A Thiadiazolopyrimidinone-Based Molecule Targeting Annexin A6 Impairs Cell Motility and Epithelial-to-Mesenchymal Transition in Pancreatic Cancer Cells Lacking Annexin A1

Highlights

What are the main findings?

• Annexin A6 is upregulated in the absence of ANXA1. ANXA6 expression increases in ANXA1 knockout human pancreatic cancer MIA PaCa-2 cells, sustaining aggressive behavior.
• LAM20 targeting Annexin A6 impairs cell aggressiveness. Inhibition of Annexin A6 by LAM20 reduces cell motility and markers of epithelial-to-mesenchymal transition without affecting proliferation.

What are the implications of the main findings?

• Dual therapeutic potential against aggressive pancreatic cancer. Targeting ANXA6, in addition to ANXA1, may provide a complementary approach to limit pancreatic cancer invasiveness.

Abstract

Pancreatic carcinoma (PC) is the most lethal malignancy due to its aggressive behavior and limited therapeutic response. Among the annexin family, Annexin A1 (ANXA1) is documented to promote PC aggressiveness, and conversely, the role of Annexin A6 (ANXA6) is less explored. Here, we report that ANXA6 is significantly upregulated in ANXA1 knockout (KO) MIA PaCa-2 cells. Using LAM20, our previously identified ANXA6 modulator, we show that inhibition of this protein impairs cell motility, and epithelial-to-mesenchymal transition markers, without affecting 2D/3D cell proliferation. ANXA6 siRNA-mediated knockdown reproduces LAM20 effects, suggesting a relationship with their impact on ANXA6. Interestingly, in ANXA1 KO cells, LAM20 reduced the migration/invasion rate differently from the ANXA1 inhibitor heparan sulfate, which retains effects on the wild-type (WT) MIA PaCa-2 counterpart. These findings suggest that in cells lacking ANXA1, ANXA6 plays a compensatory role in sustaining the aggressive phenotype, albeit to a lesser extent than in WT cells. Thus, LAM20 represents a promising therapeutic strategy to impair PC aggressiveness. Our study provides new insights into ANXA1/ANXA6 crosstalk and introduces a novel approach to disturb PC pro-invasive mechanisms. Targeting ANXA1 and ANXA6 is relevant because, where ANXA1 is downregulated/absent, ANXA6 expression can be restored in a compensatory manner, partially sustaining tumor progression.

1. Introduction

Pancreatic carcinoma (PC) represents the seventh leading cause of cancer-related deaths worldwide, primarily due to late detection, aggressive local invasion, early metastasis, and resistance to conventional therapies [1,2]. The comprehensive understanding of the molecular drivers of tumor progression remains a challenge to improve diagnostic/prognostic and therapeutic strategies. To date, epithelial–mesenchymal transition (EMT) and cellular motility are some aspects deserving attention [3]. Members of the annexin family are increasingly recognized as regulators of cancer-related processes, such as membrane dynamics, intracellular trafficking, cytoskeletal organization, and signal transduction. Notably, annexins display context-dependent action either as tumor suppressors or promoters depending on the tissue type and disease stage [4]. Within this family, Annexin A1 (ANXA1) and Annexin A6 (ANXA6) have attracted interest due to their roles in sustaining the aggressive phenotype. It is known that they bind the cell membrane in a calcium-dependent manner and are implicated in multiple cell functions not completely overlapping [4]. Also, the ANXA1 structure differs from ANXA6 since the first one contains a central core composed of four domains of five α-helices about 70 amino acids in size. This core is repeated twice and connected by an extended polypeptide linker in ANXA6 whose molecular weight is about 68 kDa, almost double the 37 kDa of ANXA1 [5].

