Assessing the Antitumor Potential of Variants of the Extracellular Carbohydrate Polymer from Synechocystis ΔsigF Mutant

Cancer is a leading cause of death worldwide with a huge societal and economic impact. Clinically effective and less expensive anticancer agents derived from natural sources can help to overcome limitations and negative side effects of chemotherapy and radiotherapy. Previously, we showed that the extracellular carbohydrate polymer of a Synechocystis ΔsigF overproducing mutant displayed a strong antitumor activity towards several human tumor cell lines, by inducing high levels of apoptosis through p53 and caspase-3 activation. Here, the ΔsigF polymer was manipulated to obtain variants that were tested in a human melanoma (Mewo) cell line. Our results demonstrated that high molecular mass fractions were important for the polymer bioactivity, and that the reduction of the peptide content generated a variant with enhanced in vitro antitumor activity. This variant, and the original ΔsigF polymer, were further tested in vivo using the chick chorioallantoic membrane (CAM) assay. Both polymers significantly decreased xenografted CAM tumor growth and affected tumor morphology, by promoting less compact tumors, validating their antitumor potential in vivo. This work contributes with strategies for the design and testing tailored cyanobacterial extracellular polymers and further strengths the relevance of evaluating this type of polymers for biotechnological/biomedical applications.


Introduction
Cancer is a leading cause of death worldwide, with an estimation of 19.3 million new cases and almost 10 million deaths in 2020, thus having a huge societal impact and economical relevance. Based on the GLOBOCAN project in 2020, these numbers are expected to grow to 28.4 million new cancer cases in 2040, a 47% rise from 2020 [1]. These projections increase the demand of early cancer detection and the development of advanced and innovative therapies. Due to the limitations and negative side effects of chemotherapy and radiotherapy, researchers are focused on finding alternative therapeutics that are clinically effective and less expensive, namely anticancer agents from natural sources [2][3][4]. In this context, biopolymers of polysaccharidic nature embody promising therapeutic alternatives, by preventing tumor development and inhibiting proliferation, invasion, adhesion, metastization, and/or angiogenesis [3,5]. These biopolymers can also act in combination with conventional anticancer drugs, improving their activity through the enhancement of tumor sensitivity and patient immune response [3,6]. The molecular mechanisms underlying the biopolymers' antitumor effects include cell cycle arrest, activation of the mitochondrial-mediated apoptotic pathway, production and activation of nitric oxide pathway or immunomodulatory pathways [7]. While some biopolymers may act through defined pathways, for others the mechanism of action is still not clear, but may include production of reactive oxygen species, as well as inhibition of proteins involved in angiogenesis and metastization, among other processes [3]. In recent years, cyanobacteria have emerged as a promising source of bioactive polysaccharides, since most strains produce extracellular polymeric substances (EPS), mainly composed by heteropolysaccharides, with a distinctive set of desirable characteristics, including a (i) large number of different monosaccharides, allowing a wide range of structural rearrangements, (ii) strong anionic character due to the presence of uronic acids and sulfate groups and (iii) presence of deoxyhexoses and peptides that contribute to hydrophobicity and, thus, amphiphilic behavior [8,9]. These peculiar features prompted significant advances on their characterization [10][11][12] and validation of their potential for biomedical applications [13][14][15][16][17][18][19]. However, studies on the antitumor potential of cyanobacterial EPS are still scarce and mainly based on tumor cell lines [20][21][22][23][24][25], with few in vivo studies using tumor xenograft mice models [26].
Previously, we showed that a sigma factor mutant (∆sigF) of the model unicellular cyanobacterium Synechocystis sp. PCC 6803 had the secretion mechanisms significantly altered, being able to release up to four-fold more polysaccharides (RPS -released polysaccharides) compared to the wild-type [27,28]. These carbohydrate polymers displayed a strong antitumor activity towards human tumor cell lines, namely melanoma (Mewo), thyroid carcinoma (8505C) and ovarian carcinoma (A2780), being the ∆sigF polymer easier to isolate and recover from the culture medium and more potent than the one from the wild-type [20]. We also showed that the antitumor effect of the ∆sigF polymer was associated with the induction of apoptotic pathways via upregulation of p53 protein levels and caspase-3 activation [20]. Importantly, the ∆sigF polymer induced one of the highest levels of apoptosis (~40%) observed among natural polymers [20,23,29,30]. Despite these results, it is still unclear which are the polymer features/constituents that mostly contribute to its antitumor activity.
The major aim of this study was to manipulate the Synechocystis' ∆sigF polymer in order to obtain variants with increased antitumor activity (in vitro and/or in vivo), gaining insight into key features that contribute to the polymer bioactivity.
The ∆sigF mutant was maintained in BG11 medium supplemented with kanamycin (100 µg/mL), while the experiments were performed in the absence of selective pressure.

