Introduction

Despite the continuous progress in its prevention, diagnosis and treatment, cancer remains a major public health problem worldwide. According to GLOBOCAN data released by the International Agency for Research on Cancer (IARC), 19.2 million new cancer cases and 9.9 million deaths from cancer were reported in 2020 [1]. Traditional cancer therapies, including surgery, chemotherapy, radiotherapy, and the new modality of immunotherapy, have been explored in attempts to improve the quality of life and survival of cancer patients. However, current cancer therapies still have major limitations, such as systemic toxicity, drug resistance, the risk of recurrence and low response rates. Therefore, these limitations emphasise the need for novel development in reliable and nontoxic cancer therapy.

Nuclear factor κB (NF-κB) is a sequence-specific DNA-binding transcription factor that operates in physiological and pathological processes. It regulates target genes by binding to κB sites in the genome [2]. NF-κB is overactivated in the context of chronic inflammation and ultimately leads to cancer; activated NF-κB can promote cancer cell proliferation, invasion, and metastasis and prevent cancer cell apoptosis [3]. Various studies have also shown that most diseases are associated with aberrant activation of NF-κB and its signalling pathway. For instance, NF-κB is abnormally activated in many human cancers, promoting cell survival and malignancy by upregulating antiapoptotic genes. As it is constitutively overactivated in almost all cancer cells, NF-κB is an attractive target for cancer therapy [4], and various NF-κB inhibitors have been developed. However, NF-κB has been described as a double-edged sword [5] because its fundamental activity is indispensable for normal cellular functions such as tissue homoeostasis and immunity. Therefore, effective NF-κB inhibitors have failed in clinical trials because of their excessive inhibition of its activity. To overcome this limitation, we previously developed a cancer cell-specific therapeutic strategy exploiting NF-κB activity in cancer cells. For this strategy, an NF-κB-responsive promoter consisting of five high-affinity κB sites that act as decoys and a minimal promoter to control the expression of exogenous effector genes was constructed [6].

Over the past few years, there has been considerable enthusiasm about precision medicine approaches for cancer therapy [7]. The advent of CRISPR–Cas9 gene editing technology has changed genetic engineering [8]. Technical developments in CRISPR‒Cas gene editing could herald the next generation in gene therapy tools for human diseases, including cancer [9]. In contrast to the RNA-mediated DNA-targeting Cas9 protein, an RNA-mediated RNA-targeting CRISPR-related protein, Cas13a, also known as C2c2 [10], was reported in 2016; subsequently, the ‘collateral cleavage’ activity of the CRISPR–Cas13a system was used to develop the RNA detection system SHERLOCK [11]. Notably, Cas13a has no collateral cleavage activity in cells and retains only a highly specific RNA-targeted cleavage function. Studies have also confirmed that LwaCas13a derived from Leptotrichia wadei can be heterologously expressed in mammalian cells and can be used to effectively knock down reporter genes and endogenous genes under the guidance of a guide RNA (gRNA), and its efficiency is comparable to that of RNAi technology [12, 13]. In addition, LwaCas13a has significantly lower off-target activity, and this system is thus a more flexible platform for gene therapy for human diseases such as cancer. Despite the various advantages of this system, such as flexibility, safety and efficiency, there are currently few reports on cancer gene therapy strategies using the CRISPR–Cas13a system to target oncogene mRNAs [14].

In this project, we selectively target the NF-κB activity in cancer cells by utilising its specific promoter to activate the exogenous effector CRISPR-Cas13a gene. The objective is to suppress certain oncogenes at the mRNA level, followed by the intravenous injection of adeno-associated virus (AAV) as a gene delivery vector into a colorectal cancer mouse model to achieve cancer gene therapy in vivo.

Results

Effectively knocking down exogenous reporter genes via the HOPE vector

In view of the advantages and limitations of current NF-κB inhibition strategies, the NF-κB-activated, high-specificity, oncogene prevention element, referred to as the HOPE vector, was constructed in this study. Via the HOPE vector, effector genes can be simultaneously expressed under the control of the previously constructed NF-κB-specific promoter DMP, and gRNAs can be expressed under the control of the U6 promoter; the promoter DMP consists of five high-affinity NF-κB binding sites, referred to as decoys, and a minimal promoter.

To evaluate the exogenous gene knockdown efficiency of the HOPE strategy in vitro, gRNAs targeting the reporter gene ZsGreen, whose expression is characterised by the emission of green fluorescence, and mCherry, whose expression is characterised by the emission of red fluorescence, were designed, cloned and inserted into the pHOPE vector to generate the pHOPE-ZsGreen and pHOPE-mCherry plasmids, respectively. HEK-293T cells with NF-κB activity confirmed in previous studies were first cotransfected with pZsGreen, pHOPE-ZsGreen, pmCherry, and pHOPE-mCherry, with the ZsGreen transcript serving as the target and the mCherry transcript serving as the negative control. The results of microscopic imaging indicated that pHOPE-ZsGreen plasmid transfection significantly decreased the expression of the reporter gene ZsGreen, attenuating green fluorescence, but did not affect the expression of mCherry, as determined by measurement of red fluorescence (Fig. 1A). Moreover, when the mCherry transcript was used as the target and the ZsGreen transcript was used as the negative control, pHOPE-mCherry plasmid transfection obviously decreased the intensity of red fluorescence resulting from mCherry reporter gene expression but did not affect the intensity of green fluorescence resulting from ZsGreen expression (Fig. 1A). Then, the cells exhibiting green and red fluorescence generated by ZsGreen and mCherry expression were counted. The results suggested that the number of cells with green fluorescence after pHOPE-ZsGreen transfection was much lower than that after pZsGreen transfection alone (p < 0.0001). Correspondingly, compared with pmCherry transfection alone, pHOPE-mCherry transfection caused a significant decrease in the number of cells with red fluorescence (p < 0.0001) (Fig. 1B). In addition, a gRNA targeting no transcript (NT) in mammalian cells was designed and used to construct the pHOPE-NT plasmid, which served as a negative control. pHOPE-NT transfection affected neither the intensity of green fluorescence driven by ZsGreen nor the intensity of red fluorescence driven by mCherry.

Fig. 1: Interference with exogenous reporter gene expression in HEK-293T cells via the HOPE strategy.
figure 1

A Representative fluorescence images of HEK-293T cells at 24 h post transfection. pHOPE-NT/mCherry/ZsGreen, HOPE plasmids targeting no transcript and the mCherry and ZsGreen reporter gene transcripts. Phase, bright field; ZsGreen/mCherry, green fluorescent/red fluorescent field; scale bar, 50 μm. B Bar chart showing the numbers of cells exhibiting fluorescence. Cells exhibiting fluorescence were counted in the corresponding fluorescence images with ImageJ2 software (n = 3 images). C The mRNA expression of reporter genes was measured by qPCR. ZsGreen and mCherry gene expression was measured by qPCR (n = 3 wells). RQ = 2–ΔΔCt; RQ relative quantification.

