Asporin increases the extracellular matrix cross-links and inhibits the cancer cell migration

Abstract

Background

Migrating strategies of the triple-negative breast cancer (TNBC) together with its role in the establishment of tumor microenvironment (TME), supporting metastasis, have been extensively studied. Extracellular matrix (ECM) is a major player for the TME, establishing the 3D spatial networks with interconnected pores necessary for the mechano-physiological function of the cells. Certain collagen aligners and cross-linkers which are necessary for the formation and the stabilization of ECM networks, however, have not been studied either in normal or in abnormal tissues. Complexities in cell-cell and cell-matrix interactions, and different in types and ratios of ECM proteins in a TME challenge to reveal the precise function of a particular protein that is exhibited by special cells and if specifically present in insignificant amount. Cancer-associated fibroblasts (CAFs) predominantly occupy the major stroma of a solid tumor where they deposit extracellular proteins in the excessive amount compared to other tumor-associated cells. For example, the TNBC tumor itself is positive for asporin (ASPN) since CAFs are major ASPN exhibitors. However, the TNBC cells express it insignificantly.

Objective

The increase in ECM and its networks suppresses the metastasis.

Methods

Here, we studied the expression of collagen type I and ASPN in CAFS and MDA-MB-231 (MM231), and evaluated the role of ASPN in collagen alignment and crosslinking.

Results

TNBC cells have an insignificant expression of ASPN and scanty collagen fibers, some of which aggregate to form the stiff deranged fibers, forming large-size pores in ECM of cancer-cell-dominant outer core of TNBC that support cancer cell invasion and metastasis. Exogenous ASPN and fibroblast-ASPN supported for the collagen alignment and crosslinking that established the small-size pores in the ECM, inhibiting the cancer cell invasion.

Conclusions

The collagen aligner and the cross-linker, ASPN increases the ECM networks and decreases the migration, and this preliminary study provides the hope that ASPN might be used as an anti-metastatic drug after its confirmation through extensive studies in animal, and positive outcomes through preclinical trials.

Introduction

Extracellular matrix (ECM) consists of both basement and interstitial matrix of a solid tissue. It is an external factor that plays the crucial role in maintaining the microenvironment. ¹ Basement ECM is mainly composed of collagen IV, nidogen and laminins, creating a stable sheet-like environment for tissue growth. Interstitial ECM, however, has major collagen types I, III and IV, establishing the 3D structural integrity of the tissue. ² The major ECM producers in the interstitial matrix are fibroblasts in both normal and tumor tissues.^3,4 A key factor in normal tissue fidelity is that collagen should be relaxed, properly aligned and cross-linked to establish the scaffolding networks to every cell in normal microenvironment (NME). However, it is highly linear and orientated, causing stiff ECM with large-size pores that facilitates breast cancer invasion in breast tumor microenvironment (TME).^2,5 Classical fibrillar collagens (90% of total collagens) are major ECM proteins that establish stability and maintain integrity of normal tissues in one hand, but increase the aggressiveness of a triple-negative breast cancer (TNBC) in another hand. ⁶ In TME, cancer-associated fibroblasts (CAFs) lead the ECM deposition followed by other cancer-associated cells, for example, cancer cells, immune cells and fatty cells, for tumor growth and metastasis.^5,7 Contribution of cancer cells for collagen and certain proteoglycans is minimal compared to that of CAFs; however, it is enough to divert biochemical properties toward oncogenic pathways.^8,9