Polymer Isolation
The polymer was isolated according to [32,33]. Briefly, the Synechocystis ∆sigF cultures were dialyzed (12-14 kDa of molecular weight cut-off; Medicell International, London, UK) against a minimum of 10 volumes of deionized water for 48 h with continuous stirring. Then, the cultures were centrifuged at 20,000× g for 25 min at 8 • C, and 2 volumes of 96% ethanol were added to the supernatant. After an incubation at 4 • C overnight, the suspension was centrifuged at 20,000× g for 25 min at 6 • C, and the supernatant was discarded. The pellet was resuspended in 1 mL of autoclaved type II water and lyophilized. The dried polymer was stored at room temperature (RT) until further use.

Polymer Hydrolysis and Molecular Mass Analysis
Hydrolysis of the polymers was performed with either 2 M trifluoroacetic acid (TFA) (VWR Chemicals, Radnor, PA, USA) at 120 • C or in 2 M hydrochloric acid (HCl) (Merk, Darmstadt, Germany) at 100 • C for 10 min, 25 min, 45 min, 1 h, or 2 h. After hydrolysis, the samples were evaporated two times, and the polymers resuspended in ultrapure water before being analyzed by size exclusion chromatography (SEC).
The apparent molecular masses (hereafter referred to as MM) were determined according to a previously reported method [34], with some changes. Briefly, samples were dissolved in high-performance liquid chromatography (HPLC) grade water at a concentration of 5 mg/mL and analyzed using a Varian ProStar HPLC chromatograph (Varian, Palo Alto, CA, USA) equipped with a 355 RI (refractive index) detector and two columns for SEC, Polysep-GFC-P 6000 and 4000 (Phenomenex, Torrance, CA, USA) connected in series. The analyses were performed with runs of 70 min and with HPLC grade water as eluent at a flow rate of 0.4 mL/min, using Dextran (Sigma-Aldrich, Burlington, MA, USA) at different MM (2000 kDa, 1100 kDa, 410 kDa, 150 kDa, and 50 kDa) and saccharose (0.34 kDa) as standards.

Quantification of Sulfate and Peptide Contents
For sulfate quantification, lyophilized RPS were hydrolyzed in 2 M HCl at 100 • C for 10 min to 2 h, centrifuged after cooling, and the supernatant was analyzed by ionexchange chromatography [35,36]. The analysis was performed using a Dionex ICS-2500 system chromatograph equipped with a continuously regenerated anion-trap column, a continuous anionic self-regenerating suppressor, a conductivity detector (ED50), an IonPac PA11 4 × 250 mm column (Dionex, Sunnyvale, CA, USA), and a reagent-free Dionex system producing high-purity 50 mM KOH at a flow rate of 2 mL/min. Sulfate solutions (1 to 10 mg/L, Fluka, Switzerland) were used as standards.
To evaluate the peptide content, the lyophilized polymer was resuspended in deionized water and quantification was performed by the Lowry method [37].

