Early diagnosis of rabies virus infection by RPA-CRISPR techniques in a rat model

Abstract

Rabies, which is caused by rabies virus (RABV), poses an ever-present threat to public health in most countries of the world. Once clinical signs appear, the mortality of rabies approaches 100%. To date, no effective method for early rabies diagnosis has been developed. In this study, an RPA-CRISPR nucleic-acid-based assay was developed for early rabies diagnosis by detecting viral RNA shedding in the cerebrospinal fluid (CSF) of rats. This method can detect a single copy of RABV genomic RNA in 1 μL of liquid. RABV genomic RNA released from viral particles in the CSF could be detected via RPA-CRISPR as early as 3 days postinfection in a rat model. This study provides an RPA-CRISPR technique for early detection of RABV with potential application in the clinical diagnosis of human rabies.

Fig. 1

Validation and optimization of the RPA-CRISPR system. (A) The process of RPA-CRISPR detection. (B) Titration of crRNA and (C) CRISPR-LwCas13a for optimizing the RPA-CRISPR reaction. Gradient dilutions of crRNA mixture (160 ng, 320 ng, 480 ng, 640 ng, 800 ng, 960 ng, 1120 ng) and CRISPR-LwCas13a (350 ng, 700 ng, 1050 ng, 1400 ng, 1750 ng, 2100 ng, 2450 ng) were used for optimizing the RPA-CRISPR reaction system. PBS was used as a control in the NTC groups. The bar graph shows the mean ± stardard deviation (SD) (n = 3). *, P < 0.05; **, P < 0.01; ***, P < 0.001

Fig. 2

Sensitivity comparison between RPA-CRISPR and qPCR. (A) Titration of RABV vRNA and (B) ssRNA for determining the detection limits of RPA-CRISPR. Tenfold gradient dilutions of RABV vRNA and ssRNA were used for RPA-CRISPR sensitivity validation. (C) Titration of RABV vRNA and (D) ssRNA for determining the detection limits of qPCR. Tenfold gradient dilutions of RABV vRNA and ssRNA were used for qPCR sensitivity validation. PBS was used as a control in the NTC groups. The bar graph shows the mean ± SD (n = 3). *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001; ns, no significant difference

Fig. 3

Pairing the HUDSON method and RPA-CRISPR. (A) RABV vRNA and (B) viral particles were serially diluted tenfold using CSF collected from specific-pathogen-free (SPF) Sprague-Dawley (SD) rats as dilution buffer for RPA-CRISPR detection. (C) The specificity of RPA-CRISPR was tested using different RABV strains (CVS-B2c, SAD-B19, DRV-Mexico, SHBRV, DRV-AH08 and DRV-HuNPN01), canine viruses (canine distemper virus [CDV] and canine parvovirus [CPV]), other RNA viruses (Sendai virus [SeV], Japanese encephalitis virus [JEV], vesicular stomatitis virus [VSV]), and human embryonic kidney 293T cells (293T) and mouse neuroblastoma cells (NA). PBS was used as a control in the NTC groups. The bar graph shows the mean ± SD (n = 3). ****, P < 0.0001; ns, no significant difference

Fig. 4

Early diagnosis of RABV infection by RPA-CRISPR in a rat model. (A) Schematic of early diagnosis of RABV infection in a rat model. Thirty-three six-week-old SD rats were randomly divided into 11 groups with three rats per group. (B, C) Five groups of rats were i.m. inoculated with 100 LD₅₀ of lab-attenuated RABV strain CVS-B2c (B). Another five groups of rats were i.m. inoculated with 100 LD₅₀ of wild-type RABV strain DRV-Mexico (C). The mock infection group was i.m. inoculated with the same volume of DMEM. From 3 to 7 dpi, one group of infected rats was taken every other day to collect CSF (50 μL per rat). At each of the indicated days postinfection, CSF from a group of rats was subjected to RPA-CRISPR analysis. PBS was used as control in the NTC group. The bar graph shows the mean ± standard error (SE) (n = 3).