TNBC cells, for example, MDA-MB-231 (MM231), could remain dormant for years and resume growth with the tendency to leave the primary tumor, disseminating to distant sites at suitable environment.^10,11 It is known that MM231 cell secretes higher concentration of proteoglycans and lesser concentration of collagens compared to CAFs.^9,12 It is however interesting that it secretes significantly less or none of asporin (ASPN), a small leucine-rich proteoglycan (SLRP). ¹³ ASPN in the ECM not only participates in regulating fibrillogenesis, but helps in aligning and crosslinking of collagens through its direct binding to the type I collagens.^9–11 SLRP regulates collagen fibrillation and supports binding of ligands, receptors and transforming growth factor-β (TGF-β) to the collagen fibers, maintaining ECM homeostasis. ¹⁴ ASPN is similar to decorin regarding the amino acid sequences; however, it has a unique asparate-rich N terminus. ¹⁵ Previous studies found exogenous ASPN activated NF-β/p65 and enhanced the EMT process via the CD44/AKT/Erk/NF-kβ pathway in pancreatic cancer cells. ¹⁶ In contrast to these results, it has been reported that ASPN derived from fibroblasts inhibits the TGF-β1 and suppresses the EMT process, inhibiting the breast cancer invasion. ¹⁷ Similar to the proteomics result from natural MM231 tumor, the 3D ECM scaffold-based MM231 in-vitro tumor has also shown negligible expression of ASPN.^9,18

The ECM with a more fibrillar network of collagen, as seen in a normal tissue or the inner core of solid tumor tissue, inhibits the metastasis whereas ECM with stiff collagen, as seen in the outer core of TNBC, enhances the metastasis.^19,20 It has been reported that ASPN, as a controller, determines the collagen assembly. In its absence, collagen becomes stiff, influencing cancer prognosis and recurrence. In contrast, it increases the collagen fibrillar networks, inhibiting metastasis.^13,20,21 Majority of the tumors express normal or more ASPN as a whole; therefore, the expression of ASPN in MM231 dominant zone, for example, outer core of a TNBC tumor tissue where ECM is stiff with large-size pores, remains unnoticed. ¹⁷

We hypothesized that realignment and crosslinking of collagen fibrils of 3D ECM breast tumor by ASPN halts the MM231 migration. We prepared the 3D ECM scaffolds from MM231-ECM and CAF-ECM from collected MM231 and CAFs, respectively, with/without decellularization. We analyzed ASPN secreted by both MM231 and CAFs on their respective 3D ECM scaffolds at certain intervals. In addition, the cell growth rate, collagen alignment, crosslinking and invasion have been reported before and after the addition of recombinant ASPN during the 3D cultures.

Materials and methods

Coverslip culture and ASPN determination

MDA-MB-231 cells (MM231) (HTB-26) and cancer-associated fibroblasts (CAFs) (CCD-1129SK) were purchased from the American Type Cell Collection (ATCC). They were cultured in high-glucose Dulbecco’s Modified Eagle Medium (DMEA) till 80%–90% confluence. They were trypsinized and collected. Decellularized hydrogels from both MM231 and CAFs were prepared by freeze-thaw cycles with subsequent nuclease treatment to remove the DNA contents and grinding of decellularized cells using mortar and pestle in the presence of liquid nitrogen with some modifications as described.^9,22 Briefly, the cells in working PBS were centrifuged at 10,000 rpm for 5 minutes at 4°C after 10 minutes in the 37°C water bath. Process was repeated 4 times, and finally nuclease was added and kept the tube at 37°C for 4 hours to completely remove the DNA contents. DNA content was measured and less than 1% DNA content was considered complete decellularization. Decellularized pellets were washed with working PBS three times, and it was processed for the hydrogel preparation by grinding under liquid nitrogen. 10% of ECM hydrogel in 1x PBS from both MM231 and CAFs was prepared and kept in −20°C for a week or in −70°C for a long-term storage till use. Working concentration of hydrogel for slide coating was 0.1% in 0.01 M HCL. 20 µL of cold hydrogel was homogeneously distributed over the clean and sterile coverslip. Coverslip was kept in the 37°C incubator for 30 min for the crosslinking of ECM and its attachment, and washed with working PBS to remove excess and uncoated hydrogel. MM231 cells (10⁵ cells/20 µL) were cultured on coverslip coated with MM231-ECM hydrogel and CAFs on CAF-ECM coated coverslip. Both cells were cultured till the 95% confluence. The cells were fixed and preserved by using 4% formaldehyde. It was washed with working cold 1x PBS twice and followed the protocol for immunostaining. Briefly, anti-ASPN antibody (1:500, # 102105-420 VWR) and anti-collagen I antibody (1:500, #10708-718 VWR) were applied and incubated for an hour in room temperature. Coverslip was washed three-time with working cold PBS and secondary conjugated antibodies, anti-rabbit Alexa Fluor 488 (anti-rabbit, Southern Biotechnology) and anti-IgG Goat Polyclonal antibody (Cy5, anti-Mouse IgG, Novus), were applied for collagen I and ASPN, respectively. After three-time washed with cold 1x PBS, coverslips were mounted using fluoro-mounting media and images were captured using the fluorescence microscope (ECHO Revolve).