In the context of cancer, ANXA1 has been recognized as a multifunctional protein involved in inflammation and cell motility and differentiation. Depending on its intracellular or extracellular localization, ANXA1 may contribute to the occurrence of EMT and the modulation of the tumor microenvironment [6,7]. Specifically, in PC, ANXA1 is frequently overexpressed and associated with EMT, increased motility, and enhanced metastatic potential [8,9,10]. In our previous studies, we used the Clustered, regularly interspaced, short palindromic repeat (CRISPR)/CRISPR-associated protein 9 (Cas9) genome editing system to generate ANXA1 knockout (KO) MIA PaCa-2 cells, which showed a reverted aggressive phenotype. With this model, we showed that intracellular ANXA1 participates in PC progression by maintaining cell cytoskeleton integrity, leading to a more aggressive phenotype, independently of the known partner Formyl Peptide Receptor (FPR) [8].

About ANXA6, its lower levels correlate with worse survival in triple-negative breast cancer, gastric cancer, and hepatocellular carcinoma, suggesting that it normally inhibits tumor spread and proliferation. On the other hand, in ovarian, cervical and esophageal cancers, its elevated levels are linked to progression, metastasis, and drug resistance [11,12]. Compared to ANXA1, the role of ANXA6 in PC remains less characterized. In this cancer type, ANXA6 is described as able to contribute to cell survival, migration, and metastatic competence, particularly through its involvement in membrane organization, intracellular Ca²⁺ handling, and extracellular vesicle (EV)-mediated communication between cancer cells and the surrounding stroma [13,14,15]. However, unlike ANXA1, the direct role of ANXA6 in EMT regulation within PC cells has not been clearly established.

It is important to consider the broader concept of functional interplay among annexin family members, mainly in PC, where ANXA1 and ANXA6 are both linked to the promotion of aggressive behavior. The possibility that different annexin family members participate in coordinated or compensatory regulatory networks has been proposed, although direct functional relationships remain poorly defined. Based on the use of various ANXA-deficient mice strains, researchers have assessed important biological functions of individual annexins and potential redundancy within the family [16]. In detail, ANXA1^−/− mice have demonstrated compensatory changes in the expression of other annexins, including ANXA6, in the lung and spleen, whereas in the ovary, heart and anterior pituitary glands, it has been found to be down-modulated. Finally, its levels have been assessed to be largely unchanged in the thymus, gut and testis [17]. Supporting this concept, shared expression trends among several annexins, including ANXA1 and ANXA6, have been reported in bladder cancer and correlated with the tumor stage [18]. Nevertheless, how this kind of coordinated regulation may occur in PC, and how it may influence tumor aggressiveness, remains unknown. Based on these considerations, the present study was designed to explore the potential functional interplay between ANXA1 and ANXA6 in our two complementary in vitro models, of wild-type (WT) MIA PaCa-2 and the ANXA1 KO counterpart. Thus, these cell lines helped us to evaluate how ANXA6 expression and function may be modulated in the absence of ANXA1 and how such changes affect EMT-associated features and cellular motility. To further investigate these effects, we used LAM20, a [1,3,4]thiadiazolo[3,2-a]pyrimidinone-based small molecule (referred to as compound 1 in our previous study) identified through a multiparametric experimental approach as a consistent and direct cellular interactor of ANXA6 (Figure 1). This binding was validated by DARTS, SPR, and CETSA and localized to functionally relevant Ca²⁺-regulated regions of the protein [19]. These previous results indicate that LAM20 binds directly to ANXA6 in a targeted manner, focusing on regions critical for its conformational and regulatory properties. Thus, based on this knowledge, our study aims to provide new insights into the roles of ANXA1 and ANXA6 in PC.

2. Materials and Methods

2.1. Synthesis of LAM20

LAM20 was obtained following the synthetic strategy previously reported by us as a brown solid (120 mg, 50% yield after HPLC purification) [19]. RP-HPLC t_R = 25.9 min, gradient condition: from 5% acetonitrile ending to 100% acetonitrile in 40 min, flow rate of 3.5 mL/min, λ = 240 nm. ¹H NMR (400 MHz, CD₃OD): δ_H = 7.98 (dt, J = 8.7, 2.1 Hz, 2H), 7.67–7.62 (m, 2H), 7.42–7.36 (m, 2H), 6.94 (d, J = 8.6 Hz, 1H), 6.30 (s, 1H), 3.97 (q, J = 7.0 Hz, 2H), 2.36 (s, 3H), 1.29 (t, J = 7.0 Hz, 3H). ¹³C NMR (100 MHz, (CD₃)₂SO): δ = 163.5, 161.8, 158.1, 156.5, 155.3, 141.5, 132.8, 132.6, 130.9, 130.8 (2C), 127.6 (2C), 127.5, 115.7, 112.7, 107.3, 64.6, 23.7, 14.9. ESI-MS: calculated for C₂₀H₁₆BrN₃O₂S 441.01; found m/z = 442.0921 [M + H]⁺, 444.0921 [M + H + 2]⁺.

