CRISPR-based rapid and ultra-sensitive diagnostic test for Mycobacterium tuberculosis

Abstract

Rapid and simple-to-use diagnostic methods for tuberculosis are urgently needed. Recent development has unveiled the diagnostic power of the CRISPR system in the detection of viral infections. However, its potential use in detecting the Mycobacterium tuberculosis complex (MTB) remained unexplored. We developed a rapid CRISPR-based assay for TB detection and conducted a retrospective cohort study of 179 patients to evaluate the CRISPR-MTB test for identifying MTB in various forms of direct clinical samples. Its diagnostic performance was compared, in parallel with culture and the GeneXpert MTB/RIF assay (Xpert). The CRISPR-MTB test is highly sensitive with a near single-copy sensitivity, demands less sample input and offers shorter turnaround time than Xpert. When evaluated in the clinical cohort of both pulmonary and extra-pulmonary tuberculosis, the CRISPR-MTB test exhibited an overall improved sensitivity over both culture (79% vs 33%) and Xpert (79% vs 66%), without comprise in specificity (62/63, 98%). The CRISPR-MTB test exhibits an improved overall diagnostic performance over culture and Xpert across a variety of sample types, and offers great potential as a new diagnostic technique for both pulmonary and extra-pulmonary tuberculosis.

Keywords: complex (MTB); CRISPR-MTB; GeneXpert MTB/RIF; diagnosis; tuberculosis.

Figure 1.

Scheme of CRISPR-MTB. Samples were added into LY buffer with microbeads, and were vortexed/heated. Supernatants were subjected to CRISPR-MTB system. Positive fluorescent signals were captured when probes were cleaved by activated Cas12a under target sequence recognized by gRNA.

Figure 2.

Performance identification of CRISPR-MTB. (a) Diameters of the beads between 0.17 and 0.11 mm are suitable for MTB broken. Each type of beads was added into lyse buffer followed same extraction and CRISPR process. No beads were served as negative control. (b) MTB DNA from optimized recipe of rapid extraction yields similar signal comparing to DNA purified by column. All three rapid extraction recipe shared same component including EDTA and Tris with indicated NaOH, CHAPS or SDS + NP40. Commercially available rapid extraction kit from Biochian was compared. DNA from column-based extraction obtained by Sangon bacteria genomic DNA purification kit was served as positive control. DNA from M. kansasii strain was rapidly extracted by SDS + NP40 and served as negative control. (c) CRISPR-MTB can detect MTB DNA close to single copy level. (d) CRISPR-MTB yields specific signal from MTB and BCG but no other DNAs. All bacteria were purified as standard manual of bacteria genomic DNA purification kit. 5 ng genomic DNA of bacteria or 50 ng hDNA was subjected to CRISPR-MTB. ** P < 0.005, **** P < 0.0001

Figure 3.

Study design.

Figure 4.

Cut-off value determination. (a) Typical fluorescent signal curve was presented by ABI7500 QPCR machine. (b) Fluorescent signals from each samples and positive controls were normalized by negative control of same batch. The folds changed were put together according indicated groups. “Non TB”, patients diagnosed with no active MTB infection. “Micro-confirmed TB”, patients diagnosed with active MTB infection by either culture or Xpert or both. “Clinical TB”, patients diagnosed with active MTB infection following Diagnostic criteria and principles of management of infectious MTB without evidence of either culture or GeneXpert.

Figure 5.

The sensitivity according to specimen type. * McNemar test, P < 0.01; ** McNemar test, P < 0.001

Figure 6.

Venn diagram of overlap in TB diagnostic tests in micro-confirmed TB cases and clinically diagnose TB.

Figure 7.

Kaplan–Meier curve of MTB positive rate by CRISPR-MTB, Xpert and culture. * Log-rank test, P < 0.0001.