Synthesizing AND gate minigene circuits based on CRISPReader for identification of bladder cancer cells

Abstract

The logical AND gate gene circuit based on the CRISPR-Cas9 system can distinguish bladder cancer cells from normal bladder epithelial cells. However, the layered artificial gene circuits have the problems of high complexity, difficulty in accurately predicting the behavior, and excessive redundancy, which cannot be applied to clinical translation. Here, we construct minigene circuits based on the CRISPReader, a technology used to control promoter-less gene expression in a robust manner. The minigene circuits significantly induce robust gene expression output in bladder cancer cells, but have nearly undetectable gene expression in normal bladder epithelial cells. The minigene circuits show a higher capability for cancer identification and intervention when compared with traditional gene circuits, and could be used for in vivo cancer gene therapy using the all-in-one AAV vector. This approach expands the design ideas and concepts of gene circuits in medical synthetic biology.

Fig. 1. Design and construction of the AND gate minigene circuits.

a The UPII promoter drove the transcription of Cas9 mRNA, while the TERT promoter was used to promote the transcription of sgRNA targeting LacI. The output Renilla luciferase gene was regulated by a LacI-controlled CMV promoter. The luciferase was expressed only when both UPII promoter and TERT promoter were both activated. In the design of the minigene circuit, the UPII and TERT promoters were replaced by their respective transcription factor binding elements. Both c-Myc and Get1, only in bladder cancer cells, had a relative high expression level at the same time. After initial expression of sgRNA1 and sgRNA2, they could further bind upstream of their own transcription initiation sites and amplify the transcription signals of c-Myc and Get1 through the positive feedback mechanism to amplify their downstream genes transcription, respectively. Furthermore, the LacI gene was knocked out by sgRNA2, and luciferase reporter gene was activated by transcription. In normal bladder epithelial cells, luciferase could not be effectively transcribed and was further silenced by trace amounts of LacI expressed at the background level. b The traditional gene circuit and the minigene circuit were compared by detecting the expression levels of luciferase in different bladder cancer cells and normal bladder cells, and the cancer diagnostic efficacy was analyzed. All transient transfections were performed in the presence of PGL3-TK-Fluc vector. Each experiment was independently performed in triplicate five times. The small purple dots represent the relative luciferase level (Rluc/Fluc expression ratio) of each cell line, while the black lines show mean ± SD. ^*p < 0.05 and ^**p < 0.01 between the groups using ANOVA. Exact p-values for asterisks (from left to right): 0.0038, 0.0045, 0.0018, 0.033, 0.0045, and 0.0025. Source data are provided as a Source Data file.

Fig. 2. The apoptosis levels of bladder cell lines transfected with either the minigene circuit or the traditional circuit.

The level of cell apoptosis was calculated using the Cell Death Detection ELISA assay. Results are shown as the mean ± SD. The minigene circuit without output gene was used as the negative control. Each experiment was performed in triplicate for five independent times. Each error bar indicates the variation (standard deviation) between the means of five independent experiments. The small purple dots represent the mean of each independent experiment. The numerical category label represented the number of the corresponding cell line in each group. mU = absorbance [10⁻³]. ^*p < 0.05 and ^**p < 0.01 between the groups using a two-tailed t-test. Exact p-values for asterisks (from left to right): 0.0002, 0.0056, 0.0043, 0.0005, 0.0041, 0.0061, 0.0001, 0.0047, 0.0052, 0.0002, 0.0045, 0.0058, 0.0006, 0.0008, 0.0005, 0.0009, 0.0005, 0.0007, 0.0003, 0.0004, 0.025, 0.031, and 0.035. Source data are provided as a Source Data file.

Fig. 3. Growth curves of bladder cell lines transfected with either the minigene circuit or the traditional circuit.

Proliferations of the transfected cells “high grade-1” (a), “high grade-2” (b), “high grade-3”(c), “high grade-4” (d), “low grade-1” (e), “low grade-2” (f), “low grade-3”(g), “low grade-4” (h), “normal-1” (i), “normal-2” (j), and “normal-3” (k) were measured using the CCK-8 assay at different time intervals. The minigene circuit without output gene was used as the negative control. The curves of cell proliferations were compared using two-way analysis of variance (ANOVA). The results at each time point are shown as the mean ± SD. Each experiment was independently performed in triplicate five times. Each error bar indicates the variation between the means of five independent experiments. ^*p < 0.05 and ^**p < 0.01 between the groups. Exact p-values for asterisks (from left to right): 0.0008, 0.0059, 0.0061, 0.0007, 0.0034, 0.0011, 0.0057, 0.0019, 0.0063, 0.0023, 0.0021, 0.0064, 0.0062, 0.0055, 0.0054, 0.0059, 0.0053, 0.034, 0.034, 0.036, 0.036, 0.041, and 0.038. OD optical density. Source data are provided as a Source Data file.

Fig. 4. Migrations of bladder cell lines transfected with either the minigene or traditional circuits.

Migrations of the transfected cells “high grade-1” (a), “high grade-2” (b), “high grade-3” (c), “high grade-4” (d), “low grade-1” (e), “low grade-2” (f), “low grade-3”(g), “low grade-4” (h), “normal-1” (i), “normal-2” (j), and “normal-3” (k) were measured using the scratch assay. Data are expressed as the mean ± SD (l). The minigene circuit without output gene was used as the negative control. Relative cell migration (%) in each group was determined by the migration distance normalized to that of the negative control group. Each experiment was independently performed in triplicate five times. Each error bar indicates the variation between the means of five independent experiments. The small black dots represent the mean of each independent experiment. ^*p < 0.05 and ^**p < 0.01, between the groups, using a two-tailed t-test. Exact p-values for asterisks (from left to right): 0.0033, 0.027, 0.0087, 0.0042, 0.045, 0.0081, 0.0043, 0.025, 0.0096, 0.0052, 0.031, 0.0093, 0.0036, 0.0045, 0.043, 0.0049, 0.0068, 0.042, 0.0043, 0.0062, 0.044, 0.0054, 0.0074, 0.046, 0.021, 0.033, 0.017, 0.029, 0.024, and 0.036. Source data are provided as a Source Data file.

Fig. 5. Minigene circuits specifically and efficiently inhibited in vivo tumor growth and metastasis.

a The design of minigene circuits targeting cellular genes p21 and E-cadherin. Among them, p21 and E-cadherin are located on the recipient cell genome, while other components are located on the AAV vector. b The tumor volume was calculated once every 5 days after injecting AAVs. The minigene circuit with sgRNA control that had no targeted intracellular gene was used as the negative control circuit. Tumors treated with the minigene circuit (n = 5 animals) targeting p21 were dramatically smaller than those treated with the negative control (n = 5 animals). Data are shown as means of tumor weight ± SD. The small blue dots represent the weight of each sample. ^**p < 0.01, relative to the negative control using the two-tailed t-test. Exact p-values for asterisks (from left to right): 0.0006, 0.0004, and 0.037. Source data are provided as a Source Data file. c Quantification for bioluminescence imaging of a metastatic model. The minigene circuit with sgRNA control that had no targeted intracellular gene was used as the negative control circuit. The luminescent signal intensities for either minigene circuit targeting E-cadherin (n = 5 animals) or negative control circuit (n = 5 animals) are shown. Only tumors in lung regions were quantified by intraperitoneal D-luciferin administration, and then the signal intensities in lungs were calculated. The small blue dots represent the mean signal intensity of each mouse. ^**p < 0.01 (p = 0.029), relative to the negative control using the two-tailed t-test. AAV adeno-associated virus. Source data are provided as a Source Data file.