Introduction

Colorectal cancer (CRC) is the third most common cancer diagnosed globally and is one of the leading causes of cancer-associated mortality. Its progression is classified into four stages, with stages 3 and 4 being marked by an increased rate of malignancy1,2. If the cancer is diagnosed in the early stages, the tumor can be removed surgically; however, as the tumor gains metastatic ability, the overall survival rate of patients decreases drastically. Therefore, understanding the molecular mechanisms underlying CRC progression, which ultimately leads to metastasis, is crucial for developing effective therapeutic strategies.

Metastasis begins with the migration and invasion of cancer cells from the original tumor site to a distant organ. This process involves complex interactions between the signaling pathways and various proteins. Migration and invasion enhancer 1 (MIEN1) is an important protein involved in this process. Previously the protein was referred to by other names, such as C17orf37 and C35. The gene for this protein is located on chromosome 17q12 positioned between ERBB2 oncogene and GRB7 in a tail-to-tail orientation, adjacent to the ERBB2/HER2 oncogene. It encodes a 112-amino acid protein of 15kD. Unlike many oncogenes driven by gene amplification, MIEN1 overexpression appears to be primarily regulated at the transcriptional level, including epigenetic changes such as promoter hypomethylation. The MIEN1 protein is overexpressed across various cancer types, significantly augmenting invasive and migratory phenotypes in breast, prostate, oral, and non–small cell lung cancers3,4,5, but has no detectable expression in normal or noncancerous adjacent tissues. Immuno-histochemical analysis of CRC patient samples showed similar MIEN1 overexpression, which is associated with cancer aggressiveness and metastasis6.

In this study, we elucidated the role of MIEN1 in CRC by employing promoter knockout studies to harness its potential for therapeutic intervention. Using the highly tumorigenic HT29-human colorectal carcinoma cell line, we generated a stable MIEN1 knockout cell line using CRISPR-Cas9 gene editing technology. We characterized the resulting phenotypic alterations and gene expression changes to elucidate the role of MIEN1 in CRC cell migration and invasion. Additionally, we examined the effects of MIEN1 knockout on actin cytoskeleton dynamics, which are pivotal determinants of cell migration. Our findings demonstrate the critical involvement of MIEN1 in CRC cell migration and invasion.

Results

CRISPR-Cas9–mediated MIEN1 promoter ablation does not affect ERBB2 amplicon

MIEN1 is positioned between GRB7 and ERBB2 in a tail-to-tail orientation, adjacent to the ERBB2/HER2 oncogene (Fig. 1A). Its expression is regulated transcriptionally rather than through alterations in the copy number of the associated amplicon75,8. Hence, to investigate the effect of the MIEN1 protein on CRC cell migration and invasion, we used the CRISPR-Cas9 technology to knock out its promoter region, ensuring that the ERBB2 amplicon remained unaffected.

Fig. 1
figure 1

MIEN1 promoter knockout and cell line characterization. (A) Schematic representation of chromosome 17 of the human genome, depicting the location of the MIEN1 gene. (B) Genomic DNA sequencing of HT29 cells following the deletion of the MIEN1 promoter region. (C) Reverse transcriptase polymerase chain reaction of MIEN1 in both HT29 cell variants, WT and KO, observed at 200 bp, and the GAPDH gene observed at 300 bp as a control. (D) Western blot image representative of the whole-cell lysate of HT29 WT and KO cells showing the expression of MIEN1, ErbB2, and GRB7 proteins, which represent the protein products of genes adjacent to MIEN1. The Hsc70 protein was used as a loading control. (E) Quantification of the band density of western blot images for proteins to compare HT29 WT and KO cell expression. The paired t-test was used to perform statistical analysis, with n = 5 in each group. ns, non-significant; ***, p < 0.0005; MIEN1, migration and invasion enhancer 1; bp, base pairs; WT, wild-type; KO, knockout.

Full size image

Deletion of the minimal promoter region of 856 base pairs (bp) upstream of the ATG start codon (+ 1 position) (Fig. S1) for MIEN1 in HT29 human CRC cells was confirmed by sequencing genomic DNA from a single-cell clone (Fig. 1B) after puromycin selection. The cell line lacking the MIEN1 promoter region was named HT29-pMIEN1-X (referred to as HT29 KO).