Peptide Removal and Analysis
Peptide removal was performed by trichloroacetic acid (TCA) (Fisher Bioreagents, Bremen, Germany) precipitation [35,36]. Briefly, an aqueous solution polymer (5 mg/mL) was incubated with 15% TCA for 15 or 30 min at room temperature. After incubation, the mixture was centrifuged at 11,000× g for 20 min at 4 • C. The supernatant was dialyzed against a minimum of 10 volumes of deionized water for 48 h with continuous stirring and lyophilized. The precipitation efficiency was assessed measuring the peptide content in the polymer suspension before and after TCA treatment using the Lowry method [37].

Cell Viability Assays
Mewo cells were plated on 96-well plates at 1 × 10 4 cells/well and allowed to adhere for 24 h at 37 • C. Cells were then treated for 24 and 48 h with: (i) supplemented media (Blank), (ii) ∆sigF polymer variants (resuspended in non-supplemented media) at 0.7, 1, and 1.5 mg/mL, or (iii) with polymer vehicle as control (consisting of supplemented media containing the equivalent amount of non-supplemented medium used in the polymer treatments). To assess cell viability after treatment, a PrestoBlue TM cell viability assay was carried out as previously described [39]. Briefly, cells were washed three times with the respective non-supplemented medium and further incubated for 45 min with 10% PrestoBlue TM reagent (Life Technologies, Carlsbad, OR, USA) in supplemented medium. Fluorescence was measured (excitation 560 nm; emission 590 nm) on a Synergy HT Multi-Mode Microplate Reader (BioTek Instruments Inc., Santa Clara, CA, USA). All samples and controls were analyzed with four biological and five technical replicates. Cellular viability was determined by analyzing the fluorescence values of each sample as a percentage in relation to control samples (cells treated with medium only), after removing the background values.

CAM Assay
The effect of polymers on tumor growth and angiogenic activity was evaluated in vivo using the chick chorioallantoic membrane (CAM) assay [40][41][42]. Fertilized chick eggs (Gallus gallus) were incubated horizontally at 37 • C in a humidified atmosphere and referred to the embryonic development day 0 (EDD0). On EDD3, eggs were prepared (window opening) to allow detachment of the developing CAM from the shell. On EDD9, CAMs were inoculated with 1 × 10 6 exponentially growing Mewo cells resuspended in Matrigel (1:1) together with (i) 0.7 mg/ml of ∆sigF polymer, (ii) its variant obtained after 15 min of TCA treatment (∆sigF.pep-), or (iii) with their vehicle solvent (DMEM), into a 5-mm silicone ring under sterile conditions. After resealing, eggs were returned to the incubator. At EDD13, CAMs were fixed in 10% neutral-buffered formalin, the ring was removed, and CAMs were excised from the embryos and photographed ex ovo under a stereoscope at 20× magnification (Olympus, Tokyo, Japan, SZX16 coupled with a DP71 camera). The number of new vessels growing towards the inoculation site, delimited by the ring mark were counted (less than 20 µm in diameter) and the area of the tumors (dense areas) were determined using the "Cell A Olympus" program, as described previously [40][41][42].

Tumor Histology and Immunohistochemical Analysis
Excised CAMs (fixed in 10% neutral-buffered formalin) were processed for paraffinembedding. Serial 4-µm-thick sections were cut for hematoxylin and eosin (H&E) stain and for immunohistochemistry evaluation of Ki-67 and cleaved caspase-3. Briefly, for immunohistochemistry, after deparaffinization and rehydration, samples were subjected to heat-induced antigen retrieval in citrate buffer, pH 6 (Richard-Allan Scientific, Kalamazoo, MI, USA) for Ki-67 or in Epitope Retrieval solution pH 9.0 (Novocastra; Leica Biosystems, Wetzlar, Germany), for cleaved caspase 3 evaluation, for 45 min at 90 • C in a steamer. Quanto Detection System HRP; Thermo Scientific, CA, USA). Detection was performed using 3,3 Diaminobenzidine (DAB Quanto Chromogen and Substrate; Thermo Scientific, CA, USA) and counterstaining with Mayer's Haematoxylin (Thermo Scientific, CA, USA). Tonsil tissue sections were used as a positive control for staining. Negative controls (in which the primary antibody was omitted) were also included. Slides were evaluated using the light microscope Olympus DP 25 Camera coupled to the Cell B Olympus software (Tokyo, Japan). The percentage of Ki-67 positive stained cells was determined by evaluating at least 2900 tumor cells per slide.

