Using tiny particles to deliver genetic material for gene editing in Pneumocystis murina

Our previous work demonstrated that Pneumocystis carinii actively uptakes native host lung EVs labeled with the lipophilic dye PKH26, suggesting a mechanism for nutrient acquisition from the host (20). Building on this, we conducted an initial feasibility study to determine if P. murina could internalize EVs containing exogenous nucleotides. EVs transformed with TxRed-labeled siRNA were co-cultured with P. murina for 24 h. Fluorescent microscopy analysis (Fig. 1) demonstrated clear uptake of the TxRed signal by P. murina organisms. In the control group, where P. murina was co-cultured with siRNA-TxRed but no EVs, no TxRed signal was observed within the organisms. Similarly, P. murina co-cultured with EVs containing no cargo also showed no TxRed signal. In contrast, P. murina co-cultured with EVs loaded with siRNA-TxRed exhibited distinct punctate TxRed fluorescence localized within the fungal cells, indicating successful internalization of the EV cargo. This confirms that P. murina can actively take up EVs and their contents, establishing a foundational step for EV-mediated genetic delivery.

The initial pSS1 plasmid (Fig. 2A) was constructed, incorporating an expression cassette with a flanking P. murina Msg promoter and PmNamp8 terminator to ensure the expression of introduced genes within the fungal host. P. murina codon-optimized mNeonGreen or Dhps^ARS were cloned into the coding region.

P. murina organisms co-cultured with EVs containing either pSS1-mNeonGreen or pSS1-Dhps^ARS demonstrated successful gene delivery and subsequent transcriptional activity in vitro after 24 h. Relative to P. murina exposed to empty EVs (control), the mNeonGreen and Dhps^ARS groups showed strong transcriptional activity, with average 2^−ΔCt values of 51,333 and 58.95, respectively. These correspond to log2-transformed values of approximately 15.7 and 5.9, confirming successful expression of the plasmid-delivered genes (Fig. 2B and C). Interestingly, endogenous Dhps^WT transcription was not detected in pSS-Dhps^ARS transformed Pm, suggesting a feedback mechanism regulating gene expression. Western blot analysis did not detect mNeonGreen protein expression (data not shown), likely due to the lack of a long-term culture system. Under these conditions, P. murina is unable to maintain cellular homeostasis, thereby likely limiting its capacity for efficient protein translation.

This initial in vitro success established the feasibility of EV-mediated gene delivery for Pneumocystis and foundational proof of concept for the entire study. It demonstrates that EVs can indeed deliver genetic material to P. murina and, crucially, that this material is transcriptionally active within the fungal cells. Without this initial success in vitro, the more complex in vivo experiments and CRISPR applications would not be feasible.

To enable stable and long-term gene expression within the host, the pSS2.1 plasmid (Fig. 3A) was engineered. This improved construct incorporates a blasticidin resistance (blasticidin S deaminase [Bsd]) gene for selective pressure and, critically, a truncated centromeric region, PmCen15 (23). This centromeric region was incorporated to ensure stable maintenance and accurate segregation of the plasmid within P. murina cells in vivo, mimicking chromosomal behavior and preventing plasmid loss, a common challenge with traditional episomal plasmids in eukaryotic systems in the absence of continuous selective pressure.

Following in vitro blasticidin selection of P. murina transformed with pSS2.1-mNeonGreen-containing EVs and subsequent infection of mice, mNeonGreen transcript expression was detected by reverse transcription quantitative PCR (RT-qPCR) and subsequent agarose gel electrophoresis (Fig. 3B). While mNeonGreen was indetectable by immunofluorescence, likely due to low expression, mNeongreen protein expression was successfully detected from Pm lysates using enzyme-linked immunosorbent assay (ELISA) (Fig. 3C). Both mNeonGreen mRNA and protein expression were sustained in lung samples collected at 1 week, 3 weeks, and 5 weeks post-infection. This sustained expression over an extended period clearly indicates successful long-term maintenance and active expression of the pSS2.1 plasmid within the P. murina population residing in the host lung environment.

Two distinct sets of crRNAs designed to target the Dhps gene were validated through in vitro cleavage assays. Agarose gel electrophoresis and subsequent densitometry analysis (Fig. 4A and B) confirmed efficient cleavage of the Dhps target DNA, demonstrating the functionality of the designed crRNAs and the Cas9 enzyme in a cell-free system. This in vitro validation was a prerequisite for proceeding with in vivo applications.

