2-Deoxy-D-Ribose (2Ddr) Upregulates Vascular Endothelial Growth Factor (VEGF) and Stimulates Angiogenesis T ⁎ Serkan Dikicia, Anthony J

Microvascular Research 131 (2020) 104035

Contents lists available at ScienceDirect

Microvascular Research

journal homepage: www.elsevier.com/locate/ymvre

2-deoxy-D-ribose (2dDR) upregulates vascular endothelial growth factor (VEGF) and stimulates angiogenesis T ⁎ Serkan Dikicia, Anthony J. Bullocka, Muhammad Yarb, Frederik Claeyssensa, Sheila MacNeila, a Department of Materials Science & Engineering, Kroto Research Institute, University of Sheﬃeld, Sheﬃeld, UK b Interdisciplinary Research Centre in Biomedical Materials (IRCBM), COMSATS University Islamabad Lahore Campus, Lahore 54000, Pakistan

ARTICLE INFO ABSTRACT

Keywords: Background: Delayed neovascularisation of tissue-engineered (TE) complex constructs is a major challenge that 2-deoxy-D-ribose (2dDR) causes their failure post-implantation. Although significant progress has been made in the field of angiogenesis, 2-deoxy-L-ribose (2dLR) ensuring rapid neovascularisation still remains a challenge. The use of pro-angiogenic agents is an effective D-Glucose approach to promote angiogenesis, and vascular endothelial growth factor (VEGF) has been widely studied both Angiogenesis at the biological and molecular levels and is recognised as a key stimulator of angiogenesis. However, the Neovascularisation exogenous use of VEGF in an uncontrolled manner has been shown to result in leaky, permeable and haemor- Vascular endothelial growth factor (VEGF) Chick chorioallantoic membrane (CAM) assay rhagic vessels. Thus, researchers have been actively seeking alternative agents to upregulate VEGF production Endothelial cells rather than exogenous use of VEGF in TE systems. We have previously revealed the potential of 2-deoxy-D-ribose (2dDR) as an alternative pro-angiogenic agent to induce angiogenesis and accelerates wound healing. However, to date, there is not any clear evidence on whether 2dDR influences the angiogenic cascade that involves VEGF. Methods: In this study, we explored the angiogenic properties of 2dDR either by its direct application to human aortic endothelial cells (HAECs) or when released from commercially available alginate dressings and demonstrated that when 2dDR promotes angiogenesis, it also increases the VEGF production of HAECs. Results: The VEGF quantification results suggested that VEGF production by HAECs was increased with 2dDR treatment but not with other sugars, including 2-deoxy-L-ribose (2dLR) and D-glucose (DG). The stability studies demonstrated that approximately 40–50% of the 2dDR had disappeared in the media over 14 days, either in the presence or absence of HAECs, and the reduction was higher when cells were present. The concentration of VEGF in the media also fell after day 4 associated with the reduction in 2dDR. Conclusion: This study suggests that 2dDR (but not other sugars tested in this study) stimulates angiogenesis by increasing the production of VEGF. We conclude 2dDR appears to be a practical and effective indirect route to upregulating VEGF for several days, leading to increased angiogenesis.

1. Introduction 1999) and in vivo (Ribatti et al., 2006; Roma-Rodrigues et al., 2016). However, researchers also reported that the exogenous use of VEGF in The lack of blood vessels in current tissue-engineered (TE) con- an uncontrolled manner could result in leaky (Yancopoulos et al., structs is one of the most important challenges in the survival of TE 2000), permeable (Cao et al., 2004) and haemorrhagic (Cheng et al., constructs (Langer, 2000; Rivron et al., 2008). The use of pro-angio- 1997) vessels which can be observed in tumour angiogenesis (Oka genic agents to stimulate angiogenesis is an effective approach to cir- et al., 2007). cumvent slow neovascularisation of TE constructs (Dew et al., 2015; Although significant progress has been made, and the studies on Richardson et al., 2001), and vascular endothelial growth factor (VEGF) angiogenesis and VEGF have significantly increased over the last two is recognised as a key stimulator of angiogenesis. It has been proven to decades, there is still a need for alternative pro-angiogenic agents to have important roles in critical steps of the angiogenic process in- promote angiogenesis in an inexpensive and safe manner (Ferrara, cluding vasodilation and permeability, destabilisation of vessels and 2007). Our group has been seeking alternative agents to increase an- degradation of extracellular matrix (ECM), proliferation and migration giogenesis, and we have explored deoxy sugars for their potential to of endothelial cells (ECs), lumen formation and vessel stabilisation in promote angiogenesis. The literature on small sugars is very limited and vitro (Obi and Toda, 2015; Poldervaart et al., 2014; Vernon and Sage, contradictory. We have recently revealed 2-deoxy-D-ribose (2dDR)'s

