Nocodazole (NZ; M1404), fibronectin (FN; F0895), protease inhibitor cocktail (P8340), phosphatase inhibitor cocktail II (P5726), polybrene (TR-1003), KAPA SYBR FAST qPCR Master Mix (2X) kit (KK4601) and antibodies raised against SALL2 (HPA004162), α-Tubulin (T902), FLAG (M2, F1804) and γ-Tubulin (T6557) were obtained from Sigma-Aldrich (St. Louis, MO United States). Lipofectamine 2000 (11668019), Alexa Fluor 488, 555 and 643 conjugated secondary antibodies (goat anti-mouse), Alexa Fluor 488 phalloidin (A12379) and the PE-conjugated CD29 antibody (eBioscience, #12-0291-82) were purchased from Invitrogen (Carlsbad, MI, United States). FAK Inhibitor compound PF562,271 was from Laviana pharma corporation (JS, China). Doxycycline (#14422) was from Cayman chemical (Ann Arbor, MI, United States). Puromycin dihydrochloride (sc-108071), normal mouse IgG2A (sc-3878), Integrin β1 (M-106, sc-8978), Integrin β1 (A-4, sc-374429), Integrin α4 (B-2, sc-376334), Integrin α6 (F-6, sc-374057), Calpain 2 (E-10, sc-373966), FAK (H-1, sc-1688), PTEN (A2B1, sc-7974), Talin (C-9, sc-365875), Integrin β3 (N-20, sc-6627) and Vinculin (7F9, sc-73614) antibodies were obtained from Santa Cruz Biotechnology (Dallas, TX, United States). Phospho-FAK Tyr397 (#3283) was purchased from Cell Signaling Technology (Danvers, MA, United States). Histone H3 (Ab1791) antibody was obtained from Abcam (Cambridge, United Kingdom). Acetyl-Histone H4 (06-598) antibody was purchased from Millipore (Burlington, MA, United States). Horseradish peroxidase–conjugated secondary antibodies (donkey anti-goat; goat anti-rabbit and goat anti-mouse) and Hoechst 33342 were obtained from Bio-Rad (Hercules, CA, United States).

The SV40 large T antigen expression pBSSVD2005 was a gift from David Ron (Addgene plasmid # 21826; http://n2t.net/addgene:21826; RRID: Addgene_21826). The pCW57-MCS1-2A-MCS2 (Addgene plasmid # 71782; http://n2t.net/addgene:71782; RRID: Addgene_71782) and pCW57-MCS1-P2A-MCS2 (Hygro) (Addgene plasmid # 80922; http://n2t.net/addgene:80922; RRID:Addgene_80922) were a gift from Adam Karpf, (Barger et al., 2019). For lentiviral infection, the pCMV-dR8.2 dvpr (Addgene plasmid # 8455; http://n2t.net/addgene:8455; RRID: Addgene_8455) and pCMV-VSV-G (Addgene plasmid # 8454; http://n2t.net/addgene:8454; RRID: Addgene_8454) were a gift from Bob Weinberg (Stewart et al., 2003). The EFIa-iTGB1 plasmid was a gift from Joan Massague (Addgene plasmid # 115799; http://n2t.net/addgene:115799; RRID:Addgene_115799) (Er et al., 2018). The pCDNA.3 FLAG_SALL2 was previously generated using the pCMV2-FLAG_SALL2 plasmid (Pincheira et al., 2009). To generate pCW57 Tet-On FLAG mSall2_E1A plasmid, FLAG mSall2_E1A was PCR amplified from pCMV2(NH)-FLAG mSall2_E1A, previously generated using the full-length mouse Sall2 coding sequence synthetized by GeneScrip (Escobar et al., 2015), using the following oligonucleotides: forward, 5′-GCA​GGT​CGA​CAT​GGA​CTA​CAA​AGA​CGA​TGA​C -3′ and reverse, 5′-GCA​GCC​TAG​GTC​ATG​GCA​TGG​TGG​GGT​CAT​CTT​T -3′ and then was subcloned into pCW57-MCS1-2A-MCS2 using SalI and AvrII restriction sites. The ITGB1 promoter_pGL4.30 was kindly provided by Dr. Masakazu Toi (Kyoto University, Japan; MTA (2019) to RP, Universidad de Concepción, Chile).