Reverse transcriptase polymerase chain reaction (RT-PCR) in both HT29 cell variants, showed minimal MIEN1 transcript levels in the HT29 KO cells. Additionally, a minor PCR product of approximately 300 bp, corresponding to GAPDH, resulted in an ambiguous PCR product, likely because of nonspecific primer binding. (Fig. 1C). To further validate the knockout at the protein level, immunoblotting was performed, which confirmed the complete absence of the MIEN1 protein in knockout cells, whereas wild-type (WT) cells had abundant levels of the protein (Fig. 1D).

To verify whether MIEN1 promoter deletion affected neighboring genes, we performed immunoblot analysis of ErbB2 and GRB7 proteins and observed no changes in their expression (Fig. 1D). Additionally, to confirm the establishment of a stable HT29 KO cell line, we performed immunoblot analysis of MIEN1 and the proteins encoded by the neighboring genes across multiple passages. Our observations indicated a consistent absence of MIEN1 protein in HT29 KO cells, with unaltered expression in HT29 WT cells throughout numerous subcultures (Fig. 1E). Moreover, no substantial reduction in the expression of proteins encoded by the neighboring genes ERBB2 and GRB7 was observed over successive passages (Fig. 1E).

Collectively, these results confirmed the establishment of a stable MIEN1 knockout in HT29 cells and indicated a lack of impact on neighboring genes, such as ERBB2 and GRB7.

MIEN1 knockout results in transcriptomic alterations

To understand the alterations in gene expression patterns following the knockout of MIEN1, mRNA sequencing was performed. Principal component analysis (PCA) revealed distinct expression profiles, with PC1 accounting for 56% of the variance, highlighting significant global differences in gene expression (Fig. 2A). Differential expression analysis revealed that, out of 1262 differentially expressed genes in HT29 KO cells, 494 genes were upregulated with a fold change greater than 1.0, whereas 768 genes were downregulated with a fold change of less than 1.0, with a P-value cut-off of 0.05 (Fig. 2B).

Fig. 2
figure 2

HT29 mRNA sequencing and enrichment analysis. (A) Principal component analysis depicting expression patterns between HT29 WT and KO cells, with PC1 capturing the variance between the two groups and PC2 capturing the variance in each group. (B) Volcano plot, with each symbol depicting each gene identified in mRNA sequencing. The gray points represent genes with log2[fold change] in the range of − 1.0 to + 1.0. Data points in green represent genes downregulated upon MIEN1 knockout with log2[fold change]  less than- 1.0, and those in red represent genes upregulated upon MIEN1 knockout with log2[fold change] greater than + 1.0. (C) Growth tumor pathway showing various processes with predicted inhibition (blue) or activation (orange). Processes of interest are outlined in red. (D) Bar graph representing the processes that were significantly modulated upon MIEN1 knockout. The − log10[P-value] indicates the extent of significance, as observed using the entire mRNA sequencing dataset. (E) Clustered bar graph showing the subset of functional processes observed relevant to this study derived from IPA software. The PCA plot, volcano plot, and pathway analysis were generated using IPA software, and bar graphs were generated using GraphPad Prism 10 software. PCA, principal component analysis; IPA, ingenuity pathway analysis; MIEN1, migration and invasion enhancer 1; WT, wild-type; KO, knockout; TME, tumor microenvironment; EMT, epithelial-mesenchymal transition.

Full size image

Using ingenuity pathway analysis (IPA; Qiagen), we identified the key pathways and biological processes affected by MIEN1 knockout. Among the disease pathways predicted by machine learning, tumor growth has been highlighted as one of the major inhibited pathways. Several key processes, such as the extension of cellular protrusions, movement of tumor cells, ruffling, transactivation, immortalization, and invasion of tumor cells, were predicted to be suppressed (Fig. 2C). Three major pathways—CRC metastasis signaling, regulation of epithelial-mesenchymal transition (EMT) by the growth factor pathway, and the tumor-microenvironment (TME) pathway—were among the downregulated canonical pathways. The CRC metastasis signaling pathway revealed that cell proliferation and survival are inhibited by the Wnt signaling pathway. Apoptosis was highlighted as an activated process, whereas migration, metastasis, cell proliferation, and angiogenesis were predicted to be indirectly inhibited because of stunted growth factor signaling in colorectal tumorigenesis (Fig. S2). In the regulation of EMT by the growth factor pathway, cytoskeleton reorganization, which indirectly influences migration, cell invasion, and disassembly of cell–cell junctions, was suppressed through receptor tyrosine kinase signaling. This suppression resulted in the overall inhibition of the EMT process, significantly affecting cell migration, lesion healing, and metastasis. Similarly, transforming growth factor-β (TGF-β) signaling was predicted to inhibit the overall EMT process, further suppressing cell invasion, migration, lesion healing, and metastasis. Cell–cell adhesion was predicted to be activated via both receptor tyrosine kinase and TGF-β signaling (Fig. S3). The TME pathway has also been predicted to inhibit tumor cell invasion and metastasis. In contrast, tumor cell proliferation, survival, angiogenesis, and apoptosis were predicted to be activated (Fig. S4).

