Pro-Apoptotic Activity and Cell Cycle Arrest of Caulerpa sertularioides against SKLU-1 Cancer Cell in 2D and 3D Cultures

Cancer is a disease with the highest mortality and morbidity rate worldwide. First-line drugs induce several side effects that drastically reduce the quality of life of people with this disease. Finding molecules to prevent it or generate less aggressiveness or no side effects is significant to counteract this problem. Therefore, this work searched for bioactive compounds of marine macroalgae as an alternative treatment. An 80% ethanol extract of dried Caulerpa sertularioides (CSE) was analyzed by HPLS-MS to identify the chemical components. CSE was utilized through a comparative 2D versus 3D culture model. Cisplatin (Cis) was used as a standard drug. The effects on cell viability, apoptosis, cell cycle, and tumor invasion were evaluated. The IC50 of CSE for the 2D model was 80.28 μg/mL versus 530 μg/mL for the 3D model after 24 h of treatment exposure. These results confirmed that the 3D model is more resistant to treatments and complex than the 2D model. CSE generated a loss of mitochondrial membrane potential, induced apoptosis by extrinsic and intrinsic pathways, upregulated caspases-3 and -7, and significantly decreased tumor invasion of a 3D SKLU-1 lung adenocarcinoma cell line. CSE generates biochemical and morphological changes in the plasma membrane and causes cell cycle arrest at the S and G2/M phases. These findings conclude that C. sertularioides is a potential candidate for alternative treatment against lung cancer. This work reinforced the use of complex models for drug screening and suggested using CSE’s primary component, caulerpin, to determine its effect and mechanism of action on SKLU-1 in the future. A multi-approach with molecular and histological analysis and combination with first-line drugs must be included.


Introduction
Cancer is the second-leading cause of death worldwide after cardiovascular diseases [1]. It is characterized by the uncontrolled and continuous growth of cells that form masses of tumors with the ability to generate metastases. By 2040, it is estimated that the incidence will increase by 47% compared to 2020 [2]. GLOBOCAN reports that lung cancer is the leading cause of death among different types of cancer, with 1,796,144 deaths reported for both sexes, representing 18.2% of cancer-related deaths worldwide and a prevalence of 5.9% over 5 years [3]. Moreover, in 2022, the United States registered 1,918,030 new cancer cases and 609,360 cancer deaths, including approximately 350 deaths per day from

HPLC-MS Analysis of Extract Identified Compounds of CSE in Negative and Positive Ionization Mode
As shown in Tables 2 and 3, the HPLC-MS profile of CSE allowed the tentative identification of phenolic compounds, including phenolic acids, flavonols, flavones, phlorotannins, phenolic diterpenes, and alkaloids. The identification of the phytochemical compounds was established based on the peak area and retention time [17,18]. The main chemical compound identified was caulerpin, which showed ionization in both negative mode (m/z = 397) and positive mode (m/z = 399) and presented the same retention time (17.7 min) in both polarities (Tables 2 and 3). The spectrum mass of the principal compound is shown in Figure 1, and some compounds' chemical structures are in Figure 2.

Identified Compounds of CSE in Negative and Positive Ionization Mode
As shown in Tables 2 and 3, the HPLC-MS profile of CSE allowed the tentative identification of phenolic compounds, including phenolic acids, flavonols, flavones, phlorotannins, phenolic diterpenes, and alkaloids. The identification of the phytochemical compounds was established based on the peak area and retention time [17,18]. The main chemical compound identified was caulerpin, which showed ionization in both negative mode (m/z = 397) and positive mode (m/z = 399) and presented the same retention time (17.7 min) in both polarities (Tables 2 and 3). The spectrum mass of the principal compound is shown in Figure 1, and some compounds' chemical structures are in Figure 2.    Compounds detected in a positive ionization mode. RT: Retention time.

CSE Decreased Cell Viability in a 2D Culture Model
Cell viability is the percentage of healthy cells within a population [19]. Different methods allow the evaluation of cellular health. However, evaluating membrane integrity is the most effective test for detecting dying cells because the latter is crucial for viable cells. In contrast, cells with compromised membranes are considered dead [20].
This work evaluated the cell viability in a 2D SKLU-1 lung cancer cell culture with the exclusion dye Sytox Green. CSE decreased the cell viability in 2D model culture with an IC50 (maximum inhibitory concentration) of 80.28 μg/mL after 24 h treatment exposure ( Figure 3).

Phase 1. Cytotoxic Effects of CSE and Cis in a 2D Culture Model of SKLU-1 Lung Cancer Cell CSE Decreased Cell Viability in a 2D Culture Model
Cell viability is the percentage of healthy cells within a population [19]. Different methods allow the evaluation of cellular health. However, evaluating membrane integrity is the most effective test for detecting dying cells because the latter is crucial for viable cells. In contrast, cells with compromised membranes are considered dead [20].
This work evaluated the cell viability in a 2D SKLU-1 lung cancer cell culture with the exclusion dye Sytox Green. CSE decreased the cell viability in 2D model culture with an IC 50  Ayoub et al. [21] consider a promising IC50 for botanicals/crude extract should be up to 100 μg/mL. Therefore, according to these criteria, an IC50 of 80.28 μg/mL, registered by CSE after 24 h exposure to SKLU-1, is within this parameter. These findings are the first promising results of the macroalgae C. sertularioides on SKLU-1 lung cancer cells.  Ayoub et al. [21] consider a promising IC 50 for botanicals/crude extract should be up to 100 µg/mL. Therefore, according to these criteria, an IC 50 of 80.28 µg/mL, registered by CSE after 24 h exposure to SKLU-1, is within this parameter. These findings are the first promising results of the macroalgae C. sertularioides on SKLU-1 lung cancer cells. The 3D cell culture model is more resistant and complex than the 2D model ( Figure 4A-C). Because of their ability to accurately mimic a tumor's natural microenvironment, response to stimuli, cell polarity, cell differentiation, protein synthesis, cell morphology, nutrient diffusion gradients, cell-cell and cell-extracellular matrix interaction, and drug metabolism ( Figure 4A) [22], in vitro 3D studies have revealed information about tumorigenesis that has not been detectable with traditional 2D models [23,24].  2D culture models allow initial evaluation of the efficacy of anticancer drugs. However, in these experiments, all cells are uniformly exposed to nutrients, oxygen, and drugs, cell differentiation could be better, and the junctions are less accurate than the real junctions ( Figure 4B) [26]. In contrast, in vivo, tumors expand as three-dimensional (3D) multicellular masses, where cells have variable and limited access to nutrients, metabolites, and drugs ( Figure 4B,C). Therefore, the response of the 3D model to therapeutic interventions is different from that of two-dimensional models (2D) ( Figure 4C) [27].
On the other hand, in vivo mouse models are more relevant than the 2D culture model due to complex biology. However, research on living animals raises ethical issues addressed through the 3Rs framework (Replacement, Reduction, Refinement) to reduce animal use [28,29]. Moreover, animal models are expensive, time-consuming, and require  [25]. (A-C), schematics of the 2D monolayer and 3D cell culture, structural, morphologic, and physiologic differences.
2D culture models allow initial evaluation of the efficacy of anticancer drugs. However, in these experiments, all cells are uniformly exposed to nutrients, oxygen, and drugs, cell differentiation could be better, and the junctions are less accurate than the real junctions ( Figure 4B) [26]. In contrast, in vivo, tumors expand as three-dimensional (3D) multicellular masses, where cells have variable and limited access to nutrients, metabolites, and drugs ( Figure 4B,C). Therefore, the response of the 3D model to therapeutic interventions is different from that of two-dimensional models (2D) ( Figure 4C) [27].
On the other hand, in vivo mouse models are more relevant than the 2D culture model due to complex biology. However, research on living animals raises ethical issues addressed through the 3Rs framework (Replacement, Reduction, Refinement) to reduce animal use [28,29]. Moreover, animal models are expensive, time-consuming, and require resources [30,31]. According to the Food and Drug Administration (FDA) report in 2004, less than 8% of medicinal compounds entering Phase I trials reach the market [25], which might be because mouse models still need to reproduce the complexity of human physiology and metabolism fully. Alternatives to the 2D culture and in vivo mouse models include clinical samples. Still, their limited use due to intratumoral heterogeneity and their many federal regulations [32] makes them routinely low-yielding and challenging to use. Therefore, the 3D model was used to preserve the geometry of typical tumors in vivo.
In this work, cell viability in a 3D SKLU-1 lung cancer cell culture was evaluated with the exclusion dye Sytox Green, registering a decrease in cell viability of 47, 71, and 77% by 500, 800, and 1000 µg/mL of the CSE, respectively, and 97% for 9.94 µg/mL of cisplatin (Cis) (positive control) ( Figure 5A). As shown in Figure 5B, cell death is distinguished by the penetration of the dye, which occurs when the cell membrane is compromised, beginning a death process. Considering the results obtained in this experiment, the two concentrations that generated the most significant cytotoxic effect were chosen for the subsequent tests. (Cis) (positive control) ( Figure 5A). As shown in Figure 5B, cell death is distinguished by the penetration of the dye, which occurs when the cell membrane is compromised, beginning a death process. Considering the results obtained in this experiment, the two concentrations that generated the most significant cytotoxic effect were chosen for the subsequent tests.
(B) Figure 5. Decreased cell viability in SKLU-1 spheroids (3D). Cell viability is determined by membrane integrity through the exclusion dye Sytox Green. Spheroids were exposed to 500, 800, and 1000 μg/mL of CSE and 9. 94  In this work, the apoptosis induced by CSE was evaluated by measuring the translocation of Annexin V-FITC and analyzed by flow cytometry.
As shown in Figure 6A,B, the exposure of CSE to 800, 1000 μg/mL, and 9.94 μg/mL of cisplatin for 24 h presented a total of 97, 98 and 99% of SKLU-1 apoptotic cells, respec-  In this work, the apoptosis induced by CSE was evaluated by measuring the translocation of Annexin V-FITC and analyzed by flow cytometry.
As shown in Figure 6A,B, the exposure of CSE to 800, 1000 µg/mL, and 9.94 µg/mL of cisplatin for 24 h presented a total of 97, 98 and 99% of SKLU-1 apoptotic cells, respectively, compared to the control (<2.5%). Less than 1% of the population, including the control, showed necrotic cells.