⁎ Corresponding author. E-mail address: s.macneil@sheffield.ac.uk (S. MacNeil). https://doi.org/10.1016/j.mvr.2020.104035 Received 18 May 2020; Received in revised form 12 June 2020; Accepted 16 June 2020 Available online 25 June 2020 0026-2862/ © 2020 Elsevier Inc. All rights reserved. S. Dikici, et al. Microvascular Research 131 (2020) 104035 highly angiogenic potential in vitro (Dikici et al., 2019a), in chick replaced every 3 days until they reached ~80–90% confluency. chorioallantoic membrane (CAM) assay (Dikici et al., 2019c; Yar et al., HAECs were used between passages 2 and 6. 2017), and in vivo (Azam et al., 2019; Yar et al., 2017). Similarly, several other groups have also reported that 2dDR is a promoter of 2.2.1.2. Preparation of sugar solutions. Working solutions of all angiogenesis in vitro (Nakajima et al., 2009; Uchimiya et al., 2002) and substances were prepared fresh at the beginning of each experiment. in vivo (Haraguchi et al., 1994; Matsushita et al., 1999). However, the 10 mM stock solutions of 2dDR, 2dLR, and DG were prepared by levo isomer of 2dDR, 2-deoxy-L-ribose (2dLR) has been reported to be dissolving the sugars in low serum (2% FCS) EC GM and filtered using a anti-angiogenic in vitro (Ikeda et al., 2002; Nakajima et al., 2004, 2006, 0.2 μm syringe filter, and the stock solutions were then diluted 100× in 2009; Uchimiya et al., 2002) and in vivo (Uchimiya et al., 2002). Yet sterile EC GM to prepare 100 μM sugar solutions. 80 ng/ml VEGF-165 other studies demonstrated that 2dLR is a promoter of angiogenesis in (VEGF-165 will be mentioned as VEGF in the rest of the text) solution the CAM assay (Yar et al., 2017) and in vivo (Sengupta et al., 2003). D- was prepared in sterile EC GM and used as a positive control. Low glucose, as the primary energy source of cells, has also been reported as serum EC GM was used as a control in the experiments. both promoting (Madonna et al., 2016; Shigematsu et al., 1999) and inhibiting (Jiraritthamrong et al., 2012; Teixeira and Andrade, 1999; 2.2.1.3. Effect of 2dDR, 2dLR, and DG on the metabolic activity of Yu et al., 2006) angiogenesis. Finally, 2-deoxy-D-glucose (2dDG), a HAECs. AlamarBlue cell metabolic activity assay was used to evaluate glucose molecule which has the 2-hydroxyl group replaced by hy- the change in the metabolic activity of HAECs in response to different drogen, has been reported to be anti-angiogenic in vitro (Chuang et al., sugar treatments (Dikici et al., 2020a). The principle of this assay is the 2015; Kovacs et al., 2016; Merchan et al., 2010; Tagg et al., 2008) and reduction of non-fluorescent resazurin to highly fluorescent resorufin in vivo (Merchan et al., 2010) by several groups. upon entering cells. The viable cells convert resazurin to resorufin As can be seen from this literature on sugars, in contrast to VEGF, continuously. the mechanism of their action is still essentially unclear. Accordingly, in Once the cells reached confluence, HAECs were trypsinised and this study, we aimed to explore some important pro-angiogenic end- seeded into 48-well plates with a seeding density of 1 × 104 HAECs/ points in the relationship between 2dDR and VEGF. First, we in- cm-1. HAECs were cultured with EC GM, either containing different vestigated the effective pro-angiogenic concentrations of 2dDR in vitro concentrations of 2dDR or VEGF. AlamarBlue assay was performed at and in ex-ovo CAM assay either when administered directly to ECs or days 1, 4, and 7 or 1, 3, and 5 depending on the experiments. Briefly, when released from commercially available alginate dressings. Then, 0.1 mM AlamarBlue working solution was prepared by 10× dilution of we compared 2dDR with other small sugars, including 2-deoxy-L-ribose the 1 mM AlamarBlue stock solution with EC GM. Growth medium was (2dLR) and D-glucose (DG) in terms of increasing the metabolic activity removed, and the cells were washed with phosphate buffered saline of human aortic endothelial cells (HAECs). Finally, we evaluated the (PBS). 1 ml of AlamarBlue working solution was added to each well and effect of direct application of 2dDR and of 2dDR released from alginate incubated at 37 °C for 4 h. After an incubation period, 200 μl of the dressings on the concentration of VEGF in the media of HAECs and we solution was transferred into a 96-well plate, and the fluorescence looked at the stability of 2dDR in culture medium in the presence and readings were done at an excitation wavelength of 540 nm and an absence of HAECs. emission wavelength of 635 nm.

2. Experimental 2.2.1.4. Fluorescent staining of the cells. Fluorescent staining was performed by labelling F-actin and cell nuclei of HAECs as described 2.1. Materials previously (Dikici et al., 2020b). Cells were washed with PBS before (once) and after (three times) fixing them in 4% PFA for 15 min. 0.1% 2-Deoxy-D-ribose (2dDR), 2-deoxy-L-ribose (2dDR), D-glucose (DG), (v/v) Triton 100× (in PBS) was added on samples, and the samples human vascular endothelial growth factor 165 (VEGF-165) expressed in were incubated for 20–30 min at room temperature (RT). After three Escherichia coli (E. coli), orcinol, 3.7% formaldehyde (FA) solution, times washing with PBS, Alexa Fluor 594 Phalloidin (1:40 diluted in AlamarBlue cell metabolic activity assay, Triton X-100, ferric chloride PBS from stock solution) solution was added to cells in order to stain F- hexahydrate (FeCl3.6H2O) and hydrochloric acid (HCl) were purchased actin filaments of cells and incubated for 30 min at RT in the dark. Cells from Sigma Aldrich (St. Louis, Missouri, USA). Human Aortic were then washed three times with PBS. In order to stain cell nuclei, Endothelial Cells (HAECs) and EC growth medium (EC GM) were pur- DAPI solution (1:1000 diluted in PBS), which strongly binds the chased from PromoCell (Heidelberg, Germany). 4′,6-diamidino-2-phe- adenine-thymine rich regions of DNA, was added and incubated for nylindole (DAPI) solution and Alexa Fluor 594 Phalloidin were pur- 10–15 min RT in the dark, and cells were then washed 3 times with chased from ThermoFisher Scientific (San Jose, CA, USA). Human VEGF PBS. Cells were then examined with a fluorescent microscope ELISA MAX™ Deluxe Set was purchased from Biolegend (San Diego, CA, (Olympus, Japan). The number of cells was quantified using DAPI USA). Alginate dressings were obtained from Advanced Medical stained well-plates at the end of the incubation period using ImageJ Solutions (Winsford, UK). software as described previously (Dikici et al., 2019a).

2.2. Methods 2.2.2. Assessment of the VEGF-dependency of 2dDR-related angiogenic activity 2.2.1. Evaluating the angiogenic potential of 2-deoxy-D-ribose (2dDR), 2- Human VEGF ELISA MAXTM Deluxe Set was used for the quantifi- deoxy-L-ribose (2dLR), and D-glucose (DG) in vitro cation of VEGF production of HAECs either with the direct addition of 2.2.1.1. Culture of human aortic endothelial cells (HAECs). HAECs 2dDR at 100 μM and 1 mM concentrations in comparison with VEGF removed from liquid nitrogen immediately thawed at 37 °C and (80 ng/ml) and controls or when 2dDR (5% and 10%) were released transferred to a container with 10 ml of endothelial cell growth from commercially available alginate dressings in comparison with medium (EC GM). Cells were centrifuged, and the cell pellet was re- 2dLR (5% and 10%), DG (10%) and control dressings. suspended in EC GM supplemented with a total of 5% FBS and growth medium 2 kit supplements (5 ng/ml EGF, 10 ng/ml bFGF, 20 ng/ml 2.2.2.1. Direct addition of 2dDR to the EC GM. To investigate the VEGF insulin-like growth factor, 0.5 ng/ml VEGF, 1 μg/ml ascorbic acid, production of HAECs in response to 2dDR treatment, cells were 22.5 μg/ml heparin, 0.2 μg/ml hydrocortisone). The cell suspension incubated in low serum EC GM supplemented with 100 μM of 2dDR, 2 was then transferred into 75 cm flasks and incubated at 37 °C, 5% CO2 1 mM of 2dDR, and 80 ng/ml VEGF. Non-supplemented low serum EC (Thermo Scientific, Massachusetts, USA). The culture media was GM was used as a control. Growth media of the cells were collected at