2.2. Cell Cultures

MIA PaCa-2 cells (American Type Culture Collection (ATCC)^® CRL-1420; Manassas, VA USA) were maintained in Dulbecco’s Modified Eagle Medium (DMEM) supplemented with L-glutamine 2 mM, 10% heat-inactivated fetal bovine serum (FBS), 10,000 U/mL penicillin, and 10 mg/mL streptomycin (Euroclone; Milan, Italy). ANXA1 (KO) MIA PaCa-2 cells were previously generated as described [8] and were continuously cultured under selective pressure with 700 μg/mL neomycin (Euroclone; Milan, Italy). Cells were incubated at 37 °C in a humidified atmosphere containing 5% CO₂.

2.3. Western Blotting

For the preparation of total protein lysates, WT and ANXA1 KO MIA PaCa-2 cells were directly lysed with 1x gel-loading buffer (LB 1×) (50 mM Tris–Cl pH 6.8; 2% w/v sodium dodecyl sulphate (SDS); 0.1% bromophenol blue; 10% v/v glycerol; and 100 mM β-mercaptoethanol) [20]. Protein samples were separated by SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred onto nitrocellulose blotting membranes (Amersham™ Protran™, Cytiva; Little Chalfont, UK). Membranes were blocked with 5% non-fat dry milk in TBS-Tween 20 (0.1% v/v). Protein detection was performed by chemiluminescence (Amersham™, Cytiva; Little Chalfont, UK) following overnight (O/N) incubation at 4 °C with primary antibodies against ANXA6 (mouse monoclonal, 1:1000; Santa Cruz Biotechnologies, Dallas, TX, USA), ANXA1 (rabbit polyclonal, 1:10,000; Invitrogen; Carlsbad, CA, USA), tubulin (mouse monoclonal, 1:1000; Cohesion Biosciences, London, UK) and actin (mouse monoclonal, 1:1000; Cohesion Biosciences, London, UK). Membranes were then incubated for 1 h at room temperature (RT) with the appropriate HRP-conjugated secondary antibodies (rabbit or mouse, 1:10,000; Jackson ImmunoResearch; Philadelphia, PA, USA). Immunoreactive bands were acquired using Las4000 and Amersham™ Imager 680 systems (GE Healthcare Life Sciences; Little Chalfont, UK). Densitometric analysis of band intensities, expressed as optical density (OD), was carried out using ImageJ software (version 2.14.0; NIH, Bethesda, MD, USA).

2.4. Confocal Microscopy

WT and ANXA1 KO MIA PaCa-2 cells were plated onto cover glasses (6 × 10⁴ cells/well) and subjected to treatment for 24 h. Cells were fixed with 4% v/v p-formaldehyde (in PBS 1×; Lonza; Basilea, Swiss) for 10 min, permeabilized with 0.5% v/v Triton X-100 (in PBS 1×; Lonza; Basilea, Swiss) for 5 min, and finally, blocked with goat serum (20% v/v in PBS 1x; Lonza; Basilea, Swiss) for 20 min. Cells were incubated O/N at 4 °C with primary antibodies against ANXA6 (mouse monoclonal, 1:100; Santa Cruz Biotechnologies, Dallas, TX, USA), vimentin (mouse monoclonal, 1:250; Santa Cruz Biotechnologies, Dallas, TX, USA), and N-cadherin (rabbit polyclonal, 1:100; Elabscience; Houston, TX, USA). Subsequently, cells were incubated with fluorophore-conjugated secondary antibodies and with 4′,6-diamidino-2-phenylindole (DAPI; 1:1000; Thermo Fisher Scientific; Waltham, MA, USA) to stain nuclei, for 2 h at RT in the dark. Confocal imaging was performed using a TCS SP8 confocal microscope (Leica Microsystems; Wetzlar, Germany). The quantification of the mean fluorescence intensity was performed using ImageJ software version 2.14.0 (NIH, Bethesda, MD, USA). Values are reported as the normalization between the protein signal and the DAPI signal (in arbitrary units, AU).