For in vivo Dhps editing, we engineered a Cas9 expression cassette encoding a guide RNA targeting the Dhps locus, flanked by a hammerhead (HH) ribozyme at the 5′ end and a hepatitis delta virus (HDV) ribozyme at the 3′ end (Fig. 5A). This ribozyme-flanked gRNA design was inserted into the CDS region of plasmid pSS2.1 and is based on a previously described self-processing architecture (24). In this system, the HH and HDV ribozymes mediate RNA self-cleavage (25, 26), resulting in the generation of mature Cas9 mRNA and Dhps-targeting gRNA. The resulting construct (pSS2.1-Cas9-HH-gRNA^Dhps-HDV) was delivered to P. murina via EVs, along with either sense or antisense single-stranded DNA (ssDNA) donor templates encoding the Dhps^ARS mutation. Both orientations of the donor were tested to account for potential strand bias during homology-directed repair (HDR).

Following in vitro blasticidin selection to enrich for transformed organisms, mice were infected with the treated P. murina organisms. After 5 weeks post-inoculation, the animals were treated with TMP-SMX for 2 weeks. This treatment regimen applied strong selective pressure, first with blasticidin, for which resistance is encoded on the plasmid, and then with TMP-SMX, where only Pneumocystis organisms that had successfully incorporated the Dhps^ARS mutation were resistant to SMX exposure (9–13).

After 5 weeks of infection and 2 weeks of treatment, quantitative PCR revealed that Dhps^ARS copy number was detectable in both donor-treated groups compared to Dhps^WT, which was not detected (Fig. 5B). These results suggest that delivery of either sense or antisense donor templates promoted maintenance or enrichment of the Dhps locus under selective pressure, consistent with successful in vivo editing and expansion of resistant organisms. However, treatment with TMP-SMX did not result in a significant decrease in fungal organisms. However, it is plausible that blasticidin selective pressure was sufficient in maintaining only P. murina organisms that successfully received pSS2.1 and the encoded Bsd resistance.

To confirm this hypothesis, Sanger sequencing was performed on amplified Dhps genomic DNA to provide definitive evidence of precise genetic editing. Primers were designed to amplify regions flanking the donor DNA, thereby eliminating the possibility of amplifying residual plasmid or unincorporated donor template. Alignment of the Dhps [1493–1633] region from untreated organisms (P. murina Dhps^WT) was compared to sequences from P. murina organisms treated with EVs containing pSS2.1-Cas9-HH-gRNADhps-HDV and either antisense (AS) or sense (S) single-stranded DNA (ssDNA) donor templates. The alignment (Fig. 5C) revealed that 10 out of 10 treated animals harbored fungal organisms with successful HDR, displaying precise genomic integration of the Dhps^ARS mutant sequence. In addition to the TRP-to-ARS amino acid substitution, the edited sequences contained silent point mutations engineered into the donor DNA to permit specific quantitative PCR (qPCR) detection and reduce the likelihood of repeated Cas9 ribonucleoprotein (RNP) binding. Together, these findings provide evidence of accurate and durable editing of the P. murina genome in vivo.

BALB/c male mice (20–25 g, 5–6 weeks old) were immunosuppressed for the duration of the study with dexamethasone (4 mg/L) in drinking water, provided ad libitum. Mice were infected via intranasal inoculation with 10⁶P. murina, cryopreserved from previously infected mice. The inoculum was administered under isoflurane anesthesia. Mice were euthanized humanely, and their lungs removed for isolation and quantification of fungal organisms (n = 5). To quantify fungal burden, lungs were homogenized in phosphate buffered saline (PBS) using gentleMACS (Miltenyi Biotec, Auburn, CA, USA). The lung homogenate was sequentially filtered through 70 µm then 30 µm filters. The resulting filtrate was then centrifuged at 300 × g for 10 min to pellet host cells. The supernatant was then pelleted at 3,500 × g for 10 min and resuspended in 10 mL. Samples were placed onto a slide, heat fixed, then stained with a modified Diff-Quik staining to visualize the nuclei for microscopic enumeration, as previously described in Cushion et al. (38).

The mouse lung EVs used in this study are those isolated in Sayson et al. (20). Briefly, BALF was collected from uninfected, immunocompetent BALB/c mice using cold PBS (1 mL × 3), and the cellular debris was removed by centrifugation at 3,400 × g for 15 min. Size exclusion chromatography was performed on BALF using qEV10 columns and the Automatic Fraction Collector (Izon Science, Medford, MA, USA) to isolate EVs from fractions 1–5. EVs were quantified using the Micro BCA Protein Assay Kit (Thermo Scientific, Rockford, IL, USA). Quality control measures were performed as previously described (20), which included nanoparticle tracking analysis for size distribution and concentration, electron microscopy for morphological assessment, and Western blot analysis for common EV markers (e.g., CD9, CD63, TSG101) and absence of cellular contaminants (data not shown; see reference 20).