IPA also highlighted a range of biological processes affected by MIEN1 knockout, including disruptions in cell signaling, cellular movement, molecular transport, tumor morphology, and cellular assembly and organization (Fig. 2D). The clustered bar graph in Fig. 2E highlights the molecular and cellular processes affected by MIEN1 knockout. Cell movement and migration, tumor frequency, transdifferentiation, cell invasion, tumor incidence, EMT, tumor cell proliferation, actin filament quantity, and cytoskeletal rearrangement had negative Z-scores, indicating inhibition, whereas tumor cell death, necrosis, apoptosis, and cell–cell adhesion were predicted to be activated, with positive Z-scores.

Overall, IPA analysis emphasized that the knockout of MIEN1 in HT29 cell line significantly impaired critical processes, such as migration, invasion, and cytoskeleton rearrangement, while promoting cell–cell adhesion and apoptosis.

MIEN1 deletion alters actin cytoskeleton rearrangement

Gene ontology enrichment analysis indicated that MIEN1 knockout negatively affected cell motility, migration, locomotion, and the formation of membrane ruffles (Fig. S3). Cell migration is a complex process orchestrated by the coordinated polymerization of actin filaments, leading to the formation of protrusive structures, known as lamellipodia, at the leading edges of migratory cells. Previously, we demonstrated the pivotal role of the actin cytoskeleton in prostate and breast cancer metastases.3,4 HT29 WT cells displayed enhanced staining at their cell peripheries, suggesting the formation of lamellipodium ruffles, along with fewer stress fibers within the cell body, indicating decreased cell adhesion (Figs. 3A, S5, S6 and S13). Conversely, HT29 KO cells exhibited a notable accumulation of stress fibers across their cell bodies and minimal F-actin build-up at the cell periphery, signifying enhanced cellular adhesion and reduced motility (Figs. 3C, S7, S8 and S14). The distance-versus-phalloidin intensity graphs corroborated our findings, revealing a three-fold increase in phalloidin intensity at the cell peripheries of WT cells compared to that in their cell bodies (Figs. 3B, S9, S10 and S15). This suggests the presence of lamellipodia at the cell edges and fewer stress fibers in the cell body. Conversely, the knockout cells displayed increased intensities uniformly across the entire cell, indicating the prevalence of stress fibers throughout the cell and the absence of lamellipodium ruffles (Figs. 3D, S11, S12 and S16).

Fig. 3
figure 3

Effect of MIEN1 knockout on actin cytoskeleton structure. (A and C) Immunofluorescence microscopy images showing phalloidin staining of F-actin in HT29 WT and KO cells. The first column shows a merged image of the DAPI nuclear stain (second column) and phalloidin stain (third column). (B and D) Fluorescence intensity of the image around the yellow arrow (phalloidin, third column) is represented in the graph. (E and F) Images represent zoomed sections of the merged images from Fig. 3A and C (first column, gray box) with their respective intensities of the area indicated by the yellow arrow on the edge of the cells. Images were acquired using the Zeiss LSM 880 Confocal with Airyscan microscope at 63 × magnification. Fluorescence intensities were measured using ImageJ software. Scale bars: 5 µm (3A and C), 1 µm (3E and 3F). WT, wild-type; KO, knockout; MIEN1, migration and invasion enhancer 1.

Full size image

When examining the phalloidin staining pattern along the cell edges, we observed that phalloidin intensity increased exponentially, maintaining a steady level within a specific distance range. This indicated increased intensity, specifically in the lamellipodium region, tapering off on either side of the WT cells (Figs. 3E and S17). In contrast, the edges of the knockout cells displayed a graph similar to that observed in the cell body, showing irregular staining patterns and distinctly emphasizing the absence of lamellipodium ruffles (Figs. 3F and S18).