2 S. Dikici, et al. Microvascular Research 131 (2020) 104035 day 1, 3, and 5, centrifuged at ~150 relative centrifugal force (RCF) for of Bial's reagent was added. The solution was heated to boil using a hot 5 min, and stored at −20 °C. plate, and the sample tubes were then submerged into the boiling Bial's For quantification of the samples, first, 1 ml of VEGF stock solution reagent solution for 1 min. Absorbance was measured using a UV-VIS was prepared as 1500 pg/ml in 1× assay diluent A, and then six serial spectrophotometer at 630 nm. dilutions were done down to 750, 375, 187.5, 93.8, 46.9, and 23.4 pg/ ml in order to obtain a reference curve for converting readings to 2.2.4. Evaluation of the angiogenic potential of 2dDR, 2dLR, and DG in ex- concentrations. ovo CAM assay On day 1, 100 μl of 1× capture antibody was added to each well to The CAM assay was performed to evaluate the angiogenesis either coat the well plate with the anti-VEGF antibody. The plate was then with the direct application of 2dDR (200 μg/day/embryo) in compar- sealed using parafilm and incubated overnight at 4 °C. ison with VEGF (80 ng/ml) and controls (PBS) or when 2dDR (5% and On day 2, the plate was washed 4 times with 1× wash buffer, and 10%), 2dLR (5% and 10%), DG (10%) were released from alginate blocked by adding 200 μl of 1× assay diluent A to each well and in- dressings. cubated at RT for 1 h on a shaker. The plate was then washed 4 times with 1× wash buffer, and 50 μl of 1× assay diluent D was added to 2.2.4.1. Direct application of 2dDR to CAM. All CAM experiments were each well. 50 μl of the samples (growth media from HAECs) and 50 μlof carried as described previously (Aldemir Dikici et al., 2019; Dikici et al., the prepared VEGF concentrations (reference standards) were added to 2019c; Mangir et al., 2019). Briefly, fertilised chicken eggs (Gallus wells, and the plate was then sealed and incubated at RT for 2 h on a domesticus) were purchased from Henry Stewart & Co. MedEggs shaker. The plate was washed 4 times with 1× wash buffer, and 100 μl (Norwich, UK) and cleaned with 20% industrial methylated spirit of 1× detection antibody was added to each well and incubated at RT solution. Eggs were incubated at 37.5 °C for 3 days in a rocking egg for 1 h on a shaker. The plate was washed 4 times with 1× wash buffer, incubator (RCOM King SURO, P&T Poultry, Powys, Wales). On and 100 μl of 1× avidin-HRP was added to each well and incubated at embryonic development day (EDD) 3, the embryos were transferred RT for 30 min on a shaker. The plate was washed 5 times with 1× wash gently into sterile petri dishes and incubated at 38 °C in a cell culture buffer (at least for 30 s for each wash), and 100 μl of 1× substrate incubator (Binder, Tuttlingen, Germany). The CAM assay was solution D was added to each well and incubated at RT in the dark for conducted according to the Home Office, UK guidelines. 10 min. 100 μl of 1× stop solution was then added to each well to stop Plastic rings (~6.5 mm in diameter) were placed on the CAM the reaction. The absorbance was read at 450 nm and 570 nm, the halfway between the embryo and the outer border of the CAM and reading at 570 nm was subtracted from that at 450 nm. The absorbance between two large vessels as described previously (Mangir et al., 2019). values were converted to concentrations using the reference curve The substances to be examined were applied as 20 μl volume into the drawn from the standard VEGF concentrations. rings placed on the CAM twice a day for 3 days starting from EDD 7. On EDD 14, images of the CAM area circumscribed by the plastic rings 2.2.2.2. Release of the agents from alginate dressings. Alginate dressings were acquired using a digital microscope and embryos were sacrificed (size: 10 × 10 cm2, thickness: 2 mm) were loaded with 2dDR (5% and immediately after image acquisition. Digital images were used for 10%), 2dLR (5% and 10%), and DG (10%) as described previously quantification of the results. Multiple image processing steps were ap- (Andleeb et al., 2020; Azam et al., 2019). Briefly, 1 cm2 patches of pre- plied following previously described protocols (Brooks et al., 1999; sterilised alginate dressings (11 mg approx.) were cut and placed into a Ribatti et al., 2006; Roma-Rodrigues et al., 2016). Firstly, the internal sterile multi-well plate. Syringe filtered solutions of 2dDR, 2dLR, and area of the ring was cropped, and the raw image was split to its three DG were loaded onto alginate patches by pipetting 100 μl of the sugar main colour channels (red, green and blue (RGB) channels) using Adobe solution onto the dressing. After ensuring that the liquid had soaked Photoshop CS6 (ADOBE Systems Inc., San Jose, California, USA). Only through the dressings, they were left in a sterile environment at RT for the green channel was exported as an image file and then imported to 24 h to allow the water to evaporate. Once dry, the dressings were used ImageJ (Wayne Rasband, National Institutes of Health, USA) for further in experiments. analysis including unsharp mask filtering, enhancing the local contrast, To investigate the VEGF production of HAECs in response to the noise removal, converting the image to binary and segmentation. Fi- release of the agents from the dressings, cells were incubated in low nally, the average vessel lengths were quantified using ImageJ software serum EC GM in 24-well plates, and the dressings were placed in cell as described previously (Aldemir Dikici et al., 2020; Dikici et al., 2019b, culture inserts. Growth media of the cells were collected at day 1, 3, and 2019c). 7, centrifuged at ~150 RCF for 5 min, and stored at −20 °C. For quantification of the samples, Human VEGF ELISA MAXTM Deluxe Set 2.2.4.2. The release of sugars from the alginate dressings to evaluate was used as described above. angiogenesis on CAM. The 2dDR (5% and 10%), 2dLR (5% and 10%), and DG (10%) loaded alginate dressings were prepared as described in 2.2.3. Assessment of the stability of 2dDR in the presence and absence of Section 2.2.2.2. Fertilised eggs were incubated at 37.5 °C for 3 days in a HAECs rocking egg incubator (RCOM King SURO, P&T Poultry, Powys, Wales). The stability of 2dDR was assessed using Bial's Orcinol assay, as On EDD 3, the embryos were transferred gently into sterile petri dishes described previously (Azam et al., 2019). Briefly, two defined con- and incubated at 38 °C in a cell culture incubator (Binder, Tuttlingen, centrations (100 μM and 1 mM) of 2dDR were prepared in EC GM. The Germany). On EDD 7, 1 cm2 the alginate dressings were implanted to solutions were incubated at 37 °C in the presence and absence of CAMs for a further 7 days. On EDD 14, images of the CAMs with the HAECs. For this, HAECs were seeded in 24-well plates with a seeding implants were acquired using a digital microscope and embryos were density of 4 × 104 cells/cm2 and incubated with EC GM supplemented sacrificed immediately after image acquisition. with 2dDR. On days 1, 4, 7, and 14, the growth medium was collected, The total length of small blood vessels was calculated by using and Bial's assay was performed (Patterson and Mura, 2013). The prin- binary image histograms with known pixel/mm ratios in ImageJ as ciple of this test is the formation of furfural (an organic compound: described previously (Dikici et al., 2019c).

