performed in accordance with Emory University’s IACUC. All animals were housed in the Emory University Winship Cancer Center Animal Facility. NU/J mice (The Jackson Laboratory, stock #002019, 6-8 weeks, female), NOD.Cg-Prkd^scid, IL2rg^tm1Wjl/SzJ (NSG) mice (The Jackson Laboratory, stock #005557, 6-8 weeks, female), and C57BL/6J mice (The Jackson Laboratory, stock #000664, 6-8 weeks, female) were used throughout this article. At least N = 2–6 female mice/group were used for all studies, unless noted otherwise. NSG and NU/J mice were housed in a microisolator on static racks in animal rooms maintained at 22 ± 2 °C, and 40–50% humidity, with 12-hour light/dark cycle. Both NSG and NU/J mice were fed with sterile Rodent Lab Diet (Laboratory Rodent Diet no. 5R53) and sterile water. The C57BL/6J mice were housed in a microisolator in animal rooms maintained at 22 ± 2 °C and 40-50% humidity, with a 12-hour light/dark cycle. C57BL/6J mice were fed with Rodent Lab Diet (Laboratory Rodent Diet no. 5053).

Ethics statement

All animal experiments were performed in accordance with Emory University’s IACUC. In all cases, the animal was euthanized when the maximal tumor diameter reached 20 mm (TBS of 3, see Tumor Volume Measurement and Clinical Scoring section below). PDX models were obtained from the NCI Patient-Derived Models Repository (PDMR; https://pdmr.cancer.gov; NCI-Frederick, Frederick National Laboratory for Cancer Research, Frederick, MD). Specimens were collected under NCI-sponsored tissue procurement protocols with institutional review board approval; investigators obtained written informed consent from each participant for the use of their delinked specimens to genetically characterize and generate patient-derived models and to make these models available to researchers along with limited clinical information.

FaDu, PDX, and MOC1 tumor inoculation

FaDu head and neck squamous cell carcinoma cells (ATCC, catalog #HTB-43) were subcutaneously injected into NU/J mice to induce bilateral xenograft hindleg tumors (1 × 10⁶ cells per unilateral inoculation). Tumor volumes were monitored twice a week starting on day 7, once FaDu masses reached between 150 and 250 mm³ in size. Patient-derived xenograft (PDX) cells were obtained as mixed/crude (non-clonal) tumor samples (e.g., 328373-195-R-J1-PDC) extracted from a lateral neck mass according to protocols at the National Cancer Institute (NCI) patient-derived models repository (PDMR) (https://pdmr.cancer.gov/). Cancer models from PDMR undergo rigorous quality control profiling and exhibit suitable doubling time and carcinogenic pathology for assessing antitumor activity in preclinical studies with reliable fidelity. NSG mice were injected subcutaneously with 1 × 10⁷ PDX cells, as recommended by PDMR/NCI to model an aggressive form of HNSCC. The additional 3 PDX lines tested later (PDX2, PDX3, PDX4; Supplementary Fig. 10a) were also obtained from the NCI PDMR. MOC1 cells (Kerafast, catalog #EWL001-FP) were subcutaneously injected into C57BL/6 J mice to induce unilateral xenograft hindleg tumors (1 × 10⁶ cells per unilateral inoculation). PDX and MOC1 tumor volumes were monitored twice a week starting on day 7 after tumor inoculation, and mice underwent experimental procedures when masses reached a size between 150 and 250 mm³. All cell lines used in vivo tested negative for mycoplasma contamination, as required by Emory University School of Medicine’s IACUC. Specifically, PCR-based mycoplasma testing was performed on FaDu, MOC1, and PDX1 cell lines, since these 3 lines were used in animal studies. Authentication of cell lines was performed by the commercial provider ATCC in the case of FaDu and MOC1 cells, per its commercial guidelines, and the National Cancer Institute (NCI) in the case of the PDX cells, per the PDMR guidelines.

Cell isolation and staining

Cells were isolated from digested tissues, which were harvested 16 hours after intratumoral injection with LNPs, unless otherwise noted. Mice were perfused with 20 mL of 1X PBS through the right atrium when harvesting non-tumor tissues. Tumors or livers were finely minced and transferred into a digestive enzyme solution with collagenase type I (Sigma-Aldrich), collagenase IV (Sigma-Aldrich), collagenase XI (Sigma-Aldrich), and hyaluronidase (Sigma-Aldrich), and incubated at 37 °C with controlled shaking at 550 rpm for 45 minutes. Cell suspensions were then filtered through a 70 μm mesh. Next, cells were stained to identify specific cell populations and gate out lysed cells and red blood cells. The target CD47+ aVHH+ cell populations were then sorted using the BD FACS Fusion cell sorters in the Georgia Institute of Technology Cellular Analysis Core. The antibody clones used were anti-CD31 (390, BioLegend), anti-CD45.2 (104, BioLegend), anti-hCD47 (CC2C6, BioLegend), and MonoRab^TM rabbit anti-camelid VHH antibody iFluor647 (A01994, GenScript), all at 1:500 dilution ratios in 10% fetal bovine serum (FBS) in PBS solutions, where the single-cell suspension samples were dissolved for flow cytometry. Representative flow gates are located in Supplementary Fig. 1, Supplementary Fig. 13.

