Targeted delivery of diverse biomolecules with engineered bacterial nanosyringes

Plasmid construction

The wild-type PVCpnf locus was constructed previously¹³ and is publicly available as separate structural (pvc1–pvc16) and cargo/regulatory (Pdp1, Pnf and regulatory genes PAU_RS16570–PAU_RS24015) plasmids (Addgene, 198271 and 198272, respectively). Routine plasmid manipulations involved PCR amplification using KOD ONE polymerase (Sigma, KMM-101NV; 6-min extension time for structural plasmids and 2-min extension time for cargo/regulatory plasmids), assembly with Gibson Assembly Master Mix (New England Biolabs, E2611L) and either electroporation into EPI300 electrocompetent cells (Lucigen, EC300110; for structural plasmids) or transformation into chemically competent Stbl3 cells (Thermo Fisher, C737303; for cargo/regulatory plasmids). Novel designs were structurally validated with AlphaFold3 and visualized with PyMOL (2.5.2) and ChimeraX (1.8). A list of amino acid/DNA sequences for constructs used throughout this study can be found in Supplementary Tables 1–10.

PVC expression and purification

PVC samples were obtained using a previous method^3,13 with modifications. Briefly, one variant each of the PVC structural plasmid and cargo/regulatory plasmid were electroporated into EPI300 cells, after which colonies were inoculated into 5 ml of Terrific Broth (TB; US Biological, T2810) and shaken at 37 °C for 16 h. Starter cultures were then inoculated (at 1:100) in 500 ml of TB medium and shaken at 24 °C for an additional 48 h. Cultures were then spun for 25 min at 4,000g, resuspended in 22 ml of lysis buffer (25 mM Tris-HCl (pH 7.5; Thermo Fisher, 15567027), 140 mM NaCl (AmericanBio, AB01915), 3 mM KCl (Sigma-Aldrich, P9541), 5 mM MgCl₂ (Sigma-Aldrich, M4880), 200 μg ml⁻¹ lysozyme (Thermo Fisher, 89833), 50 μg ml⁻¹ DNase I (Sigma-Aldrich, DN25), 0.5% Triton X-100 (Sigma-Aldrich, 93443) and 1× Protease Inhibitor Cocktail (MedChem Express, HY-K0010)) and subsequently shaken at 37 °C for 60 min to promote lysis. Lysates were then pelleted at 5,000g for 30 min at room temperature (RT) to remove bulk cell lysate. Supernatants were then spun in an ultracentrifuge at 120,000g for 1 h at 4 °C to pellet PVC protein complexes. Pellets were hydrated overnight at 4 °C in 1 ml of PBS (Life Technologies, 10010049), resuspended and spun at 16,000g for 15 min at 16 °C to remove residual solid lysate. Supernatants were then applied to 22 ml of cold PBS before repeating the ultracentrifuge spin (120,000g for 1 h) and resuspending in 100–500 µl of cold PBS (with the final volume depending on pellet size). Final resuspensions were again clarified at 16,000g for 15 min at 16 °C, and PVC yields were quantified by A280 measurement on a NanoDrop instrument (Thermo Fisher). For mouse experiments, endotoxin was then removed from the final PVC samples using a detergent-based method used previously¹³. All PVC samples containing spike–cargo fusions were used immediately; all other PVCs were stored at 4 °C for a maximum of 2 weeks before use. Long-term storage of PVCs was achieved by adding 5% (final concentration) glycerol to the PVC suspension and storing at –80 °C without flash-freezing.