C4H3OCHO) by the dehydration of pentoses (sugars with five carbon atoms) with Bial's reagent. Furfural will then react with orcinol and 2.2.5. Statistical analysis generate a substrate with a blue colour. Briefly, Bial's reagent was Statistical analyses were performed using one-way analysis of var- prepared by combining 0.4 g of orcinol, 200 ml of 37% concentrated iance (ANOVA) for CAM assay results and the quantification of DAPI hydrochloric acid and 0.5 ml of a 10% solution of ferric chloride. A 2 ml stained HAECs. The results of the AlamarBlue metabolic activity mea- of media collected from 6-well plate was placed in a test tube, and 2 ml surements, the VEGF quantification results, and the 2dDR stability

3 S. Dikici, et al. Microvascular Research 131 (2020) 104035

Fig. 1. The effect of 2dDR on the metabolic activity and proliferation of HAECs. (A) Phalloidin-TRITC & DAPI stained HAECs cultured over 5 days. (B) The results of AlamarBlue assay showing the change in the metabolic activities of HAECs over 5 days in response to 2dDR and VEGF treatment. (C) The quantification of the number of HAECs at the end of day 7 from DAPI staining. Scale bars represent 200 μm. results were analysed by two-way ANOVA. Error bars indicate standard 3. Results deviations in the graphs unless otherwise stated. 3.1. Comparative effects of 2dDR, 2dLR, and DG on endothelial cells

Two aspects of endothelial cell biology were measured. (i) the cell viability of HAECs when treated with diﬀerent concentrations of 2dDR over 7 days as shown in Fig. 1 and (ii) cell viability of HAECs in

4 S. Dikici, et al. Microvascular Research 131 (2020) 104035

Fig. 2. Comparison of 2dDR with other sugars (2dLR and DG) in terms of increasing the metabolic activity of HAECs over 7 days in comparison with VEGF and controls. (***p ≤ 0.001, *p ≤ 0.05, not significant (ns) p ≥ 0.05, n = 3). The statistical comparison of the results is given in the table on the right. response to the treatment with an effective dose of 2dDR in comparison of the time points, where HAECs were incubated with non-supple- with other sugars as shown in Fig. 2. mented EC GM. The addition of 2dDR and VEGF caused an increase in metabolic The VEGF production by HAECs in response to 2dDR treatment was activity of HAECs over 7 days. None of the other small sugar molecules statistically significant even at day 1. The amount of VEGF rose to increased the activity of HAECs over 7 days. 871.4 ± 361.7 pg/ml and 1284.5 ± 683.1 pg/ml, respectively for At day 1, none of the groups showed any significant difference in 100 μM and 1 mM 2dDR treated groups when compared to controls the activity of cells while VEGF and 2dDR (100 μM and 1 mM) showed a (average VEGF amount: 18.3 ± 1.9 pg/ml). significant increase in the metabolic activity of HAECs. A further in- By day 5, the amount of VEGF was significantly increased to crease was observed on day 7. No statistically significant difference was 1140.6 ± 441.8 pg/ml and 1233.4 ± 384.1 pg/ml, respectively when found between VEGF treated and 2dDR treated groups (Fig. 1B). treated with 100 μM and 1 mM of 2dDR whereas the control group This data was further confirmed with the fluorescent staining of the showed almost no change compared to VEGF production on day 1 HAECs to quantify cell number by staining for DAPI at the end of day 7 (15.3 ± 4.4 pg/ml). (Fig. 1C). The highest number of cells was in the VEGF treated groups, The group to which VEGF was added had a significantly higher and the number of HAECs treated with 100 μM and 1 mM 2dDR were amount of VEGF in the media than the other groups at all time points as also significantly higher than controls. expected (3669.2 ± 559.0 pg/ml and 3930.9 ± 850.5 pg/ml of VEGF The results of the comparison of 2dDR with other sugars (2dLR and by days 1 and 5, respectively). DG) in terms of increasing the metabolic activity of HAECs over 7 days As for the direct treatment of HAECs with 2dDR, the release of 2dDR in comparison with VEGF and controls are given in Fig. 2. from the alginate dressings enhanced the production of VEGF over 7 days (Fig. 3B). On day 4, the results showed that 5% and 10% 2dDR loaded algi- 3.2. 2dDR increases the VEGF production of HAECs nate dressings significantly increased the production of VEGF to 1.75 ng/ml and 5.25 ng/ml, respectively when compared to controls The results of VEGF quantification demonstrated that the adminis- (0.05 ng/ml). Although the 10% 2dLR groups also showed an increase tration of either 100 μM or 1 mM of 2dDR to HAECs growth media led in the production of VEGF (0.75 ng/ml), this was not statistically dif- an increase in the amount of VEGF produced by the cells over 5 days ferent from controls. (Fig. 3A). No VEGF production was observed in the control group at any

Fig. 3. (A) Quantiﬁcation of VEGF production by HAECs in response to direct 2dDR treatment (100 μM and 1 mM). (B) VEGF production of HAECs when 2dDR, 2dLR, and DG was released from the alginate dressings. (***p ≤ 0.001, **p ≤ 0.01, *p ≤ 0.05, not signiﬁcant (ns) p ≥ 0.05, n = 3).