Statistical Analysis
Data were plotted and statistically analyzed using GraphPad Prism version 9.0 (Graph-Pad Software, San Diego, CA, USA) using an analysis of variance (ANOVA), followed by Bonferroni's multiple comparisons test. Regarding the CAM assays, statistical analysis was performed using one-way ANOVA test with IBM SPSS statistics.

Results and Discussion
We previously demonstrated that the extracellular carbohydrate polymers released by the unicellular cyanobacterium Synechocystis sp. PCC 6803 and its ∆sigF mutant exhibited a strong antitumor activity against several human tumor cell lines, with the latter being more effective [20]. The present study goes a step forward by manipulating several features of the ∆sigF polymer to obtain variants, which were tested in vitro in the human melanoma (Mewo) cell line. The most promising variant was then evaluated in vivo, using the chick embryo chorioallantoic membrane (CAM) model.

Polymer Hydrolysis and Evaluation of Antitumor Activity In Vitro
Since previous studies reported a correlation between the molecular mass (MM) of algal polysaccharides and their antitumor activity [43][44][45], modification of the MM distribution of Synechocystis ∆sigF polymer was firstly achieved through trifluoroacetic acid (TFA) hydrolysis ( Figure 1). A gradual increase of the abundance of lower MM fractions was observed by increasing the hydrolysis time. This stepwise effect was expected since TFA disrupts glycosidic bonds selectively, at different rates depending on the linked monosaccharide [46]. For example, the disruption of bonds from amino sugars takes a longer time, while pentoses are among the first monosaccharides to be released and degraded under longer reaction times (>2 h).
Subsequently, the effect of two polymer variants (obtained after 25 min and 2 h of TFA hydrolysis) on the viability of Mewo cells was evaluated. Overall, our results showed that both variants had less antitumor activity than the original ∆sigF polymer, and that after 48 h of treatment there was a significant decrease in bioactivity for the three polymer concentrations tested ( Figure 2) [33]. Moreover, the variant with lower abundance of high MM fractions (2 h hydrolysis) had the lowest effect on Mewo cells viability, indicating that the high MM fractions play an important role in its antitumor activity.
One should bear in mind that although hydrolysis with TFA is considered mild and believed not to cause significant desulfation, this effect cannot be completely ruled out [47,48]. Considering this, the ∆sigF polymer was also subjected to hydrolysis with HCl, which led to the formation of polymer variants mainly composed by low MM fractions (<342.30 Da) ( Figure S1). As expected, the disruption of glycosidic bonds by HCl was quite fast, and HCl had a stronger action than TFA. The polymer variants obtained not only had alterations in their MM distribution, but also differed in sulfate content ( Figure S2), being the complete desulfation achieved after 2 h of HCl hydrolysis. This is in agreement with the saturation point described for the HCl hydrolytic reaction but, as previously reported, the degradation of some sugars such as uronic acids might also occur [49,50]. The polymer variants obtained after HCl hydrolysis were also tested regarding their antitumor potential in Mewo cells. The results were similar to those obtained with the TFA variants, with the HCl variants having the same or lower antitumor activity compared to the original polymer Subsequently, the effect of two polymer variants (obtained after 25 min and 2 h of TFA hydrolysis) on the viability of Mewo cells was evaluated. Overall, our results showed that both variants had less antitumor activity than the original ΔsigF polymer, and that after 48 h of treatment there was a significant decrease in bioactivity for the three polymer concentrations tested ( Figure 2) [33]. Moreover, the variant with lower abundance of high MM fractions (2 h hydrolysis) had the lowest effect on Mewo cells viability, indicating that the high MM fractions play an important role in its antitumor activity.     This is in accordance with a previous study that reported that the hydrolysis of the EPS from Synechocystis aquatilis with TFA led to the loss of its bioactivity as inhibitor of the human complement system [47]. Furthermore, recent studies showed that hydrolyzed fucoidans from different sources maintained or reduced their antitumor activity compared to the original polymer [44]. On the other hand, our findings do not support the hypothesis that polymer variants with lower MM fractions could have higher antitumor activity by being more easily internalized by the cells [43,51].

