Naegleria fowleri encodes a diverse repertoire of small RNAs.

(A) Schematic showing EV small RNA identification pipeline. (B) Top 3 most prevalent precursor and mature microRNAs identified in whole cell small RNA sequencing with miRDeep2. (C) Top 4 most prevalent tRNAs identified in EV small RNA sequencing with tRNAscan-SE. (D) Top 6 highly prevalent small RNAs in EV RNAs identified manually via read-stacking with IGV. SmallRNA-5 is a tRNA fragment of s68.tRNA2-Glu. The 2 most prevalent small RNAs were identified first with Shortstack and confirmed manually with IGV. RNA secondary structures were generated with RNAfold v2·5·1(33). Schematic in panel A was generated with BioRender (https://BioRender.com/j76b02m).

Detection of smallRNA-1 and smallRNA-2 in the EVs of various free-living amoebae.

(A) Standard curve for smallRNA-1 assay (consisting of 675 standards across 43 qPCR plates) was used to calculate copy number for qPCR reactions. (B) smallRNA-1 detection in RNA of EVs extracted via ultracentrifugation (UC) across 7 different clinical isolates of Nf. Cq values ranged from 10·5-16·3. (C) Standard curve for smallRNA-2 assay used to calculate copy number. (D) smallRNA-2 detection in RNA of EVs extracted via UC across 4 different isolates of Nf. Cq values ranged from 13·2-16·3. (E) smallRNA-1 detection in RNA of amoebae EVs extracted via extraction reagent from various species of free-living amoebae. For FBS and non-Naegleria species Cq values ranged from 33 to no signal. For Naegleria species Cq values ranged from 14·1-27·4. Each data point and/or bar in panels B-D consists of 3 technical replicates in RT-qPCR assay. Statistical significance was determined for panel E using One-way ANOVA test in GraphPad Prism v10.0.0 (GraphPad, La Jolla, CA, USA).

Detection of smallRNA-1 via RT-qPCR in small volumes of culture media and amoebae pellets.

(A) Dilution series of Villa Jose (VJ) amoebae cultured for 24h followed by centrifugation and media extraction. Cq range: 24·8 to no signal. (B) Whole cell detection in dilution series of amoebae pellets. Cq range for 1 amoeba: 29-35·5. (C) Schematic showing experimental set-up for 24h plate assay data generated from cell culture media and displayed in panel D. (D) smallRNA-1 levels detected in media extracted from wells cultured for 24h containing: 1-VJ feeding on Vero cell monolayer (Cq range: 19-22), 2-axenically cultured VJ (Cq range: 24·9-27·4), 3-axenic VJ media mixed with Vero media (Cq range: 23·1-25·9), 4-Vero media (Cq range: 41·8 to no signal). Each data point in panels A, B and D represents a single well in a 96-well plate, or a single amoeba pellet with 3 technical replicates in RT-qPCR assay. Statistical significance was determined for panel D using Unpaired T tests in GraphPad Prism v10.0.0 (GraphPad, La Jolla, CA, USA). Schematic in panel C was generated with BioRender (https://BioRender.com/4rj2hdk).

Detection of smallRNA-1 and -2 in the plasma of PAM-infected mice at the end-stage of infection.

(A) Schematic showing infection process, plasma extraction via cardiac puncture followed by RNA extraction and RT-qPCR. (B) Assaying for smallRNA-1 in plasma of mice infected by 6 different Nf clinical isolates provided 100% positivity in our assay compared to the uninfected mouse and human plasma. (C) Assaying for smallRNA-2 provided similar results to smallRNA-1 albeit at slightly lower concentrations per µL of plasma. All N. fowleri-infected mice were confirmed positive for amoebae by culturing brains and observing amoebae. Each data point in panels B-D is representative of 3 technical replicates in RT-qPCR assay. Schematic in panel A was generated with BioRender (https://BioRender.com/5xr668r).

Detection of smallRNA-1 in the plasma, serum, and urine of PAM-infected mice at various timepoints post-infection.

(A) Schematic showing infection process followed by urine collection at various timepoints and blood collection at the end-stage of infection followed by RNA extraction and RT-qPCR. (B) A cohort of 64 mice was infected with 1,000 VJ amoebae (fed over Vero cells 7 times) and cohorts of 8 mice each were sacrificed at various timepoints to extract blood. Urine was extracted from mice throughout infection. Mean time to death was 116·6h post infection. (C) A cohort of 10 mice was infected with 5,000 Nf69 amoebae (fed over Vero cells 8 times) and urine was either collected before infection or at various timepoints post-infection with plasma being extracted with euthanasia. Mean time to death was 140h post infection. The infection status of each cohort for panels B-C was determined by culture positivity of mouse brains and is shown under graphs. Each data point in panels B-C is representative of 3 technical replicates in RT-qPCR assay. “Positive” signals (>35 copies/µL) were defined as Cq values of 19·2 to 35·1 for panel B and 17·1 to 35·1 for panel C. Schematic in panel A was generated with BioRender (https://BioRender.com/19mfh55).

Determination of sensitivity of smallRNA-1 assay in various human biofluids compared to water and detection in PAM-infected human cerebrospinal fluid and whole blood.

(A) SmallRNA-1 spike-ins into various biofluids indicate that CSF, plasma, and urine provide the most consistent results compared to H2O spike-ins, while serum seems to inhibit the assay. (B) Detection of smallRNA-1 in PAM-infected (n=6), Ac-infected (n=2), Bm-infected (n=2), and uninfected human CSF (n=2) (Cq range for PAM-infected CSF: 24-30·8). (C) Detection of smallRNA-1 in PAM-infected, uninfected, Bm-infected, and Ac-infected human whole blood, plasma, and serum (Cq range for PAM-infected whole blood:32·4-32·6). We adopted a “positivity” cut-off of <33 cycles for urine, and <35 cycles for other biofluids; equivalent to > ∼100 copies per µL of biofluid). Each data point is representative of 3 technical replicates in RT-qPCR assay.