2.5. Wound-Healing Assay

WT and ANXA1 KO MIA PaCa-2 cells were seeded into a 24-well plate at 1 × 10⁵ cells per well. After 24 h, when full confluence was reached, a scratch wound was created across the monolayer using a sterile p10 pipette tip. After washing with PBS 1×, treatments were administered, including LAM20 in both WT and ANXA1 KO MIA PaCa-2 cells at different concentrations, either in the presence or absence of heparan sulfate HS (10 µg/mL; Laboratori Derivati Organici—LDO—S.p.a.; Trino, VC, Italy). In all experimental conditions, mitomycin C (0.5 μg/mL, Sigma-Aldrich; Saint Louis, MO, USA) was added to inhibit cell proliferation. Images of the wounds were captured at 0 and 24 h using an Axiovert 5 microscope with Axiocam 208 color (Carl Zeiss; Oberkochen, Germany), using a 10× phase-contrast objective. Quantitative data shown in the graphs were obtained through ImageJ software analysis (version 2.14.0; NIH, Bethesda, MD, USA) by averaging the measured wound distances (expressed in µm) at multiple cells’ positions along the wound edges.

2.6. Invasion Assay

The invasive capacity of WT and ANXA1 KO MIA PaCa-2 cells was evaluated using trans-well inserts (12 mm diameter, 8.0 µm pore size; Corning Incorporated; New York, NJ, USA), as previously reported [21]. Cells were seeded in the upper chamber, while treatments identical to those used in the wound-healing assay were added to the lower chamber. Mitomycin C (0.5 μg/mL; Sigma-Aldrich; Saint Louis, MO, USA) was included to prevent cell division. After 24 h, trans-wells were washed twice with PBS 1×, and fixed with 4% v/v paraformaldehyde (in PBS 1×, Lonza; Basilea, Swiss) for 10 min, followed by fixation in 100% methanol (PanReac Applichem; Darmstadt, Germany) for 20 min. Cells were then stained with 0.5% crystal violet solution (crystal violet powder; Merck Chemicals; distilled water containing 20% methanol) for 30 min. Trans-wells were imaged using an Axiovert 5 microscope equipped with an Axiocam 208 color camera (Carl Zeiss; Oberkochen, Germany), using a 10× phase-contrast objective. Quantification was performed by counting the number of invaded cells, and results were reported graphically.

2.7. Fluo-4 a.m. Flow Cytometry Assay for Detection of Ca²⁺ Mobilization

Changes in intracellular Ca²⁺ concentration were monitored using the fluorescent probe Fluo-4 a.m. (Thermo Fisher Scientific; Waltham, MA, USA). ANXA1 KO MIA PaCa-2 cells were trypsinized, washed, and placed in 1.5 mL tubes at 5 × 10⁵/mL and then incubated with LAM20 (25, 50 and 75 μM final concentrations) in PBS 1× for 30 min at RT. Then, the intracellular Ca²⁺ concentrations [Ca²⁺] were measured using the fluorescent Fluo-4 a.m. probe (Molecular Probes, Thermo Fisher Scientific, Waltham, MA, USA), as previously described [22]. The fluorescence in each sample was analyzed by a BD FACSCalibur cytometer (Becton Dickinson FACScan, Franklin Lakes, NJ, USA). Ca²⁺-ionophore (ionomycin 1 mM; Sigma-Aldrich; St. Louis, MO, USA) was used as a positive control to induce Ca²⁺ influx and release from intracellular stores, whereas the chelating agent EDTA (15 mM, 15 min; Sigma-Aldrich; St. Louis, MO, USA) served as a negative control to chelate extracellular Ca²⁺ only, as EDTA does not penetrate intact membranes.