CSE Causes Morphological Changes Characteristic of Apoptosis
Multiple biochemical and morphological changes characterize the apoptotic event. The latter is characterized by cell contraction, pyknosis, karyorrhexis, and apoptotic bodies [33]. During cell contraction, the cells become smaller, making the cytoplasm more compact. The chromatin is condensed in pyknosis, and the nucleus often has a slightly irregular outline. Karyorrhexis consists of the fragmentation of the nucleus and the separation of cell fragments into apoptotic bodies consisting of cytoplasm with compacted organelles with or without nuclear fragments [34].
As shown in Figure 7, Hoechst stain 33258 indicated that the cells had shrunken, con-

CSE Causes Morphological Changes Characteristic of Apoptosis
Multiple biochemical and morphological changes characterize the apoptotic event. The latter is characterized by cell contraction, pyknosis, karyorrhexis, and apoptotic bodies [33]. During cell contraction, the cells become smaller, making the cytoplasm more compact. The chromatin is condensed in pyknosis, and the nucleus often has a slightly irregular outline. Karyorrhexis consists of the fragmentation of the nucleus and the separation of cell fragments into apoptotic bodies consisting of cytoplasm with compacted organelles with or without nuclear fragments [34].
As shown in Figure 7, Hoechst stain 33258 indicated that the cells had shrunken, condensed, and fragmented nuclei after exposure to 800 and 1000 µg/mL and 9.94 µg/mL cisplatin for 24 h. The cell morphology of SKLU-1 was severely distorted, and the cells presented very irregular, contracted shapes and showed pyknosis. Unlike the treated cells, the untreated cells emitted a blue fluorescence with an intensity consistent with normal nuclei, a round shape with an intact and healthy membrane, and uniform chromatin. These results indicate that CSE generated morphological changes typical of apoptosis in SKLU-1 cells.
Molecules 2023, 28, x FOR PEER REVIEW 9 of 29 These results indicate that CSE generated morphological changes typical of apoptosis in SKLU-1 cells. The duplication of genetic material characterizes the cell cycle and results in two identical daughter cells, each with an exact copy of the genetic material. It consists of two phases: interface and mitosis. This event is negatively regulated by activated checkpoints when DNA damage (ATM/ATR) is detected and by deprivation of growth factors, cytokines, insufficient cell size, lack of molecules for the next phase, and telomere length, among others. The cell cycle checkpoints are critical because the cell can enter the next phase of the cell cycle only through checkpoint testing [35]. The main control points are at the end of the G1 phase (G1/S), in the S phase, at the end of the G2 phase, and at the end of the M phase (M/G1) [36]. Different families of proteins play an essential role in the control points by positively regulating the cell cycle. Of these, cyclins and cyclin-dependent kinases (CDKs) can be mentioned, with the complex cyclin D/CDK 4/6 and cyclin E/CDK2 being the main ones of the G1 phase, cyclin A/CDK2 the main one of S phase, and cyclin A/B/CDK1 the main one of the G2 phase.
On the other hand, there are cyclin-dependent kinase inhibitors (CDKIs), including the INK4 protein family (p14, p15, p16, p18, and p19) that specifically inhibit CDK4 and CDK6 and the CIP/KIP family (p21, p27, and p57) that regulates the cyclin E/CDK2, cyclin A/CDK2, and cyclin B/CDK1 complexes [37]. To determine whether CSE treatment resulted in impaired cell-cycle progression, the cell-cycle patterns of SKLU-1 cells were examined. Compared to the control group, cells treated with 800 and 1000 μg/mL of CSE accumulated in the cell cycle's S and G2/M phases (Table 4) ( Figure 8). The most significant percentage of cells was accumulated in the S phase for cisplatin.  The duplication of genetic material characterizes the cell cycle and results in two identical daughter cells, each with an exact copy of the genetic material. It consists of two phases: interface and mitosis. This event is negatively regulated by activated checkpoints when DNA damage (ATM/ATR) is detected and by deprivation of growth factors, cytokines, insufficient cell size, lack of molecules for the next phase, and telomere length, among others. The cell cycle checkpoints are critical because the cell can enter the next phase of the cell cycle only through checkpoint testing [35]. The main control points are at the end of the G1 phase (G1/S), in the S phase, at the end of the G2 phase, and at the end of the M phase (M/G1) [36]. Different families of proteins play an essential role in the control points by positively regulating the cell cycle. Of these, cyclins and cyclin-dependent kinases (CDKs) can be mentioned, with the complex cyclin D/CDK 4/6 and cyclin E/CDK2 being the main ones of the G1 phase, cyclin A/CDK2 the main one of S phase, and cyclin A/B/CDK1 the main one of the G2 phase.
On the other hand, there are cyclin-dependent kinase inhibitors (CDKIs), including the INK4 protein family (p14, p15, p16, p18, and p19) that specifically inhibit CDK4 and CDK6 and the CIP/KIP family (p21, p27, and p57) that regulates the cyclin E/CDK2, cyclin A/CDK2, and cyclin B/CDK1 complexes [37]. To determine whether CSE treatment resulted in impaired cell-cycle progression, the cell-cycle patterns of SKLU-1 cells were examined. Compared to the control group, cells treated with 800 and 1000 µg/mL of CSE accumulated in the cell cycle's S and G2/M phases (Table 4) ( Figure 8). The most significant percentage of cells was accumulated in the S phase for cisplatin.