Affinity purification of spike–cargo fusions and binding domain–SC fusions

Several proteins used to functionalize PVCs in vitro (such as the Pvc10–Cas9^RNP protein used in Fig. 2c and all cargo/targeting proteins used in Extended Data Fig. 5b) were purified in isolation of PVCs using affinity chromatography. Plasmids carrying the desired open reading frames with N-terminal affinity tags (6×His-TwinStrep-Sumo) were transformed into Rosetta 2 (DE3) cells (Sigma-Aldrich, 71401M). Each colony was inoculated into 10 ml of TB medium and grown at 37 °C for 14–18 h before being inoculated into an additional 500 ml of TB. Cultures were grown at 37 °C until they reached an optical density at 600 nm of 0.6–0.8, induced with 1 mM isopropyl β-d-1-thiogalactopyranoside (GoldBio, I2481C) and subsequently allowed to grow at 18 °C for an additional 16–18 h. Cells were then collected via centrifugation at 4,000g for 30 min, resuspended in 15 ml of cold lysis buffer (50 mM Tris-HCl (pH 7.5; Thermo Fisher, 15567027), 280 mM NaCl (AmericanBio, AB01915), 3 mM KCl (Sigma-Aldrich, P9541), 5 mM MgCl₂ (Sigma-Aldrich, M4880), 20 mM imidazole (VWR, IC102033), 1× cOmplete Protease Inhibitor Cocktail (Sigma-Aldrich, 11836170001) and 50 µg ml⁻¹ DNase I (Sigma-Aldrich, DN25); 0.004 U µl⁻¹ SUPERase-In RNase Inhibitor (Thermo Fisher, AM2694) if purifying RNPs) and passed twice through a Microfluidics M110P microfluidizer. Lysates were then clarified at 19,000g for 20 min at 4 °C, after which they were incubated with 2 ml of 50% Ni-NTA agarose (Qiagen, 30210; pre-equilibrated in lysis buffer) with shaking for 1 h at 4 °C. Following incubation, Ni-NTA beads were pelleted by centrifugation at 1,000g for 3 min at 4 °C, transferred to a column (Thermo Fisher, 29922), washed with 10 ml of cold lysis buffer, washed with 10 ml of cold wash buffer 1 (40 mM imidazole in PBS; Life Technologies, 10010049), washed with 10 ml of cold wash buffer 2 (60 mM imidazole in PBS) and finally eluted with 2 ml of cold elution buffer (250 mM imidazole in PBS) for 5 min. The elution was collected, suspended in excess cold PBS and concentrated using an Amicon Ultra centrifugal filter unit (molecular weight cutoff of 10 kDa or 50 kDa; Sigma-Aldrich, UFC9050). Purified protein yield was quantified by measuring A280 on a NanoDrop instrument (Thermo Fisher). For cargos containing sgRNA, 1 U µl⁻¹ SUPERase-In was also added to prevent RNase activity. To remove the 6×His-TwinStrep tag, purified proteins were treated with Ulp1 protease (a gift from S. Vo, Broad Institute of MIT and Harvard) at 0.4 mg of Ulp1 per 1 mg of cargo at RT for 1.5 h.

In vitro expression of spike–cargo fusions and binding domain–SC fusions

Several proteins used to functionalize PVCs in vitro (such as the Pvc10 protein in Fig. 2b, the anti-MHC class II/anti-EGFR proteins in Fig. 2j and the Ad5–SC protein in Extended Data Fig. 3c,d) were expressed using an in vitro transcription/translation (IVTT) system. The desired open reading frames were synthesized as gBlocks (IDT) containing 5′ and 3′ extensions required for expression in a T7-based IVTT system (5′ T7 promoter and ribosome binding site; 3′ T7 terminator) according to the manufacturer’s guidelines. gBlocks were enriched using PCR with NEBNext polymerase (New England Biolabs, M0541), and the resulting amplicons were purified using a PCR purification kit (Qiagen, 28104) and eluted in distilled water. One microgram of each purified amplicon was then spiked into PURExpress reactions (New England Biolabs, E6800; 50 µl final volume) and incubated at 37 °C for 3 h to promote expression of the desired protein. These reactions were then used as raw inputs for in vitro complementation reactions (see ‘In vitro cargo complementation’ and ‘In vitro retargeting’ below).

In vitro cargo complementation

To complement PVCs with spike–cargo fusion proteins in vitro, 150 µg of PVCs was incubated with either 15 µg of affinity-purified spike–cargo protein in a 50-µl reaction (in PBS) or with 50 µl of a raw IVTT reaction containing a spike–cargo protein; these reactions were incubated for 1–2 h at 24 °C. For cargos containing Cas9 RNPs, 1 U µl⁻¹ SUPERase-In (final concentration) was also added before the addition of cargo. Reactions were then diluted to 100 µl with PBS and purified with 100 µl of a 50% slurry (in PBS) of Capto Core 700 resin (Cytiva, 17548101). Reactions were incubated with resin for 30 min at RT with rotation, after which they were transferred to a filter plate (Harvard Apparatus, 74-5650) and eluted by centrifugation at 1,500g for 3 min. An amino acid sequence for an engineered Pvc10 construct designed to load Cas9 onto PVCs can be found in Supplementary Table 1.