2.8. Colony-Forming Assay

ANXA1 KO MIA PaCa2 cells (5 × 10³ cells/well) were seeded in 6-well plates and incubated at 37 °C. After 24 h from seeding, LAM20 75 µM was added for 24 h; then, growth medium was refreshed and left for ten days. At the end of the experimental time, cells were fixed with 1 mL di 4% v/v p-formaldehyde (Lonza, Basel, Swiss) for 10 min; dehydrated for 20 min with pure methanol (PanReac Applichem, Darmstadt, Germany); and stained with 0.5% w/v crystal violet in a 20% v/v methanol solution (Merck Chemicals, Darmstadt, Germany) for 30 min at RT. The wells were washed with deionized water and photographed before dissolving crystals in 1% SDS and measuring the absorbance at 570 nm, with a microplate spectrophotometer (Titertek Multiskan MCC/340; Labsystems, Midland, ON, Canada), as previously described [23].

2.9. Spheroid Generation and Area Analysis

The extracellular 3D cellular models were performed in 96-well U-bottom plates (#3799, Corning^® Costar^®, New York, NY, USA) as reported in [23]. In these plates, 5 × 10³ ANXA1 KO MIA PaCa-2 cells were seeded per well and were incubated for 10 days at 37 °C with 5% CO₂ until their formation. At this time, their images were captured using an Axiovert 5 microscope with Axiocam 208 color (Carl Zeiss; Oberkochen, Germany). A 5× phase contrast objective was used. Moreover, spheroids were treated for 24 h with LAM20, and then fresh growth medium was replaced for 3 days. At this further time point, spheroids were photographed. The dimensions of at least 10 spheroids for each cell line and experimental point were measured with ImageJ software version 2.14.0 (NIH, Bethesda, MD, USA), and the relative areas (area at experimental time [1 day with treatment + 3 days with growth medium]/initial area at 10 days from seeding) of the spheroids were analyzed by comparing the treated versus control spheroids.

2.10. Small Interfering (si)RNA Transfection

siRNA sequences against ANXA6 and siRNA Oligo-Scrambled (siRNA ctrl) were purchased from Santa Cruz Biotechnology (Dallas, TX, USA) [19,24] and used at a final concentration of 100 nM. ANXA6-siRNAs were transfected using Lipofectamine 2000 Transfection Reagent (Thermo Fisher Scientific; Waltham, MA, USA), according to the manufacturer’s instructions. Cells were harvested after 24 h from transfection. The analysis of calcium mobilization and cell migration and invasion was performed with transfected cells as reported above.

2.11. Statistical Analysis

Data analyses and statistical evaluation were carried out using Microsoft Excel (version 2601) and GraphPad Prism 10. The independent experiments were repeated at least three times, and all results were shown as the mean ± standard deviation (SD). The different groups were compared using one-way ANOVA with Tukey’s post hoc and two-tailed t tests when appropriate. Differences were considered significant if p ≤ 0.05.

3. Results

3.1. ANXA6 Expression Appeared Strongly Enhanced in ANXA1 KO MIA PaCa-2 Cells

In order to investigate the role of ANXA6 in our PC in vitro models, we first evaluated its expression levels on WT and ANXA1 KO MIA PaCa-2 cells. Thus, by Western blotting (Figure 2A) and by confocal analysis (Figure 2B), we revealed a strong enhancement of ANXA6 expression in the absence of ANXA1 compared to the WT cell counterpart.

3.2. LAM20 Negatively Affected Migration and Invasion Rate of ANXA1 KO MIA PaCa-2 Cells and the Expression of Mesenchymal Markers

