Tumor Growth and Metastasis in LoVo Cells infected with hGeCKO Library lentivirus
We utilized a pooled genome-wide human sgRNA library (termed human genome-scale CRISPR knockout library, or hGeCKO) containing 123411 sgRNAs targeting 19050 genes and microRNA precursors in human genome[15]. The library also contains 1,000 control sgRNAs (termed non-targeting sgRNAs)[15, 21]. We transduced the colorectal cancer cell line LoVo with the hGeCKO library in three independent infection replicate experiments. for each replicate, the library representation (cells per lentiviral CRISPR construct) was greater than 500×.
After in vitro culture for 1 week, we injected 5×106 cells into the spleen of BALB/c nude mice. (Fig. 1A). We transplanted the cells from each hGeCKO transduced replicate into three mice, and untransduced cells as control group. After 6 and 8 weeks of injection, we imaged the mice and assessed tumor fluorescence intensity in liver metastases and spleen primary tumors. The hGeCKO-transduced group exhibited a tumor fluorescence intensity in liver metastases and p spleen primary tumors comparable to that of the untransfected group. (Fig. 1B-C). Subsequently, at the culmination of the eighth week, mice were sacrificed, and their spleens and livers were harvested for histopathological examination using H&E staining. The H&E staining results provided unequivocal evidence of tumor presence in both the hGeCKO-transduced group and the untransfected groups, with a remarkably similar number of liver metastatic lesions (Fig. 1D-E). These findings underscore the negligible impact of lentiviral infection on tumor initiation and metastatic potential. Further investigations entailed sequencing analysis of the spleen and liver metastases from the hGeCKO-transduced group mice to elucidate the enrichment patterns of sgRNA and identify pivotal regulatory genes that functionally drive liver metastasis.
Representation of hGeCKO Library at primary Tumor and Metastasis
To investigate the sgRNA representation of tumors from spleen and liver and identify genes where loss of function confers metastatic phenotype, we used deep sequencing to readout the sgRNA representation. At 8 weeks post transplantation, we sequenced the spleen primary tumors and liver metastases of each of the hGeCKO-transduced mice (Fig. 1). In parallel, we also sequenced the pre-transplantation hGeCKO-transduced LoVo cell pool (cultured in vitro for 7 days after transduction). In three spleen samples, the sgRNA representations showed high concordance in biological replicates (correlation, r = 0.75 on average, n = 3) (Fig. 2A). Three liver samples also showed high concordance of sgRNA representations and clustered tightly with each other (correlation, r = 0.77 on average, n = 3) (Fig. 2A). Meanwhile, the sgRNA representations originating from the liver and spleen of the same murine displayed pronounced positive correlation (correlation, r = 0.86 on average, n = 3), indicative of the source of sgRNA in liver metastases originated from the spleen (Fig. 2A). To investigate the sgRNA library dynamics in different sample types (pre-transplantation cell pool, spleen primary tumors, and liver metastases), we compared the overall distributions of sgRNAs from all samples sequenced. Compared to the pre-transplantation cell pool sample, only a small fraction of sgRNA were detected in the spleen primary tumor samples (Fig. 2B-C). Remarkably, the sgRNA abundance observed in liver tumor specimens closely parallels that detected in corresponding spleen tumor specimens derived from the identical murine model (Fig. 2B). Despite a marked reduction in the diversity of sgRNAs observed in both spleen and liver tumor samples, there was a striking surge in the normalized read counts of sgRNAs (Fig. 2C). The distribution patterns of sgRNA in tumor samples exhibit distinct variations compared to the pre-transplantation cell pool samples, as clearly demonstrated by the pronounced shifts observed in their respective cumulative distribution functions (Fig. 2D). The plausible mechanism can be attributed to the exclusive derivation of liver metastatic tumors from the spleen primary tumors. As a result, sgRNAs that expeditiously drive the metastatic genes exhibit a conspicuous enrichment within the spleen primary tumors, while sgRNAs that exert inhibitory effects on the metastatic genes accumulate within the liver metastases. Remarkably, genes devoid of any association with the metastatic cascade manifest simultaneous presence both within the spleen primary tumors and liver metastases.