In vitro ssDNA complementation

To load ssDNA cargos onto PVCs in vitro, 150 µg of PVCs (containing Pvc10–HUHe) was mixed with 10 pmol of DNA (containing an HUHe recognition sequence: AAGTATTACCAGC; synthesized by Azenta) in a 50-µl reaction in HUH buffer (50 mM HEPES (pH 8), 50 mM NaCl, 1 mM MgCl₂ and 1 mM MnCl₂), which was incubated at 37 °C for 2 h to promote conjugation of HUHe to the DNA cargo. Protein–DNA conjugation reactions were then purified using Capto Core 700 pre-equilibrated with HUH buffer before DNA delivery experiments in live cells. An amino acid sequence for an engineered Pvc10 construct harboring HUHe, as well as a DNA sequence for a representative HUHe-specific ssDNA cargo, can be found in Supplementary Tables 6 and 7.

In vitro retargeting

To alter PVC target specificity in vitro, 150 µg of PVCs was incubated with either 5 µg of affinity-purified SC/BG-labeled binding protein (purified as described in ‘Affinity purification of spike–cargo fusions and binding domain–SC fusions’ or provided by B. Lash, M. Segel and D. Strebinger, Broad Institute of MIT and Harvard) in a 50-µl reaction (in PBS) or with 50 µl of a raw IVTT reaction containing a binding domain–SC protein (see ‘In vitro expression of spike–cargo fusions and binding domain–SC fusions’); these reactions were incubated for 1 h at RT. Retargeted PVCs were then purified using Capto Core 700 (see ‘In vitro cargo complementation’) before experiments in live cells. Amino acid/DNA sequences for an ST-functionalized PVC tail fiber (Pvc13–ST), a binding domain–SC protein used to retarget Pvc13–ST and a DNA amplicon used to express a binding domain–SC protein via IVTT can be found in Supplementary Tables 8–10.

Biochemical assessment of cargo loading

Loading of cargo domains onto Pvc8/Pvc10 was verified using immunoblots. Ten micrograms of purified PVCs (containing spike–cargo fusion proteins) was mixed with NuPAGE LDS Sample Buffer (Thermo Fisher, NP0008) and NuPAGE Sample Reducing Agent (Thermo Fisher, NP0009), both to a final concentration of 1×, and were subsequently boiled at 95 °C for 10 min. The denatured samples were then run on NuPAGE Bis-Tris 1–12% protein gels (Thermo Fisher, NP0321) for 20 min at 200 V in 1× MOPS buffer (Thermo Fisher, NP000102) and were blotted onto PVDF membranes using an iBlot 2 instrument (Thermo Fisher). Membranes were blocked for 1 h at RT in blocking buffer (5% BLOT-QuickBlocker (VWR, 786-011) in 1× TBS-T (Thermo Fisher, 28360)), stained for 1 h at RT with mouse anti-Flag (Sigma, F3165; diluted 1:1,000 in blocking buffer), washed three times with TBS-T (5 min each), stained for 30 min with anti-mouse IgG (Cell Signaling, 7076; diluted 1:1,000 in blocking buffer) and washed another three times with TBS-T (5 min each). Finally, membranes were incubated for 1 min with chemiluminescence substrate (Thermo Fisher, 32209) and imaged with a Bio-Rad ChemiDoc instrument. For ssDNA loading experiments, the Cy3 channel was also imaged alongside chemiluminescence. Pvc12 was used as a loading control for immunoblots; this protein was visualized via the addition of a C-terminal Flag tag.

Electron microscopy

Loading of Cas9 onto the PVC spike complex was visualized using negative-stain TEM. Briefly, PVC samples were diluted to 100–500 ng µl⁻¹ in PBS and applied to a glow-discharged 200-mesh carbon-coated copper TEM grid (VWR, 100489-722) for 60 s before removing excess liquid with a light-duty tissue wiper (VWR, 82003). Grids were then treated with clarified 2% uranyl formate stain (incubating with gentle agitation for 5 s, 5 s, 10 s, 20 s and 20 s on five separate 30-µl uranyl formate droplets positioned on Parafilm) and allowed to dry at RT. TEM imaging was performed using an FEI Tecnai (G2 Spirit TWIN) microscope at 120 kV equipped with a Gatan Orius SC1000B camera located in the MIT.nano Characterization Facilities at Massachusetts Institute of Technology.