LAM20 inhibited HeLa cells’ motility by binding ANXA6 [19]. Thus, we first investigated these processes on ANXA1 KO MIA PaCa-2 cells considering their high levels of ANXA6. Surprisingly, this compound, when used at the same concentration already assessed [19], did not induce any kind of effect on cell migration (Figure S1A) and invasion unlike the behavior observed in the HeLa cell line (Figure S1B). For this reason, we increased concentrations at 50 and 75 µM, with which we witnessed a significant decrease in migration and invasion rate (Figure 3A and Figure 3B, respectively). Moreover, since ANXA6 is known to mediate many of its cellular functions in a Ca²⁺-dependent manner [25,26], we assessed calcium transition through the Fluo-4 a.m. cytofluorimetric assay. In Figure 3C, we report that LAM20 was able to inhibit the transition of this electrolyte in a concentration-dependent manner, from 25 µM to 75 µM. As reported in Section 2, ionomycin and EDTA were used as technical controls. Furthermore, the role of ANXA6 in the EMT program remains controversial and tissue-specific [27,28]. Here, we showed that the treatment with LAM20 at 75 µM induced a notable decrease in vimentin and N-cadherin levels, compared to non-treated cells, as presented by the immunofluorescence images in Figure 3D and Figure 3E, respectively.

3.3. The Treatment with LAM20 Showed No Effects on Cell Proliferation in Both 2D and 3D Models

Next, we assessed the colony formation ability of ANXA1 KO MIA PaCa-2 cells in the presence or absence of LAM20 75 µM, used as reported in Section 2. The absence of any effect of LAM20 is reported in Figure 4A,B, which show representative images of the ANXA1 KO MIA PaCa-2 colonies, and the related spectrophotometric reading of crystal violet dissolved with SDS, respectively. LAM20 did not affect the growth of spheroids used as a 3D model, able to mimic the typical spatial structure of solid tumors [23] (Figure 4C and Figure 4D for spheroids’ representative images and the measure of area calculated as reported in Section 2, respectively).

3.4. The siRNA-Mediated Down-Modulation of ANXA6 Levels Induced the Decrease in MIA PaCa-2 ANXA1 KO Cells’ Motility and of the Mesenchymal Markers’ Expression

In order to further investigate the role of ANXA6 in our PC in vitro model, we induced protein expression knockdown by transfecting ANXA1 KO MIA PaCa-2 cells with ANXA6 siRNAs. Once we proved the successful silencing of the protein of interest by Western blotting, as shown in Figure 5A, at 24 h from transfection, wound-healing and invasion assays were performed. In both cases, we demonstrated a significant slow-down in migration (Figure 5B) and invasion (Figure 5C) rates on the ANXA6 knocked-down cells compared to non-transfected cells and to those treated with only Lipofectamine 2000 and with siRNA ctrl (Figure S3B,C). These last were used as technical controls to ensure safe experimental conditions. In the same conditions, through a confocal analysis, we first confirmed the ANXA6 silencing (Figure 5D), and then we were able to correlate this with the decrease also in vimentin and N-cadherin levels (Figure 5E and Figure 5F, respectively), displaying a similar trend to that observed after LAM20 treatments.

Moreover, when LAM20 was added to cells knocked down for ANXA6, no further significant migratory and invasive activity was revealed (Figure S3B and S3C, respectively) compared to ANXA6 siRNA alone.

3.5. The Inhibition of ANXA1 by HS Improved the Motility Slow-Down on WT MIA PaCa-2 Cells Together with LAM20

To investigate the interplay between ANXA6 and ANXA1 in our PC model, ANXA6 was modulated using LAM20, and ANXA1 was targeted using HS, whose effect on WT MIA PaCa2 cells has already been reported [22]. First, we proved that at low concentrations (6.25, 12.5 and 25 µM), LAM20 was able to inhibit the cell migration and invasion rate (Figure S2A and Figure S2B, respectively). Then, we used HS 10 µg/mL, able to block the ANXA1 action [22]. In the graph of Figure 6A, we displayed the inhibitory effects of both HS and LAM20 on WT MIA PaCa-2 cells, taken as single treatments. Interestingly, when used together, a more relevant additional effect was obtained. Furthermore, on ANXA1 KO MIA PaCa-2 cells, only LAM20, alone or in combination with HS, induced the slow-down of cell migration differently from HS, which confirmed the lack of effects due to the absence of ANXA1 protein in this cell line (Figure 6B). Compared to the wound-healing results, very similar trends were observed in the analysis of invasion rates on both cell lines (Figure 6C and Figure 6D, for WT and ANXA1 KO MIA PaCa-2 cells, respectively).