Reduction of Peptide Content and Evaluation of the Antitumor Activity In Vitro
To evaluate the impact of peptide content on the antitumor activity of Synechocystis ∆sigF polymer (~27% w/w polymer dry weight, [27]), an aqueous solution of the polymer was incubated for 15 or 30 min with trichloroacetic acid (TCA). After either 15 or 30 min, the amount of peptides was reduced by approximately 40% (Figure 3a), with a minimal loss (~5%) in the yield of the polymer (Figure 3b) [36].
EPS from Synechocystis aquatilis with TFA led to the loss of its bioactivity as inhibitor of the human complement system [47]. Furthermore, recent studies showed that hydrolyzed fucoidans from different sources maintained or reduced their antitumor activity compared to the original polymer [44]. On the other hand, our findings do not support the hypothesis that polymer variants with lower MM fractions could have higher antitumor activity by being more easily internalized by the cells [43,51].

Reduction of Peptide Content and Evaluation of the Antitumor Activity In Vitro
To evaluate the impact of peptide content on the antitumor activity of Synechocystis ΔsigF polymer (~27% w/w polymer dry weight, [27]), an aqueous solution of the polymer was incubated for 15 or 30 min with trichloroacetic acid (TCA). After either 15 or 30 min, the amount of peptides was reduced by approximately 40% (Figure 3a), with a minimal loss (~5%) in the yield of the polymer (Figure 3b) [36]. Although TCA precipitation is usually highly efficient and widely used to remove peptides from polysaccharides [52][53][54], the efficiency reported here is lower than previously reported for polysaccharidic polymers from heterotrophic bacterial and plant sources [55,56], but in the same range as the ones reported for other cyanobacterial polymers [57,58]. In some cases, such as for polysaccharides from Spirulina platensis, higher TCA concentrations had to be used to achieve a precipitation efficiency similar to Although TCA precipitation is usually highly efficient and widely used to remove peptides from polysaccharides [52][53][54], the efficiency reported here is lower than previously reported for polysaccharidic polymers from heterotrophic bacterial and plant sources [55,56], but in the same range as the ones reported for other cyanobacterial polymers [57,58]. In some cases, such as for polysaccharides from Spirulina platensis, higher TCA concentrations had to be used to achieve a precipitation efficiency similar to the one reported here (~41% less peptides) [57]. These differences can be attributed to the high complexity of the cyanobacterial polymers and/or to a stronger association between the peptides and the polymer. Nevertheless, the amount of ∆sigF polymer that was lost during the TCA treatment in our study was among the lowest reported, even when compared with similar treatments using polymers from non-cyanobacterial sources [55]. Since the decrease in peptide content was similar in the two treatments (Figure 3a), the polymer variant obtained after 15 min (∆sigF.pep-) was the one selected to proceed.
The antitumor potential of ∆sigF.pepvariant was tested in vitro in Mewo cells and compared to the ∆sigF polymer with intact peptide content (Figure 4) [36]. Both polymers induced a strong decrease in cells viability in a time-and dose-dependent manner being this effect significantly stronger following treatment with the ∆sigF.pepvariant compared to the treatment with the original ∆sigF (Figure 4). At the lowest concentration tested (0.7 mg/mL), the ∆sigF polymer reduced Mewo cells viability in approximately 9% and 28% after 24 h and 48 h of treatment respectively, while the cells treated with ∆sigF.pepshowed a viability decrease of about 47% and 76%, respectively. This is in line with the results obtained by Garbacki et al. [59], which