Full size image

To subsequently confirm the exogenous gene knockdown efficiency of the HOPE system at the transcriptional level, the expression of the reporter genes ZsGreen and mCherry was measured by qPCR. The results indicated that the expression level of the ZsGreen gene was significantly lower in the pHOPE-ZsGreen-transfected cells than in the pZsGreen-transfected cells (p < 0.0001). Similarly, the expression level of the mCherry gene was significantly lower in the pHOPE-mCherry-transfected cells than in the pmCherry-transfected cells (p < 0.0001) (Fig. 1C). These results verified the high efficiency and high targeted gene knockdown ability of the HOPE strategy, consistent with our expectations and the results of previous studies showing that CRISPR–Cas13a can strongly and specifically knock down its target RNA.

In vitro cancer cell killing effect of the HOPE strategy

To evaluate whether the HOPE strategy can efficiently knock down endogenous oncogenes and induce cancer cells killing effect, gRNAs targeting the oncogenes TERT, PLK1, MYC, and KRAS and targeting these four oncogenes were designed. Human cancer cells, including SiHa, BGC823, and A549 cells, with NF-κB overactivation and high expression of target oncogenes (Supplementary Fig. S1), were transfected separately with the pHOPE-hTERT, pHOPE-hPLK1, pHOPE-hMYC, pHOPE-hKRAS, and pHOPE-hTPMK plasmids. Control cells, including HL7702, GES-1, and MRC-5 cells, with low NF-κB activity and low expression of target oncogenes (Supplementary Fig. S1), were transfected with the same plasmids at the same time. All cell lines were transfected with pHOPE-NT as a negative control. The AO/EB staining results revealed that compared to pHOPE-NT control group, fewer SiHa, BGC823, and A549 cells survived after transfection with pHOPE plasmids targeting the oncogenes TERT, PLK1, MYC, and KRAS and especially after transfection with pHOPE-hTPMK, than after transfection with pHOPE-NT; that is, the HOPE strategy can knock down oncogenes and elicit cancer cell-killing effects, and the killing efficiency of the HOPE system targeting four oncogenes was greater than that of the HOPE system with a single oncogene target (Fig. 2A). Moreover, the results of live cell counting indicated that fewer SiHa, BGC823, and A549 cells survived after transfection of pHOPE plasmids targeting either a single oncogene or multiple oncogenes than after pHOPE-NT transfection (p < 0.01). The HOPE strategy caused a significant decrease in the number of cancer cells and exhibited cytotoxic effects (Fig. 2B) on cancer cells. Notably, the pHOPE plasmid targeting multiple oncogenes had stronger cancer cell-killing effects than did those targeting a single oncogene in SiHa, BGC823, and A549 cells (p < 0.001) (Fig. 2B).

Fig. 2: Interference with endogenous oncogene expression via the HOPE strategy in vitro.
figure 2

A Representative fluorescence images of stained cancer cells at 24 h post transfection. Stain, acridine orange/ethidium bromide (AO/EB); scale bar, 50 μm. B Bar chart showing the numbers of viable cancer cells after transfection with the HOPE plasmids. SiHa, BGC823, and A549 cells were counted in the fluorescence images of AO/EB-stained cells with ImageJ2 software (n = 3 images). C Representative fluorescence images of AO/EB-stained normal cells at 24 h post transfection. Scale bar, 50 μm. D Bar chart showing the numbers of viable normal cells after transfection with the HOPE plasmids. HL7702, GES-1 and MRC-5 cells were counted in the fluorescence images of AO/EB-stained cells with ImageJ2 software (n = 3 images). pHOPE-NT/hTERT/hPLK1/hMYC/hKRAS, HOPE plasmids targeting no transcript and the human oncogenes TERT, PLK1, MYC, and KRAS.

Full size image

In addition, the results of AO/EB staining revealed that the transfection of neither the pHOPE plasmids targeting single oncogenes (TERT, PLK1, MYC, and KRAS) nor the pHOPE plasmid targeting multiple oncogenes affected the normal human cell lines HL7702, GES-1 and MRC-5 relative to the corresponding pHOPE-NT-transfected control groups of cells (Fig. 2C). There were no significant differences in the numbers of HL7702, GES-1 and MRC-5 cells between the pHOPE-NT control group and the pHOPE-targeting oncogene treatment groups (p > 0.05) (Fig. 2D). The negative control plasmid pHOPE-NT had no effect on cancer cells or normal cells. The above results indicated that the efficiency and specificity of the HOPE strategy for targeting oncogenes led to effective induction of cancer cell death but had no effect on normal cells and that the HOPE plasmid targeting multiple oncogenes had a more significant cytotoxic effect on cancer cells.

Killing colorectal cancer cell via the HOPE strategy

To investigate the ability of the HOPE strategy to kill colorectal cancer cells in vitro and to provide a foundation for colorectal cancer gene therapy in vivo, the mouse colorectal carcinoma cell line CT-26 and the human colorectal adenocarcinoma cell line HT-29, both of which exhibit NF-κB overactivation and high expression of target oncogenes (Supplementary Fig. S2A), were transfected separately with the pHOPE-NT plasmid and pHOPE plasmids targeting the mouse and human oncogenes TERT, PLK1, MYC, and KRAS for 24 h to 72 h. The results of AO/EB staining demonstrated that fewer CT-26 and HT-29 cells remained in the targeting groups than in the pHOPE-NT control groups and that the pHOPE plasmids targeting these oncogenes resulted in time-dependent cell death (Supplementary Fig. S2B); in particular, few CT-26 and HT-29 cells survived after transfection with the pHOPE plasmid targeting multiple oncogenes (Fig. 3A). Moreover, the results of cell counting indicated that compared with negative control pHOPE-NT transfection, transfection of pHOPE plasmids targeting either single or multiple oncogenes caused a significant decrease in the number of colorectal cancer cells (p < 0.001). At 72 h post-pHOPE plasmid transfection, fewer CT-26 and HT-29 cells remained viable than after pHOPE-NT transfection; in particular, the pHOPE plasmid targeting multiple oncogenes, namely, TERT, PLK1, MYC, and KRAS, had significantly stronger cytotoxic effects on colorectal cancer cells (p < 0.0001) (Fig. 3B). Moreover, the results of cell viability assays also revealed that the transfection of pHOPE plasmids targeting oncogenes decreased CT-26 and HT-29 cell growth compared with that after negative control pHOPE-NT transfection (Fig. 3C). Transfection of the pHOPE plasmid targeting multiple oncogenes induced significant decreases in the growth and viability of colorectal cancer cells. In addition, transfection with the negative control pHOPE-NT had no such effects on colorectal cancer cells (Fig. 3C).