5 S. Dikici, et al. Microvascular Research 131 (2020) 104035

Fig. 4. Bial's Orcinol Assay for the assessment of the stability of 2dDR. (A) 100 μM 2dDR and (B) 1 mM 2dDR in the presence or absence of HAECs over 14 days. (**p ≤ 0.01, *p ≤ 0.05, n = 3).

On day 7, a decrease was observed in the amount of VEGF present in respectively when CAMs were treated with 5% 2dLR, 10% 2dLR, and the media. All other groups, apart from the 10% 2dDR, were not sta- 10% DG loaded alginate dressings. tistically diﬀerent from controls. The amount of VEGF found in the 10% 2dDR loaded alginate dressing groups was 2.75 ng/ml. 4. Discussion

3.3. The stability of 2dDR in the presence and absence of HAECs VEGF is a very well-established pro-angiogenic agent with important roles in stimulating angiogenesis (Hoeben et al., 2004). It is The results of stability studies showed that there was a decrease of considered as the gold standard for inducing angiogenesis in vitro and in 2dDR in the media over 7 days, either in the presence or absence of vivo (Olsson et al., 2006; Wang et al., 2008). However, the uncontrolled HAECs. The percentage reductions of 2dDR on both day 4 and 7 were administration of exogenous VEGF has been shown to promote the higher in the presence of HAECs when compared to control groups (no formation of leaky (Yancopoulos et al., 2000), permeable (Cao et al., cells). 2004), and haemorrhagic (Cheng et al., 1997) blood vessels. For 100 μM 2dDR treatment, on day 4, the amount of 2dDR was To date, an extensive number of studies have explored the signalling reduced by 23.6 ± 8.2% and 11.5 ± 2.7% in the presence and ab- pathway of VEGF in the angiogenic process (Hoeben et al., 2004; Koch sence of HAECs, respectively. On day 7, this was further reduced by et al., 2011). VEGF stimulates proliferation of ECs via the activation of 37.4 ± 9.5% (with HAECs) and 21.5 ± 6.6% (without HAECs). On VEGF receptor 2 (VEGFR2) which triggers RAS/ Rapidly Accelerated day 14, the percentage reductions in the amount of 2dDR were Fibrosarcoma (RAF) /extracellular signal-regulated kinase (ERK)/ Mi- 41.2 ± 9.9% and 37.1 ± 3.5%, respectively, for cellular and acellular togen-activated protein kinase (MAPK) pathway stimulates prolifera- groups with no statistically significant difference (Fig. 4A). tion of ECs (Meadows et al., 2001; Zachary and Gliki, 2001). It induces For 1 mM 2dDR treatment, on day 4, the amount of 2dDR was re- the migratory response of ECs with the activation of either focal ad- duced by 28.7 ± 7.7% and 12.8 ± 6.3% in the presence and absence hesion turnover (Parsons, 2003) or actin reorganisation (Holmqvist of HAECs, respectively. On day 7, this was further reduced by et al., 2003; Lamalice et al., 2006) signalling pathways. VEGF also 31.5 ± 4.4% (with HAECs) and 18.5 ± 5.9% (without HAECs). By regulates survival of ECs via VEGFR-2 dependent activation of phos- day 14, the percentage reductions in the amount of 2dDR were phoinositide 3 kinase (PI3K)-AKT8 virus oncogene cellular homolog 47.1 ± 10.2% and 37.2 ± 10.3%, respectively, when HAECs existed (AKT) signalling pathways to inhibit apoptosis via B cell leukaemia 2 and not existed in the well plates. (Fig. 4B). protein (BCL-2) associated death promoter (BAD) and caspase 9 (Cardone et al., 1998) pathways. It also plays a key role in the per- 3.4. 2dDR treatment increases the angiogenic activity on CAM either when meability of ECs via endothelial nitric oxide synthase (eNOS)-mediated added directly or when released from alginate dressings generation of nitric oxide (NO) (Fukumura et al., 2001). 2dDR, one of the degradation products of thymidine, is an in- The quantification of the macro images of CAMs with the direct expensive, commercially available and effective alternative to VEGF treatment of 2dDR (200 μg/day) showed that 2dDR promoted angio- which can induce angiogenesis. Recently, our group demonstrated that genesis and significantly increased the total vessel length 1.2-fold when 2dDR, when released from a hydrogel, is not only angiogenic but also compared to controls. VEGF treatment increased the total vessel length stimulates wound healing in a rat skin wound model (Yar et al., 2017). 1.4-fold compared to controls (Fig. 5). In addition, our group has recently revealed the angiogenic potential of The CAM assay results of the alginate dressings loaded with 2dDR, 2dDR using well-established in vitro models (Dikici et al., 2019a), an ex- 2dLR and DG demonstrated that only 2dDR (10%) significantly im- ovo CAM assay (Dikici et al., 2019c), and in a comparatively difficult to proved angiogenic activity and increased the total length of the small heal diabetic rat model (Azam et al., 2019). However, all of these blood vessels (Fig. 6). The total small vessel length per image was studies focused on the biological activity of 2dDR and did not attempt 95.7 ± 5.6 mm and 103.0 ± 3.4 mm, respectively for 5% and 10% to clarify how this small sugar stimulates angiogenesis. loaded groups while the vessel length was 51.8 ± 2 mm in the non- Accordingly, in this study, we aimed to explore how 2dDR takes part loaded dressing group. Although 2dLR and DG loading increased the in the angiogenic pathway and what is the relationship between VEGF total blood vessel length on CAM over 7 days, these were not statisti- and 2dDR. Our results demonstrated that 2dDR could be used to induce cally significant. The total small blood vessel lengths were increased to angiogenesis in vitro and in vivo either when administered directly at a 67.5 ± 30.4 mm, 91.9 ± 16.7 mm, and 73.6 ± 21.2 mm, concentration range of 100 μM to 1 mM or when released from 10%

6 S. Dikici, et al. Microvascular Research 131 (2020) 104035

Fig. 5. CAM assay results showing the angiogenic activity of 2dDR in comparison with VEGF and PBS. (***p ≤ 0.001, **p ≤ 0.01, *p ≤ 0.05, n = 3). Scale bars represent 2 mm.