showed that TCA peptide precipitation from isolated cyanobacterial capsular polysaccharides (CPS) from two Phormidium strains increased the anti-inflammatory activity of these polymers in~60%. One possible explanation for the stronger bioactivity of the ∆sigF.pepcompared to ∆sigF may be the increased exposure of bioactive component(s). However, in our study we cannot exclude that at least some of the bioactive components could be the remaining peptides, since only a partial peptide removal was achieved with TCA treatment. A previous work reported that specific protein fractions in the EPS from the cyanobacterium Gloeocapsa sp. might be involved in their antitumor activity towards human cervical carcinoma (HeLa) cells [21]. Furthermore, other studies have shown that peptides present in the EPS from microalgae are important for their biological activity [60][61][62], namely a glycoprotein found in the EPS from Chlorella vulgaris that was associated with their antitumor activity [63].
polymer variant obtained after 15 min (ΔsigF.pep-) was the one selected to proceed.
The antitumor potential of ΔsigF.pep-variant was tested in vitro in Mewo cells and compared to the ΔsigF polymer with intact peptide content (Figure 4) [36]. Both polymers induced a strong decrease in cells viability in a time-and dose-dependent manner being this effect significantly stronger following treatment with the ΔsigF.pep-variant compared to the treatment with the original ΔsigF (Figure 4). At the lowest concentration tested (0.7 mg/mL), the ΔsigF polymer reduced Mewo cells viability in approximately 9% and 28% after 24 h and 48 h of treatment respectively, while the cells treated with ΔsigF.pep-showed a viability decrease of about 47% and 76%, respectively. This is in line with the results obtained by Garbacki et al. [59], which showed that TCA peptide precipitation from isolated cyanobacterial capsular polysaccharides (CPS) from two Phormidium strains increased the anti-inflammatory activity of these polymers in ~60%. One possible explanation for the stronger bioactivity of the ΔsigF.pep-compared to ΔsigF may be the increased exposure of bioactive component(s). However, in our study we cannot exclude that at least some of the bioactive components could be the remaining peptides, since only a partial peptide removal was achieved with TCA treatment. A previous work reported that specific protein fractions in the EPS from the cyanobacterium Gloeocapsa sp. might be involved in their antitumor activity towards human cervical carcinoma (HeLa) cells [21]. Furthermore, other studies have shown that peptides present in the EPS from microalgae are important for their biological activity [60][61][62], namely a glycoprotein found in the EPS from Chlorella vulgaris that was associated with their antitumor activity [63].

Assessment of the Antitumor Activity In Vivo
To validate in vivo the antitumor potential of our original polymer (ΔsigF) and its most promising variant (ΔsigF.pep-), the chick embryo chorioallantoic membrane (CAM) assay was used. This is a low-cost, reproducible, and reliable preclinical cancer model that allows investigation of tumor growth and angiogenesis in vivo, being often used to evaluate potential anticancer drugs [64]. Therefore, and following the in vitro results Figure 4. Effect of the Synechocystis ∆sigF polymer (∆sigF) and ∆sigF polymer with reduced peptide content (∆sigF.pep-) on the viability of human melanoma (Mewo) cell line, analyzed using Presto-Blue™ viability assay. Cells were treated with 0.7, 1 or 1.5 mg/mL of polymer for 24 or 48 h (T24 and T48, respectively). Cell treatment with polymer vehicle showed no differences compared to Blank (data not shown). Results are express in relation to Blank and are represented as mean ± STD of three independent experiments (**** p ≤ 0.0001).