Phosphorylation of essential actin cytoskeleton proteins FAK and cofilin is altered owing to the absence of MIEN1

To validate actin cytoskeleton impairment in HT29 KO cells, we examined the key actin cytoskeleton proteins—FAK and cofilin, specifically their phosphorylated forms, pFAK-Y397 and pCofilin-S3 by immunofluorescent staining and immunoblotting. In WT cells, pFAK-Y397 was localized to the cellular edges where lamellipodium ruffles were evident, whereas in HT29 KO cells, pFAK was present within the stress fibers in the cell body (Fig. 4A). Upon quantifying the fluorescence intensity of pFAK-Y397 , we observed a 2.5-fold higher localization of pFAK-Y397 at the cell periphery in WT cells compared to MIEN1 KO cells. Whereas MIEN1 KO cells exhibited a pronounced 20-fold increase in pFAK-Y397 localization within the cell body compared with that of WT cells (Fig. 4B). These quantifications clearly illustrate the disruption in the distribution pattern of active FAK upon MIEN1 knockout. WT cells, with significantly higher peripheral pFAK, reflect active focal adhesions at the cell edge, a hallmark of efficient migration. Conversely, the substantial increase of active FAK in the cell body in MIEN1 KO cells indicates mislocalization, potentially causing impaired focal adhesion dynamics and inhibited cell motility. Immunoblotting analysis showed a significant reduction in the pFAK Y397 to total FAK ratio compared to WT cells, indicating decreased FAK activation. (Fig. 4E and F).

Fig. 4
figure 4

MIEN1 knockout modulates key actin cytoskeleton proteins. (A) Immunofluorescence staining images of the actin cytoskeleton and pFAK. (B) Quantification of the intensity of pFAK-Y397 at the cell periphery and in the cell body of the same images, performed using ImageJ software. The paired t-test was used to compare the groups, with n = 2 in each group. (C) Immunofluorescence staining images of the actin cytoskeleton and pCofilin-S3. (D) Quantification of the intensity of pCofilin-S3 in HT29-WT and KO cells, performed using ImageJ software. The two-tailed unpaired t-test was used to compare the groups, with n = 6 in each group. (E) Western blot image representative of the whole-cell lysate of HT29 WT and KO cell lines, showing the expression of key cytoskeletal proteins, with Hsc70 as a control. (F) Quantification of the band density of western blot images for proteins to compare HT29 WT and KO cell expression, as performed using the ImageJ software. The paired t-test was used to perform statistical analysis, with n = 3 in each group. GraphPad Prism 10 was used to generate the graph and perform statistical analysis, with *, p < 0.05; **, p < 0.005; ***, p < 0.0005. Images were acquired using the Zeiss LSM 880 Confocal with Airyscan microscope at 63 × magnification. Fluorescence intensities were measured using ImageJ software. MIEN1, migration and invasion enhancer 1; KO, knockout; WT, wild-type.

Full size image

Cofilin, an actin-depolymerizing protein, is inactivated upon phosphorylation at the serine residue in the third position. Staining for pCofilin-S3 showed a higher intensity in HT29 KO cells than in WT cells (Fig. 4C). Quantification of fluorescence intensity showed a 40% increase in KO cells compared with that in WT cells (Fig. 4D). Immunoblotting for cofilin and its phosphorylated form at S3 showed a visible increase in phosphorylated Cofilin levels., although the pCofilin/Cofilin ratio showed an upward trend in KO cells relative to WT. While the pCofilin/Cofilin ratio exhibited an upward trend compared to WT, the difference did not reach statistical significance (Fig. 4E and F). These observations demonstrated cofilin inactivation upon MIEN1 knockout, suggesting decreased actin depolymearization.

Collectively, our data provides evidence of impaired actin cytoskeleton dynamics, which are critical for cell motility machinery (lamellipodia and filopodia), in the absence of MIEN1 protein. These results emphasize the importance of MIEN1 in CRC cell migration.

MIEN1 knockout affects the migration and invasion potential of CRC cells

We performed various functional assays to validate the observed changes in the actin cytoskeleton. Cell–cell adhesion was evaluated using the hanging-drop experiment. HT29 KO cells formed single spheroids within 24 h of incubation. In contrast, HT29 WT cells required more than 72 h to form a single spheroid, along with multiple smaller cell aggregates. By day 4, HT29 KO cells formed elliptical spheroids with a distinct border and a denser core, whereas HT29 WT cells struggled to create a singular spheroid with clear boundaries (Fig. 5A). These results suggest that the knockout of MIEN1 enhances cellular adhesion among HT29 KO cells, which likely leads to a reduction in migration potential. These results imply that MIEN1 knockout amplifies cellular adhesion in HT29 KO cells, likely resulting in a diminished migration potential.