Fig. 3: Colorectal cancer cell killing effect of the HOPE strategy in vitro.
figure 3

Colorectal cells, including CT-26 and HT-29 cells, were transfected with various HOPE plasmids and evaluated at 24, 48, and 72 h post transfection. A Representative fluorescence images of acridine orange/ethidium bromide (AO/EB)-stained CRC cells. Scale bar, 50 μm. B Bar chart showing the numbers of viable colorectal cancer cells after transfection with the HOPE plasmids. CT-26 and HT-29 cells were counted in the fluorescence images of AO/EB-stained cells with ImageJ2 software (n = 3 images). C Proliferation curve of colorectal cancer cells. The proliferation of CT-26 and HT-29 cells was evaluated by the CCK-8 method (n = 3 wells). D Relative mRNA expression of oncogenes. The expression of the TERT, PLK1, MYC, and KRAS genes was measured by qPCR at 48 h post transfection (n = 3 wells). RQ = 2–ΔΔCt; RQ relative quantification. pHOPE-NT/hTERT/hPLK1/hMYC/hKRAS, HOPE plasmids targeting no transcript and the human oncogenes TERT, PLK1, MYC, and KRAS. pHOPE-mTERT/mPLK1/mMYC/mKRAS, HOPE plasmids targeting the mouse oncogenes TERT, PLK1, MYC, and KRAS.

Full size image

To further verify the colorectal cancer cell killing efficiency of the HOPE strategy at the transcriptional level, the expression of the target genes, i.e., TERT, PLK1, MYC, and KRAS, was measured via qPCR. The results demonstrated that the mRNA expression levels of the targeted oncogenes, i.e., TERT, PLK1, MYC, and KRAS, were significantly lower in CT-26 cells transfected with pHOPE plasmids targeting the corresponding oncogenes (p < 0.001) than in those transfected with the negative control pHOPE-NT (Fig. 3D). The mRNA expression levels in cells transfected with the pHOPE plasmid targeting multiple oncogenes were lower than those in cells transfected with the pHOPE plasmids targeting a single oncogene. Similarly, the corresponding mRNA expression levels of the targeted oncogenes were obviously decreased in HT-29 cells transfected with pHOPE-hTERT, pHOPE-hPLK1, pHOPE-hMYC, pHOPE-KRAS, and pHOPE-hTPMK (p < 0.0001). Notably, the corresponding mRNA expression levels of the targeted oncogenes in HT-29 cells transfected with pHOPE-hTPMK were lower than those in cells transfected with the pHOPE plasmids targeting a single oncogene (Fig. 3D). Additionally, the mRNA expression levels of the targeted oncogenes were not affected by pHOPE-NT transfection in CT-26 and HT-29 cells.

Taken together, these results indicated that the HOPE strategy effectively knocked down the oncogenes TERT, PLK1, MYC, and KRAS at the transcriptional level and resulted in significant death of colorectal cancer cells with NF-κB overactivation and that the colorectal cancer cell killing efficiency of HOPE targeting multiple oncogenes was greater than that of HOPE targeting a single oncogene.

Colorectal cancer gene therapy using rAAV-HOPE

To further explore the effect of the HOPE strategy for cancer gene therapy in vivo, we selected PLK1, MYC, KRAS, and TPMK as the target genes due to the increased cancer cell killing efficiency of the corresponding HOPE plasmids observed in vitro and treated mice bearing colorectal cancer (CRC) cell homograft with rAAV-HOPE. CT-26 cells were subcutaneously transplanted into BALB/c mice to generate the CRC mouse model. On the 8th day post-tumour cell inoculation, the tumour-bearing mice were randomly divided into five groups and then intravenously administered PBS (n = 8) or rAAV-HOPE-mPLK1 (n = 8), rAAV-HOPE-mMYC (n = 8), rAAV-HOPE-mKRAS (n = 8), or rAAV-HOPE-mTPMK (n = 8) at a dose of 1 × 1010 vg/mouse (Fig. 4A). Compared to PBS treatment, rAAV-HOPE treatment inhibited tumour growth, as indicated by the slow and stable increase in body weight (Fig. 4B and Supplementary Fig. S3). The tumours in the mice injected with rAAV-HOPE-mPLK1, rAAV-HOPE-mMYC, rAAV-HOPE-mKRAS, or rAAV-HOPE-mTPMK were obviously smaller (p ≤ 0.05) than those in the mice in the PBS control group (Fig. 4C). Moreover, the photographs of the excised tumours suggested that rAAV-HOPE exerted a clear tumour growth inhibition effect, with smaller tumours in the rAAV-HOPE groups than in the PBS control group (Fig. 4D), and the corresponding measured tumour size in each rAAV-HOPE group was significantly smaller than that in the PBS control group (p ≤ 0.01) (Fig. 4E). The tumour weight results further confirmed that the rAAV-HOPE treatments effectively inhibited tumour growth, with obviously lower tumour weights in the rAAV-HOPE treatment groups than in the PBS control group (p ≤ 0.01) (Fig. 4F). According to the images of the harvested spleens, most of the spleens in the PBS control group exhibited swelling (Fig. 4G). The corresponding spleen weights in the rAAV-HOPE-mPLK1 and rAAV-HOPE-mKRAS treatment groups were lower (p ≤ 0.05) and those in the rAAV-HOPE-mMYC and rAAV-HOPE-mTPMK treatment groups were significantly lower (p ≤ 0.01) than those in the PBS control group (Fig. 4H).

Fig. 4: Treatment with rAAV-HOPE in a colorectal cancer mouse model.
figure 4

A Schematic diagram of the establishment of the CT-26 colorectal cancer (CRC) mouse model and treatment schedule. The CRC mouse model was established by subcutaneous (s.c.) injection of CT-26 cells, and the mice were then subjected to intravenous (i.v.) injection of phosphate-buffered saline (PBS) (n = 8), rAAV-HOPE-mPLK1 (n = 8), rAAV-HOPE-mMYC (n = 8), rAAV-HOPE-mKRAS (n = 8), or rAAV-HOPE-mTPMK (n = 8). B Body weight changes in CRC model mice post-tumour cell inoculation. C Tumour size in CRC model mice intravenously injected with PBS or various types of rAAV-HOPE. D Images of tumours from CRC model mice intravenously injected with PBS or various types of rAAV-HOPE. E Tumour size as determined from harvested tumour tissues. F Tumour weights as determined from harvested tumour tissues. G Images of spleens from colon cancer model mice intravenously injected with PBS and various types of rAAV-HOPE. H Weights of harvested spleens. I Measurement of serum biochemical indicators of kidney function. J Measurement of serum biochemical indicators of liver function. mPLK1/mMYC/mKRAS/mTPMK, rAAV-HOPE-mPLK1/rAAV-HOPE-mMYC/rAAV-HOPE-mKRAS/rAAV-HOPE-mTPMK.