Assessment of the Antitumor Activity In Vivo
To validate in vivo the antitumor potential of our original polymer (∆sigF) and its most promising variant (∆sigF.pep-), the chick embryo chorioallantoic membrane (CAM) assay was used. This is a low-cost, reproducible, and reliable preclinical cancer model that allows investigation of tumor growth and angiogenesis in vivo, being often used to evaluate potential anticancer drugs [64]. Therefore, and following the in vitro results showing that the variant ∆sigF.pephad stronger antitumor potential, CAM assays were performed [35]. For this purpose, CAMs were inoculated with Mewo cells at embryonic development day 9 (EDD9), together with the ∆sigF polymer, its ∆sigF.pepvariant or with their vehicle (DMEM, control). The lowest concentration (0.7 mg/mL) was chosen for this analysis. A dose-response curve for the effect of the ∆sigF.pepvariant in Mewo cells is shown in Figure S4, while for the ∆sigF polymer it has been previously published in [20].
After four days of treatment, CAMs were retrieved and analyzed ex ovo to evaluate vessel formation (angiogenesis) and tumor growth ( Figure 5). Results regarding the number of blood vessels counted on the region of interest showed no significant differences between polymer and control treatments, although a tendency for a slight decrease in vessel count was observed with ∆sigF.pepvariant, compared to the control (9.5 ± 1.2 and 10.7 ± 1.2, respectively) ( Figure 5b). On the other hand, evaluation of tumor size showed that treatment with both polymers (∆sigF polymer and ∆sigF.pepvariant) resulted in a significant decrease of tumor area (7.2 ± 2.1 mm 2 and 7.1± 3.7 mm 2 , respectively) compared to the control condition (9.9 ± 2.1 mm 2 ) (Figure 5c). These results, demonstrating in vivo antitumor activity of both polymers, are in accordance with previously in vitro results ( Figure 4). However, no significant differences were observed when comparing the effect of the two polymers on tumor growth in vivo.
(9.5 ± 1.2 and 10.7 ± 1.2, respectively) (Figure 5b). On the other hand, evaluation of tumor size showed that treatment with both polymers (ΔsigF polymer and ΔsigF.pep-variant) resulted in a significant decrease of tumor area (7.2 ± 2.1 mm 2 and 7.1± 3.7 mm 2 , respectively) compared to the control condition (9.9 ± 2.1 mm 2 ) (Figure 5c). These results, demonstrating in vivo antitumor activity of both polymers, are in accordance with previously in vitro results ( Figure 4). However, no significant differences were observed when comparing the effect of the two polymers on tumor growth in vivo. To gain insight into potential mechanisms involved in the effect of these polymers in tumor growth, CAM xenografted tumors were further evaluated histologically. Hematoxylin-eosin staining showed differences in tumor cell morphology ( Figure 6). While the control tumors (treated with DMEM) exhibited a more compact and cellular structure, the tumors treated with the polymers (in particular with ΔsigF polymer) were To gain insight into potential mechanisms involved in the effect of these polymers in tumor growth, CAM xenografted tumors were further evaluated histologically. Hematoxylineosin staining showed differences in tumor cell morphology ( Figure 6). While the control tumors (treated with DMEM) exhibited a more compact and cellular structure, the tumors treated with the polymers (in particular with ∆sigF polymer) were less compact (with more stroma and less cells within the same area). Since we previously showed that short term exposition to the ∆sigF polymer increased apoptosis markers [20], here we proceed to evaluate proliferation and apoptosis in the CAM tumors. No differences were observed regarding the percentage of Ki-67-stained cells per slide, with polymer treated tumors presenting similar levels compared to controls (∆sigF: 70.3 ± 4%; ∆sigF.pep-: 71.9 ± 4% Control: 73.8 ± 2%, Figure 6 middle panel). Moreover, no evident differences in caspase 3 cleavage were observed in any of the conditions (Figure 6 lower panel). We cannot exclude the possibility that alterations in apoptosis (or proliferation, for that matter) may be occurring in polymer-treated cells at earlier time points. It is possible that the observed differences in tumor cell morphology may result from an initial effect of the polymers on the inoculated cells, which was not evident in the proliferative or apoptotic profiles of the surviving cells at EDD13. In addition, in our previous study [20], the experiments were carried out only in in vitro conditions and using different methodologies (Annexin V/PI staining by flow cytometry and cleaved caspase-3 by Western blot). Finally, it is important to consider that the use of different models (in vitro vs. in vivo) may also impact the results, as in vitro experiments do not take into account the complex interactions between tumor cells and the surrounding stroma. of the polymers on the inoculated cells, which was not evident in the proliferative or apoptotic profiles of the surviving cells at EDD13. In addition, in our previous study [20], the experiments were carried out only in in vitro conditions and using different methodologies (Annexin V/PI staining by flow cytometry and cleaved caspase-3 by Western blot). Finally, it is important to consider that the use of different models (in vitro vs. in vivo) may also impact the results, as in vitro experiments do not take into account the complex interactions between tumor cells and the surrounding stroma.