Fig. 5
figure 5

Migratory and invasive potential of HT29 cells is reduced upon MIEN1 knockout. (A) Representative brightfield microscopy images of spheroids from days 0 to 4 in HT29 WT and KO cells. The number of spheroids formed each day was quantified manually and is represented by a bar graph. A multiple comparison test was performed to analyze the statistical differences between the groups, with n = 3 in each group. The area of spheroids on each day was measured using ImageJ software and is represented on the bar graph, with n = 3 (top right-end) Scale bars: 100 µm. (B) Representative two-dimensional migration assay images depicting the scratch at the time points mentioned above. The percentage of wound closure was measured using ImageJ software, and statistical analysis between the groups for each day was performed using two-way analysis of variance, with n = 5. (C) Representative image of the Boyden chamber insert used for the transwell invasion assay of HT29 WT and KO cells. The bar graph represents the quantified number of cells in each image captured and averaged. The two-tailed unpaired t-test was used to compare the groups, with n = 3 in each group. GraphPad Prism 10 was used to generate graph and perform statistical analysis with *, p < 0.05; **, p < 0.005; ***, p < 0.0005, ****, p < 0.0001 (microscope software details, magnification or zoom). MIEN1, migration and invasion enhancer 1; KO, knockout; WT, wild-type.

Full size image

To further investigate the effect of MIEN1 knockout on cell migration, we performed a traditional wound-healing assay. A significant 0.25-fold reduction in wound closure was observed in KO cells compared with that in WT cells (Fig. 5B).

We also examined the impact of MIEN1 knockout on the invasion potential of these cells. Using the Boyden chamber invasion assay, we found that HT29 KO cells displayed a remarkable 40% reduction in invasion potential compared with HT29 WT cells (Fig. 5C).

Collectively, these findings suggest that MIEN1 knockout increases cell–cell adhesion, severely limiting the migration and invasion capabilities of HT29 cells.

Discussion

MIEN1 is overexpressed in various cancers, including breast, prostate, oral, ovarian, and colorectal cancers. Our previous investigations revealed the active involvement of MIEN1 in promoting prostate cancer metastasis by activating the Akt/NF-kB pathway, thereby upregulating the transcription of matrix metalloproteases and angiogenic factors5. Additionally, RNA interference–mediated downregulation of MIEN1 in prostate and breast cancer results in alterations in the actin cytoskeleton, impacting cellular movement4,5. However, there are limited studies on the role of MIEN1 in CRC metastasis. This study aimed to address this gap by investigating the contribution of MIEN1 in CRC metastasis using a gene knockout model.

Previous studies have emphasized the significance of the MIEN1 promoter region in regulating its expression. Evans et al.7 reported that MIEN1 overexpression in cancer cell lines was attributed to transcriptional control rather than gene amplification. Rajendiran et al.8 further demonstrated that hypomethylation of the SINE Alu region of the MIEN1 promoter results in its overexpression in cancer cells. Hence, we developed a strategy to delete the minimal promoter region of MIEN1 in the HT29 CRC cell line, ensuring the specificity of the knockout while avoiding interference with important neighboring genes such as ERBB2 and GRB7. This contrasts with a method previously reported by Van Treuren et al., which involved generating a CRISPR-mediated MIEN1 knockout in triple-negative breast cancer cells by deleting a specific segment of MIEN1 within the ERBB2 amplicon9. The absence of the MIEN1 transcript and protein in KO cells, in contrast to its abundant expression in WT cells, validates the efficacy of our knockout strategy. While this strategy provides an effective means of investigating MIEN1 function, complementary studies in additional CRC cell lines would strengthen the robustness of the findings.

Subsequent transcriptomic analysis revealed significant alterations in the gene expression patterns following MIEN1 knockout. The distinct global expression patterns between WT and KO cells emphasize the profound impact of MIEN1 deletion on the cellular transcriptome. Various physiological functions, such as endothelial cell development, transactivation, immortalization, cytoskeletal rearrangement, and EMT, are directly or indirectly associated with malignancy advancement. Notably, reduced sperm motility was predicted to be inhibited by MIEN1 knockout, indicating its involvement in cellular movement beyond cancer.