Full size image

To explore the biosafety of the HOPE strategy for cancer gene therapy in vivo, blood and serum were collected from the mice. Then, the serum biochemical indicators UREA, UA, and CREA, which indicate kidney function, and ALP, ALT, AST, and γ-GT, which indicate liver function, were evaluated. Compared with the rAAV-HOPE treatment group, especially the rAAV-HOPE-TPMK group, the PBS non-treated group exhibited kidney damage, as indicated by the significantly higher levels of UREA, UA, and CREA (p ≤ 0.05) (Fig. 4I). Similarly, compared with the rAAV-HOPE-TPMK treatment group, the PBS-treated group also exhibited liver damage, as indicated by the significantly higher ALT, AST, and γ-GT levels (p ≤ 0.05) (Fig. 4J). The above results indicated that when AAV was used for gene delivery, the HOPE strategy was effectively applied for CRC gene therapy, with good biosafety in vivo.

Immune system activation and stemness reduction in cancer cells by the HOPE strategy

To explore the cellular diversity in tumour tissue, we integrated the scRNA-seq data of two rAAV-HOPE-treated samples and one PBS-treated sample (Fig. 5A). By analysis of marker gene expression, a total of 10 cell types in the CRC mouse model microenvironment were annotated and included T cells, NK cells, mast cells, DCs, B cells, neutrophils, macrophages, endothelial cells and fibroblasts (Fig. 5B). Then, the genes specifically expressed in each cell type were further identified (Fig. 5C). Based on the chromosomal landscape inferred through the scRNA-seq data, malignant cells were distinguished from non-malignant cells in the microenvironment (Supplementary Fig. S4). To investigate the differences in the tumour microenvironment between the two groups, the proportions of the cell types were compared. Compared with the PBS-treated samples, the rAAV-HOPE-treated samples exhibited a lower proportion of cancer cells and a greater proportion of immune cells, such as T cells and B cells, indicating a high level of immune infiltration (Fig. 5D, E). To investigate the reason that rAAV-HOPE treatment of cancer could have this strong effect on the mouse immune system, the differentially expressed genes in each cell type were identified and compared between the groups. A large number of genes were significantly more highly expressed in the rAAV-HOPE-treated samples than in the PBS-treated samples, especially in cancer cells (Fig. 5F and Supplementary Fig. S5). Based on these differentially expressed genes, the enriched pathways were identified to further investigate the differences between the two groups. Surprisingly, the significantly enriched pathways were dramatically different. In the rAAV-HOPE-treated group, the significantly enriched pathway was TNFα Signalling via NF-κB, which is a well-known pathway closely related to inflammation and inflammation-related immune responses [15], and was also closely related to the necroptosis [16], inducing the necroptosis might be activated in HOPE-treated group, which could further induce the death of cancer cells during the treating process (Fig. 5G and Supplementary Fig. S6). The other enriched pathways, such as the oestrogen response-related pathway and p53 pathway, play important roles in CRC (Fig. 5G). Pathways enriched in the PBS group were related to tumour progression, for example, the epithelial–mesenchymal transition pathway, which has been reported to play a role in the metastasis of colorectal cancer (Fig. 5G) [17]. The difference between the activated pathways might be caused by differences in transcription factor (TF)-based regulation in the two groups. For this reason, regulon analysis was performed to investigate the differences in gene expression regulation. As expected, the specifically upregulated TFs in the different cell types were dissimilar (Supplementary Fig. S7 and Supplementary Table S5–S14). In addition, the candidate TFs in the two groups were closely related to the enriched pathways shown in Fig. 5G. For example, Twist1 [18], Irf7 [19] and Sox4 [20] in PBS-treated cancer cells are crucial in the EMT pathway. In the rAAV-HOPE-treated group, several NF-κB pathway-related TFs, such as RelA, Rel and RelB, were highly expressed. We also investigated differences in cell‒cell communication among various cell types between the two groups to identify differences in the immunological microenvironment. The results showed that the number of interactions between cell types in the rAAV-HOPE-treated group was greater than that in the PBS-treated group (Supplementary Fig. S8), especially for interactions among cancer cells, T cells and B cells (Fig. 5H). Considering these results collectively, we concluded that the HOPE strategy can activate the immune response and decrease the percentage of cancer cells.

Fig. 5: Effect of HOPE treatment on cancer stem cells in the TME.
figure 5

A Schematic workflow for the data collection and scRNA-seq analysis to define the changes in the TME caused by HOPE treatment. B UMAP plot of cell types derived from the HOPE-treated and PBS non-treated samples. The cells are coloured according to cell type. C Bubble plot showing genes (columns) that were differentially expressed across various cell types (rows). The scale bar depicts low expression in white and high expression in red. The percentages of cells expressing specific genes are represented by the sizes of the circles. D UMAP plot of the HOPE-treated and PBS non-treated samples. The cell types are represented by different colours. E Sankey diagram showing the proportion of each cell type in the HOPE-treated and PBS non-treated samples. F Comparison of significant DEGs in each cell type between the HOPE-treated and PBS non-treated samples. The significantly upregulated genes in the HOPE-treated samples are shown with red dots. The blue dots represent genes significantly upregulated in the non-treated samples. G The pathways significantly enriched with the DEGs. Biological pathways in the MSigDB Hallmark database were used for enrichment analysis. Blue represents pathways enriched in PBS non-treated samples, and red represents pathways enriched in the HOPE-treated samples. H The numbers of intercellular communication interactions among cancer cells and other cell types in the HOPE-treated and PBS non-treated samples. The line thickness represents the number of interactions. HOPE-treated, rAAV-HOPE-treated; non-treated, PBS-treated.

Full size image

To investigate whether the composition of cancer cells is altered by rAAV-HOPE treatment, subclusters of cancer cells were defined. A total of five subclusters of cancer cells were defined (Fig. 6A). After comparing the proportions of cells in each subcluster in the two groups, we found that there was a unique subcluster in both groups (subcluster 4 in the PBS-treated group; subcluster 3 in the rAAV-HOPE-treated group) (Fig. 6B). The proportion of cells in subcluster 4 in the PBS-treated group was lower than that in subcluster 3 in the rAAV-HOPE-treated group (Fig. 6C). To explore the relationships among these cancer cell subclusters, pseudotime analysis was performed, and the cellular differentiation states were inferred from the stemness scores. The trajectories of the five subclusters were plotted: subcluster 3 was the end of the differentiation trajectory of the cancer cells, with the lowest stemness score, and subcluster 4 was also found to have a weak differentiation ability compared with that of the other subclusters (Fig. 6D, E and Supplementary Fig. S9). Based on these results, we inferred that the HOPE strategy can reduce the stemness of cancer cells. Consistent with this assumption, the proportion of cancer cells with low stemness scores was increased after rAAV-HOPE treatment, and not only the cancer cells belonging to subcluster 4 but also other cancer cells were induced to differentiate into cells with low stemness; these cells were defined as subcluster 3 in the rAAV-HOPE-treated samples. To validate this hypothesis, the distribution of the expression of the HOPE-targeted oncogenes TERT, PLK1, MYC, and KRAS was determined. Compared with those in the other cell subclusters, the expression levels of these oncogenes were significantly lower in subcluster 3 (Fig. 6F). Based on these results, we concluded that the knockdown of these oncogenes via the HOPE strategy reduced the stemness of cancer cells.