Fabrication of 3D ECM scaffolds

Similar in coverslip culture, we collected the sufficient cell-pellets from both MM231 and CAFs after trypsinization and centrifugation. Here, the decellularization process was not processed, instead cell-pellets were transferred into the mold and kept at 4°C refrigerator to create the scaffold that could mimic the natural tumor tissue integrity. Since cells were dead and any role of nuclei or DNA of dead cells was not expected in vitro culture. 3D scaffolds were prepared with slight modification of the method we previously described. ⁹ Briefly, it was washed with 1x PBS three-time, kept fist at −70°C for half an hour in a 2 mm × 2 mm mold and transferred into the lyophilizer with the temperature set at - 45°C for 48 hours. It was then treated with tyrosinase solution (0.01%) and kept into the 37°C incubator for 3 hours for crosslinking the ECM proteins. They were treated with absolute ethanol for an hour at 4°C with frequent exposure (3–4 times at every 15–20 min) in UV rays, followed by three-time washed with 1x PBS. This process helped convert the compact cell-pellet into the 2 mm × 2 mm 3D scaffold with maximum integrity and stability. Scaffolds were then kept inside the refrigerator in a closed container with 1x PBS till use.

3D culture

3D scaffolds prepared from the MM231-ECM and CAF-ECM were used for the 3D cultures. GFP labeled-MDA-MB-231 (GFP-MM231) (SC040-bsd GenTarget Inc), 10³ cells/scaffold, were seeded onto the scaffolds and let them grow for 14 days, changing the media (high-glucose DMEM media from Sigma with 1% penicillin/streptomycin and 10% FBS) in every alternate day. Cell migration assay, relative proliferation assay and western blot testing were performed.

Cell migration assay

Cell migration assay of both MM231 and CAFs cells was performed in both 2D and 3D cultures in the presence or absence of recombinant ASPN. Non-coated and hydrogel-coated (0.1%) culture plates were applied for the migration studies in 2D culture. We applied 1 mm width sterile adhesive tape through the central of 35 mm culture plates from one end to another. After exposure to UV rays for 15 min for sterilization, we dispensed 10⁵ cells/mL on the plate. We aspirated the media after 30–45 min and washed with 1x PBS twice to remove any unattached cells. We carefully removed the tape and added the fresh media with ASPN (0.01 µg/mL, MBS1292257 MYBioscience) in the experiment plates and without ASPN in control plates. Plates were incubated for 4-day, taking pictures at every 12 hours. We noticed that there was not a difference in migration ability of CAFs by ASPN; however, there was a significant decrease, especially in hydrogel-coated 2D culture plates, in migration of MM231 cells in the presence of recombinant ASPN. The migration of CAFs was almost similar in both experiment and control plates. It might be because of no effect on CAFs by ASPN added into the media since they secrete ASPN themselves in higher concentration. ²³

Cell migration assay in 3D culture

After finding the results from 2D cultures that there was a significant decrease in MM231 cell migration, 3D culture with GFP-MM231 had been performed. The purpose of 3D culture was to observe the migration ability of cancer cells after addition of recombinant ASPN. 10³ cell/scaffold was loaded on the MM231-ECM and CAF-ECM scaffolds, letting them culture for 14 days with refreshing media every alternate day. ASPN (0.01 µg/mL) had been added to the experiment group at day first and no ASPN was in the control group. After 14 days, the scaffolds were preserved in 4% formaldehyde and cyrosectioned for the ASPN immunostaining and nuclei with DAPI. The sections from the middle area of the scaffolds were selected for the migration study. After immunostaining, the slides were mounted with fluoro-mounting media and were observed under the fluorescent microscope.