Sall2 KO mice (Sato et al., 2003) were obtained by collaboration with Dr. Ruichi Nishinakamura (Kumamoto University, Japan; MTA to RP, Universidad de Concepción, Chile). Sall2 ^+/+ and Sall2 ^−/− mouse embryonic fibroblasts (MEFs) were isolated and genotyped as previously described (Escobar et al., 2015).

Primary Sall2 ^+/+ and Sall2 ^−/− MEFs (passage 4) were immortalized using simian virus 40 (SV40) large T antigen based on a modified protocol from Zhu et al. (Zhu et al., 1991). For transfection, we used Lipofectamine 2000 and 2 μg SV40 large T antigen expression vector. After cell transfection, we proceeded to select for low density. To complete the immortalization process, 5-6 post-transfection passages were carried out. The inducible reconstitution of Sall2 in immortalized Sall2 ^−/− MEFs, was produced by transient transfection of pCMV-dR8.2 dvpr, pCMV-VSV-G, and pCW57 Tet-On FLAG mSall2_E1A into HEK293T cells in a 10 cm dish with Lipofectamine 2000 reagent according to the manufacturer’s instructions. Viral supernatants were collected at 24 h and passed through a 0.2 μm filter. Sall2 ^−/− iMEFs were infected with viral supernatants containing polybrene at 8 μg/ml. For positive clone selection, a fresh medium containing 5 μg/ml of puromycin was used for 72 h. For Sall2_E1A induction, iMEFs were treated with doxycycline (1,000 ng/ml) for 24–48 h before each experiment. Like the generation of the inducible mSall2_E1A cellular model, the integrin β1 overexpression in Sall2 ^−/− iMEFs was performed by transfecting HEK293T cells with the EFIa-iTGB1 plasmid or empty vector (negative control). For positive clone selection, a fresh medium containing 400 μg/ml of hygromycin was used for 96 h. After selection, the generation of the stable models was evaluated through western blot.

Sall2 ^+/+ and Sall2 ^−/− iMEFs, Tet-On Sall2 iMEFs, HEK293T (ATCC, Manassas, VA, United States; CRL-3216) and a previously generated SALL2 KO HEK293 cell model (Hermosilla et al., 2018) were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Corning) supplemented with 10% (v/v) fetal bovine serum (FBS; Hyclone), 1% glutamine (Invitrogen) and 1% penicillin/streptomycin (Invitrogen). The cell lines used in this study were regularly tested (6 months) for mycoplasma using EZ-PCR Mycoplasma Test Kit (Biological Industries).

iMEFs were seeded at 1 × 10⁵ cells per 12-well plate. After 24 h, cells were serum-starved for 16 h. Survival rate was determined by using the trypan blue exclusion method. All experiments were performed in triplicate.

Cell culture dishes were previously covered with fibronectin (FN) 2 μg/ml, iMEFs were seeded at 4 × 10⁴ cells per 24-well plate. After 48 h, cells were serum-starved overnight and wounded with a pipette tip. Phase contrast images were acquired at 0 and 16 h. Wound closure percentage was calculated by the change in wound area between 0 and 16 h. All experiments were performed in quintuplicate. For the integrin β1 blocking study, Sall2 ^+/+ iMEFs were incubated at the time of wounding with 20 mg/L of anti-integrin β1 antibody (CD29) or non-immune IgG (mouse IgG2A). Non-immune IgG was used as a negative control. IncuCyte S3 system was used to perform a time-lapse video of the wound closure by iMEFs, every 30 min for a total period of 19 ^½  h.

Transwell inserts (Costar, 6.5-mm diameter, 8 um pore size) were previously covered with FN (2 μg/ml). iMEFs (2 × 10⁴) were seeded into the upper chamber in serum-free medium, complete medium was placed into the bottom chamber. After 4 h, the cells remaining at the upper membrane surface were removed and migrating cells (on the lower membrane surface) were fixed with 10% v/v acetic acid, 10% v/v methanol solution, and stained with 0.4% (w/v) crystal violet for 10 min at room temperature. All experiments were performed in triplicate.

iMEFs (4 × 10⁴) were seeded on coverslips previously covered with FN (2 μg/ml). After 48 h, a wound was performed as an activator of migration. After wound completion (4 or 16 h), cells were fixed with 4% paraformaldehyde, permeabilized with 0.1% Triton X-100 and incubated overnight at 4°C with primary antibodies (in blocking buffer, 3% bovine serum albumin in phosphate-buffered saline). After washing, fixed cells were incubated with Hoechst 33342 and Alexa fluor-conjugated secondary antibodies for 2 h. Images were obtained with LMS 780 spectral confocal system (Zeiss). For cell polarity analysis, the microtubule-organizing center (MTOC) was detected using γ-tubulin antibody. To measure the protrusion formation, actin filaments were stained with Alexa Fluor 488 phalloidin. For FAs analysis, vinculin protein was detected using a specific antibody. Identical exposure times and zoom (40x) were used for comparisons and quantification.