Fig. 6: The HOPE strategy reduced the stemness of cancer cells.
figure 6

A UMAP plot of subclusters derived from the HOPE-treated and non-treated samples. The subclusters are coloured according to cell type. B UMAP plot of cancer cell subclusters in the HOPE-treated and non-treated samples. The subclusters of cancer cells are represented by different colours. C Sankey diagram showing the proportions of the subclusters of cancer cells in the HOPE-treated and non-treated samples. D Pseudotime trajectories of the cancer cell subclusters. The differentiation direction was defined by the stemness score. The cancer cell subclusters are represented by different colours. E Comparison of stemness scores among the cancer cell subclusters. F The expression of oncogenes targeted by the HOPE strategy in HOPE-treated samples in the cancer cell subclusters. HOPE-treated, rAAV-HOPE-treated; non-treated, PBS-treated.

Full size image

Discussion

With the rapid development of genomic tools, functional genomics has become a research focus in the life sciences, and one of its central topics is gene expression regulation. Therefore, gene therapy has made a turnaround and is now used to treat various diseases, especially rare diseases and cancers [21]. A total of 48 gene therapy drugs had been approved for marketing worldwide as of 2023, but there is no true sense of the utility of cancer gene therapy drugs in vivo because of the impossibility of accurately controlling the number of introduced exogenous genes and because of the lack of endogenous gene targeting. Most cancers are not caused by mutations in coding sequences but by changes in gene expression levels. The most obvious feature of gene expression regulation is that it can result in the activation of specific genes at specific times and in specific cells to permit ‘predetermined’, orderly, and irreversible biological processes. Therefore, investigating gene expression regulation is very important for analysing the pathogenesis of diseases and developing new technologies for disease treatment. In this work, we developed a novel cancer gene therapy strategy named HOPE by exploiting the constitutive activity of NF-κB in cancer cells. In the HOPE strategy, a previously constructed NF-κB-responsive promoter is used to control the expression of the effector gene CRISPR-Cas13a [22], after which the expression of oncogenes is knocked down, ultimately leading to apoptosis in cancer cells. We confirmed that the HOPE strategy can specifically and efficiently knock down the expression of endogenous and exogenous target genes at the transcriptional level when the expression of oncogenes was inhibited, inducing a cancer cell killing effect in vitro but not affecting normal cells. The HOPE strategy provides a new tool for NF-κB-based cancer gene therapy.

In the HOPE strategy, a total of four well-known oncogenes were chosen as target genes for cancer gene therapy. TERT gene encodes telomerase, its transcriptional activity is most likely the key determinant in regulation of telomerase activity, which is specifically activated in almost all cancer cells, but is silent in most normal cells [23]. There is investigation reports that TERT level increases with the adenoma-carcinoma sequence, and telomerase activity present in 83% colorectal cancers (CRCs) but absent or at very low level in normal mucosa [24]. In a series of CRCs, the TERT mRNA level in plasma was related to tumours, stage II CRC patients with high TERT level showed significantly worse median overall survival and disease-free survival than patients with low TERT level [25]. Thus the TERT gene has become an attractive target for CRC and other cancers therapy; thus, the TERT gene has become an attractive target for cancer therapy [26]. PLK1 is overexpressed in primary CRC specimens as compared with normal colon mucosa, its overexpression is of pathogenic, prognostic and proliferative importance [27], and inhibition of PLK1 could lead to death of CRC cells by interfering with multiple stages of mitosis [28]. There has study reveals that PLK1 overexpression is significantly associated with CRC stage, particularly cancer progression to invasion and lymph node metastasis [29], evaluation of PLK1 mRNA expression in cancer patients belonging to post-surgical stage I or II provides an early parameter to judge their prognosis. Thus, PLK1 gene is expected to be a potential target for CRC therapy. MYC is one of the most frequently deregulated driver genes in cancers, which plays a pivotal role in the development, proliferation, metastasis and trigger chemoresistance of CRC, a high expression level of MYC is a negative prognostic marker in CRC [30]. The well-known Wnt, Notch and other signalling pathways closely related to the occurrence of CRC have been found to be associated with expression of the MYC gene [31]. Previous studies have found that the prognosis of CRC patients with MYC gene mutation was worse than that of patients without mutation [32]. Because of the high prevalence of deregulation and its causal role in CRC, MYC gene is theoretically an attractive target for treating CRC. KRAS is a commonly mutated oncogene in CRC, with mutation ratio of approximately 40% in all CRC cases [33]; its mutations result in constitutive activation of the KRAS protein, which acts as a molecular switch to persistently stimulate downstream signalling pathways, including cell proliferation and survival, thereby leading to tumorigenesis [34]. Studies have found that KRAS-mutant CRC harbour a more immunosuppressive microenvironment, which significantly limits the use of immune checkpoint inhibitors in these CRC patients. Patients with KRAS-mutant CRC have a poorer prognosis than those with KRAS wild-type. As KRAS gene plays a critical role in the prognosis, diagnosis and treatment of CRC, new technologies for CRC therapy focus on the directly targeting KRAS itself or RAS‑binding pocket, indirectly targeting downstream its signalling pathways, inhibiting the nucleotide exchange cycle, disruption of KRAS processing and membrane localisation [35].

In recent years, major breakthroughs have been made in research using CRISPR technology to construct disease models [36], in gene therapy for genetic diseases [37], and in modified CAR-T cells [38]. At the same time, there are technical bottlenecks, such as the inability to completely overcome the off-target effects of Cas9 activity, which also restricts the clinical application and the future of gene editing in vivo. The important innovation of CRISPR–Cas13a technology is that unlike CRISPR–Cas9, it can edit genes at the transcriptional level. If the DNA is a ‘blueprint’ for building cell morphology and function, the RNA is an ‘engineer’ that turns the ‘blueprint’ into reality. Compared with CRISPR–Cas9, CRISPR–Cas13a does not require a homology-guided repair mechanism or a PAM sequence at the target site, making it more flexible; it can process mature gRNA intrinsically, making it more advanced; and it can target multiple genes or a single gene with different transcripts simultaneously, making it more efficient. The most important feature is that RNA editing involves no changes to the DNA sequence of the coding gene but rather involves editing of the gene transcript, making it more controllable than DNA editing in terms of time and efficiency and safer for public use. Because of these advantages, CRISPR–Cas13a RNA editing technology has broad application prospects in gene therapy for diseases and is expected to become the core tool in the next generation of biotechnological applications after CRISPR–Cas9. In the HOPE strategy, CRISPR–Cas13a, which can effectively knock down target genes at the mRNA level, was chosen as an exogenous effector gene. It was worthy to note that once the construction of the presented HOPE backbone expression vector (including the U6 promoter, gRNA, DMP and Cas13a) is completed in this study, when targeting different genes or different transcripts of the same gene, it only needs to synthesise the corresponding two oligonucleotides, annealed and ligated by BbsI enzyme digestion, which has much simple operation and lower cost.