2dDR loaded alginate dressings. However, no other sugar molecules, VEGF-165, VEGF-121 and VEGF-145 can be secreted extracellularly including 2dLR and DG, were found to be as effective in terms of in- whereas VEGF-189 and VEGF-206 are stored in ECM and need to be creasing the metabolic activity of HAECs and promoting angiogenesis in released by their cleavage from the ECM (Holmes and Zachary, 2005). CAM assay over 7 days. The VEGF quantification results suggested that 2dDR plays a key We have previously reported the suitability of small-diameter blood role in the stimulation of angiogenesis by increasing the production of vessel cells to be used in several angiogenesis applications such as to VEGF by HAECs. No significant increase in the amount of VEGF that create endothelial monolayers to study angiogenesis (Dew et al., 2016; HAECs produced was observed when 2dLR and DG at any concentra- Dikici et al., 2020b; Ortega et al., 2015), to promote angiogenesis as a tions were released from alginate dressings. The VEGF production by prevascularisation approach (Dikici et al., 2019; Dikici et al., 2020a). HAECs in response to direct treatment with 2dDR was statistically Although they have been found to be very suitable for many angio- significant even at day 1. The amount of VEGF increased by approxi- genesis applications, they were also very difficult to grow under la- mately 48-fold and 70-fold, respectively, with the addition of 100 μM boratory conditions and required surface coating or the presence of a and 1 mM 2dDR. At the end of day 5, VEGF production of HAECs was periendothelial type of cell (as a helper cell) to attach and grow on increased by 74-fold and 80-fold, respectively with the addition of synthetic (Dew et al., 2016; Dikici et al., 2020b) or natural (Dikici et al., 100 μM and 1 mM 2dDR. Similarly, the release of 10% 2dDR from al- 2019b) polymeric surfaces. On the other hand, we have recently re- ginate dressings increased the amount of VEGF approximately 30-fold ported that HAECs were very easy to grow on tissue culture plastic as and 48-fold by days 3 and 7, respectively, when compared to control well as polymeric surfaces, and they respond positively to the treatment dressings. of pro-angiogenic agents and flow in terms of increased proliferation, The stability studies demonstrated that there was a gradual decrease migration and tube formation (Dikici et al., 2019a). When the aim of in the amount of 2dDR present in the growth medium over 14 days, this study and the pros and cons of small-diameter blood vessel cells either in the presence or absence of HAECs. This decrease was greater and big diameter blood vessel cells were considered, we selected HAECs when HAECs were present in the well plates. This is likely to be due to to be used to investigate whether there is any evidence for VEGF pro- the 2dDR being internalised by HAECs and used to increase the production in response to 2dDR treatment. duction of VEGF. At the end of day 14, approximately 40% to 50% of The VEGF-ELISA kit used in this study has a reactivity of 100% to the 2dDR had disappeared from the culture medium in the presence of VEGF-165 and 7.2% to VEGF-121 and is specific to bind recombinant HAECs. The drop in 2dDR amount in the culture medium over 7 days is human VEGF-A. Although there are several isoforms of VEGF-A asso- consistent with the decrease in the amount of VEGF produced by HAECs ciated with the angiogenic cascade (i.e. VEGF-121 VEGF-145, VEGF- in response to 2dDR treatment from day 4 to 7 (Fig. 3). Thus, the ad- 165, VEGF-189, and VEGF-206) VEGF-165 has been demonstrated as dition of 2dDR leads to a temporary increase in VEGF over a few days. It the prototypical pro-angiogenic VEGF-A isoform because of its optimal does not lead to a permanent or sustained increase which might be bioavailability and biological activity and is consequently widely used worrying because of the known association between the addition of in studies focusing on promoting angiogenesis (Ferrara, 2001, 2009; exogenous VEGF and the production of leaky blood vessels such as Soker et al., 1998). A major difference between these isoforms is that those found in tumours. The fact that the levels of 2dDR and of VEGF

7 S. Dikici, et al. Microvascular Research 131 (2020) 104035

Fig. 6. Evaluation of the angiogenesis on CAM when 2dDR (5% and 10%), 2dLR (5% and 10%), and DG (10%) loaded alginate dressings were implanted. The graph shows the quantification of the total length of small blood vessels. (***p ≤ 0.001, **p ≤ 0.01, *p ≤ 0.05, not significant (ns) p ≥ 0.05, n = 4). Scale bars represent 1 mm. were increased for only a few days strongly suggests that both are 2002). Nakajima et al. reported that 2dLR inhibits VEGF production subject to endogenous regulation in these studies. The most probable and TP-related angiogenesis in metastatic nodules (Nakajima et al., explanation for the reduction in the amount of VEGF is the low stability 2004) and inhibits Matrigel invasion of tumours (Nakajima et al., of VEGF as solution at RT (The manufacturer suggests that reconstituted 2006). Ikeda et al. showed that 2dLR promotes hypoxia-induced VEGF could be stored at 2–8 °C for up to one week). In addition, the apoptosis of HL-60 cells (Ikeda et al., 2002). Uchimiya et al. demon- density of immobilized VEGF (as a solution) has been reported to de- strated the inhibition of angiogenesis in a rat corneal assay by 2dLR crease by almost 40% in 5 days (Mizumachi and Ijima, 2013). Another (Uchimiya et al., 2002). Although all these studies reported the anti- possible reason is the rate of VEGF uptake by HAECs is faster than the angiogenic activity of 2dLR, Yar et al. previously showed the promotion production of it. Also, since the amount of 2dDR in media reduces of angiogenesis by the release of 2dLR from hydrogels in CAM assay significantly from day 3 to 7, 2dDR stimulation of the production of (Yar et al., 2017). Similarly, Sengupta et al. also suggested that 2dLR VEGF will reduce with time. showed an angiogenic response in sponge granuloma model of angio- In contrast to VEGF, the literature on small sugars is very limited genesis (Sengupta et al., 2003). Similar conflicting literature can be and conflicting. Uchimiya et al. demonstrated that TP-induced angio- seen for DG, which has previously been reported to induce prolifera- genesis could be reversed by the addition of 2dLR (Uchimiya et al., tion, migration and tube formation of ECs by many groups (Madonna

8 S. Dikici, et al. Microvascular Research 131 (2020) 104035

Table 1 Summary of the studies on some small sugars that have been found to promote or inhibit angiogenesis.