iMEFs were seeded at 1 × 10⁵ cells per 12-well plate. After 24 h, cells were incubated in ethylenediaminetetraacetic acid (EDTA) 0.05 mM at 37°C for 5, 10, and 15 min, respectively. As a control, untreated cells were used. Cells were fixed and stained with crystal violet for 10 min at room temperature. Then, crystal violet was extracted in 10% (v/v) acetic acid, and the absorbance at 590 nm was measured. The adherent cells were expressed as percentages, where 100% is A₅₉₀, corresponding to the number of adherent cells before EDTA-detachment, and 0% is A₅₉₀ of an empty well (no cells attached). All experiments were performed in triplicate.

iMEFs (4 × 10⁵) in suspension were allowed to attach and spread onto FN-coated 60-mm dishes (2 μg/ml) for the indicated periods (0, 10, 15, and 30 min). Samples were prepared for western blot analysis.

For focal adhesion synchronization, iMEFs (2 × 10⁴) were seeded on coverslips previously covered with FN (2 μg/ml). After 24 h, cells were treated with 10 μM nocodazole (NZ) in serum-free medium for 2 ^½  h. NZ was washed-out with serum-free medium at 5, 10, 15, and 30 min. Subsequently, cells were fixed and prepared for immunofluorescence and stained with anti-vinculin antibody (Alexa 488). For inhibition of FAK autophosphorylation, PF562,271 was used at 1 μM concentration in conjunction with NZ treatment. The number of focal adhesions per cell was quantified at all time points. Identical exposure times and zoom (63x) were used for comparisons and quantification.

Proteins from cell lysates (30-50 μg of total protein) were fractionated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred for 16 h at 30 V to polyvinylidene fluoride (PVDF) membrane (Millipore; IPVH00010) using a wet transfer system. The PVDF membranes were blocked for 1 h at room temperature in 5% nonfat milk in TBS-T (TBS with 0.1% Tween) and incubated with primary antibody at an appropriate dilution at 4°C overnight in blocking buffer. After washing, the membranes were incubated for 1 h at room temperature with horseradish peroxidase-conjugated secondary antibodies diluted in TBS-T buffer. Immunolabeled proteins were visualized by ECL (RPN2209) or ECL prime (RPN2232) (Cytiva, Marlborough, MA, United States). For protein expression in vivo, proteins were extracted from brain tissues of 6–8 weeks old Sall2 ^+/+ and Sall2 ^−/− mice as previously described (Hermosilla et al., 2018).

For integrin β1 expression at the membrane surface, 2 × 10⁵ iMEFs were washed, detached in 0.25% trypsin, and collected in 200 μL of cold sorting buffer (2% bovine serum albumin in phosphate-buffered saline). Then, the cells were incubated with 1 μg of PE-conjugated CD29 antibody in the dark for 30 min at 4°C. After incubation, cells were washed, resuspended in 500 μL of sorting buffer, and sorted using a BD FACSAria III cell sorter (BD Biosciences).

Identification of putative SALL2/Sall2 sites was performed using a previously reported binding site matrix (consensus sequence GGG (T/C)GGG) (Gu et al., 2011) in Transcriptional Regulatory Element Database (TRED) (http://rulai.cshl.edu/cgi-bin/TRED/tred.cgi?process=analysisMatrixForm) with a cutoff score = 7.5 (Jiang et al., 2007). Sequences analyzed [–2000 bp from transcription start site (+1)] were obtained from Eukaryotic Promoter Database (EPD) (http://epd.vital-it.ch/). The comparative analysis of ITGB1 promoter derived from ChIP-seq data was performed as previously described (Farkas et al., 2021).