Proliferation assay

We performed the proliferation assay in both 2D and 3D cultures. Since the growth of the cells was rapid in 2D cultures, we performed proliferation assay each day for 4 days. However, we performed proliferation assay at 1^st, 3^rd, 7^th and 14^th day in 3D cultures because of the slow growth rate. We used 10³ cells /scaffold for 3D cultures and 10⁵ cells for 2D cultures. At every time points, we fixed the cells with the 4% paraformaldehyde in 48-well plate, and stained with the Ki-67 (rabbit, Novus, 20 µL in 100 µL working PBS) for 5 min to 2D and 20 min to 3D culture at room temperature in the dark. After washing with staining buffer twice, 10 µl of Propidium Iodide staining solution was added and waited for 5 minutes and flow cytometric analysis was proceeded for the counting of viable and dead cells. We repeated the experiments three times. Additionally, we performed colorimetric proliferation assay three times by cell proliferation kit I (MTT) (Roche 11465007001). The results from both tests were consistent.

Western blot

We prepared the cell lysate after their collection by using cold PBS and the cell scraper, keeping the condition cold by using the ice throughout the collection time. Protease inhibitor cocktail (20 μL fresh protease in 180 μL ice cold cell lysis buffer) has been used to lyse the cells. Supernatant was collected after incubation of cell-pellets in lysing buffer for 30 minutes on ice and centrifugation for 10 minutes at 12,000 RPM in cold room and stored frozen on ice and measured the protein concentration. We performed the western blot through the standard procedure as described with slight modification. ²⁴ Briefly, 10% stacking gel and 6% separating gel of polyacrylamide were prepared for the electrophoresis. Total 15 μL of sample after addition of sample buffer (50 μg of final total protein) was loaded in to each well along with the load marker (6 μL). After electrophoresis, proteins were transferred from the gel into the PVDF membrane (82021–258, VWR) using transfer buffer at 4°C overnight. Membrane was then blocked with 5% skim milk in working TBST for an hour. Primary anti-ASPN antibody in 5% BSA (1:500 dilution) was applied and kept on the horizontal rotator for 12 hr at 4°C. After three-round (10 minutes each round) of washing in TBST, we applied the HRP-conjugated secondary antibodies (prepared in working PBS, 1:1000 dilution) for an hour at room temperature with shaking. Membrane was washed with TBST three times, 5 minutes each wash. After application of freshly prepared substrate working solution (equal parts of the peroxide and Luminol Enhancer Solution), the membrane was incubated for a minute at room temperature and exposed under the ChemiDoc-It imager (BIO-RAD). Bands were then captured and analyzed.

Statistical analysis

Data are expressed as a mean ± SD of three independent experiments. The statistical significance of the differences among group was analyzed using One-Way ANOVA test through the software GraphPad Prism 9.5.1.

Results

MM231-ECM expressed negligible ASPN; however, CAF-ECM expressed significantly high ASPN

2D cultures and sections from 3D cultures (in MM231-ECM and CAF-ECM scaffolds) were used for the immunofluorescence staining and western blotting. Consistent with previous studies, our study has shown that ASPN has been expressed highly in CAFs and has shown little expression in MM231^9,13 (Figure 1(A)–(C)). CAFs expressed more ASPN in 2D cultures than in 3D cultures. It might be due to the increased proliferation of CAFs in 3D culture to CAFS expressed comparately more ASPN in both 2D and 3D cultures. The ASPN from MM231 remained insignificant regardless of the difference in cultural substrata, 2D or 3D (Figure 1(A), (C), and (D)).

Figure 1.

Expression of the Asporin (ASPN) by Extracellular Matrix (ECM) of MDA-MB-231 (MM231) and cancer-associated fibroblasts (CAFs). ASPN (yellow color in figure (A) and (B), shown by a white arrow-head) expressed by CAFs (B) were more compared to the MM231 cells (A). Western blot analysis had shown the ASPN expression in MM231 cell was significantly less compared to the CAF cells (C) and (D). The green-florescent tagged MM231 invasion was analyzed (E) and further showed their invasion toward the center of 3D scaffold (2 mm × 2 mm) made of MM231-ECM (F) and CAF-ECM (G). Invasion of the MM231 was more in MM231-ECM scaffold (F) compared to the CAF-ECM scaffold (G). ASPN expression was quantitatively measured in IB using Image software (public domain, BSD-2, National Institutes of Health) and expressed in pixels. Error bars in figures represent the SD of the means of the three independent experiments. Scale bar = 50 µm and p ≤ 0.05.