4. Discussion

In PC, the role of ANXA1 has been extensively investigated, whereas ANXA6 has received comparatively less attention; nevertheless, recent evidence suggests it plays a complementary yet distinct role. In this regard, Leca et al. (2016) [15] showed that cancer-associated fibroblasts (CAFs) secrete EVs enriched in ANXA6, which, upon uptake by PC tumor cells, enhance survival, migration, and metastatic ability. Notably, as happens for ANXA1 [9], elevated levels of ANXA6-positive EVs in patients’ sera correlated with a severe tumor grade. Moreover, the authors described an extracellular activity of ANXA6 mediated through the LDL receptor-related protein 1/thrombospondin 1 complex, influencing PC aggressiveness [15]. These observations indicate that ANXA6 may influence tumor progression through mechanisms distinct from those described for ANXA1 [29,30]. In this context, ANXA1 and ANXA6 may emerge as interrelated contributors to PC aggressiveness and as promising biomarkers for diagnosis and prognosis, as well as potential therapeutic targets [22,31]. However, important questions remain regarding their comparative expression patterns across PC stages. Furthermore, while both the intracellular and extracellular roles of ANXA1 have been extensively explored in PC, current knowledge on ANXA6 remains largely limited to its involvement in stromal modulation.

For all these reasons, in this work, we focused on the potential interconnection between ANXA1 and ANXA6. It is important to underscore that the imbalanced levels of the two annexins appear inconsistent with our previous work, in which we reveal no significant changes in the levels of ANXA2, ANXA4, ANXA5, ANXA6, and ANXA11 following ANXA1 removal [8]. This apparent discrepancy could be due to the use of global liquid chromatography–tandem mass spectrometry proteomic analysis, an excellent method for identifying large sets of proteins to obtain a broad profile of protein expression. Nevertheless, its level of sensitivity and detection is described as not sensitive enough if compared to Western blot and immunofluorescence techniques, as target-specific immunoassays also able to distinguish smaller relative variations in proteins and amplify those rather than total abundance [32,33]. Based on these considerations, the two proteins of interest, in addition to being members of the same family, appeared also to be related by a kind of compensatory upregulation of ANXA6 led by ANXA1 KO.

The presence of compensatory expression suggests that in biological networks, some sort of adaptation may exist when one of the members is absent. Interestingly, the ANXA6 upregulation we observed in ANXA1 KO MIA PaCa-2 cells finds a partial precedent in murine models. Indeed, with the establishment of ANXA1^−/− mice, the authors have reported compensatory changes in the expression of other annexins, including ANXA6, in several tissues. Importantly, these studies did not include an analysis of the pancreas, leaving it unclear whether a similar compensatory mechanism occurs in pancreatic tissue [17]. Overall, overlapping or compensatory functions among the members of the annexin family have been suggested as a phenomenon of redundancy further justified by a high similarity among the structures (mainly for ANXA1, A2, A4, A5, A6 and A7) and at the level of processes such as membrane organization, Ca²⁺ signaling, or vesicular trafficking rather than through direct transcriptional compensation [16,34].

In this scenario, the parallel between these murine observations and our findings in PC cells supports the notion that ANXA6 may act as a functionally adaptive player when ANXA1 levels are reduced, contributing to the maintenance of cellular motility. Thus, the ANXA1 KO model therefore remains highly motivating for our research, as it consistently provides biologically relevant and intriguing data. In the present study, the dysregulation of ANXA6 observed in this model offers additional insights relevant for PC progression. In line with this suggested adaptive mechanism, our experiments using the ANXA6 specific modulator LAM20 demonstrated that targeting this protein in ANXA1 KO cells reduced migration/invasion and EMT markers, with stronger effects than inhibition of ANXA1 alone. Notably, in ANXA1-KO cells, the higher ANXA6 may explain the less sensitive cell functions to modulation of LAM20 at low concentrations. In contrast, in WT cells with lower basal ANXA6 levels, even a modest perturbation of ANXA6 activity by LAM20 has been sufficient to significantly impair motility processes, resulting in a more pronounced inhibitory phenotype.