Total RNA was extracted from cells with TRIzol reagent (Invitrogen; 15596026) according to the manufacturer’s instructions. RNA was treated with Turbo Dnase (Invitrogen; AM1907) to eliminate any residual DNA from the preparation. RNA (1 μg) was reverse transcribed using the Maloney murine leukemia virus reverse transcriptase (Invitrogen; 28025-01) and 0.25 μg of Anchored Oligo (dT) 20 Primer (Invitrogen; 12577-011). Real-time PCR was performed using KAPA SYBR FAST qPCR Master Mix (2X) kit and the AriaMX Real-Time PCR System (Agilent, Santa Clara, CA, United States) according to the manufacturer’s instructions. The thermal cycling variables used were as follows: 40 cycles at 95°C for 5 s and 60°C for 20 s. To control specificity of the amplified product, a melting-curve analysis was carried out. No amplification of unspecific product was observed. Amplification of RNA polymerase II (Polr2) was carried out for each sample as an endogenous control. Primers used for real-time reactions are summarized in Supplementary Table S1. The relative expression ratio of the Itgb1 gene was calculated using the standard curve method, using untreated Sall2 ^+/+ iMEFs as reference.

To evaluate ITGB1 promoter transcriptional activity, SALL2 KO HEK293 cells were transiently co-transfected with 1 μg of ITGB1 promoter, 0.125 μg of RSV-β-galactosidase (β-Gal), and 2 μg of FLAG_SALL2 (pCDNA.3 FLAG SALL2) or 1 μg of vector control per well. After 24 h, the transfected cells were lysed with reporter lysis 5X buffer (E4030; Promega, Madison, WI, United States) and centrifugated at 14000 × g to pellet cell debris. The supernatant was then assayed for luciferase and β-Gal activity using the manufacturer’s suggested protocols (Promega; E1483 and E4740, respectively). Luminescent reporter activity was measured using a Luminometer (Victor3; Perkin- Elmer). All transfections were normalized to β-Gal activity and performed in triplicate. Luciferase values were expressed as fold induction relative to the pGL3 vector control.

SALL2 KO HEK293 cells were seeded at 1 × 10⁶ cells per 100-mm dish. After 24 h, cells were transiently transfected with 2 μg of pCDNA.3 FLAG_SALL2. Cell nuclei were sonicated in 300 μL of sonication buffer, using a Bioruptor plus sonicator (Diagenode) (40 times, 15 s on/15 s off each time, high potency), obtaining DNA lengths between 300 and 600 bp. Immunoprecipitations were carried out overnight at 4°C using 2 μg of FLAG (anti-FLAG antibody), 1 μg of H3 (anti-histone H3; Abcam), or 1 μg of acH4 (anti-histone H4 acetylated; Millipore) and 25 μg of chromatin. After crosslinking reversion, DNA was purified using DNA Clean & Concentrator™-25 kit (D4033, Zymo Research, Irvine, CA, United States). DNA was analyzed by PCR directed to ITGB1 promoter SALL2-specific proximal I and proximal II regions (-251/-127) and (-631/-297). The ITGB1 promoter region (-1986/-1878) was used as a negative control for SALL2 binding. Primers used for qPCR reactions are summarized in Supplementary Table S2. All PCR reactions were performed using KAPA SYBR FAST qPCR Master Mix (2X) kit containing 1 μL of input and 4 μL of IP samples.

Confocal immunofluorescence and phase contrast images were manually analyzed using the ImageJ software. Data were analyzed using unpaired t-test performed with GraphPad Prism 9.3.1 software. Values averaged from at least three independent experiments were compared. A p-value of <0.05 was considered significant.

We investigated the role of Sall2 in cell migration by comparing the migration of iMEFs isolated from isogenic Sall2 knockout (Sall2 ^−/− ) and Sall2 wild-type (Sall2 ^+/+ ) embryos immortalized with the large T antigen (Hermosilla et al., 2018). Figure 1A shows the expression of Sall2 in the Sall2 ^+/+ iMEFs. We subjected iMEFs to an in vitro wound healing assay and evaluated cell migration 16 h post-scratching. As shown in Figure 1B and Supplementary movies S1, S2, Sall2 ^−/− iMEFs exhibited a lower percentage of wound closure than the Sall2 ^+/+ iMEFs (about 76% and 90%, respectively). There were no differences in cell viability between both genotypes at the time and in the conditions of the wound closure analysis (Supplementary Figure S1).