The success of cancer gene therapy depends in part on the delivery vehicle. The clinical use of CRISPR technology for treating cancers is still in its infancy, mostly due to the limitations of delivery vectors, and the delivery of vectors may still be the greatest bottleneck in gene editing in somatic cells [39]. Studies have shown that AAV, which is the only gene delivery vector approved for the delivery of CRISPR-based gene therapy systems directly into the human body, is characterised by high safety, a wide host cell range, low immunogenicity, strong targeting ability, and long-term stable expression of exogenous genes in vivo. There are currently AAV vectors carrying the CRISPR–Cas9 gene editing system for gene therapy for diseases such as Duchenne muscular dystrophy [40], progressive muscle atrophy [41], and motor nerve regeneration research [42]. The US FDA has approved gene therapies for haemophilia and spinal muscular atrophy, and AAV vectors have been used in clinical trials [43, 44]. More than 320 gene therapy clinical trials using AAVs as vectors have been registered at ClinicalTrials.gov as of 2023, and twenty-one of these therapies have been approved for clinical use in patients. Despite their various advantages, the major limiting factor of AAV vectors is their limited viral packaging capacity (~4.7 kb). Therefore, a previously developed cancer gene therapy named the Tage system had to be packaged into three AAVs [45], which not only limited the maximum dose of each AAV but also limited the cancer gene therapy effect of Tage to a certain extent. In this work, because of the use of a smaller Cas13a protein in the CRISPR family with a coding sequence of approximately 3.5 kb, the coding sequences of DMP, the U6 promoter and gRNA were also shorter, and we were able to package the HOPE vector targeting four oncogenes into one AAV. The design of this single rAAV-HOPE strategy is beneficial for increasing the dose, and the application of this rAAV is more effective than the current in vivo applications of multiple rAAV combinations. In addition, the promoter DMP used in the HOPE strategy is a weak promoter; it helps rAAV-HOPE to achieve long-term and stable transgene expression, which in turn facilitates the maximum expression of the Cas13a protein and maximises its ability to inhibit oncogene expression. Moreover, the HOPE strategy can completely treat colorectal cancer in mice through intravenous administration. It is a non-invasive technology for in vivo cancer gene therapy that circumvents the need for certain injurious treatment modalities, such as surgery, radiotherapy, and chemotherapy.

Colorectal cancer (CRC) is one of the most common malignant tumours of the digestive tract. There were 1.93 million new cases of CRC and 935,200 CRC-related deaths worldwide in 2020; CRC has the third-highest incidence and second-highest mortality rate among all malignant tumours [46]. The primary treatment for unresectable metastatic CRC is systemic biological therapy, such as antibodies against cellular growth factors, immunotherapy, gene therapy, and combinations thereof. However, for the 35% to 40% of patients with genetic mutations, effective targeted therapies are not yet available. In this work, we evaluated the therapeutic efficacy of rAAV-HOPE in a CRC mouse model in vivo and found that it achieved effective tumour inhibition with biological safety. According to the results of scRNA-seq of CRC tumour tissues, rAAV-HOPE treatment of tumour tissues could activate immune responses and decrease the proportion of cancer cells and that the differentially expressed genes after rAAV-HOPE treatment were immune-related genes enriched in the TNFα signalling via NF-κB pathway and other pathways important to CRC. Moreover, we found a unique subcluster defined as cancer cells in the rAAV-HOPE-treated group; the cells in this subcluster were at the end of the differentiation trajectory of cancer cells and had the lowest stemness scores. Moreover, the expression of the HOPE-targeted oncogenes TERT, PLK1, MYC, and KRAS was inhibited in this subcluster. That is, the HOPE strategy can reduce the stemness of cancer cells by knocking down its target oncogenes. Cancer stem cells (CSCs) are known to be the main cause of tumour recurrence and metastasis, and they limit the possibility for a complete cure. Many studies have shown an association between the tumour microenvironment (TME) and the acquisition of stemness. Our development of the HOPE strategy not only provides a novel technology for cancer gene therapy but also might offer a new approach for the successful treatment of cancer through modification of the TME by regulating CSCs.

Conclusion

In summary, we developed an NF-κB-activated cancer gene therapy named HOPE by the use of an NF-κB-specific promoter to activate exogenous Cas13a via a CRISPR system, resulting in the knockdown of the oncogenes TERT, PLK1, MYC, and KRAS. We confirmed that the HOPE strategy effectively killed cancer cells by inhibiting oncogene expression at the transcriptional level without affecting normal cells in vitro. In addition, using an AAV as a gene vector, we applied rAAV-HOPE to treat CRC in mice and verified that the HOPE strategy inhibited the growth of tumours in vivo. rAAV-HOPE also activated the immune system and decreased the proportion of cancer cells, specifically by reducing the stemness of cancer cells, based on our scRNA-seq analysis. These results elucidate the important role of the HOPE strategy in inhibiting cancer cell growth and provide a novel tool for cancer gene therapy.

Materials and methods

Vector construction

The NF-κB-specific promoter, named DMP, which consists of five high-affinity NF-κB binding sites and one minimal promoter sequence, was amplified from the pTsgRNA-DMP-Cas9 plasmid constructed previously [22] and was then cloned and inserted into the pAAV-MCS vector (Stratagene) using EcoRI and BamHI restriction enzymes (Thermo Fisher, Waltham, USA) to generate the pAAV-DMP plasmid. The artificially synthesised U6-gRNA sequences consisting of the U6 promoter sequence and gRNA direct repeat spacer separated by BbsI restriction sites were cloned and inserted into the pAAV-DMP vector using MIuI and EcoRI restriction enzymes to generate the pAAV-U6-gRNA-DMP backbone plasmid. The Cas13a coding sequence was amplified from the pET-His6-TwinStrep-SUMO-LwCas13a vector (Addgene, MA, USA) and was then cloned and inserted into the pAAV-U6-gRNA-DMP backbone vector using BamHI and SalI restriction enzymes to generate the pAAV-U6-gRNA-DMP-Cas13a plasmid (referred to as pHOPE). The primers used for pHOPE backbone vector construction are listed in Supplementary Table S1.