Rearrangement of the actin cytoskeleton is closely associated with cancer cell migration and invasion. A previous study by Dasgupta et al. identified the presence of a post-translationally modified isoprenyl group at the carboxy terminus of the MIEN1 protein, a modification also observed in other proteins involved in actin assembly, such as the Rho family4. Furthermore, Kpetemey et al. demonstrated that MIEN1 overexpression leads to increased lamellipodia formation compared to cells with normal MIEN1 levels3. Consistent with our previous study, we report for the first time that MIEN1 knockout in CRC cells results in a significant reduction in lamellipodia formation and an increase in stress fiber accumulation in the cell bodies of HT29 cells. The substantial presence of stress fibers in MIEN1 KO cells, as opposed to WT cells (Fig. 3A–D), indicating enhanced cell adhesion to the substratum, which inhibited cell motility. Moreover, increased phosphorylation of cofilin at serine 3 (S3) was observed, indicating its inactivation (Fig. 4E and F). This inactivation led to the inhibition of actin depolymerization, thereby promoting stress fiber formation, as demonstrated by immunofluorescence staining (Fig. 4C) These results provide key insights into the cytoskeletal functions of MIEN1; however, a more detailed exploration of its downstream signaling pathways, such as the Rho GTPase and PI3K/Akt axes, would help refine our understanding of its molecular role in CRC metastasis. Another critical actin cytoskeleton protein, FAK, is phosphorylated at tyrosine 397, and its phosphorylation is significantly affected by the absence of MIEN1. In this study, we observed a marked difference in the subcellular localization of pFAK-Y397 . In WT cells, pFAK-Y397 was prominently localized at the cellular edges, particularly along the lamellipodia. This observation is consistent with the increased phosphorylation observed in WT cells, as shown by immunoblotting. Increased phosphorylation indicates FAK activation, which in turn activates the Rac pathway, enhancing lamellipodia formation and, thereby, increasing cellular migration and invasion1011. This explains the recruitment of pFAK-Y397 to the cellular edges in WT cells. In contrast, pFAK-Y397 was predominantly localized within the cell body in KO cells around abundant stress fibers. Immunoblot analysis of KO cells revealed reduced FAK phosphorylation at Y397, indicating FAK inactivation. Inactive FAK maintains Rho GTPases in an inactive state, and the downregulation of the Rho GTPase pathway can inhibit cell migration and invasion, which are often associated with FAK localization in stress fibers12. Therefore, in the absence of lamellipodia, pFAK-Y397 is recruited to the cell body, where stress fibers are present, rather than to the cell edges. However, MIEN1 knockout may trigger compensatory mechanisms that alter the cellular phenotype over time. Phosphoproteomics or functional rescue experiments, could help determine whether other molecules compensate for MIEN1 loss. Additionally, the alterations in actin cytoskeleton rearrangement were consistent with the results of functional assays, elucidating the impact of MIEN1 knockout on cell–cell adhesion, migration, and invasion. The enhanced cell-to-cell adhesion efficiency, coupled with the reduced wound closure and decreased invasion potential observed in KO cells, provides substantial evidence for the significant implications of MIEN1 deletion on CRC migration and invasion potential. Importantly, a recent study by Tripathi et. al demonstrated that short peptides designed to target MIEN1 exert potent anticancer activity, supporting the promise of MIEN1 as a therapeutic target and underscoring the need for further preclinical modeling13. Importantly, a recent study by Tripathi and colleagues (2024) demonstrated thatshort peptides designed to target MIEN1 exert potent anticancer activity, supporting the promise of MIEN1 as a therapeutic target andunderscoring the need for further preclinical modeling. These findings suggest that MIEN1 deletion may inhibit CRC metastasis. While our in vitro observations provide strong mechanistic insights, further validation in in vivo models would be crucial to assess the translational relevance of targeting MIEN1 in CRC.

In conclusion, our findings highlight the crucial role of MIEN1 in regulating the migration and invasion potential of CRC cells by influencing actin cytoskeleton dynamics. Further investigation of the compensatory signaling pathways, and validating the functional role of MIEN1 in vivo would enhance our understanding of CRC metastasis and facilitate the development of novel treatment strategies for CRC.

Materials and methods

Cell culture

HT29 cells were obtained from the ATCC (Manassas, VA, USA) (Cat. no. HTB-38). The cells were maintained at 37 °C in a 5% CO2 incubator in HyClone McCoy’s 5A medium (Cat. no. #SH30200FS) supplemented with 10% fetal bovine serum (FBS) and 1% antibiotic–antimycotic (Gibco).