We performed transwell migration assays further to analyze the effects of Sall2 in cell migration. As expected, Sall2 ^−/− iMEFs exhibited fewer migratory cells than Sall2 ^+/+ cells (Figure 1C) (about 21% less). To confirm the role of Sall2 in cell migration, we restored its expression in the Sall2 ^−/− iMEFs. Since Sall2 ^+/+ iMEFs mainly express the Sall2 E1A isoform (Hermosilla et al., 2018), we used a Tet-On system to generate a Sall2 E1A inducible cellular model (Tet-On Sall2 iMEFs) (Figure 1D). Consistent with the above results, induction of Sall2 expression by doxycycline increased cell migration (wound closure percentage from 65% to 86%, Figure 1E) and increased the number of migratory cells (about 20% more) (Figure 1F). Altogether, these results demonstrated that Sall2 promotes cell migration in iMEFs cells.

To determine how Sall2 promotes cell migration, we evaluated the requirement of Sall2 expression on the different steps involved in cellular movement, including cell polarity, membrane protrusion formation, and cell adhesion.

First, we assessed cell polarization. Like other studies in MEFs, we evaluated polarity establishment at 4 h after wounding (Grande-García et al., 2007; Tomar et al., 2009). The maintenance of cell polarity was evaluated at 16 h (Li and Gundersen, 2008), when the wound edges were still distinguishable, and migration was almost complete (Supplementary Figures S2B,E). Cell polarity was measured by analyzing the localization of the microtubule-organizing center (MTOC) relative to the nucleus following the induction of cell migration. The MTOC positioning was grouped into four grades, as previously described (Silverman-Gavrila et al., 2011), where grade 1 corresponds to a highly polarized state, grade 2 polarized, grade 3 non-polarized, and grade 4 to a highly non-polarized state (Supplementary Figure S2A). Our results showed no significant differences between Sall2 genotypes in cell polarization at 4 and 16 h (Supplementary Figures S2C,F). Similar results were observed when comparing highly polarized grade (G1) with medium polarized grade (G2). As shown in Supplementary Figures S2D,G, there were no significant differences between Sall2 ^−/− and Sall2 ^+/+ cells. All these results indicated that Sall2 is not required to induce cell polarity, polarization grade, and cell polarity maintenance over time.

We next evaluated whether Sall2 affects the membrane protrusions. We stained actin filaments after cell migration induction (Figure 2A) and analyzed the formation of lamellipodium and filopodium structures by quantifying their number, size, and polarization state. As shown in Figures 2B,C, Sall2 ^−/− iMEFs depicted a higher lamellipodia area (mean = 352 versus 301 mm²) and exhibited a slightly lower filopodia number (mean = 23 versus 25) than Sall2 ^+/+ cells. Still, there were no significant differences in the lamellipodia number, lamellipodia polarization state, filipodia length, and filipodia polarization state when comparing genotypes (Supplementary Figures S3A–D). Our data suggest that Sall2 mainly regulates the area of the lamellipodia.

Additionally, we evaluated whether Sall2 regulates the cell adhesion process. We assessed cells binding to the extracellular matrix by comparing the detachment rate induced by adding EDTA. As shown in Figure 2D, iMEFs treated with an EDTA-containing solution resulted in a time-dependent detachment of cells. Interestingly, Sall2 ^−/− cells exhibited a higher percentage of adherent cells (80% at 5 min, 69% at 10 min, and 53% at 15 min) than Sall2 ^+/+ iMEFs (58% at 5 min, 46% at 10 min and 34% at 15 min) (Figure 2E). Accordingly, induction of Sall2 expression in the Tet-On Sall2 iMEFs significantly decreased the percentage of adherent cells at all times of EDTA treatment (47% at 5 min, 38% at 10 min, and 29% at 15 min), as observed in Sall2 ^+/+ cells (Figure 2F). Our results support that Sall2 promotes cell detachment, suggesting a novel role of Sall2 in cell adhesion.