Recombinant ASPN changed the cellular morphology through the induction of cytoplasmic protrusions

We performed the 2D cultures to observe whether any morphological changes occur in the presence of the ASPN in both MM231 and CAF cells. We particularly observed for the cytoplasmic protrusions and noticed that GFP-MM231 had almost none of cytoplasmic extensions when they were cultured without recombinant ASPN; however, they were noticed when cultured in the presence of ASPN, showing the morphological transformation of the cells (Figure 2(A) and (B)).

Figure 2.

Demonstration of cell morphology and cytoplasmic protrusions. ASPN expression by MM231 and CAF cells was analyzed before and after the addition of recombinant ASPN in the 2D culture media. The changes in the MM231 morphology (A) by an expression of cytoplasmic protrusions after the culture in the presence of ASPN (B) were noticed and statically analyzed for the total numbers of protrusions (C) and lengths (D). ASPN has induced to increase more cytoplasmic protrusions in MM231 cells (B) when compared to their morphology without ASPN (A). Changes in cytoplasmic protrusions in CAFs by the addition of recombinant ASPN have not seen significantly different (F) when compared to their morphology in absence of ASPN (E) (Red arrows show the cytoplasmic protrusions in the cells). Error bars in figures represent the SD of the means of the three independent experiments. Scale bar = 20 µm and p ≤ 0.05.

We have analyzed and observed that not only protrusion numbers had been increased, but protrusion length also increased significantly in the culture with ASPN (Figure 2(C) and (D)). The CAFs had cytoplasmic protrusions before the ASPN treatment and there were not significant changes in the protrusions after the treatment of the cells with ASPN (Figure 2(E) and (F)). The morphological transformation in MM231 cells by ASPN leads to the enhancement of cell-ECM and cell-cell interactions and their participation in ECM networks. However, there were not significant changes in the proliferation of both MM231 and CAF cells by recombinant ASPN. Therefore, physical environment changes due to the enhancement in cell-ECM or cell-cell interactions by induced protrusions play the vital role in the cancer cell migration as nicely described by Hoang Anh Le and Roberto mayor ²⁵ (Figures 3 and 4).

Figure 3.

MM231 cell migration test in absence or presence of the ASPN in 2D cultures. The green fluorescent MM231 cells were cultured in absence and presence of recombinant ASPN and monitored for 24 hours. Migration studies were performed using coverslip culture in no-ECM coated and ECM coated (MM231-ECM or CAF-ECM) coverslips. After noticing the delay of cell migration in ECM coated coverslip cultures, cultures in MM231-ECM and CAF-ECM coated coverslips without ASPN (A) and (B) were performed and statistically analyzed for cell migration in terms of distance (C) and number of cells migrated (D) with comparison to results of cultures in presence of ASPN (F) and (F). Cells without ASPN treatment migrated faster (A) and (B) in terms of distance (C) and cell population (D) when compared to the ASPN treated cells (E) and (F). Error bars in figures represent the SD of the means of the three independent experiments. Scale bar = 100 µm and p ≤ 0.05.

Figure 4.

Extraneous recombinant ASPN induced deposition and crosslinking of collagen by both MM231 and CAF cells. MM231 cells expressed less collagen (A) compared to CAF (B). Because of the less collagen for the ECM networks, large-size pores had been noticed in MM231-ECM (A) compared to CAF-ECM (B). The more deposition of the collagen and transformation of the large-pore ECM scaffold (A) to the small-pore ECM scaffold (D) had been statistically analyzed (C) after 14 days culture of MM231 cells in MM231-ECM 3D scaffolds and compared to ECM of CAF cells in CAF-ECM 3D scaffold (B), (C), and (E). There was not much significant different in the collagen deposition and crosslinking in the ECM of CAFs after culture of cells in the presence of ASPN (0.01 µg/mL). Error bars in figures represent the SD of the means of the three independent experiments. Scale bar = 50 µm and p ≤ 0.05.