CSE Decreased ATP Level in a 3D Culture Model
Adenosine triphosphate (ATP) is the primary energy source for cellular reactions [38] and a critical molecule that maintains and drives the life process by actively participating in metabolic processes. In carcinogenesis, there is a great demand for energy on the part of the cells to guarantee their growth [39].
There are different methods for assessing cell viability. However, the determination based on the quantification of ATP is used the most due to the metabolic dysregulation of tumor cells, one of the hallmarks of cancer [40]. This bioluminescence assay measured ATP, with the light detected directly proportional to the ATP content and indicating the presence of metabolically active cells.
In this work, concentrations of 800 and 1000 µg/mL of CSE showed a significant decrease in the ATP level after 24 h treatment exposure to the control. The decline in the ATP level indicates a reduction in cell viability. The 800 µg/mL concentration decreased the ATP level by 82%; meanwhile, 1000 µg/mL of CSE and 9.94 µg/mL of cisplatin decreased Adenosine triphosphate (ATP) is the primary energy source for cellular reactions [38] and a critical molecule that maintains and drives the life process by actively participating in metabolic processes. In carcinogenesis, there is a great demand for energy on the part of the cells to guarantee their growth [39].
There are different methods for assessing cell viability. However, the determination based on the quantification of ATP is used the most due to the metabolic dysregulation of tumor cells, one of the hallmarks of cancer [40]. This bioluminescence assay measured ATP, with the light detected directly proportional to the ATP content and indicating the presence of metabolically active cells.
In this work, concentrations of 800 and 1000 µg/mL of CSE showed a significant decrease in the ATP level after 24 h treatment exposure to the control. The decline in the ATP level indicates a reduction in cell viability. The 800 µg/mL concentration decreased the ATP level by 82%; meanwhile, 1000 µg/mL of CSE and 9.94 µg/mL of cisplatin decreased the ATP level by 87 and 93%, respectively ( Figure 9). The results in this work show that CSE is a potential therapeutic adjuvant to reduce ATP levels in addition to cell viability in lung cancer spheroids SKLU-1. Apoptosis is a regulated process of cell death executed by two alternative pathways: extrinsic and intrinsic. Inducing apoptosis is one of the essential goals of cancer therapy because cancer cells develop abilities to evade cancer mechanisms [41].
The mitochondria are the organelle that generates most of the energy needed for the cell's biochemical reactions, which is stored in ATP. Mitochondrial membrane potential (MMP) provides information on cellular health. It is essential for ATP production, so a decrease or loss of ATP results in lower ATP production and the release of apoptotic factors that lead to cell death [42].
There are different ways to monitor the state of the mitochondria. One of the most frequent is the use of fluorescent dyes that accumulate in healthy mitochondria. This work used rhodamine 123 (Rh123), a cationic dye attracted to the electronegative interior of the mitochondria of viable cells where the fluorescence intensity indicates the mitochondria condition [43]. CSE at 800 µg/mL and cisplatin at 9.94 µg/mL generated an MMP loss of 74%, while CSE at 1000 µg/mL induced a loss of 76% after 24 h of exposure to SKLU-1 spheroids. On the other hand, no change in MMP was observed in the control group. As shown in Figure 10, CSE generated changes in mitochondrial membrane permeability, which correlated with increased cell death and decreased ATP. Apoptosis is a regulated process of cell death executed by two alternative pathways: extrinsic and intrinsic. Inducing apoptosis is one of the essential goals of cancer therapy because cancer cells develop abilities to evade cancer mechanisms [41].
The mitochondria are the organelle that generates most of the energy needed for the cell's biochemical reactions, which is stored in ATP. Mitochondrial membrane potential (MMP) provides information on cellular health. It is essential for ATP production, so a decrease or loss of ATP results in lower ATP production and the release of apoptotic factors that lead to cell death [42].
There are different ways to monitor the state of the mitochondria. One of the most frequent is the use of fluorescent dyes that accumulate in healthy mitochondria. This work used rhodamine 123 (Rh123), a cationic dye attracted to the electronegative interior of the mitochondria of viable cells where the fluorescence intensity indicates the mitochondria condition [43]. CSE at 800 µg/mL and cisplatin at 9.94 µg/mL generated an MMP loss of 74%, while CSE at 1000 µg/mL induced a loss of 76% after 24 h of exposure to SKLU-1 spheroids. On the other hand, no change in MMP was observed in the control group. As shown in Figure 10, CSE generated changes in mitochondrial membrane permeability, which correlated with increased cell death and decreased ATP. olecules 2023, 28, x FOR PEER REVIEW Figure 10. Loss of mitochondrial membrane potential in SKLU-1 spheroids (3D mo termined by Rh123. Spheroids were exposed to 800 and 1000 µg/mL of CSE and 9.9 platin for 24 h. Mean values and standard deviation of relative units of fluorescence b hoc test (p ≤ 0.05) n = 12. Different letters represent significant differences between typ CSE Induced Apoptosis in a 3D Model by Caspases-3/7, -8, and -9 Activatio Apoptosis is programmed cell death that is executed by extrinsic and i ways. This event is mainly triggered by the regulatory activity of caspases, zymes of the cysteine-protease family capable of hydrolyzing tetrapeptides aspartic acid residue [44]. The activation of caspases results in a chain react to the activation of other caspases downstream and cell death [45]. Depen function, caspases are divided into initiating caspases: -2, -8, -9, and -10, w propagation signals of apoptosis, and effector caspases: -3, -6, and -7, whic process. Caspase-8 plays a vital role in the extrinsic pathway, while caspa essential role in the intrinsic pathway, where both pathways converge in ac tor caspases [46]. To investigate the pathway and mechanism by which CSE i tosis, the activity of caspases -8 and -9 was evaluated. Figure 11 shows an increase in the initiating caspases (caspases -8 extri -9 intrinsic pathway) and effectors caspases (-3/7) after being exposed to µg/mL of CSE and 9.94 µg/mL of cisplatin. The treatments induced apopto spheroids after 24 h exposure. CSE Induced Apoptosis in a 3D Model by Caspases-3/7, -8, and -9 Activation Apoptosis is programmed cell death that is executed by extrinsic and intrinsic pathways. This event is mainly triggered by the regulatory activity of caspases, which are enzymes of the cysteine-protease family capable of hydrolyzing tetrapeptides containing an aspartic acid residue [44]. The activation of caspases results in a chain reaction that leads to the activation of other caspases downstream and cell death [45]. Depending on their function, caspases are divided into initiating caspases: -2, -8, -9, and -10, which form the propagation signals of apoptosis, and effector caspases: -3, -6, and -7, which execute this process. Caspase-8 plays a vital role in the extrinsic pathway, while caspase-9 plays an essential role in the intrinsic pathway, where both pathways converge in activating effector caspases [46]. To investigate the pathway and mechanism by which CSE induces apoptosis, the activity of caspases -8 and -9 was evaluated. Figure 11 shows an increase in the initiating caspases (caspases -8 extrinsic pathway, -9 intrinsic pathway) and effectors caspases (-3/7) after being exposed to 800 and 1000 µg/mL of CSE and 9.94 µg/mL of cisplatin. The treatments induced apoptosis in SKLU-1 spheroids after 24 h exposure.
Molecules 2023, 28, x FOR PEER REVIEW 13 of Figure 11. Induction of apoptosis in SKLU-1 spheroids (3D model). Induced apoptosis was det mined by caspases-3/7, -8, and -9 activity. Spheroids were exposed to 800 and 1000 µg/mL of C and 9.94 µg/mL of cisplatin for 24 h. Mean values and standard deviation of relative units of flu rescence by Dunnett post hoc test (p ≤ 0.05) n = 12. Different letters represent significant differenc between types of treatment.