To further investigate the role of Sall2 in cell adhesion, we evaluated whether Sall2 regulates focal adhesions by quantifying their number, length, and fluorescence intensity. After induction of cell migration in Sall2 ^+/+ and Sall2 ^−/− iMEFs and the Tet-On Sall2 iMEFs model, we measured focal adhesion (FA) by immunocytochemistry of vinculin (Figures 3A,F), a protein involved in FA formation (Legate and Fässler, 2009; Wehrle-Haller, 2012a). As shown in Figure 3B, the FA number in Sall2 ^−/− iMEFs was significantly higher than in Sall2 ^+/+ cells (a mean of 15 versus 12 FA/cell). In addition, FA length (mean = 3.5 versus 2.7 mm) and FA fluorescence intensity (mean = 5,034 versus 4,088 a.u.) in Sall2 ^−/− iMEFs were also significantly higher in comparison with Sall2 ^+/+ cells (Figures 3C,D). Accordingly, Sall2 induction by doxycycline in Tet-On Sall2 iMEFs decreased the number, length, and fluorescence intensity of FAs, like the Sall2 ^+/+ cells (Figures 3G–I). No significant differences in the polarization state of FAs between Sall2 ^−/− and Sall2 ^+/+ cells (Figure 3E) or after induction of Sall2 expression (Figure 3J) were observed. Since FA formation is a dynamic process where the FAs can disassemble or mature into stabilized complexes characterized by increased size (Gardel et al., 2010; Wehrle-Haller, 2012b), our results suggested that Sall2 negatively regulates the maturation of focal adhesions, as well as their number.

Considering these observations, we next evaluated whether Sall2 regulates FA dynamics. FAs were stabilized by microtubule depolymerization using nocodazole (Mendoza et al., 2013) at the concentration and time previously optimized (Figure 4A; Supplementary Figure S4). As expected, nocodazole increased the number of mature FAs compared with the non-treated cells ‘NT’. In contrast, nocodazole wash-out was followed by a progressive decrease of FA number, as observed in both genotypes (Figure 4B). Interestingly, Sall2 ^−/− iMEFs exhibited delayed kinetics of FA disassembly (about 5 min) compared with Sall2 ^+/+ cells (Figure 4C). In addition, Sall2 deficiency decreased the FA disassembly’s initial velocity (v₀-D from 0.99 ± 0.1 to 0.55 ± 0.1). Conversely, we did not observe significant differences in FA assembly (kinetics and initial velocity of FA assembly, v0-A) when comparing Sall2 genotypes (Figure 4C). Accordingly, the induction of Sall2 expression significantly accelerated FAs disassembly (v₀-D from 0.46 ± 0.1 to 0.73 ± 0.1), but not their assembly (Figure 4D). These results were supported by biochemical analysis of FAK autophosphorylation at Y397 (Hamadi et al., 2005). Phospho-Y397-FAK was evaluated after nocodazole release and during spreading on fibronectin at different time points (both approaches are widely used to assess changes in FAK activation, Nader et al., 2016; Erusappan et al., 2019). Phospho-Y397-FAK levels after 15 min of nocodazole wash-out were significantly lower in Sall2 ^−/− iMEFs than in control cells (Figure 4E ). Similarly, the phosphorylation of FAK at Y397 after 10 min of cell spreading was significantly lower in Sall2 ^−/− iMEFs than in Sall2 ^+/+ cells (Figure 4F). We further investigated the role of FAK activity in the Sall2-dependent regulation of FA dynamics. To this aim, Sall2 ^+/+ iMEFs were treated with PF562,271, a specific inhibitor of FAK. As expected, PF562,271 decreased FAK- Y397 phosphorylation (Figure 4G). Like the Sall2 ^−/− iMEFs phenotype, the treatment with PF562,271 decreased the FAs disassembly of Sall2 ^+/+ iMEFs after 10 min of nocodazole wash-out. (Figure 4H). We observed a decrease in the initial velocity of FA disassembly (v₀-D from 0.80 ± 0.1 to 0.30 ± 0.1) and assembly (v₀-A from 0.80 ± 0.1 to 0.20 ± 0.1) (Figure 4I). Altogether, these results indicate that Sall2 promotes FAK activation and, consequently, the focal adhesion turnover.

To determine the underlying mechanism by which Sall2 promotes FA dynamics, we analyzed the expression of different proteins involved in FA assembly-disassembly, such as integrins, vinculin, talin, and calpain 2. The analysis was performed under the conditions where the highest differences in FAK activation were observed (15 min after nocodazole release and 10 min of cell spreading) and under normal growth conditions. Although Sall2 ^+/+ cells depicted higher levels of FAK phosphorylation (Figures 4E,F), Sall2 deficiency was associated with higher FAK protein levels (Supplementary Figures S5B–G). These results suggested that Sall2-mediated FAK activation does not relate to positive regulation of FAK expression. In addition, we found that Sall2 ^−/− iMEFs have significantly lower protein levels of integrin β1 than Sall2 ^+/+ cells in all conditions analyzed (Figures 5A–C). Alpha subunits analysis showed slightly lower integrin α4 levels in Sall2 ^−/− iMEFs than Sall2 ^+/+ cells under normal growth conditions. However, we did not observe significative differences in the integrin α6 expression levels (Supplementary Figure S6). Additionally, we did not observe differences in the expression levels of vinculin, talin, calpain2 (Supplementary Figure S5), and integrin β3 (Supplementary Figure S6).