Recombinant ASPN had no impact on cell proliferation

A relative proliferation assay was performed using ECM coated/non-coated plates with or without ASPN. There was no significant difference in the proliferation of both MM231 and CAFs after the addition of ASPN in 2D or 3D cultures. The result was expected since ASPN interacts with the collagen of ECM to realign that affects migration but not multiplication.

Recombinant ASPN altered the cancer cell migration

Since the ASPN expression was high in ECM from CAFs and almost none from MM231 cells. We added the extraneous human recombinant ASPN, keeping all culture condition the same in both 2D and 3D cultures. We performed migration assay for 3 days in 2D culture and 14 days in 3D culture. After 24 hours of 2D culture on non-coated slides, the assay revealed that MM231 cell migration was better in the absence of ASPN in terms of the migrating distances and number of cells migrated (Figure 3(A)–(D)). Both the distance migrated by cells and their numbers were inhibited by the recombinant ASPN (Figure 3(C)–(F)). Distance and number of cancer cells were, however, less, similar patterns of migration were observed in 2D cultures on coated slides. The tendency of the migration was the same in 3D cultures where MM231 cells had shown a significant decrease in migration in the presence of ASPN compared to the migration of cells in the absence of ASPN. The decrease in migration rate is, may be due to the rearrangement of collagen fibers by ASPN leading to a more complex ECM network formation and change in the phenotypic properties in MM231 cells as we discussed above in Figure 2 since matrix stiffness and initial cell density were the same in both groups (with ASPN and without ASPN treatment). An extensive study is required to understand the reasons of how a significant decrease in the cell migration in 3D cultures happens in presence of recombinant ASPN.

Furthermore, it was an interesting that MM231 cells cultured on CAF-ECM coated coverslip had demonstrated none significant changes in the migration after addition of ASPN. It might be due to a sufficient concentration of ASPN by CAF-ECM itself that is enough to cross-link the collagen fibers and therefore, there was not any effect of extraneous ASPN on already established ECM networks fibrillar networks (Figure 3(D)).

Recombinant ASPN supported conversion of large-porous ECM to small-porous ECM

MM231 and CAF cells were allowed to confluent growth, changing the media every alternate day for 14 days in absence and presence of the recombinant ASPN. After confluent growth, they were preserved by 4% paraformaldehyde and the decellularization technique was performed to remove the nuclei, making the observation of ECM porosity relatively easier by using the 0.5% sodium dodecyl sulfate for 24 hours.

Microscopic images of cryosections after fluorescent staining showed that there were increased in cross-links among collagen fibers compared to controls (without ASPN treatment). We have noticed that large-size MM231-ECM networks (Figure 4(A)) and small-size CAF-ECM networks (Figure 4(B)) formed by cells without ASPN treatment and analyzed (Figure 4(C)). We analyzed the porosities of both MM231-ECM and CAF-ECM (Figure 4(C)) that had been changed after the addition of ASPN (Figure 4(D) and (E)). There were noticeable changes in the porosities in MM231-ECM after growing of the MM231 cells in the presence of ASPN compared to those of CAF-ECM after MM231 culture (Figure 4(D) and (E)).

Recombinant ASPN inhibited cell invasion

Both CAF and MM231 cells were treated with the ASPN (0.01 µg/mL) at the first day of the culture and allowed to culture for 4 days in 2D and 14 days in 3D cultures The human recombinant ASPN with the concentration of 0.15 µg/mL was effective for normal interstitial cells; however, we selected the concentration (0.01 µg/mL) that sufficed to express the ASPN by the cancer cells. ²⁶ Immunofluorescent (IF) staining results have shown that MM231 expressed the ASPN on their ECM after treatment with the extraneous ASPN compared to the control (no ASPN treatment) (Figure 5(A)–(D)). Western blot was performed and results were consistent with those of IF results, indicating that extraneous ASPN increased the ASPN expression from the cancer cells (Figure 5(C) and (D)).

Figure 5.