Moreover, in our findings, LAM20 did not show effects on cell proliferation. In this last case, the observed data is in line with the reported ability of most of the annexin family members to promote tumor cell proliferation, except for ANXA6 [35].

Therefore, siRNA-mediated silencing of ANXA6 inhibited cell motility and EMT features in ANXA1-KO MIA PaCa-2 cells, reproducing comparable effects observed upon LAM20 treatment. These similar consequences suggest that ANXA6 is not merely upregulated in the absence of ANXA1, but functionally contributes to the maintenance of an invasive, mesenchymal-like phenotype, albeit to a lesser extent than in cells expressing ANXA1. It should be emphasized that the present study does not provide direct evidence for molecular interaction between ANXA1 and ANXA6. The siRNA experiments were primarily used as a methodological approach to assess whether modulation of ANXA6 levels could influence EMT-associated phenotypes. Moreover, to support the hypothesis by which LAM20 exercises its activity by inhibiting ANXA6, we demonstrated that when this compound does not have the possibility to bind the protein, it has no effect on cell motility. Indeed, the analysis of the migration and invasion rate highlights no additive effects when LAM20 is administered to ANXA6-silenced cells. These findings are very similar to those obtained in HeLa cells in our previous study [19], supporting the conclusion that LAM20 functionally acts through ANXA6.

Therefore, the potential therapeutic approach becomes attractive considering that, in the case of an absence of ANXA1, cells still retain weak capabilities to move and quite detectable levels of EMT markers, while when ANXA6 is further inhibited, either via LAM20 or by siRNA, a superior negative effect of these aspects is found.

In this scenario, the relevant inhibitory effects of HS and LAM20 together on WT MIA PaCa-2 cells suggest that the simultaneous targeting with ANXA1 and ANXA6 could be useful to obtain a more effective therapeutic benefit. In addition, this can also be valuable to prevent compensatory adaptations within cancer cells whose ability in maintaining homeostatic balance and supporting their own resilience is a well-known feature [36,37]. Moreover, it is reasonable that this issue could also be addressed regarding further members of the annexin family, pursuing the aim to strongly interfere with the crucial signaling functions related to motility, membrane dynamics, vesicle externalization, and calcium homeostasis, all biological processes necessary for tumor cell survival and dissemination [35]. Furthermore, our results align with the current scientific efforts aimed at synergistically compromising tumor adaptability, for example, by reducing both the intrinsic invasive potential of cancer cells and their capacity to remodel the surroundings to the tumor microenvironment.

5. Conclusions

By inhibiting ANXA6 activity, LAM20 appeared able to impair the motility of PC cells, probably as a consequence of the interference with the calcium balance [38]. However, the specific downstream events and molecular mechanisms following this interaction remain to be fully elucidated. In the present study, LAM20 served as a useful tool for highlighting the unprecedented functional relevance of targeting ANXA6 together with ANXA1 in PC cells. The identification and characterization of the crosstalk between these two proteins could be used to expand the knowledge of cell motility and EMT in tumor progression. Nevertheless, future research should elucidate these mechanisms more in detail, or if other occurrences happen and/or molecular actors intervene in this context. Indeed, the block of ANXA6, even when the ANXA1 levels are low, must be described as a multi-faceted process with several values for novel therapeutic approaches. About the compensatory mechanisms, further investigations are necessary to establish if the imbalance is due to the genetic or post-translational modifications or to the inactivation of protein degradation. Additional studies are needed to elucidate how the other annexins, each of them and/or the interplay among them, become biologically and clinically relevant during PC progression. It will be interesting to focus on the interplay between ANXA1 and ANXA6 as modulators of PC stroma and as proteins contained in the EVs. Indeed, while ANXA1-correlated EV trafficking was extensively explored in these cell models, the potential contribution of ANXA6 was not specifically investigated here. Given the known role of ANXA6 in vesicle trafficking and signaling processes, further studies will be aimed at characterizing EVs released by ANXA1 KO PC cells to clarify the potential involvement of ANXA6 in tumor–stroma interactions. Finally, the further characterization of the anticancer functions of LAM20 may pave the way for novel therapeutic strategies against ANXA6-driven PC progression in the context of ANXA1 inhibition.