Conclusions
Overall, Synechocystis ΔsigF polymer was shown to be a powerful platform to develop polymer variants with in vitro and in vivo antitumor activity. We demonstrated that high molecular mass fractions of the ΔsigF polymer were important for its antitumor activity in human melanoma cells, and that the reduction of the peptide content generated a polymer variant with enhanced in vitro antitumor activity. Importantly, the antitumor potential of ΔsigF polymer and its ΔsigF.pep-variant was validated in vivo, with both polymers significantly decreasing xenografted CAM tumor growth and affecting tumor morphology, by promoting less compact tumors. Further studies are necessary to fully disclose the mechanisms behind the effects of these polymers in tumor growth, as well as to overcome technical issues usually associated with polymers with high molecular mass

Conclusions
Overall, Synechocystis ∆sigF polymer was shown to be a powerful platform to develop polymer variants with in vitro and in vivo antitumor activity. We demonstrated that high molecular mass fractions of the ∆sigF polymer were important for its antitumor activity in human melanoma cells, and that the reduction of the peptide content generated a polymer variant with enhanced in vitro antitumor activity. Importantly, the antitumor potential of ∆sigF polymer and its ∆sigF.pepvariant was validated in vivo, with both polymers significantly decreasing xenografted CAM tumor growth and affecting tumor morphology, by promoting less compact tumors. Further studies are necessary to fully disclose the mechanisms behind the effects of these polymers in tumor growth, as well as to overcome technical issues usually associated with polymers with high molecular mass fractions/viscosity. Nevertheless, the methodologies herein implemented represent straightforward strategies for the design and testing of tailored cyanobacterial extracellular polymers for biotechnological/biomedical applications. Moreover, this work further strengthens the relevance of evaluating the antitumor potential of this type of natural polymers both in vitro and in vivo.

Supplementary Materials:
The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/polym15061382/s1, Figure S1: Molecular mass distribution of Synechocystis sp. PCC 6803 ∆sigF polymer before and after hydrolysis with HCl for different time periods; Figure S2: Quantification of the sulfate released from Synechocystis ∆sigF polymer after HCl hydrolysis (from 10 min to 2 h), and the percentage of desulfation obtained; Figure S3: Effect of the Synechocystis ∆sigF polymer, and its variants obtained after hydrolysis with HCl (from 10 min to 2 h), on the viability of human melanoma (Mewo) cell line evaluated with the PrestoBlue TM viability assay; Figure S4: Effect of the Synechocystis ∆sigF polymer with reduced peptide content (∆sigF.pep-) on the viability of human melanoma (Mewo) cell line, analyzed using the PrestoBlue™ viability assay.