Expression of ASPN by MM231-ECM after addition of extraneous recombinant ASPN and MM231 cells invasion assay in the 3D scaffolds. Scaffolds were prepared from the MM231-ECM and MM231 cells were cultured in absence (A) and presence of ASPN (B). Extraneous ASPN induced MM231 to express the ASPN in its ECM (B) (yellowish green color as shown by white arrow-head in the picture (B). IB result showed that the extraneous ASPN induced MM231 cells to express higher ASPN (C) and (D). The cell invasion assay was performed in absence (F) and presence of ASPN (G) and was statistically analyzed (E). Cell invasion was markedly decreased in the presence of ASPN (E) and (G) in 3 weeks of culture. Error bars in figures represent the SD of the means of the three independent experiments. Scale bar = 50 µm and p ≤ 0.05.

Because of no significant changes in the CAF-ECM networks and migration of CAFs in the presence or absence of ASPN in 2D cultures as seen in Figures 3 and 4, we excluded the CAF-ECM and CAFs in 3D invasion assay. MM231-ECM had been used for this study where ASPN induced the cytoplasmic extensions or microfibers from the MM231 cells, helping cell-matrix interactions and inhibiting the cell invasion (Figure 5).^27,28

We cultured GFP-MM231 cells in MM231-ECM scaffold in presence or absence of ASPN for 14 days. At 14^th day of the culture, we fixed the scaffold using 4% paraformaldehyde and cryosectioned the 3D scaffolds. To ease the handling of fragile scaffolds and to minimize the ECM artifacts, they were treated with 10% sucrose solution before impregnated with the OCT. We kept scaffolds at −20°C to make more hardener before cryosectioning. 10 µm sections were prepared and immunofluorescence was performed for ASPN and nuclei. When GFP-MM231 cells were introduced to the 3D MM231-ECM prior to ASPN treatment, there was a high cell invasion due to scaffolds with large-size pores with the stiff collagens (Figure 5(E) and (F)). Migration of cells during culture in the presence of ASPN was reduced, as shown in the Figure 5(E) and (G). Fluorescent images showed that addition of ASPN inhibited the migration of the GFP-MM231 cells significantly compared to migration of control cells without ASPN (Figure 5(E)–(G)).

Discussion

MDA-MB-231 (MM231) cell is a highly aggressive TNBC with a high mortality rate compared to other breast cancer types because of the absence of specific targets for targeted therapy. TME associated with MM231 cancer plays a vital role in metastasis, and ECM proteins guide the TME for the tumor growth (Figure 1).^29,30

At an early tumor stage, MM231 cells aggregate together that facilitate cancer metastasis through the misaligned ECM.^31–33 Cancer cells also direct cancer-associated cells in a tumor, for example, immune cells, stromal cells, T lymphocytes, fatty cells and CAFs to secrete the soluble and insoluble proteins and factors, accelerating the cancer progression.^2,34 ECM is directed to support tumor progression, growth and metastasis in TME.^8,9

As the tumor continuously grows, the central core is occupied by CAFs with the formation of more complex networks through the deposition of high concentration of collagens together with collagen cross-linkers, for example, ASPN.^14,35 MM231 cells however are present in the outer core of the growing tumor with a very minimal number of CAFs.^9,36 Collagen secreted by MM231 cells has linear fibers in a tumor tissue, supporting for the formation of stiff ECM that enhances cancer cell migration.^18,19,37 Since MM231 secretes minimal of collagens and ASPN (Figure 1(A), (C), and (D)), collagen networks are compromised and ended with stiffed collagen fibers, leading to large-size pores (Figure 4(A) and (C)) that facilitate the migration and invasion.^7,38 The extracellular matrix usually has pores of 1 µm–20 µm in diameter; however, in the outer core of MM231, tumor has ∼10 µm or more. ³⁹ The large-size pores act as pre-existing channel-like pathways that facilitate the cell migration with minimal degradation of surrounding ECM (Figure 4(A) and (C)).^40,41 One cause of collagen linearization and reduction in ECM network formation in MM231-ECM is due to the lack of collagen aligners and cross-linkers, for example, ASPN (Figure 1(A), (C), and (D)).^20,31,42 In this study, we used the extraneous recombinant ASPN as a collagen aligner and a cross-linker. MM231 cells, which expressed significantly less ASPN, started expressing ASPN in the presence of recombinant ASPN; however, there was not significant change in ASPN expression by recombinant ASPN in CAFs (Figures 1(C) and (D); 5(C) and (D)). The realignment of collagens and transformation of the large-pore 3D ECM to small-pore 3D ECM with more complex collagen networks were consistent results from previous studies and showed the role of ASPN as the collagen aligner and the cross-linker as shown in Figure 4. ³⁹ ECM network with small-size pores (∼7 µm²) resists the cell migration and the degradation of ECM complex is required to initiate the migration (Figures 1(E) and (G); 5(B), (E), and (G)). ³⁹ ASPN reduced the collagen bundle formation (stiffness) by directing the collagen fibers in network formation and decreased the cancer cell invasion and migration by obstructing the physio-mechanical pathways of the migrating MM231 cells.^39,43,44