CRISPR-Cas9 MIEN1 promoter editing

HT29 WT cells were cultured and electroporated with a DNA mixture of plasmids expressing C1486_MIEN1_5′gRNA- 1, C1486_MIEN1_3′gRNA-1, and the cas9 gene (chimeric gRNA + hSpCas9 coexpression vector, C1486_MIEN1_5′gRNA-1:5´ GCCATCAGCACCGGGCGTGG-3′, and C1486_MIEN1_3′gRNA-2:5′ CATCGCGGCCGGCTCCGCTC-3′). The region encompassing 856 bp (− 870 to − 15 bp) was deleted using the above-mentioned guide RNAs (Fig. S1). Cell selection was performed using 1.0–1.5 µg/mL puromycin for 48 h. A small portion of the cell culture (presumably containing a mixed population) was subjected to genotype analysis. The mixed culture was diluted to less than 1 cell/200 µL of culture media and dispersed into each well of a 96-well plate. The cells were allowed to grow for 15–20 days. Cells derived from single-cell cloning were subjected to genotype analysis. Positive clones for the desired mutation were cultured-expanded, and a portion was used to confirm the predicted genotype. Genotyping involves the use of genomic DNA from single-cell clones. An approximately 400-bp DNA fragment was amplified using MIEN_2F and MIEN_2R primers (MIEN_2F: 5′-GGGATGCGCAGAAACTGTTGG-3′; MIEN_2R: 5′-GTTCACTGGGGAGTCAAGAGATGG-3′). The amplicons were then subjected to direct sequencing. The obtained cell line lacking the MIEN1 promoter region was named HT29-pMIEN1-X and is referred to as HT29 KO. Applied StemCell (Applied Stem Cell Inc., Foster City, CA, USA) was used for genome editing.

RT-PCR

Total RNA was extracted using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA). Superscript-III First Strand Synthesis kit (Invitrogen) was used for complementary DNA (cDNA) synthesis using 1 µg of total RNA, following the manufacturer’s protocol. PCR was performed to amplify the cDNA using Taq DNA polymerase and specific primers for the target genes. The PCR cycling conditions were as follows: 95 °C for 15 min, followed by 32 cycles at 95 °C for 30 s, annealing at 54 °C for 1 min, extension at 72 °C for 45 s, and held at 4 °C. PCR products were analyzed by agarose gel electrophoresis on a 1.2% agarose gel stained with ethidium bromide. Gel images were captured using a gel documentation system.

Antibodies and reagents

The rabbit anti- p-cofilin Ser 3 (cat #3311, 1:1000), rabbit anti-cofilin (cat #5175, 1:1000), rabbit anti-HER2/ErbB2 (29D8) (cat #2165, 1:1000), were procured from Cell Signaling Technologies (Danvers, MA, USA). Purchased from Invitrogen (Waltham, MA, USA) were the rabbit anti-p-FAK (cat # 700255, 1:1000), were the rabbit anti-FAK (cat # AH00502, 1:1000). Abclonal (Woburn, MA, USA) provided rabbit anti-GRB7 (cat # A5690, 1:1000). Mouse anti-MIEN1 (cat #H00084299-M02, 1:5000) was from Abnova. Rabbit anti-HSC70 ( cat # ADI-SPA-816, 1:1000) was provided by Enzo Life Sciences (Farmingdale, NY, USA). Invitrogen (Waltham, MA, USA) provided the Alexa Fluor anti Rabbit 488 antibody (cat # A11034, 1:1000), Phalloidin 594 (cat # A12381, 1:1000), and DAPI (cat # 62248, 1:1000).

Immunoblotting analysis

Proteins were extracted from cells using a radioimmunoprecipitation assay buffer (Thermo Scientific) supplemented with protease inhibitors (Millipore Sigma) and phosphatase inhibitors (Millipore Sigma), and then quantified using the BCA Protein Assay Kit (Thermo Scientific). Equal amounts of protein were loaded onto 4–12% precast SDS-PAGE gels (Invitrogen), and the samples were electrotransferred to a nitrocellulose membrane (Invitrogen) at a constant voltage of 25 V. The blots were then blocked with 5% bovine serum albumin (Thermo Scientific) and incubated with the respective primary antibodies at 4 °C for 8–10 h, followed by incubation with the respective secondary antibodies tagged with horseradish peroxidase. Protein bands were visualized using the Alpha Innotech Chemiluminescent Detection System and the iBright CL1500 Imaging System (Thermo Scientific).