Quantification of sgRNA loading onto Pvc8/Pvc10–Cas9 complexes

To confirm the presence of sgRNA in PVCs loaded with Cas9 RNPs, we performed quantitative reverse transcription PCR (RT–qPCR). PVCs were standardized to a concentration of 1,000 ng µl⁻¹ and denatured by mixing with NuPAGE Sample Reducing Agent (Thermo Fisher, NP0009; 1× final concentration) and boiling at 95 °C for 10 min. These denatured samples were then used to generate cDNA using ProtoScript II Reverse Transcriptase (New England Biolabs, M0368) and random primers (New England Biolabs, S1330), according to the manufacturer’s protocol. We then ran qPCR on the resulting cDNAs using Fast SYBR Green Master Mix (Thermo Fisher, 4385612) in a Bio-Rad CFX Opus 384 qPCR instrument. A list of RT–qPCR primers used during this study can be found in Supplementary Table 11.

Cell culture

All cell lines were maintained in T75 flasks (Thermo Fisher, 156499) at 37 °C with 5% CO₂ in either DMEM-GlutaMAX (Thermo Fisher, 10569044; for HEK293FT and A431 cells) or RPMI-GlutaMAX (Thermo Fisher, 61870127; for A549 and A20 cells). All media were supplemented with 10% fetal bovine serum (VWR, 97068-085) along with 10 µg ml⁻¹ gentamicin (Sigma-Aldrich, G1397) and 1× penicillin–streptomycin (Thermo Fisher, 15140122) to prevent bacterial growth. A list of cell lines used in this study can be found in Supplementary Table 12.

Live-cell experiments with PVCs

To assess the activity of PVCs in live cells, we first seeded target cells into clear-bottom, 96-well plates (VWR, 89091-012) and allowed them to grow to ~80% confluence. A total of 7.5 µg of purified PVCs (into 100 µl of cell culture medium) was then added to each well. For some experiments, improved efficiency was observed if this dose was spaced over 3 days (2.5 µg per day). For experiments involving co-transfection, DNA was transfected using GeneJuice Transfection Reagent (Sigma-Aldrich, 70967) either immediately after adding PVCs (for Cre reporter plasmids) or 6 h before adding PVCs (for Cas9/sgRNA or bdSENP1 plasmids; these transfections were followed by a medium exchange before adding PVCs). For ssODN delivery experiments, 1 µM (final concentration) Alt-R HDR enhancer V2 (IDT, 10007921) was also added alongside PVCs to promote HDR-mediated repair. For toxin delivery experiments, cytotoxicity was assessed using CellTiter-Glo 2.0 Cell Viability Assay (Promega, G9241) 24 h after adding PVCs; any wells exhibiting higher luminescence than the control well (PBS) were assigned a cytotoxicity value of 0% to avoid negative cytotoxicity. For assays involving Cre-driven GFP expression, cells were incubated for 3 days, imaged with a Leica DMi8 confocal microscope and analyzed by flow cytometry (see ‘Flow cytometry’). For gene editing experiments, cells were incubated for 4 days, genomic DNA was extracted with 50 µl of QuickExtract DNA Extraction Solution (Lucigen, QE09050), and indels/base substitutions were quantified by next-generation sequencing (see ‘Deep sequencing’). All numerical data from the PVC experiments were plotted with Prism (10.2.2), and figures were graphically assembled in Adobe Illustrator (27.1.1).

Flow cytometry

Cells were first collected by incubation with 30 µl of TrypLE Express dissociation reagent (Thermo Fisher, 12604) for 5 min at RT, followed by resuspension with 50 µl of cold flow cytometry buffer (PBS supplemented with 2% EDTA (Life Technologies, 15575020) and 5% fetal bovine serum (VWR, 97068-085)). Samples were then run on a Beckman Coulter Cytoflex S flow cytometer, and analysis was performed using FlowJo (10.8.2).