PCR amplification

We amplified and prepared samples for sequencing following a one-step PCR protocol⁷⁰. Specifically, 1 μL of primers (5 mM for final reverse/forward, 0.5 mM for base forward) were added to 5 μL of Kapa HiFi 2X master mix and 4 μL template DNA/water. If no clear bands were produced by the PCR reaction, we adjusted and optimized the primer concentrations, DNA template input, PCR temperature, and number of cycles for individual samples.

Deep sequencing and normalization

Illumina deep sequencing was performed on an Illumina MiniSeq^TM. Primers were designed based on Nextera XT adapter sequences. Counts for each particle, per tissue, were normalized to the barcoded LNP mixture injected into each mouse. This ‘input’ DNA provided DNA counts and was used to normalize DNA counts from cells and tissues. Sequencing results were processed using a custom Python-based tool to extract raw barcode counts from individual samples. Raw counts were then normalized with an R script prior to further assessment. Statistical analysis was conducted using GraphPad Prism 8.

Whole organ imaging

Tissues were isolated 24 hours after administration of LNPs, unless otherwise noted. To measure luminescence, mice were sacrificed and organs collected, followed by treatment with Nano-Glo Luciferase Assay Substrate (Promega) for 5 minutes before being placed on solid black paper for imaging. Luminescence was measured using an IVIS system (PerkinElmer) and quantified using Living Image software (PerkinElmer).

Single-cell RNA sequencing (scRNA-seq)

Sixteen hours following IT injection of aVHH mRNA-encapsulating LNP²⁸ or LNP^IT to PDX tumors, masses were dissected from the animals. After tissue digestion, cells were resuspended in RoboSep buffer (Stemcell Technologies) for further processing. Whole transcriptome analyses were performed using the BD Rhapsody Single-Cell Analysis System (BD Biosciences), following the manufacturer’s protocol. Briefly, dead cells and red blood cells (RBCs) were depleted by using EasySep^TM dead cell (Annexin V) and RBC (anti-TER119) removal kit (Stemcell Technologies). Cell viability and numbers were recorded for each sample, followed by tagging with TotalSeq^TM anti-human Hashtag antibody (5 µg/mL; [TotalSeq-A0251 (BioLegend 394601), TotalSeq-A0252 (BioLegend 394603), TotalSeq-A0253 (BioLegend 394605), TotalSeq-A0257 (BioLegend 394613), TotalSeq-A0258 (BioLegend 394615), TotalSeq-A0259 (BioLegend 394617)]) and oligo-tagged anti-VHH antibody (5 µg/mL; A01860, GenScript). For the preparation of oligo-tagged anti-VHH antibody, 5′ DBCO-modified oligonucleotide (CCTTGGCACCCGAGAATTCCAAAGTATGCCCTACGABAAAAAAAAAAAAAAAAAAAAAAAAAAAA*A*A, where * indicates phosphothioate bonds; GenScript) was conjugated to azide-modified rabbit anti-camelid VHH antibody (clone 96A3F5; GenScript) by click chemistry (GenScript). Samples were then pooled at the same ratio and a BD Rhapsody cartridge loaded with 40,000 cells. cDNA libraries were prepared using the BD Rhapsody Whole Transcriptome Analysis Amplification Kit following the BD Rhapsody System mRNA Whole Transcriptome Analysis (WTA) and Sample Tag Library Preparation protocol (BD Biosciences). Final libraries were quantified using a Qubit Fluorometer and the quality checked with BioAnalyzer (Agilent) by size distribution. The data were processed using STARsolo (v 2.9.7) for the RNA mapping and counting⁷¹. All samples were mapped to GRCh38.p14, and only exonic regions were counted. Output files were next loaded into Seurat (v 4.0.4) and cells log normalized to a scale factor of 10,000, then scaled using a linear transformation⁷². DoubletFinder (v3) (Ref. ⁷³), which is an algorithm for scRNA-seq datasets that predicts doublets according to each real cell’s proximity in gene expression space to artificial doublets resulting from averaging the transcriptional profile of randomly chosen cell pairs, was used to identify doublets. This was followed by PCA dimensional reduction and t-SNE clustering and further analyzed in BBrowserX⁷⁴. In BBrowserX, gene expression profiles were compared within cell types of interest. Reactome pathway analysis was performed using Reactome database 85⁵². The pathway expression levels are shown as z-score normalized values.