Enriched sgRNAs from the hGeCKO Screen in Primary Tumors and Metastasis
Using stringent criteria (false discovery rate [FDR] < 0.5%), we identified significantly enriched sgRNAs in spleen primary tumor samples and liver metastatic samples. Through systematic ranking of sgRNA counts data derived from spleen primary tumors and liver metastatic tumors originating from three distinct murine model, we have successfully identified a repertoire of high-rank and poor-rank sgRNAs and associated genes (Fig. 3A-C). The ranking and counts statistical analysis of sgRNAs yields intriguing insights into the dynamic landscape of gene regulation governing CRC metastasis. Notably, the sgRNAs of paramount genes exhibit pronounced enrichment within the spleen primary tumor while manifesting a discernible attenuation in liver metastases (Fig. 3D-F). These genes represent significant regulatory elements that drive the process of liver metastasis in CRC, such as ANKRD42, FANC, IRF2BP2 (Fig. 3A-F). Conversely, the top genes of sgRNAs exhibiting enriched in liver metastases but decreased in spleen primary tumors comprise crucial regulatory factors functioning as potent inhibitors of liver metastasis in CRC, such as TCP10L, PABPC3, PNLIP (Fig. 3G-H). Taking these results, we further grouped all the potential genes involved in CRC liver metastases identified from CRISPR knockout screening. The important genes were jointly identified from spleen primary tumors and liver metastatic tumors. Simultaneously, the sgRNA of ANKRD42 demonstrates enrichment in the spleen primary tumors of all three mice, while being lost or downregulated in liver metastatic tumors (Fig. 3I). These findings strongly indicate that ANKRD42 likely plays a pivotal regulatory role in the process of CRC liver metastasis. These results presented herein elucidate that the precise enrichment of a repertoire of key genes driving CRC liver metastasis can be achieved by the CRISPR knockout screening in vivo.
Upregulated ANKRD42 expression associates with augmented migration and invasion in colorectal cancer
In order to demonstrate the reliability of our proposed experimental setup in identifying key contributors to CRC liver metastasis, we selected the ANKRD42 gene, which emerged as one of the top hits from our screenings, for subsequent analyses. ANKRD42 exhibited widespread expression in the gastrointestinal tract and various other tissues, with its subcellular localization predominantly predicted to be within the nucleus. Notably, ANKRD42 was found to be actively involved in promoting the positive regulation of NF-κB transcription factor activity[22]. Furthermore, ANKRD42 demonstrated early predictive capabilities for the development of systemic inflammatory response syndrome or sepsis[23]. Nevertheless, the functional implications of ANKRD42 in the development and progression of CRC, and even other malignancies, have yet to be elucidated. Through in-depth analysis of The Cancer Genome Atlas (TCGA) database, it has come to light that ANKRD42 exhibits prominent upregulation across a diverse spectrum of tumor tissues, displaying notably heightened expression levels in various malignancies, most notably in colorectal cancer, pancreatic cancer, gastric cancer, and other malignancies (Fig. 4A). Furthermore, investigation employing paired analysis of tumor tissues and adjacent noncancerous tissues demonstrated overexpression of ANKRD42 in colorectal cancer (Fig. 4B). The amalgamation of these compelling database insights with the elucidating findings obtained from the CRISPR screening conducted in this study collectively substantiates the pivotal regulatory role ascribed to ANKRD42 in the intricate processes underpinning the development and progression of CRC.
Given the potential pivotal role ANKRD42 might play in the intricate cascade of CRC initiation and metastasis, it becomes critically necessary to substantiate this hypothesis with compelling experimental evidence. Herein, we embarked on an extensive and rigorous series of in vitro investigations to unravel the functional dimensions of ANKRD42 in a cellular context. First and foremost, we successfully engineered three distinct shRNA constructs targeting the ANKRD42 gene. Western blotting analysis and quantitative real-time PCR (RT-qPCR) results revealed that shRNA1# and shRNA3# emerged as the most potent in suppressing ANKRD42 expression, thus warranting their selection for subsequent experiments (Fig. 4C). Subsequently we employed shRNA-mediated knockdown to meticulously dissect the subsequent alterations in the migration and invasion of CRC tumor cells. Strikingly, relative to the corresponding control groups, ANKRD42 knockdown led to a remarkable attenuation in both migration and invasion processes in LoVo cells (Fig. 4D-E). Deepening our investigation, quantitative assessments based on statistical analyses via wound healing assays illuminated a conspicuous decline in the migratory abilities of LoVo cells following ANKRD42 depletion (Fig. 4F-G). Remarkably, our exploratory endeavors also encompassed the probing of ANKRD42's plausible impact on cellular proliferation. Intriguingly, we observed a discernible reduction in cell proliferation as an immediate repercussion of ANKRD42 knockdown (Fig. 4H). Based on these compelling findings obtained from these cellular experiments, the confluence of evidence reinforces the indispensable role of ANKRD42 in augmenting the migration and invasion propensities of CRC cells.