RNA sequencing

Library construction and sequencing

RNA sequencing libraries were prepared using the Illumina Stranded mRNA Library Prep kits (Cat. no. 20040534), following the manufacturer’s instructions. Briefly, 1 µg of total RNA was used as the starting material. Poly(A)-tailed RNA was purified using poly(T) oligo beads included in the kit. The resulting RNA samples were then fragmented and converted into double-stranded DNA for sequencing. Individual libraries were uniquely indexed using Illumina RNA UD Indexes (Cat. no. 20040555) and pooled at an equimolar ratio. The pooled libraries were sequenced on an Illumina NextSeq 550 sequencing system to obtain 75-bp paired-end reads. Raw sequence reads were deposited in the Sequence Read Archive with the reference ID PRJNA1218296 in GenBank. RNA sequencing was performed using the Genomics Core at the UNT Health Science Center.

Read mapping and gene expression analysis

The quality of the raw FASTQ files was evaluated using the Galaxy web-based platform. Subsequently, adaptor trimming was performed using Trim Galore software. The trimmed files were merged into one file for R1 and R2 using text manipulation by concatenating the datasets from tail to head. The merged FASTQ files were then aligned to the human genome build GRCh38-Ensembl using HiSat2. Mapped BAM files were run using feature counts to obtain gene counts, which were then run using DEseq2 to identify differentially expressed genes.

Gene ontology and gene set enrichment analysis

Gene ontology analysis was performed using QIAGEN’s IPA software with features such as network analysis, upstream regulator analysis, pathway analysis, and canonical pathway analysis, with a P-value cut-off set to 0.05.

Wound-healing assay

3 × 105 cells were seeded in 6-well culture plates and incubated at 37 °C and 5% CO2 until they reached 80% confluency. The cells were then serum-starved for 24 h prior to creating a wound for cell cycle synchronization. A straight wound was created using a 100-μL pipette tip. Floating cells were removed by washing with 1 × phosphate-buffered saline (PBS). The cells were incubated in McCoy’s 5A medium (HyClone) with 10% FBS and 1% antibiotic–antimycotic at 37 °C in 5% CO2. Wound-healing images were acquired over 24 h at equal intervals using a Keyence AIO BXZ-X810 inverted fluorescence microscope. ImageJ software (v.2.14.0) (https://imagej.net/ij/) was used to quantify the gaps in the cell monolayers.

Transwell invasion assay

Boyden chambers with a transwell membrane of 8-μm pore size (Corning, NY, USA) were seeded with 3 × 105 cells/well in 200-μL serum-free McCoy’s 5A medium. Next, 500 μL of McCoy’s 5A medium (HyClone) with 10% FBS and 1% antibiotic–antimycotic were added to the wells of a 24-well plate and incubated for 48 h at 37 °C in 5% CO2. The Boyden chambers were stained with 0.05% crystal violet after fixation with 4% paraformaldehyde. Uninvaded cells were removed using cotton swabs, and the number of invaded cells was counted using phase-contrast microscope – Nikon Eclipse Ti–U.

Hanging-drop assay

A total of 2 × 105 cells/mL were seeded on the lid of a 90-mm dish in six 20 μL drops. Next, 7 mL of 1× PBS was added to a 90-mm dish, and the lid was carefully inverted on the dish. Images were captured every 24 h for 4 days.

Immunofluorescence staining

The cells were seeded on 22-mm coverslips, fixed in 4% paraformaldehyde (Thermo Scientific) for 15 min at room temperature, and permeabilized with 0.1% Triton X-100 (Sigma) for 15 min at room temperature. Blocking was performed using 5% goat serum (Invitrogen) for 1 h at room temperature, followed by overnight incubation at 4 °C with the respective primary antibodies. After primary incubation, the cells were rinsed three times with 1 × PBS containing Tween 20 and subsequently incubated with the respective secondary antibodies for 1 h at room temperature. Next, the cells were incubated with 4’,6-diamidino-2-phenylindole for 10 min at room temperature, followed by mounting on slides using a mounting medium (Invitrogen). Images were acquired using an Inverted Zeiss LSM 880 Airyscan confocal microscope.

Statistical analysis

The statistical analysis of all the data was done with GraphPad Prism 10 (La Jolla, CA, USA). Every experimental result was obtained from a minimum of three separate experiments. One-way ANOVA or t-tests, as required for the experiment, were used to evaluate statistical significance. A statistically significant result was defined with reference to P < as mentioned in the figure legends.