PNP enzymatic quantification assay

PDX tumors harvested from the flanks of NSG mice were flash frozen and stored at −80 °C until preparation of extract. Crude extracts were prepared⁵³ By homogenizing fixed tumor samples in 0.01 M HEPES buffer and then incubated on ice for 15 min. Samples were then sonicated three times for 30 seconds each, centrifuged at 100,000 g for 60 minutes, and dialyzed with 50 volumes of 0.1 M HEPES buffer (1 mM DTT and 20% glycerol at pH 7.4). These crude extracts were then incubated at 25 °C with 50 mM potassium phosphate, 100 μM 6-methylpurine-2’-deoxyriboside (MeP-dR), 100 mM HEPES buffer (pH 7.4), and an amount of enzyme that yielded a linear signal during the incubation period. Reactions were stopped by boiling. The formation of 6-methylpurine (MeP) was monitored using reverse-phase HPLC (PerkinElmer). The level of MeP produced from MeP-dR is directly related to the amount of PNP expressed in the PDX cells. The PNP activity is presented as nmoles of MeP produced per mg protein per hour.

Spin concentration of PNP mRNA-carrying LNP^IT

LNP^IT-PNP formulated in the NanoAssemblr Ignite was diluted in molecular grade sterile 10 mM tris buffer pH 7.4 (Calbiochem) at a 1:40 ratio of LNP:tris buffer dilution. This was then poured into Amicon Ultra-15mL centrifugal filter tubes of 100 kDa membranes (Millipore) and centrifuged at 1,000 G for 30-minute cycles until the whole LNP in tris solution was brought to the desired concentration.

In vitro cell killing assay

MOC1 (3 × 10⁴ cells/well), FADU (2×10⁵ cells/well) or PDX (8 × 10⁴ cells/well) cells were seeded into 24-well plates, and LNP^IT carrying PNP mRNA was added 24 h after plating at 1 µg/well. MeP (6-methylpurine) or MeP-dR (9-(2-deoxy-beta-D-ribofuranosyl)-6-methylpurine) was added at 100 µM within 20 h after LNP transfection. Cells were monitored for 4 days and stained with 0.1% crystal violet to evaluate cell survival. During the assay, dead or detached cells were washed away, while attached live cells were stained with crystal violet. Clear wells indicate >95% cell killing.

In vitro PNP activity measurement

MOC1 (Kerafast, catalog #EWL001-FP), FADU (ATCC, catalog #HTB-43), or PDX cells (NCI PDMR, PDX1 – 328373-195-R-J1-PDC, PDX2 – 929823-356-R-J2-PDC, PDX3 – 958767-090-R-J1-PDC, PDX4 – 845751-090-R-J2-PDC) were seeded into 6-well plates, and LNP^IT carrying PNP mRNA at 4.5 µg/well was added 24 h after plating. The following day, cells were incubated for 4 hours with 100 μM MeP‐dR (substrate for E. coli PNP, see above), and formation of MeP (cytotoxin) was measured using reverse phase HPLC (PerkinElmer).

Tumor volume measurement and clinical scoring

Tumor volume measurements were performed using a digital caliper (Fisherbrand) and initiated 7 days after PDX tumor inoculation. Width (W, in mm) and length (L, in mm) measurements were taken, and tumor volume (V, in mm³) was calculated using the following formula:

$$V=\frac{L\,x\,{W}^{2}}{2}$$

Once tumor sizes reached between 150 and 250 mm³ in size, animals were divided into experimental groups and treated as noted in the text. Tumor volume and body weight were measured for each mouse twice a week for the remainder of the study or until the animals reached endpoint conditions and were euthanized. Clinical scores were calculated based on the tumor burden (size and ulceration) and body condition and mobility (Supplementary Fig. 9). Tumor diameter was measured, and a tumor burden score (TBS) was assigned following Emory University’s IACUC policy 304. A TBS of 0 was assigned when the tumor diameter was less than 18 mm, TBS 1 when the tumor diameter was between 18 and 20 mm, and TBS 3 when the diameter was larger than 20 mm. When the maximal TBS of 3 was reached, the animal was euthanized.