CSE Inhibited Invasion in a 3D Model of SKLU-1 Cells
The invasion process is carried out through three main stages: invasion, intravas tion, and extravasation. Invasion is caused by the reduction or loss of intercellular adh sion, allowing the dissociation of an individual or group of cells to the primary tum mass and by the changes generated in the cell-matrix interaction where the cells acqui abnormally high motility invading the surrounding stroma [47]. This process is characte ized by the secretion of substances that degrade the basement membrane (BM) and t extracellular matrix (ECM), showing characteristic markers of epithelial-to-mesenchym transition (EMT) [48]. Invasion and metastasis are considered hallmarks of cancer becau they represent the aggressive nature of cancer [40]. Therefore, inhibiting tumor invasio is critical in discovering drugs and/or therapeutic adjuvants. In this work, cell invasio was determined by the area and perimeter of the invaded matrigel.
As shown in Figure 12A, the invaded area decreased significantly after exposing t spheroids of SKLU-1 cells for 24 h at 800 and 1000 µg/mL; this decrease was even mo significant than cisplatin. In Figure 12B, the invaded perimeter shows that CSE concentr tions inhibited tumor invasion. The changes that originated in the spheroids due to t cytotoxicity caused by the treatments can be observed in Figure 12C. As shown in Figu  12C, the beginning of the angiogenesis process in the control group can be noticed, co firming the dissociation of the tumor cells from the spheroid, which crossed the BM an ECM gaining new territories. When observing the images of the treated cells, it was co cluded that CSE did not allow the detachment of cells from the tumor. These results ind cated that CSE could directly inhibit the invasive potential of SKLU-1 cells. Figure 11. Induction of apoptosis in SKLU-1 spheroids (3D model). Induced apoptosis was determined by caspases-3/7, -8, and -9 activity. Spheroids were exposed to 800 and 1000 µg/mL of CSE and 9.94 µg/mL of cisplatin for 24 h. Mean values and standard deviation of relative units of fluorescence by Dunnett post hoc test (p ≤ 0.05) n = 12. Different letters represent significant differences between types of treatment.

CSE Inhibited Invasion in a 3D Model of SKLU-1 Cells
The invasion process is carried out through three main stages: invasion, intravasation, and extravasation. Invasion is caused by the reduction or loss of intercellular adhesion, allowing the dissociation of an individual or group of cells to the primary tumor mass and by the changes generated in the cell-matrix interaction where the cells acquire abnormally high motility invading the surrounding stroma [47]. This process is characterized by the secretion of substances that degrade the basement membrane (BM) and the extracellular matrix (ECM), showing characteristic markers of epithelial-to-mesenchymal transition (EMT) [48]. Invasion and metastasis are considered hallmarks of cancer because they represent the aggressive nature of cancer [40]. Therefore, inhibiting tumor invasion is critical in discovering drugs and/or therapeutic adjuvants. In this work, cell invasion was determined by the area and perimeter of the invaded matrigel.
As shown in Figure 12A, the invaded area decreased significantly after exposing the spheroids of SKLU-1 cells for 24 h at 800 and 1000 µg/mL; this decrease was even more significant than cisplatin. In Figure 12B, the invaded perimeter shows that CSE concentrations inhibited tumor invasion. The changes that originated in the spheroids due to the cytotoxicity caused by the treatments can be observed in Figure 12C. As shown in Figure 12C, the beginning of the angiogenesis process in the control group can be noticed, confirming the dissociation of the tumor cells from the spheroid, which crossed the BM and ECM gaining new territories. When observing the images of the treated cells, it was concluded that CSE did not allow the detachment of cells from the tumor. These results indicated that CSE could directly inhibit the invasive potential of SKLU-1 cells.

Discussion
The total phenols content obtained from the algae C. sertularioides in this work was higher than those reported with a methanolic extract of Caulerpa racemosa and Caulerpa lentillifera (10.33 ± 0.02 and 4.52 ± 0.42 mg GAE/g dry extract, respectively) [49]. A similar phenol content was reported using a methanol extract from the algae C. racemosa (19.8 ± 2.01 mg GAE/g dry extract) [50]. Values such as 73 ± 2.08 mg GAE/g dry extract were reported using an ethanol extract from the algae C. lentillifera [51]. The flavonoids obtained in this study were between 2.85 and 14.22 times higher than that reported by [49] in the species C. racemosa and C. lentillifera (24.52 ± 2.17 and 4.93 ± 0.27 mg QE/g of dry extract, respectively) using a methanol extract. Previous studies have found significant variations in the chemical components of the same species and/or genus due to seasonal changes, geographical and environmental conditions, extraction method used, and solvent used, among others [52].