Sugars Angiogenesis assays Results Eﬀective doses Reference

2-Deoxy-D-ribose In vitro Promotes proliferation, migration and tube formation of ECs 100 μM to 1 mM (Dikici et al., 2019a) (2dDR) Inhibits hypoxia-induced apoptosis 10 μM(Ikeda et al., 2002) Induces Matrigel invasion 100 μM(Nakajima et al., 2006) Stimulates proliferation and migration of ECs 100 μM to 1 mM (Sengupta et al., 2003) Activates NOX2 which triggers NF-κB and upregulates VEGFR2 8 μM to 1 mM (Vara et al., 2018) CAM assay Promotes angiogenesis 200 μg/day (Dikici et al., 2019c) 250 μg/1 g of polymer (Dikici et al., 2019c) 1 mg/ml (Yar et al., 2017) In vivo Promotes angiogenesis and wound healing 1 mg/ml (Yar et al., 2017) 5% or 10% (w/v) in the (Azam et al., 2019) dressing Promotes angiogenesis 2 nmol (Sengupta et al., 2003) 2-Deoxy-L-ribose In vitro Suppresses migration and tube formation of ECs 100 μM(Nakajima et al., 2009; Uchimiya et al., (2dLR) 2002) Inhibits VEGF production 10 μM to 100 μM(Nakajima et al., 2004) Promotes hypoxia-induced apoptosis 30 μMto50μM(Ikeda et al., 2002) Inhibits Matrigel invasion of tumours 100 μM(Nakajima et al., 2006) CAM assay Stimulates angiogenesis 1 mg/ml (Yar et al., 2017) In vivo Inhibits angiogenesis in a rat corneal assay 200 ng/pellet (Uchimiya et al., 2002) Promotes angiogenesis in mouse sponge angiogenesis model 2 nmol (Sengupta et al., 2003) 2-Deoxy-D-glucose In vitro Inhibits the proliferation, migration and tube formation of ECs, 60 μM to 9 mM (Merchan et al., 2010) (2dDG) and induces endothelial apoptosis Inhibits proliferation of cells and reduces ATP levels 3 mM (Tagg et al., 2008) Inhibits the proliferation, migration and tube formation of ECs 50 μM to 1 mM (Chuang et al., 2015) Downregulates AKT and ERK pathways and inhibits tube 600 μM(Kovacs et al., 2016) formation of ECs Rat aortic ring Inhibits tube formation of ECs 50 μM to 1 mM (Chuang et al., 2015) In vivo Inhibits angiogenesis 6 mM (Merchan et al., 2010) D-Glucose In vitro Induces migration and tube formation of ECs 25 mM (Shigematsu et al., 1999) (DG) Inhibits proliferation, migration and tube formation of ECs in a 5mMto30mM (Yu et al., 2006) dose-dependent manner Promotes tube formation of ECs and increases COX-2 expression 25 mM to 30.5 mM (Madonna et al., 2016) Inhibits the tube formation of ECs in a dose-dependent manner 10 mM to 16 mM (Jiraritthamrong et al., 2012) In vivo Reduced angiogenesis 22 mM (Teixeira and Andrade, 1999) et al., 2016; Shigematsu et al., 1999; Yu et al., 2006). However, several concentration is due to extracellular secretion or release by the cleavage other reports assert that DG inhibits tube formation in vitro of ECM. (Jiraritthamrong et al., 2012) and is anti-angiogenic in vivo (Teixeira and Andrade, 1999). 5. Conclusion In order to better consider this conflicting literature, we summarised the reported studies on these small sugars that have been found to In conclusion, we aimed to contribute to the knowledge of how promote or inhibit angiogenesis in Table 1. 2dDR promotes angiogenesis specifically by determining whether its It was not clear how 2dDR stimulates angiogenesis. Several re- pro-angiogenic effect is VEGF-dependent or not. We showed that 2dDR searchers have attempted to clarify the role of 2dDR at the molecular was angiogenic in vitro and in an ex-ovo CAM assay either with its direct level in stimulating angiogenesis. The first proposed mechanism was application or when released from an alginate dressing. No other small the promotion of oxidative stress by the endogenous generation of sugar molecules, including 2dLR and DG, were found to induce an- 2dDR by enzymatic degradation of thymidine to thymine. This has been giogenesis in vitro and in vivo. The VEGF quantification results de- thought to promote the production and secretion of several pro-an- monstrated that only 2dDR increased VEGF production of ECs when giogenic agents such as VEGF and interleukin-8 (IL-8) by cells which added directly into the growth medium or when released from com- can then be internalised by the surrounding ECs to trigger the angio- mercially available alginate dressings. Other small sugars such as 2dLR genic pathways (Brown et al., 2000; Nakajima et al., 2009; Sengupta and DG did not increase VEGF production. The 2dDR stability studies et al., 2003). Alternatively, Vara et al. suggested that 2dDR can be showed that approximately 40–50% of the 2dDR disappeared from the generated either endogenously by cells with thymidine phosphorylase media over 14 days, either in the presence or absence of HAECs, and the (TP) activity and then be secreted to the extracellular environment or reduction was higher when cells were present. In parallel with the re- TP could be released from injured or dead cells to generate 2dDR by the duction in 2dDR over time, the presence of VEGF was also found to degradation of thymidine to thymine. They reported that 2dDR acti- decrease in the media after day 4. In conclusion, the simplest ex- vates NADPH oxidase 2 (NOX2) and triggers the nuclear transcription planation for our results is that 2dDR achieves angiogenesis by upre- factor kappa B (NF-κB), which then upregulates VEGFR2 to promote gulating the production of VEGF. angiogenesis (Vara et al., 2018). Thus, there is small literature in this area which suggests at least Ethical approval and consent to participate two routes to affect VEGF production or upregulation of VEGFR-2, and these are not mutually exclusive. The current study provides an im- The authors state that all CAM assay experiments were carried out portant contribution to our knowledge strongly suggesting that the according to the Home Office, UK guidelines. angiogenic activity of 2dDR is strongly associated with the production of VEGF. In the future, experiments need to be carried out to explore Consent for publication which specific VEGF-A isoforms are being affected by the treatment of 2dDR, which will reveal whether the increase in the VEGF Not applicable.