Statistics & reproducibility

All experiments were done at least in triplicate (n = 3). For all tumor regression studies, n = 6 was used for each group and mice were randomized, ensuring that the average tumor size was similar across all groups when beginning treatment. PDX in vivo studies were performed twice to ensure the reproducibility of the therapeutic approach, which was validated further in other in vivo tumor models. No statistical method was used to predetermine sample size. No data were excluded from the analyses. The Investigators were not blinded to allocation during experiments and outcome assessment. All measurements were taken from distinct samples; the same sample was NOT measured repeatedly in any instance.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

The scRNA-seq data generated in this study have been deposited in the SRA database [https://www.ncbi.nlm.nih.gov/sra] under BioProject ID PRJNA1090507, BioSamples accession numbers SAMN40568326 and SAMN40568327, SRA numbers SRS20810744 [https://www.ncbi.nlm.nih.gov/biosample/40568326] and SRS20810746 [https://www.ncbi.nlm.nih.gov/biosample/40568327]. Other data generated in this study are provided in the main figures, Supplementary Information or the Source data file. Source data are provided with this paper.

Code availability

All code used to analyze the data is available (https://github.com/Jack-Feldman/barcode_count).

References

Baden, L. R. et al. Efficacy and safety of the mRNA-1273 SARS-CoV-2 vaccine. N. Engl. J. Med. 384, 403–416 (2020).
Polack, F. P. et al. Safety and efficacy of the BNT162b2 mRNA Covid-19 vaccine. N. Engl. J. Med. 383, 2603–2615 (2020).
Article CAS PubMed Google Scholar
Patel, M. R. et al. A phase I study of mRNA-2752, a lipid nanoparticle encapsulating mRNAs encoding human OX40L, IL-23, and IL-36γ, for intratumoral (iTu) injection alone and in combination with durvalumab. J. Clin. Oncol. 38, 3092 (2020).
Article Google Scholar
Moderna in Dose Escalation Study of mRNA-2752 for Intratumoral Injection to Participants in Advanced Malignancies - NCT03739931 (https://www.clinicaltrials.gov/ct2/show/NCT03739931?term=NCT03739931&draw=2&rank=1; 2018).
Huayamares, S. G., Loughrey, D., Kim, H., Dahlman, J. E. & Sorscher, E. J. Nucleic acid-based drugs for patients with solid tumours. Nat. Rev. Clin. Oncol. 21, 407–427 (2024).
Liu, J.-Q. et al. Intratumoral delivery of IL-12 and IL-27 mRNA using lipid nanoparticles for cancer immunotherapy. J. Control. Rel. 345, 306–313 (2022).
Article CAS Google Scholar
Li, Y. et al. Multifunctional oncolytic nanoparticles deliver self-replicating IL-12 RNA to eliminate established tumors and prime systemic immunity. Nat. Cancer 1, 882–893 (2020).
Article CAS PubMed PubMed Central Google Scholar
Yang, J. et al. Intratumoral delivered novel circular mRNA encoding cytokines for immune modulation and cancer therapy. Mol. Ther. Nucleic Acids 30, 184–197 (2022).
Article CAS PubMed PubMed Central Google Scholar
Rosenblum, D. et al. CRISPR-Cas9 genome editing using targeted lipid nanoparticles for cancer therapy. Sci. Adv. 6, eabc9450 (2020).
Johnson, D. E. et al. Head and neck squamous cell carcinoma. Nat. Rev. Dis. Prim. 6, 92 (2020).
Article PubMed Google Scholar
Barsouk, A., Aluru, J. S., Rawla, P., Saginala, K. & Barsouk, A. Epidemiology, Risk Factors, and Prevention of Head and Neck Squamous Cell Carcinoma. Med. Sci. 11, 42 (2023).
Carvalho, A. L., Nishimoto, I. N., Califano, J. A. & Kowalski, L. P. Trends in incidence and prognosis for head and neck cancer in the United States: a site-specific analysis of the SEER database. Int J. Cancer 114, 806–816 (2005).
Article CAS PubMed Google Scholar
Behbahani, T. E., Rosenthal, E. L., Parker, W. B. & Sorscher, E. J. Intratumoral generation of 2-fluoroadenine to treat solid malignancies of the head and neck. Head. Neck 41, 1979–1983 (2019).
Article PubMed PubMed Central Google Scholar
Elgemeie, G. H. & Mohamed-Ezzat, R. A. In New Strategies Targeting Cancer Metabolism. (eds. Elgemeie G. H. & Mohamed-Ezzat R. A.) 69–105 (Elsevier, 2022).