Discussion
The total phenols content obtained from the algae C. sertularioides in this work was higher than those reported with a methanolic extract of Caulerpa racemosa and Caulerpa lentillifera (10.33 ± 0.02 and 4.52 ± 0.42 mg GAE/g dry extract, respectively) [49]. A similar phenol content was reported using a methanol extract from the algae C. racemosa (19.8 ± 2.01 mg GAE/g dry extract) [50]. Values such as 73 ± 2.08 mg GAE/g dry extract were reported using an ethanol extract from the algae C. lentillifera [51]. The flavonoids obtained in this study were between 2.85 and 14.22 times higher than that reported by [49] in the species C. racemosa and C. lentillifera (24.52 ± 2.17 and 4.93 ± 0.27 mg QE/g of dry extract, respectively) using a methanol extract. Previous studies have found significant variations in the chemical components of the same species and/or genus due to seasonal changes, geographical and environmental conditions, extraction method used, and solvent used, among others [52].
Referring to carotenoids, Balasubramaniam et al. [53] reported a lower carotenoid content (195 ± 0.00 µg β carotene/g dry extract) compared to that obtained in this study (207.56 ± 2.67 µg Eq. β carotene/g dry extract). However, these researchers used the species C. lentillifera as raw material. These variations in the content of β carotene and other pigments in some macroalgae species vary considerably, as these are affected temporally and spatially [54].
Antioxidants are essential at nutritional and physiological levels because they prevent, protect, delay, and/or eliminate oxidative damage caused by reactive oxygen species (ROS) to cell membranes, mitochondria, DNA, lipids, or proteins during aerobic cell metabolism [55]. The antioxidant capacity evaluated by the ORAC assay in this work showed that C. sertularioides 80% ethanol extract (CSE) (2171.21 ± 1.35 µmol TE/g dry extract) has greater antioxidant capacity against peroxyl radicals than an ethanolic extract of C. racemosa (683.72 ± 15.86 µmol TE/g of dry extract) [53]. Likewise, two aqueous extracts of the algae Fucus vesiculosus had an antioxidant capacity of 1540 ± 220 and 1840 ± 3 µmol TE/g of the dry extract [56]. The antioxidant capacity of the ethanolic extract is significant because one of the mechanisms that reduce cell viability and invasion is fighting the oxidative stress generated in the cellular environment, which is one of the common imbalances of chronic diseases. Different studies have shown a close relationship between the content of phenolic compounds and the high antioxidant capacity of a substance or food, which is no exception for macroalgae because these compounds have been considered one of their most effective antioxidants [52]. Carotenoids have also been recognized for their anti-oxidative properties due to their ability to eliminate and deactivate free radicals [57]. Therefore, the high antioxidant capacity of C. sertularioides is attributed to the high content of phenols, flavonoids, and carotenoids.
CSE decreased the cell viability in a 2D culture model with an IC 50 of 80.28 µg/mL after 24 h treatment exposure. The HPLC-MS analysis revealed the presence of caulerpin (CLP), an alkaloid, in 80% of the algae species of the Caulerpa genus; however, it can be found in other algae genera [58]. Various biological activities have been attributed to it, including anticancer, antiviral, and antimalarial activities [50]. Table 5 shows the cytotoxic effect of caulerpin (CLP), the main compound that was identified in CSE by HPLC-MS but isolated and purified from other seaweed species on different cancer cell lines in a 2D culture model with relatively low IC 50 (from 7.97 to 47.4 µg/mL in a dose-dependent manner) [59][60][61][62][63]. These variations in the IC 50 might depend on the cancer cell line, the initial cell seeding density, and CLP's nature, source, and extraction method. According to He et al. [64], IC 50 errors due to differences in the proliferation rates and enzyme activity of cancer cells were not mentioned in any articles. The authors reported that only 27.6% (8/29) of the manuscripts said per-well seeding numbers (i.e., cell densities), and the other papers did not provide such information. Carnosic acid was detected, the main phenolic diterpene of the Rosmarinus officinalis L. plant (commonly known as rosemary). In a dose-dependent manner, carnosic acid inhibited the cell viability of three colorectal cancer cell lines from a two-dimensional model: Caco-2, HT29, and LoVo, with IC 50 values from 7.98 to 31.91 µg/mL [65]. Likewise, the presence of dihydroxyflavanone and pinocembrin is related to the decrease in cell viability, proliferation, and autophagy in lung carcinoma cells, A549, in a dose-dependent manner.
In this work, the IC 50 in the 2D and 3D culture models was 80.28 and 530 µg/mL, respectively, after 24 h of CSE exposure on SKLU-1 cells for both models. The IC 50 obtained in the 3D model is understandable because cells in 3D models (spheroids) are more resistant to pharmacological treatments. Table 6 shows different reports on lung carcinoma cells (H460, A549, and H1650), showing that the IC 50 of the same drug on the same cell line increases significantly in the 3D model compared to the 2D model [66]. Although the IC 50 of these drugs (Table 6) is very low compared to that of CSE, they cause significant side effects such as anemia (decrease in erythrocytes), leukopenia (decrease in leukocytes), neutropenia (reduction of neutrophils), thrombocytopenia (decrease in platelets), nausea, and vomiting, among others, causing the patient to be more vulnerable to other diseases and resulting in a lower quality of life. The IC 50 fluctuates between 10.39 and 66.51 times higher when comparing both models (Table 6) while, in this work, the difference is less than ten times.  In this work, the IC50 in the 2D and 3D culture models was 80.28 and 530 µg/mL, respectively, after 24 h of CSE exposure on SKLU-1 cells for both models. The IC50 obtained in the 3D model is understandable because cells in 3D models (spheroids) are more resistant to pharmacological treatments. Table 6 shows different reports on lung carcinoma cells (H460, A549, and H1650), showing that the IC50 of the same drug on the same cell line increases significantly in the 3D model compared to the 2D model [66]. Although the IC50 of these drugs (Table 6) is very low compared to that of CSE, they cause significant side effects such as anemia (decrease in erythrocytes), leukopenia (decrease in leukocytes), neutropenia (reduction of neutrophils), thrombocytopenia (decrease in platelets), nausea, and vomiting, among others, causing the patient to be more vulnerable to other diseases and resulting in a lower quality of life. The IC50 fluctuates between 10.39 and 66.51 times higher when comparing both models (Table 6) while, in this work, the difference is less than ten times. In an additional study, the same concentration of fucosterol and its combination with 5-fluorouracil, which causes cytotoxicity in a 2D colon carcinoma model (HCT116 and HT29), was ineffective in the three-dimensional model [67]. Likewise, Malhão et al. [68] did not detect cytotoxic and antiproliferative activity in the three-dimensional breast carcinoma model (MDA-MB-231) after exposing the cells to 2.06 µg/mL of fucosterol for 96 h. In the same way, a concentration of 6.59 µg/mL of fucoxanthin did not present a cytotoxic effect on the spheroids of MDA-MB-231 cells until its combination with 0.54 µg/mL of doxorubicin, decreasing cell viability by 22% through the MTT assay. Similarly, cytotoxicity was generated on the spheroids MDA-MB-231 through the LDH assay because of  In an additional study, the same concentration of fucosterol and its combination with 5-fluorouracil, which causes cytotoxicity in a 2D colon carcinoma model (HCT116 and HT29), was ineffective in the three-dimensional model [67]. Likewise, Malhão et al. [68] did not detect cytotoxic and antiproliferative activity in the three-dimensional breast carcinoma model (MDA-MB-231) after exposing the cells to 2.06 µg/mL of fucosterol for 96 h. In the same way, a concentration of 6.59 µg/mL of fucoxanthin did not present a cytotoxic effect on the spheroids of MDA-MB-231 cells until its combination with 0.54 µg/mL of doxorubicin, decreasing cell viability by 22% through the MTT assay. Similarly, cytotoxicity was generated on the spheroids MDA-MB-231 through the LDH assay because of combining 13.18 µg/mL of fucoxanthin with 1.09 and 2.72 µg/mL of doxorubicin, releasing LDH at 56 and 77 %, respectively [69].
The translocation of phosphatidylserine (PS) from the cytoplasm to the outer layer of the plasma membrane is one of the most relevant alterations that occur on the cell surface under the induction of apoptosis. Annexin V is a 35 kDa protein with a high affinity for PS, binding to a Ca+-dependent phospholipid bilayer containing PS [70]. A total of 43% of apoptotic hepatocarcinoma cells (HepG2) after 48 h exposure to an ethyl acetate fraction from an ethanol extract (320 µg/mL) of the algae Turbinaria conoides was observed [71]. A total of 20% of apoptotic glioblastoma cells (A172) were detected after being exposed to a 250 µg/mL hexane extract from the algae C. lentillifera [72]. Arumugam et al. [73] observed 20-40% of apoptotic hepatocarcinoma cells (HepG2) using fucoidan at 50-200 µg/mL concentrations. A 96% ethanol extract of C. racemosa showed the induction of apoptosis in treated cells. It decreased HeLa cell viability at 24 h and 48 h post-treatment with a range of 50-200 µg/mL [74].
Regarding the cell cycle, it is essential to remember that DNA replication occurs in the S phase. In the G2 phase, the cell continues the biosynthetic metabolic phase, verifies the fidelity of the replicated DNA, and prepares to enter mitosis. Therefore, the arrest of the cell cycle in phase S indicates damage to the genetic material. CSE may be inhibiting that replication by interfering in its organization, inhibiting the regulatory proteins of this phase, or altering the signaling pathways. When the cell cycle stops during cell division, damage, and error are challenging to repair [75]. In addition, a marked cell arrest generated by CSE at 1000 µg/mL in phase G2/M after damage in phase S was observed; this arrest suggests that the remaining cells depended on the G2/M checkpoint to counteract and/or prevent the consequences of DNA damage occurring in phase S. All of this confirms the conclusions of Kuczler et al. [76] that the arrest of the cell cycle allows more extensive repairs of the DNA or apoptosis in the case of damages that extend beyond the point of repair. ROS is involved in cell cycle arrest by hindering the repair of damaged genetic material by inhibiting repair pathways [77]. The HPLC-MS analysis revealed the presence of different bioactive compounds that interfere with the cell cycle, for example, damnacanthal, an anthraquinone compound (alkaloid of the quinone family) with anticancer properties [78]. Li et al. [79] said damnacanthal induced cell cycle arrest by increasing p27 protein Kip1 levels in ovarian carcinoma cells SKVO3 and A2780. Kim et al. [80] showed that baicalein stopped the S-phase cell cycle of human colorectal cancer cells HCT-116. Likewise, baicalein caused cell cycle arrest in G1/S by inhibiting the Akt/mTOR pathway in H1299 and H1650 lung cancer cells, which decreased the expression of the proteins CDK2, CDK4, and cyclin E2 [81]. Likely, cell cycle arrest in the S and G2/M phases caused by CSE is due to the compounds damnacanthal and baicalein. To understand the mechanism of action of CSE on the cell cycle, the activity of the regulatory proteins of the S phases (cyclin E/CDK2) and G2 (Cyclin B/CDK1) and CDKIs, as well as the generation of ROS and the MAPK/AKT/mTOR signaling pathways, should be examined due to the essential role they play in the regulation of the cell cycle.
