Multiple TMA-aided CRISPR/Cas13a platform for highly sensitive detection of IL-15 to predict immunotherapeutic response in nasopharyngeal carcinoma

Abstract

Background: Immune checkpoint inhibitors (ICIs)-based treatments have been recommended as the first line for refractory recurrent and/or metastatic nasopharyngeal carcinoma (NPC) patients, yet responses vary, and predictive biomarkers are urgently needed. We selected serum interleukin-15 (sIL-15) out of four interleukins as a candidate biomarker, while most patients' sIL-15 levels were too low to be detected by conventional methods, so it was necessary to construct a highly sensitive method to detect sIL-15 in order to select NPC patients who would benefit most or least from ICIs.

Methods: Combining a primer exchange reaction (PER), transcription-mediated amplification (TMA), and a immuno-PER-TMA-CRISPR/Cas13a system, we developed a novel multiple signal amplification platform with a detection limit of 32 fg/mL, making it 153-fold more sensitive than ELISA.

Results: This platform demonstrated high specificity, repeatability, and versatility. When applied to two independent cohorts of 130 NPC sera, the predictive value of sIL-15 was accurate in both cohorts (area under the curve: training, 0.882; validation, 0.898). Additionally, lower sIL-15 levels were correlated with poorer progression-free survival (training, HR: 0.080, p<0.0001; validation, HR: 0.053, p<0.0001).

Conclusion: This work proposes a simple and sensitive approach for sIL-15 detection to provide insights for personalized immunotherapy of NPC patients.

Keywords: Immunoassay; Tumor Biomarkers.

Figure 1

Overview of the study design. Top, IL-15, IL-2, IL-12, and IL-8 in the serum of 62 NPC patients were analyzed using ELISA to identify predictive biomarker candidates. Middle, process, and performance of the iPTC platform. Botton, serum IL-15 was evaluated as a novel predictive biomarker for the response to ICI-related therapy in the training cohort (n=87) and validation cohort (n=43) by the iPTC platform. The discovery set is included in the training cohort. ICI, immune checkpoint inhibitor; IL, interleukin; iPTC, immuno-PER-TMA-CRISPR/Cas13a; NPC, nasopharyngeal carcinoma; PFS, progression-free survival.

Figure 2

Prediction of ICI-related therapy response by baseline sIL-15, sIL-2, sIL-12, and sIL-8. (A–D) Scatter plots of sIL-15, sIL-2, sIL-12, and sIL-8 levels in non-PD and PD patients measured by ELISA. The two groups were statistically compared by the Kruskal-Wallis test: ** and ns represent p<0.01, p>0.05, respectively. (E–H) ROC curve analysis evaluating the predictive power of sIL-15, sIL-2, sIL-12, and sIL-8 levels in differentiating PD patients from non-PD patients, with the corresponding AUC. AUC, area under the curve; PD, progressive disease; ROC, receiver operating characteristic; sIL-15, serum interleukin-15.

Figure 3

Schematic of iPTC. The target protein is captured by an ELISA double-antibody sandwich structure. After hybridization with AS/S probes containing T7 promoters, long biotinylated repetitive DNA sequences (PER concatemers) bind to the antigen-antibody complex. Then, multiple TMA rounds are performed to generate a mass of target RNA, recognized by crRNA, and trigger transcleavage by Cas13a, resulting in a significant fluorescence signal. AS/S, antisense strand/sense strand; iPTC, immuno-PER-TMA-CRISPR/Cas13a; PER, primer exchange reaction; TMA, transcription-mediated amplification.

Figure 4

The feasibility of the PTC platform performed on magnetic beads. (A) Schematic for the principle of short-strand target DNA (3a)-activated LbuCas13a after T7 transcription on streptavidin-conjugated magnetic beads (left). The real-time fluorescence kinetics measurements of CRISPR/Cas13a reactions with 3a concentrations ranging from 62.5 to 2500 fM (middle). The sensitivity of CRISPR/Cas13a reactions with 3a (right). (B) Schematic for the primer exchange reaction (PER) cycle principle. (C) The feasibility of the PER reaction was verified by urea-PAGE. (D) The length of PER products was monitored in real-time by urea-PAGE. (E) Schematic for the principle of the long-strand target DNA (PER concatemer)-activated CRISPR/Cas13a system after T7 transcription on streptavidin-conjugated magnetic beads (left). The real-time fluorescence kinetics measurements of CRISPR/Cas13a reactions with PER concatemer concentrations ranging from 0.625 to 125 fM (middle). The sensitivity of CRISPR/Cas13a reactions with PER concatemers (right). Blank-subtracted fluorescence was calculated by subtraction of blank fluorescence values. Error bars represent the SD for three independent experiments. Data are represented as the mean±SD, n=3, three technical replicates. LOD, limit of detection; PTC, PER-TMA-CRISPR/Cas13a.

Figure 5

The feasibility of the iPTC platform performed on an ELISA plate. (A) The curve slope and visual fluorescence of the CRISPR/Cas13a reaction on an ELISA plate. The top-right inset demonstrates the schematic for the principle of PTC performed on an ELISA plate, that is, 379 pM streptavidin is precoated on a 96-well plate, and 500 pM different lengths of biotinylated DNA concatemers (3a, 6a, 9a, and PER) hybridized with AS/S probes were bound for subsequent T7 transcription. The slope of the curve over 1 hour was calculated by performing a simple linear regression of data merged from three replicates and is shown as the slope±95% CI. The slopes were compared with the former group by the Kruskal-Wallis test followed by Dunnett’s multiple-comparison test: *p<0.05, **p<0.01, ****p<0.0001, respectively. The top-left insert demonstrates the fluorescence intensity of the CRISPR/Cas13a system observed under ultraviolet radiation-based visual readouts, and the tubes from left to right are 3a, 6a, 9a, and PER. (B) The corresponding curve between the IL-15 concentrations and the absorbance at 450 nm was measured by ELISA. The insert shows the linear relationship between absorbance and the concentrations of IL-15, with a LOD of 4.92 pg/mL. (C) Schematic of the principle of iPTC. (D) The corresponding curve between the IL-15 concentrations and the fluorescence intensity measured by iPTC. The insert shows the linear relationship between fluorescence intensity and the concentrations of IL-15, with an LOD of 0.032 pg/mL. Error bars represent the SD for three independent experiments. Data are represented as the mean±SD, n=3, three technical replicates. AS/S, antisense strand/sense strand; iPTC, antisense strand/sense strand; LOD, limit of detection; PER, primer exchange reaction.

Figure 6

The anti-PD-1 treatment response and outcome prediction based on baseline sIL-15 levels. (A, B) Scatter plots of sIL-15 levels in non-PD patients and PD patients in the training cohort and validation cohort determined by the iPTC platform. The two groups were statistically compared by the Mann-Whitney U test: ****p<0.0001, respectively. (C, D) ROC curve analysis evaluating the predictive power of baseline sIL-15 in differentiating PD patients from non-PD patients, with the corresponding area under the curve in the training and validation cohorts. (E) (F) Baseline sIL-15 levels measured prior to anti-PD-1 therapy were associated with PFS in NPC patients in the training and validation cohorts. Data were analyzed using the log-rank test. iPTC, immuno-PER-TMA-CRISPR/Cas13a; NPC, nasopharyngeal carcinoma; PD, progressive disease; PFS, progression-free survival; ROC, receiver operating characteristic; sIL-15, serum interleukin-15.