Parker, W. B. & Sorscher, E. J. Use of E. coli purine nucleoside phosphorylase in the treatment of solid tumors. Curr. Pharm. Des. 23, 7003-7024 (2017).
Rosenthal, E. L. et al. Phase I dose-escalating trial of Escherichia coli purine nucleoside phosphorylase and fludarabine gene therapy for advanced solid tumors. Ann. Oncol. 26, 1481–1487 (2015).
Article CAS PubMed PubMed Central Google Scholar
Fausther-Bovendo, H. & Kobinger, G. P. Pre-existing immunity against Ad vectors: humoral, cellular, and innate response, what’s important? Hum. Vaccin Immunother. 10, 2875–2884 (2014).
Article PubMed PubMed Central Google Scholar
Ahi, Y. S., Bangari, D. S. & Mittal, S. K. Adenoviral vector immunity: its implications and circumvention strategies. Curr. Gene Ther. 11, 307–320 (2011).
Article CAS PubMed PubMed Central Google Scholar
Hatit, M. Z. C. et al. Species-dependent in vivo mRNA delivery and cellular responses to nanoparticles. Nat. Nanotechnol. 17, 310–318 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Cho, K. J., Park, E. J., Kim, M. S. & Joo, Y. H. Characterization of FaDu-R, a radioresistant head and neck cancer cell line, and cancer stem cells. Auris Nasus Larynx 45, 566–573 (2018).
Article PubMed Google Scholar
Parker, W. B. et al. The use of Trichomonas vaginalis purine nucleoside phosphorylase to activate fludarabine in the treatment of solid tumors. Cancer Chemother. Pharm. 85, 573–583 (2020).
Article CAS Google Scholar
Paunovska, K. et al. A direct comparison of in vitro and in vivo nucleic acid delivery mediated by hundreds of nanoparticles reveals a weak correlation. Nano Lett. 18, 2148–2157 (2018).
Article ADS CAS PubMed PubMed Central Google Scholar
Akinc, A. et al. Targeted delivery of RNAi therapeutics with endogenous and exogenous ligand-based mechanisms. Mol. Ther. 18, 1357–1364 (2010).
Article CAS PubMed PubMed Central Google Scholar
Jayaraman, M. et al. Maximizing the potency of siRNA lipid nanoparticles for hepatic gene silencing in vivo. Angew. Chem. Int Ed. Engl. 51, 8529–8533 (2012).
Article CAS PubMed PubMed Central Google Scholar
Dong, Y. et al. Lipopeptide nanoparticles for potent and selective siRNA delivery in rodents and nonhuman primates. Proc. Natl Acad. Sci. 111, 3955–3960 (2014).
Article ADS CAS PubMed PubMed Central Google Scholar
Schoenmaker, L. et al. mRNA-lipid nanoparticle COVID-19 vaccines: Structure and stability. Int. J. Pharm. 601, 120586 (2021).
Article CAS PubMed PubMed Central Google Scholar
Love, K. T. et al. Lipid-like materials for low-dose, in vivo gene silencing. Proc. Natl Acad. Sci. 107, 1864–1869 (2010).
Article ADS CAS PubMed PubMed Central Google Scholar
Da Silva Sanchez, A. J. et al. Substituting racemic ionizable lipids with stereopure ionizable lipids can increase mRNA delivery. J. Control. Release 353, 270–277 (2023).
Article PubMed Google Scholar
Huayamares, S. G. et al. High-throughput screens identify a lipid nanoparticle that preferentially delivers mRNA to human tumors in vivo. J. Control. Release 357, 394–403 (2023).
Article CAS PubMed PubMed Central Google Scholar
Patel, S. et al. Naturally-occurring cholesterol analogues in lipid nanoparticles induce polymorphic shape and enhance intracellular delivery of mRNA. Nat. Commun. 11, 983 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Patel, S. et al. Boosting intracellular delivery of lipid nanoparticle-encapsulated mRNA. Nano Lett. 17, 5711–5718 (2017).
Article ADS CAS PubMed PubMed Central Google Scholar
Paunovska, K. et al. Nanoparticles containing oxidized cholesterol deliver mRNA to the liver microenvironment at clinically relevant doses. Adv. Mater. (Deerfield Beach, Fla.), e1807748 (2019).
Kauffman, K. J. et al. Rapid, single-cell analysis and discovery of vectored mRNA transfection in vivo with a loxP-flanked tdTomato reporter mouse. Mol. Ther. Nucleic Acids 10, 55–63 (2018).
Article CAS PubMed Google Scholar