In carcinogenesis, a tumor microenvironment rich in extracellular ATP is created, consequently increasing the interaction of tumor cells and immune cells [39]. Chauvin et al. [82] showed that after 24 h exposure of human colon cancer spheroids to a plasma-activated medium, the ATP level was decreased by 70%. The ATP level of colon cancer cell spheroids (HT-29) decreased by 62 and 80% after 24 and 48h treatment exposure, respectively, when treated with 100 µg/mL of doxorubicin. Meanwhile, the antimicrobial peptide gramicidin decreased the ATP level only by 20 and 50% at 113 µg/mL concentration in the same period reported [83]. Likewise, Martínez-Rodríguez et al. [84] indicated a significant decrease in cell viability by ATP quantification after one, four, and seven days of naringenin exposure to different concentrations in spheroids of cervical cancer cells (HeLa).
Depolarization of the mitochondrial membrane induced by CSE is characteristic of a decrease in mitochondrial membrane potential. It should be noted that no studies of MMP performed in a three-dimensional model have been reported. Therefore, the possible comparisons are with monolayer culture data. A 62% loss of MMP from MCF-7 cells has been observed after 24 h exposure to a fucoidan extract [85]. Likewise, Ryu et al. [86] reported an increase in the depolarization of the mitochondrial membrane of colon cancer cells HCT116 using an 80% ethanol extract of the algae Ulva fasciata. Sakthivel et al. [87] reported a 58% decrease in MMP from lung adenocarcinoma cells, A549, after 24 h of phytol exposure. Similarly, a reduction in MMP using 100 µg/mL of ethanolic, methanolic, and hexane extracts of the algae Enteromorpha compressa on squamous cell carcinoma of the pharynx (FaDu) and squamous carcinoma of the tongue (Cal33) was observed [88]. Similar results have been reported using a 250 µg/mL hexane extract from the algae C. lentillifera on glioblastoma A172 cells [72]. Cancer cells consume more energy to survive and continue proliferating than cells under normal conditions. It is important to remember that this alteration has been identified as a hallmark of cancer [40]. Mitochondria is the energy provider for cancer cells, so it is considered one of the critical organelles in cancer therapy [89]. It should be emphasized that mitochondria play an essential role in apoptosis because it contains different proapoptotic molecules, such as cytochrome c, that can trigger the intrinsic pathway that leads to programmed cell death and molecules such as SMAC/DIABLO (second mitochondrial activator of caspases/direct IAP binding protein with low PI) that favor apoptosis by inhibiting IAPs (inhibitory apoptosis proteins) [33]. The loss of mitochondrial membrane potential is associated with the generation of reactive oxygen species (ROS) because the oxidative damage caused by ROS is probably a major cause of mitochondrial genomic instability and respiratory dysfunction [90]. Srinivas et al. [77] stated that elevated ROS levels could trigger apoptosis by generating ROS. As a result of oxidative stress, the pores along the mitochondrial membrane can be oxidized, or the mitochondrial membrane can be depolarized, causing the release of proapoptotic compounds into the cytoplasm and thus initiating the apoptotic program [77,91]. Therefore, it is likely that CSE might increase oxidative stress beyond the limit, thus triggering mitochondria-mediated apoptosis. Consequently, it was concluded that the mitochondrial damage observed is a result of CSE-induced apoptosis accompanied by the release of mitochondrial proapoptotic molecules into the cytoplasm and that the generation of ROS in cells can be examined to strengthen the results and better understand the mechanism of action of CSE.
Studying the caspases family is essential to understanding the mechanism and pathway of triggering the apoptotic program. An upward regulation of caspases-3, -8, and -9 in a 2D model of MCF-7 cells was reported after 24, 48, and 72 h exposure to a methanolic extract of Sargassum muticum [92]. Similar results were obtained by Gomes et al. [93] after having exposed a 2D model of HeLa cells to 500 µg/mL of methanol extract from the algae Dictyota cilliolata and D. menstrualis. Pradhan et al. [89] observed a positive regulation of caspases-3/7 using 100 µg/mL of ethanolic, methanolic, and hexane extracts of Enteromorpha compressa on squamous cell carcinoma of the pharynx (FaDu) and squamous carcinoma of the tongue (Cal33) from a 2D model. Martínez-Rodríguez et al. [84], did not record caspase activity -3/7, -8 and -9 after 12, 24 and 72 h of exposure of 136.13 µg/mL of naringenin on 3D model of HeLa cells. A 96% methanolic extract of C. racemosa at 200 µg/mL significantly increased the expression of pro-apoptotic proteins Bax and cleaved caspase-3 in a 2D culture model compared to the control [74]. The HPLC-MS showed different bioactive compounds with pro-apoptotic activity, of which baicalein can be mentioned. This flavone has excellent potential to treat and prevent cancer without causing severe side effects [94]. This compound activated caspases-3 and -9 in colon carcinoma cells, HT-29, and apoptosis [95]. Likewise, Klimaszewska-Wiśniewska et al. [96] reported that quercetin, a flavonol, induced apoptosis in lung adenocarcinoma cells A549 through the negative and positive regulation of the anti-and proapoptotic proteins Bcl-2 and Bax, respectively. Pinocembrin-induced apoptosis of A549 cells is accompanied by an upward increase in caspase-3 activity [97]. The pro (Bax, Bok, Bak, Bik, Blk, Bad, Bid, Puma, and Noxa) and antiapoptotic (Bcl-2, Bcl-XL, Bcl-w, Boo, and Mcl-1) proteins of the Bcl-2 family are known to regulate the mitochondrial pathway of apoptosis. The effect of CSE on caspase-9 is more evident than on caspase-8; the intrinsic pathway would preferably carry out the apoptotic process. The induction of apoptosis through the intrinsic pathway was correlated with the results obtained from the loss of MMP because this is an event before the activation of caspase-9. This activation occurs when the proapoptotic protein Bax is translocated to the mitochondrial membrane, inducing its permeability (MOMP: mitochondrial outer membrane permeabilization). In contrast, the antiapoptotic proteins maintain control of mitochondrial permeability by blocking the activity of the proapoptotic proteins of that family. MOMP allows the release of different molecular compounds, such as cytochrome c, to the cytosol that, together with dATP and Apaf-1, bind to procaspase-9 forming the apoptosome complex and converting procaspase-9 into its active form (caspase-9) [44]. Therefore, it is likely that intrinsic pathway-mediated apoptosis is due to the balance generated by quercetin between the Bax and Bcl-2 proteins accompanied by the effect of baicalein and pinocembrin on caspases-3 and -9. Therefore, the activity of pro-and anti-apoptotic proteins, including some essential protein compounds on the inner side of the mitochondrial membrane, can be examined to better understand CSE's mechanism of action.
The greatest terror in carcinogenesis is the sequence of events that lead to metastasis. Metastasis is the spread of cancer cells from a primary tumor to secondary and tertiary tissues and/or organs. In addition, it is the main event that causes the death of most cancer patients [98]. Lee et al. [99] reported that 200 µg/mL of fucoidan extracted from the algae F. vesiculosus decreased the invasion of lung cancer cells A549 by 86% compared to the control in a 2D model after 48h of exposure. Similarly, a decrease in the invasion of colon carcinoma cells HT-29 in a 2D model by inhibiting MMP-2 after 48 h of exposure of these to 200 µg/mL of fucoidan was reported [100]. Martínez-Rodríguez et al. [84] reported that naringenin at 136.12 µg/mL significantly decreased the invasion of HeLa cancer cells in a three-dimensional model. It is important to remember that cancer cells express matrix metalloproteinases (MMPs), for example, MMP-2/-9, which degrade the ECM-generating pathways that allow migratory cells to invade freely [101]. These secrete substances that modify the expression of proteins that control motility and migration so that the tumor initiates the process of angiogenesis, without which it would not develop [47]. The HPLC-MS analysis revealed the presence of different bioactive compounds capable of inhibiting the invasive capacity of tumor cells. Li et al. [79] reported that damnacanthal inhibited migration and invasion in SKVO3 and A2780 ovarian carcinoma cells. According to Gao et al. [102], the proliferation and migration of ovarian cancer cells can be suppressed by pinocembrin through decreased expression of N-cadherin and the gamma-aminobutyric acid receptor, also confirming that treatment with pinocembrin led to a decrease in the proliferation, migration, and invasiveness of colorectal cancer cells. Barni et al. [65] reported that carnosic acid inhibited cell adhesion and migration of the colon cancer line Caco-2, possibly reducing the activity of secreted proteases such as urokinase plasminogen activators (uPAs) and metalloproteinases (MMPs). In addition, Klimaszewska-Wiśniewska et al. [96] confirmed that quercetin repressed the migration of A549 cells, proposing and explaining that the influence of quercetin disassembly on vimentin filaments, microtubules, and microfilaments accompanied by its suppressive effect on N-cadherin and vimentin expression could be responsible for reduced migration of A549 cells in response to quercetin therapy. Therefore, the decreased invasion in SKLU-1 cells might be due to these compounds' activity on ECMdegrading proteins. Consequently, it is suggested to analyze the expression of essential proteins (MMP-2/9, N-cadherin, among others) that actively participate in the invasive and metastatic process to understand better how CSE directly inhibited the invasive potential of SKLU-1 cells.