Cryopreserved P. murina organisms from infected mice were thawed briefly at 37°C, then resuspended in Roswell Park Memorial Institute (RPMI) 1640 medium. Organisms were pelleted at 300 × g for 10 min, and the resulting pellet was resuspended at a density of 2 × 10⁶ cells/mL using RPMI 1640 medium (Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (Cytiva, Marlborough, MA, USA), 1,000 U/mL penicillin, 1,000 µg/mL streptomycin (Gibco, Grand Island, NY, USA), 1% minimum essential medium (MEM) vitamin solution (Gibco, Grand Island, NY, USA), and 1% MEM non-essential amino acid solution (Gibco, Grand Island, NY, USA) (39, 40). Short-term maintenance cultures were set up by distributing 10⁶P. murina organisms in triplicate into 48-well polystyrene plates (Corning Costar, Glendale, AZ, USA) and incubated at 37°C in a 5% CO₂ atmosphere for an initial 2 h before EV treatment, as described below.

The initial pSS1 plasmid was designed and constructed from gene fragments. The pSS1 plasmid (Fig. 2A) was engineered for the transfer and expression of genetic material, encoding either Dhps^ARS or mNeonGreen. These genes were under the transcriptional control of a P. murina Msg promoter, which has been previously shown to exhibit strong transcriptional activity in heterologous systems, including Saccharomyces cerevisiae (41, 42). A 151 bp sequence downstream of the 3′ untranslated region of Namp8 (PNEG_1673) was selected and used as a transcriptional terminator.

Subsequent modifications and the generation of various derivative versions, including pSS2.1, were designed in-house and constructed by VectorBuilder (www.vectorbuilder.com↗). The pSS2.1 plasmid (Fig. 3A) was further enhanced by the incorporation of a blasticidin resistance gene, Bsd, for selection. This was under transcriptional control of a Tef1 (PNEG_01204) promoter (444 bp upstream of start codon) and terminator (500 bp downstream of stop codon). Additionally, a truncated PmCen15 centromeric region was added to ensure stable maintenance (23).

To promote efficient transgene expression, the coding sequences for Bsd, mNeonGreen, and Cas9 were codon-optimized for P. murina codon usage using the online codon optimization tool provided by NovoPro (https://www.novoprolabs.com/tools/codon-optimization↗).

Plasmids and siRNA were transformed into purified mouse lung EVs utilizing the Exo-Fect Exosome Transfection Kit, following the manufacturer's protocol (System Biosciences, Palo Alto, CA, USA). Transformation reactions were stopped, and EVs were purified using ExoQuick-TC (System Biosciences, Palo Alto, CA, USA) to remove residual nucleic acids, then resuspended in PBS. A 2 µg protein equivalent of loaded EVs or control EVs was introduced to 10⁶P. murina and incubated at 37°C in a 5% CO₂ atmosphere for a 24 h period, as described above.

Cells were fixed in 3.7% formaldehyde prepared in PBS for 15 min following a PBS wash. After fixation, cells were transferred onto microscope slides using a CytoSpin 2 cytocentrifuge (Thermo Shandon, Kalamazoo, MI, USA) at 1,000 rpm for 10 min. Non-specific binding was blocked by incubating samples in 10% goat serum for 1 h. Cells were then stained with anti-Msg primary antibodies, followed by Alexa Fluor 488-conjugated anti-rabbit secondary antibodies, each applied for 1 h. Between antibody incubations, samples were washed three times for 15 min each with PBS containing 0.1% Tween-20. Imaging was performed on an Olympus IX83 inverted fluorescence microscope.

Total RNA was extracted from P. murina cultures utilizing the ZymoResearch Direct-zol RNA Miniprep Kit (Zymo Research, Irvine, CA, USA) and treated with DNase I (Zymo Research, Irvine, CA, USA), according to the manufacturer's instructions. Genomic DNA was extracted from P. murina organisms isolated from lung samples using the ZymoResearch Quick-DNA Miniprep Kit (Zymo Research, Irvine, CA, USA).

cDNA was synthesized from isolated RNA using SuperScript IV VILO Master Mix (Invitrogen, Carlsbad, CA, USA). RT-qPCR was performed on mNeonGreen expression normalized against large subunit (LSU) rRNA using PowerUp SYBR Green Master Mix (Applied Biosystems, Waltham, MA, USA). RT-qPCR was performed on Dhps^WT or Dhps^ARS expression normalized against LSU using TaqMan Fast Advanced Master Mix (Applied Biosystems, Waltham, MA, USA). Real-time PCR assays were run on an Applied Biosciences 7500 Fast PCR system. Relative expression was calculated using the 2^−ΔCt method, normalized to the endogenous housekeeping gene LSU.