Cheng, Q. et al. Selective organ targeting (SORT) nanoparticles for tissue-specific mRNA delivery and CRISPR–Cas gene editing. Nat. Nanotechnol. 15, 313–320 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
LoPresti, S. T., Arral, M. L., Chaudhary, N. & Whitehead, K. A. The replacement of helper lipids with charged alternatives in lipid nanoparticles facilities targeted mRNA delivery to the spleen and lungs. J. Control. Rel. 345, 819-831 (2022).
Radmand, A. et al. The transcriptional response to lung-targeting lipid nanoparticles in vivo. Nano Lett. 23, 993–1002 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Tiwari, P. M. et al. Engineered mRNA-expressed antibodies prevent respiratory syncytial virus infection. Nat. Commun. 9, 3999 (2018).
Article ADS PubMed PubMed Central Google Scholar
Sago, C. D. et al. Modifying a commonly expressed endocytic receptor retargets nanoparticles in Vivo. Nano Lett. 18, 7590-7600 (2018).
Belliveau, N. M. et al. Microfluidic synthesis of highly potent limit-size lipid nanoparticles for in vivo delivery of siRNA. Mol. Ther. Nucleic Acids 1, e37 (2012).
Article PubMed PubMed Central Google Scholar
Lokugamage, M. P., Sago, C. D. & Dahlman, J. E. Testing thousands of nanoparticles in vivo using DNA barcodes. Curr. Opin. Biomed. Eng. 7, 1–8 (2018).
Article PubMed PubMed Central Google Scholar
Radmand, A. et al. Cationic cholesterol-dependent LNP delivery to lung stem cells, the liver, and heart. Proc. Natl Acad. Sci. 121, e2307801120 (2024).
Article CAS PubMed PubMed Central Google Scholar
Panaampon, J., Sasamoto, K., Kariya, R. & Okada, S. Establishment of nude mice lacking NK cells and their application for human tumor Xenografts. Asian Pac. J. Cancer Prev. 22, 1069–1074 (2021).
Article CAS PubMed PubMed Central Google Scholar
Pressnall, M. M., Huayamares, S. G. & Berkland, C. J. Immunostimulant complexed with polylysine limits transport and maintains immune cell activation. J. Pharm. Sci. 109, 2836–2846 (2020).
Article CAS PubMed Google Scholar
Pressnall, M. M. et al. Glatiramer acetate enhances tumor retention and innate activation of immunostimulants. Int. J. Pharm. 605, 120812 (2021).
Article CAS PubMed Google Scholar
Hatit, M. Z. C. et al. Nanoparticle stereochemistry-dependent endocytic processing improves in vivo mRNA delivery. Nat. Chem. 15, 508-515 (2023).
Choi, J.-H. et al. Single-cell transcriptome profiling of the stepwise progression of head and neck cancer. Nat. Commun. 14, 1055 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Stoeckius, M. et al. Simultaneous epitope and transcriptome measurement in single cells. Nat. Methods 14, 865–868 (2017).
Article CAS PubMed PubMed Central Google Scholar
Sharaf, K. et al. Discrimination of cancer stem cell markers ALDH1A1, BCL11B, BMI-1, and CD44 in different tissues of HNSCC patients. Curr. Oncol. 28, 2763–2774 (2021).
Article PubMed PubMed Central Google Scholar
Wu, Z. et al. SLA2 is a prognostic marker in HNSCC and correlates with immune cell infiltration in the tumor microenvironment. Eur. Arch. Otorhinolaryngol. 281, 427–440 (2024).
Article PubMed Google Scholar
You, G. R. et al. MYH9 Facilitates cell invasion and radioresistance in head and neck cancer via modulation of cellular ROS levels by activating the MAPK-Nrf2-GCLC Pathway. Cells 11, 2855 (2022).
Lee, M. K., Zhang, X., Kim, H. J. & Hwang, Y. S. Peroxiredoxin 5 is involved in cancer cell invasion and tumor growth of oral squamous cell carcinoma. Oral. Dis. 29, 423–435 (2023).
Article PubMed Google Scholar
Griss, J. et al. ReactomeGSA - Efficient multi-omics comparative pathway analysis. Mol. Cell Proteom. 19, 2115–2125 (2020).
Article CAS Google Scholar
Gadi, V. K. et al. In vivo sensitization of ovarian tumors to chemotherapy by expression of E. coli purine nucleoside phosphorylase in a small fraction of cells. Gene Ther. 7, 1738–1743 (2000).
Article CAS PubMed Google Scholar
Judd, N. P., Allen, C. T., Winkler, A. E. & Uppaluri, R. Comparative analysis of tumor-infiltrating lymphocytes in a syngeneic mouse model of oral cancer. Otolaryngol. Head. Neck Surg. 147, 493–500 (2012).