Collection of Macroalgae
The specimen was collected at Carreyeros Beach, Bahía de Banderas, Nayarit, Mexico, in February 2022. The collected biological material was frozen in plastic bags in the Biopolymers laboratory of the Department of Biochemical Engineering of the National School of Biological Sciences-IPN Zacatenco Unit. The macroalgae were washed with distilled water to remove all epiphytes and impurities and stored in an ultra-freezer at −74 • C for further analysis. The observation of taxonomic characteristics and a histological examination of the specimen based on dichotomous keys in the literature were conducted to identify the species. The specimen's internal and external morphology was analyzed using a microscope [103] for this identification.

Macroalgae Extract
Once the species was identified, Caulerpa sertularioides was freeze-dried for 72h at −50 • C and 0.014 mbar and sprayed to obtain a fine powder of average size less than 1/2 mm. Then, the extract was obtained using an ultrasonic bath for 1 h with cold stirring using 80% ethanol [104]. Finally, the extract was dried in centrifugal concentrators (Genevac TM miVac Duo Concentrator) for further trials.

Quantification of Total Phenolic Compounds
The content of total phenolic compounds was determined by the Folin-Ciocalteu method with some modifications. The researchers added 90 µL of 10% Folin-Ciocalteu reagent to 20 µL of C. sertularioides ethanol extract. After five min., 90 µL of Na 2 CO 3 solution (60 g/L) was added to the mixture. Subsequently, the preparation was incubated for 90 min in a microplate reader (SYNERGY H1, BioTek, Winooski, VT, USA), and the absorbance was measured at 750 nm [105]. The content of phenolic compounds was expressed as mg of gallic acid (GAE) equivalent/g of sample.

Quantification of Total Flavonoids
According to Fattahi et al. [106], the colorimetric method evaluated flavonoid content with some modifications. Briefly, 0.5 mL of C. sertularioides ethanol extract was mixed with 2 mL of distilled H 2 O and 150 µL of 5% NaNO 2 . After five min., 150 µL of 10% AlCl 3 was added, and finally, 2 mL of NaOH at 0.5 M after three min. The mixture was then incubated for 30 min. The absorbance was read at 510 nm on a spectrophotometer (GENESYS 10S, Thermo Fisher Scientific, Waltham, MA, USA). The total flavonoid content was expressed in mg of quercetin (QE) equivalent/g of the sample.

Quantification of Total Carotenoids
The total carotenoid content was evaluated by spectrophotometry according to some modifications by Osuna-Ruiz et al. [105]. Briefly, 0.01 g of powdered sample was mixed with 5 mL of acetone, leaving it to stand for 10 min. Then, 5 mL of petroleum ether was added to the mixture. The solution was washed with distilled water until the acetone was removed, and the wastewater was removed with anhydrous sulfate. Subsequently, absorbance was read at 450 nm using a spectrophotometer (GENESYS 10S, Thermo Fisher Scientific, Waltham, MA, USA). The total carotenoid content was expressed as µg/g of β carotene X (µg/g) = A × y (mL) × 10 6 /A 1%1 cm × 100 x (µg/g) = x (µg)/sample weight (g) × FD where: A: Absorbance y: Volume of the solution that gave the absorbance A 1%1 cm : Carotenoid absorption coefficient FD: Dilution Factor

Determination of Antioxidant Capacity by ORAC
With some modifications, oxygen radical absorption capacity (ORAC) was determined according to Quek et al. [107]. Briefly, 20 µL of a 10 mM fluorescein solution and 50 µL of 2,2-azobis (2-amidino-propane) dihydrochloride (AAPH) 12 mM were added to 20 µL of the C. sertularioides sample or 6-hydroxy-2,5,7,8-tetra-methylchroman-2-carboxylic acid (Trolox). The readings were performed by recording the loss of fluorescence every two minutes for two hours at an excitation length of 485 and 515 nm of emission using a microplate reader (SYNERGY H1, BioTek, Winooski, VT, USA). All samples were prepared in PBS at pH 7.4 in a 1:2000 ratio.
The ORAC value was calculated using the following equation, and the results were expressed as TEAC values (µmol of Trolox/g dry base sample). It was carried out in agreement with Chia et al. [50] with some modifications. The CSE extract was subjected to HPLC-MS after testing its antitumor activity. The samples (500 µL) were injected and analyzed with the LCMS Triple Quadrupole system, Agilent brand model G6410 with dual ESI source. The solution was filtered with a 0.45 µm Nylon syringe filter. The dimension of the column used was 4.6 × 150 mm. The binary mobile phase consisted of solvents A (water with 0.1% formic acid) and B (100% acetonitrile). The flow rate was 10 L/min. The data were analyzed with Agilent MasHunter Qualitative Analysis B.05.00 software (https://www.agilent.com/en-us/support/software-informatics/masshunterworkstation-software/, accessed on 20 April 2023). The compounds were identified by searching METLIN: Metabolite and Tandem MS Database. The HPLC-MS parameters were as follows: ionization chamber temperature, 100 • C; gas temperature, 300 • C; capillary voltage, 4 KV; fragments voltage, 95 and 135 V; and gas flow, 10 L/min. The electrospray ionization (ESI) source was established in positive and negative modes to acquire all mass spectrometric data.

Study Design
The experiments were carried out according to the study design divided into four phases, depicted in Figure 13.
In Phase 1, the cytotoxic effects of six CSE concentrations and one concentration of Cis were examined by the membrane permeability assay in the 2D culture model of lung adenocarcinoma cells SKLU-1. Considering the results obtained in Phase 1, (500, 800, and 1000 µg/mL, IC 50 x4, IC 50 x6, and IC 50 x12, respectively) were selected to be tested in a 3D culture model. According to the results obtained in Phase 2, the two best concentrations of CSE were chosen to conduct annexin V, apoptosis morphology, and cell-cycle studies in a 2D model. In Phase 4, these two concentrations were tested in 3D cultures by analyzing ATP level, mitochondrial membrane permeability, caspase activity, and cell invasion.
In all the experiments, control cells (negative control) were incubated in a culture medium with 0.1% DMSO.