It has been shown that CAFs secrete more ASPN than normal fibroblasts. ⁴⁵ Hormone receptor (HR) positive breast cancer cells express ASPN significantly; however, triple-negative breast cancer (TNBC) has the insignificant expression. ¹⁷ Simkova et al. demonstrated that transfected TNBC with recombinant ASPN had a significant growth reduction. ASPN can bind to type I and II collagen, significantly inhibiting collagen fibrillogenesis. It competes with decorin in binding sites for various signaling cascades that may result in regulating the development of the ECM. ¹³ Many studies suggest ASPN plays dual roles in the cancer pathogenesis. ASPN has been shown to act as an oncogene in cancers such as pancreatic, colorectal, and prostate but as a tumor suppressor in triple-negative breast cancer.^{16,17,46–48} In addition, it has been noted that ASPN increases the oxidative stress to the gastric cancer cells that favor the migration and invasion capacity through the induction of CD44 and activation of Rac1 and MMP9. ²³ However, ASPN associates with better prognosis of low-grade breast cancer by inhibition of the TGF-β1.^13,17 High ASPN expression is specifically associated with less aggressive TNBC tumor with its reduced growth (Figures 1(G), 3(F) and 5(G)).^9,42,49 It is known to modulate TGF- $\beta$ 1 by inhibiting its phosphorylation through binding directly and preventing downstream signaling of multifunctional regulation. ¹⁷ It has also shown to play a part in the reversal of epithelial-mesenchymal transition in breast cancer, this blocks the tumor from becoming more malignant and invasive. ⁵⁰

In addition, transition of blunt MM231 to microfiber-protruded MM231 by ASPN might support for cell-matrix interaction, delaying in the cell migration (Figure 2). The types of protrusions produced by MM231 under the influence of extraneous recombinant ASPN, for example, filopodia, pseudopodia, lamellipodia and lobopodia and their role from attachment to migration as described by Tommy et al. are not known (Figure 2). ⁵¹ However, the induced protrusions are stabilized locally to trigger the adhesion, thereby lowering the migration capacity of the cells (Figure 3). ⁵² The migration of the blunt MM231 cells through the formation of pseudopodia from pre-existing large microchannel-like pathways is faster, as seen in Figure 3 without ASPN; however, ECM networks should be degraded to get success through complex ECM networks with small channels.^39,53

Conclusion

This study revealed that extraneous ASPN induced the ASPN expression in MM231 cells. After ASPN treatment, MM231 cells had reduced in metastasis due to the rearrangement of collagen fibers, decreasing ECM stiffening and transforming from the large-pore to small-pore size ECM networks. In addition, ASPN had changed the MM231 morphological feathers from blunt to elongated protrusions, enhancing cells to cell and cell to -matrix interactions. Extensive studies are required to understand the detail mechanisms underlying its role in the metastasis inhibition. Exogenous- or fibroblast- ASPN could possibly be of the anti-cancer therapeutics option which targets tumor ECM matrix for its modification through alignment of collagen fibers, reducing the stiffness and decreasing the ECM pores that inhibit the metastasis.

Statements and declarations