For genomic DNA copy number analysis, qPCR was performed using TaqMan Fast Advanced Master Mix. A standard curve was generated using synthetic gene fragments (Azenta Life Sciences, South Plainfield, NJ, USA) of Dhps^WT and Dhps^ARS to determine absolute copy numbers. Primer, probe, and gene fragment sequences are provided in Table S1.

The presence of mNeonGreen in vivo was detected by ELISA. Lung homogenates were lysed in radioimmunoprecipitation assay buffer (RIPA) buffer (Alfa Aesar, Ward Hill, MA, USA). High-binding 96-well plates (Corning 2592; Kennebunk, ME, USA) were coated overnight with capture antibody, mNeonGreen VHH Recombinant Alpaca Monoclonal Antibody (ChromoTek CTK0203; Planegg-Martinsried, Germany), in 50 mM carbonate buffer, pH 9.4, at 4°C. Lung lysates were diluted 1:10 in PBS and then added to the plate for 1 h. After antigen capture, wells were washed three times with PBS with 0.1% Tween-20 (PBST), then detected using mNeonGreen polyclonal antibody (Proteintech 29523-1-AP, Planegg-Martinsried, Germany) and goat anti-rabbit IgG (H + L) secondary antibody, HRP (Invitrogen 31460, Waltham, MA, USA). After washing with PBST three times, detection was performed using 1-Step TMB ELISA Substrate Solutions (Thermo Scientific, Rockford, IL, USA), and absorbance at 650 nm was read using a Synergy HTX plate reader (BioTek, Winooski, VT, USA).

Two distinct sets of crRNA targeting the P. murina Dhps gene were designed using the CRISPR tool in Benchling (www.benchling.com↗). P. murina Dhps was amplified by PCR from gDNA and used for cleavage assays. CRISPR RNA (crRNA) was annealed to trans-activating CRISPR RNA (tracrRNA) (IDT, Coralville, IA, USA) at a 1:1 ratio to form a guide RNA (gRNA), then complexed with recombinant Alt-R S.p. Cas9 protein (IDT, Coralville, IA, USA). In vitro cleavage assays were conducted by incubating gRNA:Cas9 RNP complexes with Dhps^WT DNA for 60 min at 37°C. After the digestion was completed, Proteinase K (1.81 mg; Zymo Research, Irvine, CA, USA) was added to stop the reaction. Cleavage products were resolved on a 1% agarose gel in TAE buffer, then stained with GelRed Nucleic Acid Gel Stain (Biotium, Fremont, CA, USA). Gels were visualized and analyzed on an iBright CL1500 imaging system (Invitrogen, Carlsbad, CA, USA). Densitometry analysis was performed employing the iBright software to quantify cleavage efficiency. The specific crRNA sequences and PCR primers are provided in Table S1.

P. murina (1 × 10⁶) was co-cultured with either untreated EVs or transformed EVs, as described above. After 2 h of incubation, to allow expression of Bsd, blasticidin (100 µg/mL; Gibco, Grand Island, NY, USA) selection was applied in vitro for 24 h. P. murina was then washed three times in PBS and intranasally inoculated into mice. For pSS2.1-mNeonGreen experiments, lung samples were collected at 1 week, 3 weeks, and 5 weeks post-inoculation for subsequent analysis. For CRISPR/Cas9 experiments, mice were allowed to develop infection for 5 weeks, followed by oral gavage with Sulfamethoxazole and Trimethoprim Oral Suspension (125 mg/kg, 12.5 mg/kg; Aurobindo Pharma USA, East Windsor, NJ, USA) three times a week for 2 weeks (43). Lung samples were collected at the end of the 2-week treatment period.

Genomic DNA was extracted from P. murina isolated from mouse lungs and amplified using primers positioned outside the donor DNA regions to ensure amplification of genomic DNA rather than residual donor DNA. The amplified products were submitted to Genewiz (www.genewiz.com↗) for Sanger sequencing. Sequencing primer is provided in Table S1. Sequence alignments were performed using Clustal Omega to confirm successful homologous recombination (44).

Statistical analysis was performed in GraphPad Prism version 10.4.2 (534) using unpaired t-tests or two-way analysis of variance followed by Sidak's multiple comparisons post hoc test to control groups. A P value <0.05 was considered statistically significant.