The gRNAs targeting NT (no transcript), ZsGreen, mCherry, and human and murine oncogenes, including TERT (telomerase reverse transcriptase), PLK1 (polo-like kinase 1), MYC (myelocytomatosis viral oncogene homologue), and KRAS (Kirsten rat sarcoma viral oncogene homologue), were chosen with CHOPCHOP online software (http://chopchop.cbu.uib.no/). Complementary oligonucleotides containing a 28-bp gRNA target-specific region and two flanking BbsI sites were artificially synthesised, annealed into double-stranded oligonucleotides, and ligated into the pHOPE vector via the Golden Gate Assembly method. The ligation reaction contained 10 units of BbsI (NEB, Ipswich, MA, USA), 1000 units of T4 DNA ligase (NEB), 1 μL of 10 × T4 DNA ligase buffer (NEB), 1 μL of 0.1 mg/mL bovine serum albumin (BSA) (NEB), double-stranded oligonucleotides (1 nM), 50 ng of the pHOPE backbone plasmid, and nuclease-free H2O to 10 μL. Ligation was performed in a thermal cycler as follows: 10 cycles at 37 °C for 5 min, 16 °C for 10 min, then 37 °C for 30 min, and 80 °C for 5 min. The generated plasmids were named pHOPE-NT, pHOPE-ZsGreen, pHOPE-mCherry, pHOPE-hTERT, pHOPE-hPLK1, pHOPE-hMYC, pHOPE-hKRAS, pHOPE-mTERT, pHOPE-mPLK1, pHOPE-mMYC, and pHOPE-mKRAS. One plasmid coexpressing the gRNAs targeting four human oncogenes (TERT, PLK1, MYC, and KRAS) was constructed and named pHOPE-hTPMK. In addition, another plasmid coexpressing the gRNAs targeting four murine oncogenes (TERT, PLK1, MYC, and KRAS), was constructed and named pHOPE-mTPMK. The oligonucleotides used for gRNA target construction are listed in Supplementary Table S2.

The pZsGreen and pmCherry reporter plasmids, in which the expression of ZsGreen and mCherry, respectively, was under the control of the CMV promoter, were maintained in our laboratory. The functional sequences are provided in Supplementary Table S3.

Cell culture and cell transfection

The following cancer cell lines were used in this study: SiHa (human cervical carcinoma cells), BGC823 (human gastric cancer cells), A549 (human non-small cell lung cancer cells), HT-29 (human colorectal adenocarcinoma cells), and CT-26 (mouse colorectal carcinoma cells). The following normal cell lines were used in this study: HL7702 (human normal hepatocytes), GES-1 (human gastric epithelial cells), MRC-5 (human embryonic lung fibroblasts), and HEK-293T (human embryonic kidney cells). SiHa, BGC823, and HEK-293T cells were cultured in Dulbecco’s modified Eagle’s medium (DMEM) (Gibco, California, USA). A549, HT-29, CT-26, HL7702, and GES-1 cells were cultured in Roswell Park Memorial Institute (RPMI) 1640 medium (Gibco). MRC-5 cells were cultured in minimum essential medium (MEM) (Gibco). All media were supplemented with 10% fetal bovine serum (FBS) (Sigma-Aldrich, Missouri, USA), 100 units/mL penicillin and 100 µg/mL streptomycin (Gibco). The cell lines were originally purchased from the Cell Resource Center of the Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, and were incubated at 37 °C in a humidified incubator with 5% CO2.

For cell transfection, 293T, SiHa, BGC823, A549, HL7702, GES-1, MRC-5, CT-26, and HT-29 cells were seeded into 24-well plates at a density of 1 × 105 cells/well and cultured overnight. Then, the cells were transfected with 500 ng of various plasmids using Lipofectamine® 2000 (Thermo Fisher) according to the manufacturer’s instructions. After being cultured for 4 h, the medium in each well was replaced with 500 μL of fresh DMEM, RPMI 1640 medium or MEM containing 10% FBS. After transfection, the cells were washed with PBS and then stained with acridine orange/ethidium bromide (AO/EB) (Sangon Biotech, Shanghai, China) according to the manufacturer’s instructions; the live cells appeared uniformly green. The cells were observed and imaged with a fluorescence microscope (IX51, Olympus, Tokyo, Japan) and counted with ImageJ2 software.

Cell viability assay

SiHa, BGC823, A549, HL7702, GES-1, MRC-5, CT-26, and HT-29 cells were seeded into 96-well plates at a density of 5 × 103 cells/well and cultured overnight. Then, the cells were transfected with 100 ng of various plasmids using Lipofectamine® 2000. After transfection, 10 μL of Cell Counting Kit-8 (CCK-8) reagent (Yeasen, Shanghai, China) was added to the cells in each well, and the cells were cultured in a 37 °C, 5% CO2 humidified incubator for 1 h. Then, the absorbance of each well was measured at 450 nm with a microplate plate reader (Tecan, Männedorf, Switzerland). Wells without transfected cells were regarded as blank wells, and the percentage of cell viability (%) was calculated as follows: cell viability (%) = [A450 (treated wells) − A450 (blank wells)]/[A450 (control wells) − A450 (blank wells)] × 100%.

Reverse transcription quantitative PCR

293T, CT-26 and HT-29 cells were seeded into 6-well plates at a density of 1.5 × 106 cells/well and cultured overnight. Then, the cells were transfected with 2500 ng of various plasmids using Lipofectamine® 2000. Twenty-four hours posttransfection, total RNA was isolated using TRIzol® reagent (Invitrogen, MA, USA) according to the manufacturer’s instructions. The concentration of RNA was quantified using a NanoDrop 2000 spectrophotometer (Thermo Fisher), and 500 ng of total RNA was reverse transcribed into complementary DNA (cDNA) using Hifair® V One-step RT-gDNA Digestion SuperMix for qPCR (Yeasen) according to the manufacturer’s instructions in a total volume of 20 µL. The mRNA expression levels of ZsGreen, mCherry, TERT, PLK1, MYC, KRAS, and RelA were quantified via quantitative PCR (qPCR) using an Analytic Jena qTOWER3 G thermal cycler. The qPCR mixture contained 10 µL of Hieff UNICON® Universal Blue qPCR SYBR Green Master Mix (2×) (Yeasen), the forward and reverse primers (0.2 µM each), 2 µL of cDNA, and nuclease-free H2O to a final volume of 20 μL. The thermal cycling programme used for qPCR was as follows: 95 °C for 2 min followed by 45 cycles at 95 °C for 10 s, and 60 °C for 30 s. Melting curve analysis revealed a single PCR product. The Ct values of the target genes were normalised by subtracting the Ct values for GAPDH and calculated by the instrument’s software. mRNA expression levels of target genes were calculated as relative quantity (RQ) values: RQ = 2−ΔΔCt, where ΔCt = Cttarget − CtGADPH, and ΔΔCt = ΔCttreatment − ΔCtcontrol. mRNA expression levels of RelA were calculated as 2−ΔCt. Each qPCR analysis was performed with at least three technical replicates. The primers used for qPCR are listed in Supplementary Table S4.

rAAV preparation

HEK-293T cells were added to 75-cm2 flasks at a density of 5 × 106 cells/flask and cultured overnight. Then, the cells were transfected with 4 μg of pAAV-RC (Stratagene), 4 μg of pAAV-Helper (Stratagene), and 4 μg of one pAAV plasmid (pHOPE-mTERT, pHOPE-mPLK1, pHOPE-mMYC, pHOPE-mKRAS, or pHOPE-mTPMK) using Lipofectamine® 2000. Seventy-two hours posttransfection, the cells were purified with the AAVpro® Purification Kit Maxi (TaKaRa) according to the manufacturer’s instructions. The final recombinant AAV (rAAV) products were referred to as rAAV-HOPE-mTERT, rAAV-HOPE-mPLK1, rAAV-HOPE-mMYC, rAAV-HOPE-mKRAS, and rAAV-HOPE-mTPMK.