Cell Viability in a 2D and 3D Culture Model by Plasma Membrane Integrity
This trial was carried out according to Martínez-Rodríguez et al. [84] with some modifications. For the 2D model, 5.0 × 10 3 cells were seeded in 96 wells of adhesion plates (Corning ® ). SKLU-1 cells were exposed to 25, 50, 75, 100, 150, 200 µg/mL CSE and 9.94 µg/mL of Cis for 24 h. After the treatment, the dye Sytox Green ® was added, and the fluorescence was read in a microplate reader (SYNERGY H1, BioTek, Winooski, VT, USA). The 3D model was carried out the same way as the 2D model with some modifications. The researchers seeded 1.2 × 10 3 cells in ultra-low adhesion plates from 96 wells (Corning ® ). After the spheroid's formation, the SKLU-1 cells were exposed to 500, 800, and 1000 µg/mL of CSE and 9.94 µg/mL of cisplatin for 24 h. Molecules 2023, 28, x FOR PEER REVIEW 22 of 29 Figure 13. Schematic representation of the study design.
In Phase 1, the cytotoxic effects of six CSE concentrations and one concentration of Cis were examined by the membrane permeability assay in the 2D culture model of lung adenocarcinoma cells SKLU-1. Considering the results obtained in Phase 1, (500, 800, and 1000 µg/mL, IC50x4, IC50x6, and IC50x12, respectively) were selected to be tested in a 3D culture model. According to the results obtained in Phase 2, the two best concentrations of CSE were chosen to conduct annexin V, apoptosis morphology, and cell-cycle studies in a 2D model. In Phase 4, these two concentrations were tested in 3D cultures by analyzing ATP level, mitochondrial membrane permeability, caspase activity, and cell invasion.
In all the experiments, control cells (negative control) were incubated in a culture medium with 0.1% DMSO.

Cell Viability in a 2D and 3D Culture Model by Plasma Membrane Integrity
This trial was carried out according to Martínez-Rodríguez et al. [84] with some modifications. For the 2D model, 5.0 × 10 3 cells were seeded in 96 wells of adhesion plates (Corning ® ). SKLU-1 cells were exposed to 25, 50, 75, 100, 150, 200 µg/mL CSE and 9.94 µg/mL of Cis for 24 h. After the treatment, the dye Sytox Green ® was added, and the fluorescence was read in a microplate reader (SYNERGY H1, BioTek, Winooski, VT, USA). The 3D model was carried out the same way as the 2D model with some modifications. The researchers seeded 1.2 × 10 3 cells in ultra-low adhesion plates from 96 wells (Corning ® ). After the spheroid's formation, the SKLU-1 cells were exposed to 500, 800, and 1000 µg/mL of CSE and 9.94 µg/mL of cisplatin for 24 h.

Annexin V Test in a 2D Culture by Flow Cytometry
The annexin V test was performed with the Annexin V-FITC Apoptosis kit (Calbiochem ® ) following the manufacturer's specifications. Briefly, 800 × 10 3 cells per mL were

Annexin V Test in a 2D Culture by Flow Cytometry
The annexin V test was performed with the Annexin V-FITC Apoptosis kit (Calbiochem ® ) following the manufacturer's specifications. Briefly, 800 × 10 3 cells per mL were seeded in 3.5 cm plates from 6 wells (Nest ® ) exposed to 800 and 1000 µg/mL of CSE and 9.94 µg/mL of cisplatin for 24 h. The cells were stained with annexin V-FITC and propidium iodide to be later analyzed in BD LSRFortessa cytometer (BD Biosciences).

Nuclear Staining with Hoechst 33258 for the Study of Cell Morphology
This trial was performed as described by Wang et al. [95] with some modifications. Briefly, 5 × 10 3 cells were seeded in 96-well (Corning ® ) plates exposed to 800, 1000 µg/mL of CSE, and 9.94 µg/mL of cisplatin for 24 h. Subsequently, the harvested cells were washed with PBS and fixed with 4% paraformaldehyde for 20 min. The cells were stained with Hoechst 33258 for 10 min, then observed under an inverted microscope Zeiss Axiovert 25 (Carl Zeiss) at a magnification of 20× and analyzed in the ImageJ software (http://imagej. nih.gov/ij, accessed on 20 April 2023).

Cell Cycle Test in a 2D Model by Flow Cytometry
This trial was performed according to Gomes et al. [93] with some modifications. The SKLU-1 cells were seeded in a 3.5 cm plate of 6 wells (Nest ® ) (200 × 10 3 /mL). After 24 h of exposure to the treatments, these were harvested and fixed in 4% paraformaldehyde for 60 min. Subsequently, Rnase A and propidium iodide (30 µg/mL) was added. The DNA content was analyzed using a BD LSRFortessaTM (BD Biosciences) flow cytometer. A total of 20,000 events were purchased. For data analysis, FlowlogicTM Analysis Software, version 8.6, was used. The data presented are representative of those obtained in three independent experiments conducted in duplicate.

ATP Quantification in a 3D Culture Model
The cell viability assay by ATP quantification was performed with the CellTiter-Glo ® kit, Reagent 2.0 (Promega Corp., Madison, WI, USA), following the manufacturer's specifications. Briefly, 300 cells were seeded per well. After 24 h of exposure to the spheroids with 800 and 1000 µg/mL of CSE and 9.94 µg/mL of cisplatin, the CellTiter-Glo ® reagent was added and incubated for 30 min. to then quantify the luminescence in a microplate reader (SYNERGY H1, BioTek). The detection is based on using the luciferin-luciferase reaction to measure the amount of ATP of viable cells where the amount of light detected is directly proportional to the ATP content and indicates the presence of metabolically active (viable) cells.

Mitochondrial Membrane Potential (∆Ψm) Assay in a 3D Model
This trial was conducted in agreement with Wang et al. [95] with some modifications. The researchers seeded 1.2 × 10 3 cells in ultra-low adhesion plates from 96 wells (Corning ® ). After forming the spheroid, it was exposed to 800 and 1000 µg/mL of CSE and 9.94 µg/mL of cisplatin for 24 h. Subsequently, rhodamine 123 was added, and fluorescence was read in a microplate reader (SYNERGY H1, BioTek).

Caspases 3/7, -8, and -9 Test in a 3D Model
Caspase activity was determined with the Caspase-Glo ® , -3/7, -8, and -9 kit (Promega Corp., Madison, WI, USA) following the manufacturer's specifications. Briefly, 300 cells were seeded per well, and after 24 h of exposure to the spheroids with 800 and 1000 µg/mL of CSE and 9.94 µg/mL of cisplatin, the corresponding Caspase-Glo ® reagent was added. Subsequently, it was incubated for an hour and quantified the luminescence in a microplate reader (SYNERGY H1, BioTek). The luminescence detected is proportional to the amount of caspase activity present.

Invasion Test in a 3D Culture Model
The invasion test was performed as described by Martínez-Rodríguez et al. [84] with some modifications. Briefly, 1.2 × 10 3 cells were seeded per well, and after the formation of the spheroid, it was treated with 800 and 1000 µg/mL of CSE and 9.94 µg/mL of cisplatin for 24 h. Subsequently, 100 µL of Matrigel ® Matrix (Corning ® ) per well and then RPMI medium was added. The invasion was determined using an inverted Zeiss Axiovert 25 microscope (Carl Zeiss) and ImageJ software and calculated by measuring the area and perimeter between the spheroids' edge and the invading cells' edge.

Conclusions
The results found in this work show that C. sertularioides extract is a potential therapeutic adjuvant capable of significantly reducing cell viability in SKLU-1 lung cancer spheroids. In addition, it generated mitochondrial damage and induced apoptosis through both signaling pathways. CSE also caused biochemical and morphological changes typical of apoptosis and directly inhibited the invasive potential of SKLU-1 tumor cells. It stopped the cell cycle in the S and G2/M phases in response to genetic damage by preventing damaged DNA from replicating further and thus inhibiting the proliferative capacity of cells. As observed, CSE has antioxidant, antiproliferative, proapoptotic, and anti-invasive abilities, so it becomes a potential candidate for an alternative treatment with bioactive compounds that can act as adjuvants, decreasing the dose of chemotherapeutic drugs and, at the same time, the aggressivity of them. Additional studies are required to strengthen and better understand CSE's mechanism of action on lung cancer's carcinogenesis.

Recommendations for Future Work
The current study suggests using the CSE's primary component, caulerpin, to determine its effect and mechanism of action. A multi-approach with molecular and histological analysis and combination with first-line drugs must be included.