Investigating the regulatory role of ANKRD42 in the progression and metastasis of colorectal cancer in vivo
To gain further insights into the role of ANKRD42 in colorectal cancer (CRC) progression and metastasis, we conducted in vivo studies using mouse models. Initially, we employed targeted knockdown of ANKRD42 in luciferase-expressing LoVo cells, followed by the establishment of a mouse model of liver metastasis through splenic injection of these modified cells into nude mice. After a six-week observation period, a significant decrease in tumor fluorescence intensity was revealed in the ANKRD42 knockdown group compared to the control group through in vivo imaging (Fig. 5A-B). Through continuous monitoring of body weight, we observed minimal weight changes in the ANKRD42 knockdown group, whereas the control group displayed a progressive decline, potentially attributed to a higher tumor burden (Fig. 5C). Subsequent examination of the isolated mouse livers provided compelling evidence supporting our findings. The control group exhibited numerous visible tumor metastatic foci, while the ANKRD42 knockdown group displayed a remarkable absence of such foci (Fig. 5D-E). Given that tumor metastasis is a leading cause of CRC-related mortality in clinical settings, we also investigated the impact of ANKRD42 depletion on mouse survival. To accomplish this, we utilized the same LoVo cell line and spleen injection method, monitoring the mice for eight weeks. Furthermore, survival analysis based on the mouse cohorts revealed striking disparities. The control group exhibited an alarming mortality rate of 80%, while the ANKRD42 knockdown group displayed a significantly lower rate of 20%, indicating a substantial improvement in overall survival compared to the control group (Fig. 5F). Histological analysis using H&E staining of the liver metastatic lesions further confirmed the stark contrast between the groups: each mouse in the control group showed distinct metastatic foci, whereas almost no metastatic foci were observed in the ANKRD42 knockdown group (Fig. 5G). These results clearly demonstrate that ANKRD42 deficiency significantly suppresses CRC tumor metastasis to the liver. Collectively, these experimental findings shed light on the crucial role of ANKRD42 in facilitating CRC tumor metastasis and progression.
ANKRD42 orchestrates colorectal cancer metastasis by regulating EMT process
In confirmation of our previous finding, we experimentally validated the pro-metastatic role of ANKRD42 in CRC through in vivo and in vitro experiments. Moreover, we substantiated the functional significance of ANKRD42 in CRC clinical specimens across multiple dimensions. ANKRD42 mRNA levels were assessed in 38 unpaired tumor specimens using RT-qPCR, which included adjacent normal tissues, primary tumors, and distant metastatic lesions. Our results unveiled an increased expression of ANKRD42 in primary tumors, with particularly elevated levels observed in distant metastatic lesions compared to adjacent normal tissues (Fig. 6A). Additionally, the RT-qPCR analysis of 15 paired specimens, including adjacent normal tissues, primary lesions, and liver metastases, revealed higher expression levels of ANKRD42 in both primary lesions and liver metastases compared to adjacent normal tissues, with the latter demonstrating even higher levels (Fig. 6B-C). Furthermore, we assembled 12 paired specimens featuring primary tumors and peritoneal metastases, alongside an additional 7 cases exhibiting primary tumors with ovarian metastases. The ensuing RT-qPCR analysis distinctly revealed heightened expression levels of ANKRD42 in both liver and ovarian metastatic tissues compared to primary tumors (Fig. 6D-E). Collectively, these results provide additional evidence supporting the pivotal regulatory role of ANKRD42 in the distant metastasis of CRC.
As widely acknowledged, the EMT process assumes a crucial role in tumor metastasis[24]. In pursuit of this understanding, we delved into the TCGA database to scrutinize the interplay between ANKRD42 and the expression of EMT markers. Our investigation uncovered a positive correlation between ANKRD42 and the expression of several genes known to facilitate the EMT process, such as CDH2, VIM, SANI2, and ZEB1 (Fig. 6F). Concurrently, a review of the literature revealed that ANKRD42 can intricately interact with NFκB, thereby promoting the expression of pertinent genes[25]. Given NFκB 's regulatory role in EMT-associated genes, such as SANI2 and CDH2[26], we posit that ANKRD42 plays a pivotal role in driving the metastatic cascade by orchestrating the expression of EMT-related genes. To unravel the specific mechanisms underlying ANKRD42-mediated CRC metastasis, we successfully generated patient-derived organoids (PDO) from freshly obtained surgical specimens of primary tumors and paired liver metastases (Fig. 6G). Through meticulous conditions, we achieved stable transfection of GFP-shRNA into liver metastasis-derived PDOs, resulting in an effective knockdown of ANKRD42 (Fig. 6H-I). Subsequent RT-qPCR analyses unveiled a significant downregulation of EMT genes, including CDH2 and SANI2, coupled with an upregulation of CDH1, as compared to the shRNA-NC group (Fig. 6J-K). These compelling findings strongly suggest that ANKRD42 propels CRC metastasis by upregulating EMT-related genes.