Animal experiment

Four-week-old female BALB/c mice weighing 18–22 g and purchased from Changzhou Cavens Laboratory Animal Co., Ltd (China), were used in this study. All mice were maintained under specific pathogen-free conditions. Each mouse was subcutaneously injected with 1 × 106 CT-26 cells in the unilateral inner thighs to establish the CT-26 homograft model of colorectal cancer and maintained for tumour formation. On the 8th day after tumour cell implantation, the tumour-bearing mice were randomly divided into five groups—the PBS (n = 8 mice), rAAV-HOPE-mPLK1 (n = 8 mice), rAAV-HOPE-mMYC (n = 8 mice), rAAV-HOPE-mKRAS (n = 8 mice), and rAAV-HOPE-mTPMK (n = 8 mice) groups—and then injected intravenously with PBS or 1 × 1010 vg of rAAV-HOPE-mPLK1, rAAV-HOPE-mMYC, rAAV-HOPE-mKRAS, or rAAV-HOPE-mTPMK every other day for a total of three injections. Body weight and tumour size were measured every other day. The tumour volume (V) was calculated as V = (Dd2)/2, where D and d are the lengths of the major and minor tumour axes, respectively. The mice were euthanized and photographed on the 14th day after the first AAV injection. The tumour and spleen were collected from each mouse and were then weighed and photographed. Serum samples were collected for biochemical analysis, including measurement of alanine transaminase (ALT), aspartate amino transferase (AST), alkaline phosphatase (ALP), and gamma-glutamyl transferase (γ-GT) to assess liver function and urea, creatinine (CREA), and uric acid (UA) to assess kidney function.

Library preparation and scRNA-seq analysis

Fresh tumour tissues collected from mice injected with PBS, rAAV-HOPE-mPLK1, rAAV-HOPE-mMYC, rAAV-HOPE-mKRAS, or rAAV-HOPE-mTPMK were immediately placed in GEXSCOPE Tissue Preservation Solution (Singleron Biotechnology) at 2–8 °C. Then, the tissue samples were dissociated into single-cell suspensions using a Singleron PythoN™ Tissue Separator (Singleron Biotechnology) and sCelLiVE® Tissue Preservation Solution (Singleron Biotechnology) according to the manufacturer’s instructions. Finally, the tissue samples were stained with trypan blue (Sigma-Aldrich), the cell viability was evaluated under a fluorescence microscope (Olympus).

Single-cell suspensions with a concentration of 1 × 105 cells/mL were loaded into microfluidic devices provided with the Singleron Matrix Single-Cell Processing System (Singleron Biotechnology). Then, the scRNA-seq libraries were prepared using GEXSCOPE® Single-Cell RNA Library Kits (Singleron Biotechnology, #5180011) according to the manufacturer’s instructions, diluted to 4 nM and pooled for sequencing. All libraries were subjected to 150 bp paired-end sequencing on an Illumina NovaSeq 6000 instrument.

Processing and scRNA-seq data analysis

The Singleron CeleScope pipeline was used to analyse the raw reads and generate a gene count matrix with default parameters [47]. Briefly, the raw data in FASTQ format were mapped to the GRCm38 (mm10) reference genome using the STAR algorithm. Then, the unique molecular identifier (UMI) and barcode counts were determined to generate gene–barcode matrices for each sample. Genes without expression values in all cells were removed. A gene–barcode matrix containing gene expression counts was generated for subsequent analysis. All additional analyses except for CNV analysis were performed using the Seurat V5.0.0 (http://satijalab.org/seurat/) R toolkit [48]. Cells with <200 or >7500 expressed genes were excluded to eliminate the influence of low-quality cells. The maximum percentage of UMIs mapped to mitochondria was set to 10%. Finally, a total of 33,612 cells were retained after quality control for further downstream analysis, specifically, 23,707 cells from rAAV-HOPE-treated samples and 9905 cells from PBS-treated samples.

Dimensionality reduction

For each dataset, the top 2000 variable features were identified using the ‘vst’ method. The datasets were anchored and integrated with the integration procedure in the Seurat package to remove batch effects between samples. A linear scaling transformation was applied to the identified variable features using the ScaleData function with default parameters. Principal component analysis (PCA) was performed on the scaled features for dimensionality reduction. The first 30 principal components were used to identify the neighbours and cluster the cells with a resolution of 0.1. The cell clusters were visualised on 2D uniform manifold approximation projection (UMAP) plots. The genes specifically expressed in each cell cluster were identified using the FindAllMarkers function. The cell types were defined based on the expression of well-established gene markers.

Identification of cancer cells

To identify malignant and non-malignant cells, two approaches were used to distinguish malignant cells from non-malignant cells in each sample. First, we identified malignant cells using the marker genes Mki67, Sox4 and Sparc. During this procedure, we found a cell cluster with high expression of malignant epithelial cell marker genes, including Epcam, Krt8, Krt18 and Krt19. We defined the cells in this cluster as cancer cells. To verify the identified cancer cells, the inferCNV R package was used to quantify copy number variants [49], with the rest of the cell types used as references. CNV analysis was performed using Seurat v4.4.0 (http://satijalab.org/seurat/).

Identification of specifically expressed genes and enriched functions in each group

To compare the features of the rAAV-HOPE-treated and PBS-treated groups, the differentially expressed genes (DEGs) in each cell type were identified using the FindMarkers function with the default settings. The lists of significant DEGs were uploaded to the Enricher (https://maayanlab.cloud/Enricher/) online tool, and the top 10 terms according to the adjusted p value were retained.

Cell‒cell communication analysis

CellPhoneDB [50] was used to analyse and compare cell‒cell communication between rAAV-HOPE-treated and PBS-treated samples. The mouse gene names were converted to the corresponding human gene names with a script developed in house.

Differential regulatory activity analysis

Regulon analysis was performed by utilising pySCENIC [51], a computational tool for identifying regulons and assessing the activity score of regulons in individual cells, with default parameters. The differentially activated regulons were defined using the Wilcoxon rank–sum test. Fold changes in activity scores and p values were also calculated.

Characterisation of the differentiation trajectories of cancer cell subclusters

The differentiation trajectories of different cancer cell subclusters were determined using Monocle2 [52]. Genes expressed in at least 10 cells and in more than 10% of cells were used to perform pseudotime analysis.

Calculation of the stemness score

The stemness score was calculated with CytoTRACE [53], an established computational framework for stemness evaluation based on transcriptional diversity. The parameters used in this analysis were all set to the default values.

Statistical analysis

The data are expressed as the means ± standard deviations (SD). The statistical significance of differences between two groups was analysed by two-tailed unpaired Student’s t test. Comparisons among three or more groups were performed by one-way or two-way analysis of variance (ANOVA) with the Tukey test for multiple comparisons. Differences for which p < 0.05 were considered statistically significant.