Article PubMed PubMed Central Google Scholar
Hoffman, R. M. In Reference Module in Life Sciences (Elsevier, 2017).
Adams, D. et al. Patisiran, an RNAi therapeutic, for hereditary transthyretin amyloidosis. N. Engl. J. Med. 379, 11–21 (2018).
Article CAS PubMed Google Scholar
Gillmore, J. D. et al. CRISPR-Cas9 in vivo gene editing for Transthyretin Amyloidosis. N. Engl. J. Med. 385, 493–502 (2021).
Article CAS PubMed Google Scholar
Intellia Therapeutics Presents New Interim Data from First-in-Human Study of NTLA-2002 for the Treatment of Hereditary Angioedema (HAE) at the American College of Allergy, Asthma & Immunology 2022 Annual Scientific Meeting. https://ir.intelliatx.com/news-releases/news-release-details/intellia-therapeutics-presents-new-interim-data-first-human (2022).
Loughrey, D. & Dahlman, J. E. Non-liver mRNA delivery. Acc. Chem. Res. 55, 13–23 (2022).
Article CAS PubMed Google Scholar
Lokugamage, M. P. et al. Optimization of lipid nanoparticles for the delivery of nebulized therapeutic mRNA to the lungs. Nat. Biomed. Eng. 5, 1059–1068 (2021).
Article CAS PubMed PubMed Central Google Scholar
Zuckerman, J. E., Choi, C. H. J., Han, H. & Davis, M. E. Polycation-siRNA nanoparticles can disassemble at the kidney glomerular basement membrane. Proc. Natl Acad. Sci. USA 109, 3137–3142 (2012).
Article ADS CAS PubMed PubMed Central Google Scholar
Huang, A. et al. Human intratumoral therapy: Linking drug properties and tumor transport of drugs in clinical trials. J. Control. Release 326, 203–221 (2020).
Article CAS PubMed Google Scholar
Huayamares, S. G. et al. Constructing a biomaterial to simulate extracellular drug transport in solid tumors. Macromol. Biosci. 20, 2000251 (2020).
Article CAS Google Scholar
Huayamares, S. G., Zenhausern, R. & Loughrey, D. Nanocarriers for inhaled delivery of RNA therapeutics. Curr. Res. Biotechnol. 7, 100200 (2024).
Article CAS Google Scholar
BioNTech in A Clinical Trial Investigating the Safety, Tolerability, and Therapeutic Effects of BNT113 in Combination With Pembrolizumab Versus Pembrolizumab Alone for Patients With a Form of Head and Neck Cancer Positive for Human Papilloma Virus 16 and Expressing the Protein PD-L1 (AHEAD-MERIT) - NCT04534205 (https://www.clinicaltrials.gov/ct2/show/NCT04534205?term=NCT04534205&draw=2&rank=1; 2020).
NCT03754933: Safety and Efficacy of Repeat Administration of Ad/PNP and Fludarabine Phosphate in Patients With Local Head/Neck Cancer. https://clinicaltrials.gov/ct2/show/NCT03754933 (2018).
Beyersdorf, J. P. et al. Robust, durable gene activation in vivo via mRNA-encoded activators. ACS Nano 16, 5660–5671 (2022).
Article CAS PubMed PubMed Central Google Scholar
Blanchard, E. L. et al. Treatment of influenza and SARS-CoV-2 infections via mRNA-encoded Cas13a in rodents. Nat. Biotechnol. 39, 717–726 (2021).
Article CAS PubMed Google Scholar
Lindsay, K. E. et al. Aerosol delivery of synthetic mRNA to vaginal mucosa leads to durable expression of broadly neutralizing antibodies against HIV. Mol. Ther.: J. Am. Soc. Gene Ther. 28, 805–819 (2020).
Article CAS Google Scholar
Ni, H. et al. Piperazine-derived lipid nanoparticles deliver mRNA to immune cells in vivo. Nat. Commun. 13, 4766 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Benjamin, K., Dinar, Y. & Alexander, D. STARsolo: accurate, fast and versatile mapping/quantification of single-cell and single-nucleus RNA-seq data. bioRxiv, 2021.2005.2005.442755 (2021).
Hao, Y. et al. Integrated analysis of multimodal single-cell data. Cell 184, 3573–3587.e3529 (2021).
Article CAS PubMed PubMed Central Google Scholar
McGinnis, C. S., Murrow, L. M. & Gartner, Z. J. DoubletFinder: Doublet detection in single-cell RNA Sequencing Data Using Artificial Nearest Neighbors. Cell Syst. 8, 329–337.e324 (2019).
Article CAS PubMed PubMed Central Google Scholar
Tri, L. et al. BBrowser: Making single-cell data easily accessible. bioRxiv, 2020.2012.2011.414136 (2020).