Considering the functional relevance of β-subunits in the recruitment of several adapters and signaling molecules to form focal adhesions (Schnittert et al., 2018), and since we did not observe changes in the integrin β3 protein levels, we focused our study on integrin β1. We evaluated how Sall2 induction affects integrin β1 levels after 15 min of nocodazole wash-out treatment, where the highest difference in integrin β1 expression was observed. We found that Sall2 expression significantly increases integrin β1 protein level, like the Sall2 ^+/+ cells (Figure 5D). Additionally, we evaluated the integrin β1 expression at the cell surface of iMEFs by flow cytometry. As expected, Sall2 ^−/− iMEFs exhibited a lower integrin β1 expression than Sall2 ^+/+ cells (mean PE-CD29 fluorescence intensity = 35 versus 52 a.u.) (Figure 5E). Previously, we reported detectable Sall2 protein levels in mice tissues, including the brain, cerebellum, and spleen (Hermosilla et al., 2018). To evaluate the significance of the Sall2-dependent regulation of integrin β1 in vivo, we compared integrin β1 expression between brain tissues from isogenic Sall2 ^+/+ and Sall2 ^−/− mice. Consistent with our previous results (Figures 5A–D), integrin β1 protein levels were significantly lower in Sall2 ^−/− than in Sall2 ^+/+ tissues (Figure 5F).

To validate that the regulation of integrin β1 expression by Sall2 is directly involved in the cell migration phenotype, we performed functional studies with integrin β1 overexpression or the blocking of integrin β1 function. The integrin β1 overexpression in Sall2 ^−/− iMEFs increased cell migration (wound closure percentage from 37% to 63%, Figures 5G,H), similar to that observed in Sall2 ^+/+ cells. On the other hand, the integrin β1 blockage with an anti β1antibody in Sall2 ^+/+ iMEFs showed a reduced wound closure percentage (from 60% to 9%, Figure 5I). Together, our results demonstrated that Sall2 positively regulates the integrin β1 expression and promotes cell migration through this regulation.

To investigate whether Sall2 regulates integrin β1 expression, we initially analyzed a previously published ChIP-seq (Farkas et al., 2021) and compared the ITGB1 promoter data between SALL2 wild-type and SALL2 knockout HEK293 cells. We identified SALL2-binding enrichment and associated peaks in the ITGB1 promoter region in the SALL2 ^+/+ cells (Figure 6A). Additionally, using the Transcriptional Regulatory Element Database (TRED) (Jiang et al., 2007), we identified several putative binding sites in the mouse and human integrin β1 promoter (15 sites in the mouse and 30 sites in the human promoter), mainly located in the proximal promoter region, –500 bp to +1 (Figure 6B). Conservation analysis using BLASTn revealed 79% of identity between both species at the proximal region. Next, we evaluated the integrin β1 mRNA level by real-time PCR in the previous three cellular contexts. Figures 6C–E show that Sall2 ^−/− iMEFs exhibited significantly lower integrin b1 mRNA than Sall2 ^+/+ cells under nocodazole release, spreading on FN, and normal growth conditions. Like the protein analysis, the highest mRNA difference was observed under nocodazole release. Accordingly, induction of Sall2 expression increased the integrin β1 mRNA level after 15 min of nocodazole wash-out treatment (Figure 6F).

To determine SALL2 activity at the ITGB1 promoter, we used SALL2 KO HEK293 cells co-transfected with a SALL2 expression plasmid and an ITGB1 promoter reporter previously described (Itou et al., 2017). SALL2 expression significantly increased the activity of the ITGB1 promoter (Figure 6G). By chromatin immunoprecipitation (ChIP) assays in SALL2 KO HEK293 cells, we demonstrated that SALL2 binds to the proximal promoter of ITGB1, specifically to the -631/-297 region (Figures 6H,I). In contrast, no binding of SALL2 was observed in a nonrelated promoter region (NRR) and the -251/-127 region of the ITGB1 promoter. The increment of SALL2 binding was correlated with an increase in histone H4 acetylation, a transcriptional activation marker (Supplementary Figure S7). These results support that integrin β1 is a novel Sall2 transcriptional target gene.