Specificity assay with mixed populations

To assess the target specificity of in vitro-retargeted PVCs (Fig. 2j), we instead used a modified protocol involving cocultures of two cell lines. Briefly, an approximately 1:1 mixture of A431 and A20 cells (1 × 10⁶ cells each) was seeded into 2 ml of RPMI-GlutaMAX (with supplements listed in ‘Cell culture’) in six-well, flat-bottom plates (VWR, 29442), 150 µg of in vitro-retargeted PVCs was added, and the cultures were allowed to grow for 48 h. The supernatants (containing suspension cells) were then collected and combined with adherent cells (collected in a similar manner as detailed in ‘Flow cytometry’), pelleted at 300g for 5 min, resuspended in 500 µl of fresh flow cytometry buffer to remove residual TrypLE reagent, immunostained with anti-mouse CD19–Alexa Fluor 700 (Thermo Fisher, 56-0193-82; 1:100 dilution) at RT for 30 min in the dark, washed with blank flow cytometry buffer three times and analyzed on a Beckman Coulter Cytoflex S flow cytometer. The flow cytometry gating scheme used during this experiment can be found in Extended Data Fig. 3g.

Deep sequencing

We used deep sequencing to detect PVC-mediated delivery of gene editing systems (Cas9 and ZFDs). We first amplified the desired target region out of each genomic DNA extract using NEBNext (New England Biolabs, M0541) and subsequently barcoded the resulting amplicons with indexed Illumina P5 and P7 next-generation sequencing primers. Libraries were purified with a PCR purification kit (Qiagen, 28104), quantified on a NanoDrop instrument (Thermo Fisher) and sequenced on an Illumina MiSeq instrument. Indels/base substitutions were then quantified with Geneious Prime (2020.0.5). A list of primers used for deep sequencing can be found in Supplementary Table 13.

In vivo B cell depletion assay

All experiments in mice conformed to guidelines established by the National Institutes of Health and were conducted under Institutional Animal Care and Use Committee protocol 0017-09-14-3, approved by the Broad Institute of MIT and Harvard. Animals were chosen randomly for treatment with either control or experimental conditions without blinding. Female C57BL/6J mice (aged 8–12 weeks) were obtained from the Jackson Laboratory (strain 000664). All mice were maintained on a 12-h light/12-h dark cycle with ad libitum access to food and water. A total of 100 µl per mouse at 1.2 µg µl⁻¹ of either MHC class II-targeting or untargeted PVC in sterile 0.9% NaCl was injected intravenously. Animals were deeply anaesthetized 24 h after injection with CO₂, and spleens of PVC-injected mice were extracted. Single-cell suspensions were generated by mashing through 100- and 70-µm cell strainers (Greiner One-Bio, 542000 and 542070) and repeated washing with PBS. Splenocyte samples were blocked with 1:50 TruStain FcX (anti-mouse CD16/CD32) reagent (BioLegend, 101320) before antibody staining for flow cytometry and stained with fixable viability dye eFluor780 (1:1,000 dilution; Thermo Fisher Scientific), anti-mouse CD45-eFluor450 (1:100 dilution; BioLegend), anti-mouse CD3-FITC (1:100 dilution; BioLegend), anti-mouse CD19-APC (1:100 dilution; BioLegend), anti-mouse CD11b-PE-Cy7 (1:100 dilution; BioLegend) and anti-mouse IA/IE (MHC class II)-PE (1:100 dilution; BioLegend) at 4 °C for 30 min. Flow cytometry was performed on a Beckman Coulter CytoFlex S device with standard laser and detector configuration, and data were analyzed using FlowJo 10.8.1 software.

Serum neutralization assay

To assess whether mice mount a neutralizing immune response against PVCs following systemic injection, we used a serum neutralization assay. Briefly, we injected mice with either PBS (mock injection) or MHC class II-targeting PVCs (intravenously as described in ‘In vivo B cell depletion assay’), repeated the injections in one cohort at t = 7 days (to amplify the generation of neutralization antibodies), extracted serum for all cohorts at t = 14 days via submandibular bleed, incubated clarified serum with similar MHC class II-targeting PVCs (5% (vol/vol) serum and 15 µg of PVCs in each 100-µl reaction, diluted in PBS) for 1 h at RT and titrated the resulting PVC/serum mixtures on MHC class II⁺ cells (A20). Cytotoxicity was assessed as described in ‘Live-cell experiments with PVCs’.

Statistics and reproducibility

All statistical analyses were performed in Prism (10.2.2). All quantitative data are presented as mean ± s.d. with n = 5 biological replicates per condition; biological replicates represent independent treatments in separate culture wells or mice. All images and blots are representative examples from n = 3 independent experiments. Statistical significance was computed using two-sided unpaired Student’s t-test or one-way ANOVA, followed by a Bonferroni post hoc test (to correct for multiple comparisons), as indicated in the figure legends.

Reporting summary

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