Download references

Acknowledgements

The graphical schematics were generated with BioRender. We would like to thank the NCI Patient-Derived Models Repository (PDMR; https://pdmr.cancer.gov; NCI-Frederick, Frederick National Laboratory for Cancer Research, Frederick, MD) for help obtaining materials. This research was funded by the National Institutes of Health (R01DE026941, awarded to E.J.S. and J.E.D.) and the Georgia Research Alliance.

Author information

Authors and Affiliations

Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University School of Medicine, Atlanta, GA, USA
Sebastian G. Huayamares, Liming Lian, Hyejin Kim, Ryan Zenhausern, Melissa P. Lokugamage, David Loughrey, Hannah E. Peck, Elisa Schrader Echeverri, Aram Shajii, Andrea Li, Karen E. Tiegreen, Philip J. Santangelo & James E. Dahlman
Department of Pediatrics, Emory University, Atlanta, GA, USA
Regina Rab, Bhagelu R. Achyut & Eric J. Sorscher
Winship Cancer Institute, Emory University, Atlanta, GA, USA
Yuning Hou, Bhagelu R. Achyut, Melissa Gilbert Ross & Eric J. Sorscher
Petit Institute for Bioengineering and Biosciences, Georgia Institute of Technology, Atlanta, GA, USA
Afsane Radmand & Alejandro J. Da Silva Sanchez
Department of Chemical Engineering, Georgia Institute of Technology, Atlanta, GA, USA
Afsane Radmand & Alejandro J. Da Silva Sanchez

Authors

Sebastian G. Huayamares
View author publications
Search author on:PubMed Google Scholar
Liming Lian
View author publications
Search author on:PubMed Google Scholar
Regina Rab
View author publications
Search author on:PubMed Google Scholar
Yuning Hou
View author publications
Search author on:PubMed Google Scholar
Afsane Radmand
View author publications
Search author on:PubMed Google Scholar
Hyejin Kim
View author publications
Search author on:PubMed Google Scholar
Ryan Zenhausern
View author publications
Search author on:PubMed Google Scholar
Bhagelu R. Achyut
View author publications
Search author on:PubMed Google Scholar
Melissa Gilbert Ross
View author publications
Search author on:PubMed Google Scholar
Melissa P. Lokugamage
View author publications
Search author on:PubMed Google Scholar
David Loughrey
View author publications
Search author on:PubMed Google Scholar
Hannah E. Peck
View author publications
Search author on:PubMed Google Scholar
Elisa Schrader Echeverri
View author publications
Search author on:PubMed Google Scholar
Alejandro J. Da Silva Sanchez
View author publications
Search author on:PubMed Google Scholar
Aram Shajii
View author publications
Search author on:PubMed Google Scholar
Andrea Li
View author publications
Search author on:PubMed Google Scholar
Karen E. Tiegreen
View author publications
Search author on:PubMed Google Scholar
Philip J. Santangelo
View author publications
Search author on:PubMed Google Scholar
Eric J. Sorscher
View author publications
Search author on:PubMed Google Scholar
James E. Dahlman
View author publications
Search author on:PubMed Google Scholar

Contributions

S.G.H., P.J.S., E.J.S. and J.E.D. conceptualized and designed experiments, which were performed by S.G.H., L.L., R.R., Y.H., A.R., H.K., R.Z., B.R.A., M.G.R., M.P.L., D.L., H.E.P., E.S.E., A.J.D.S.S., A.S. and A.L.; data was analyzed by S.G.H., R.R., Y.H., M.G.R., H.K., E.J.S. and J.E.D.; S.G.H., K.E.T., E.J.S. and J.E.D. wrote the paper, which was reviewed by all other authors.

Corresponding authors

Correspondence to Eric J. Sorscher or James E. Dahlman.

Ethics declarations

Competing interests

S.G.H., R.R., B.R.A., M.P.L., P.J.S., E.S. and J.E.D. have filed intellectual property related to this manuscript (PCT/US2023/033245). J.E.D. is an advisor to Readout Capital, Edge Animal Health, and Nava Therapeutics. E.J.S. has an ownership interest in PNP Therapeutics, Inc., and serves on the Board of Directors for the company, which develops products used in cancer research. Dr. Sorscher is also an inventor of technology being evaluated in studies described by this report. The terms of this arrangement for Dr. Sorscher have been reviewed and approved by Emory University in accordance with its conflict-of-interest policies. All other authors declare no conflict of interest.

Peer review

Peer review information

Nature Communications thanks Cameron Alexander, Qiang Cheng, Jong Kim and the other anonymous reviewer for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information

Reporting Summary

Transparent Peer Review file

Source data

Source Data

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.

Reprints and permissions

About this article

Cite this article

Huayamares, S.G., Lian, L., Rab, R. et al. Nanoparticle delivery of a prodrug-activating bacterial enzyme leads to anti-tumor responses. Nat Commun 16, 3490 (2025). https://doi.org/10.1038/s41467-025-58548-1

Download citation

Received: 31 August 2023
Accepted: 26 March 2025
Published: 12 April 2025
DOI: https://doi.org/10.1038/s41467-025-58548-1