9 S. Dikici, et al. Microvascular Research 131 (2020) 104035

Availability of data and material Chuang, I.C., Yang, C.M., Song, T.Y., Yang, N.C., Hu, M.L., 2015. The anti-angiogenic action of 2-deoxyglucose involves attenuation of VEGFR2 signaling and MMP-2 expression in HUVECs. Life Sci. 139, 52–61. https://doi.org/10.1016/j.lfs.2015.08.002. The datasets generated and/or analysed during the current study are Dew, L., MacNeil, S., Chong, C.K., 2015. Vascularization strategies for tissue engineers. available from the corresponding author on reasonable request. Regen. Med. 10, 211–224. https://doi.org/10.2217/rme.14.83. Dew, L., English, W.R., Ortega, I., Claeyssens, F., MacNeil, S., 2016. Fabrication of biodegradable synthetic vascular networks and their use as a model of angiogenesis. Funding Cells Tissues Organs 202, 319–328. https://doi.org/10.1159/000446644. Dikici, S., Aldemir Dikici, B., Bhaloo, S.I., Balcells, M., Edelman, E.R., MacNeil, S., Reilly, Serkan Dikici was funded by Republic of Turkey - Ministry of G.C., Sherborne, C., Claeyssens, F., 2019a. Assessment of the angiogenic potential of National Education for the duration of his PhD. Financial support for 2-deoxy-D-ribose using a novel in vitro 3D dynamic model in comparison with established in vitro assays. Front. Bioeng. Biotechnol. 7, 451. https://doi.org/10.3389/ Dr. Anthony Bullock was provided by Cannenta Pty Ltd., Australia. fbioe.2019.00451. Dikici, S., Claeyssens, F., MacNeil, S., 2019b. Decellularised baby spinach leaves and their Author contributions potential use in tissue engineering applications: studying and promoting neovascularisation. J. Biomater. Appl. 34, 546–559. https://doi.org/10.1177/ 0885328219863115. The experimental study was designed by all the authors. Serkan Dikici, S., Mangir, N., Claeyssens, F., Yar, M., MacNeil, S., 2019c. Exploration of 2-deoxy- β Dikici and Dr. Anthony J Bullock contributed to performing the ma- D-ribose and 17 -estradiol as alternatives to exogenous VEGF to promote angiogenesis in tissue-engineered constructs. Regen. Med. 14, 179–197. https://doi.org/ jority of the research experiments and data analysis and generating/ 10.2217/rme-2018-0068. improving the manuscript, figures and tables. Professor Sheila MacNeil, Dikici, S., Claeyssens, F., MacNeil, S., 2020a. Pre-seeding of simple electrospun scaffolds Dr. Frederik Claeyssens, and Dr. Muhammad Yar contributed to this with a combination of endothelial cells and fibroblasts strongly promotes angiogenesis. Tissue Eng. Regen. Med. https://doi.org/10.1007/s13770-020-00263-7. in work with their supervision and critical revision of the manuscript for press. important intellectual content. Dikici, S., Claeyssens, F., MacNeil, S., 2020b. Bioengineering vascular networks to study angiogenesis and vascularization of physiologically relevant tissue models in vitro. ACS Biomater. Sci. Eng. 6, 3513–3528. https://doi.org/10.1021/acsbiomaterials. Declaration of competing interest 0c00191. Ferrara, N., 2001. Role of vascular endothelial growth factor in regulation of physiolo- – The authors declare that they have no competing interests. gical angiogenesis. Am. J. Physiol. - Cell Physiol. 280, C1358 C1366. Ferrara, N., 2007. Angiogenesis: From Basic Science to Clinical Application. CRC Press Taylor & Francis Group, New York. Acknowledgement Ferrara, N., 2009. VEGF-A: a critical regulator of blood vessel growth. Eur. Cytokine Netw. 20, 158–163. https://doi.org/10.1684/ecn.2009.0170. Fukumura, D., Gohongi, T., Kadambi, A., Izumi, Y., Ang, J., Yun, C.O., Buerk, D.G., We are grateful to the Republic of Turkey - Ministry of National Huang, P.L., Jain, R.K., 2001. Predominant role of endothelial nitric oxide synthase in Education for the funding of a PhD award to Serkan Dikici. We grate- vascular endothelial growth factor-induced angiogenesis and vascular permeability. fully acknowledge Cannenta Pty Ltd., Australia for funding Dr. Anthony Proc. Natl. Acad. Sci. U. S. A. 98, 2604–2609. https://doi.org/10.1073/pnas. 041359198. J Bullock. Haraguchi, M., Miyadera, K., Uemura, K., Sumizawa, T., Furukawa, T., Yamada, K., Akiyama, S., Yamada, Y., 1994. Angiogenic activity of enzymes. Nature 368, 198. Appendix A. Supplementary data Hoeben, A.N.N., Landuyt, B., Highley, M.S.M., Wildiers, H., Oosterom, A.T.V.A.N., Bruijn, E.A.D.E., Van Oosterom, A.T., De Bruijn, E.A., 2004. Vascular endothelial growth factor and angiogenesis. Pharmacol. Rev. 56, 549–580. https://doi.org/10.1124/pr. Supplementary data to this article can be found online at https:// 56.4.3.549. doi.org/10.1016/j.mvr.2020.104035. Holmes, D.I.R., Zachary, I., 2005. The vascular endothelial growth factor (VEGF) family: Angiogenic factors in health and disease. Genome Biol. 6, 209. https://doi.org/10. 1186/gb-2005-6-2-209. References Holmqvist, K., Cross, M., Riley, D., Welsh, M., 2003. The Shb adaptor protein causes Src- dependent cell spreading and activation of focal adhesion kinase in murine brain endothelial cells. Cell. Signal. 15, 171–179. https://doi.org/10.1016/S0898- Aldemir Dikici, B., Dikici, S., Reilly, G.C., MacNeil, S., Claeyssens, F., 2019. A novel bi- 6568(02)00076-1. layer polycaprolactone membrane for guided bone regeneration: combining elec- Ikeda, R., Furukawa, T., Kitazono, M., Ishitsuka, K., Okumura, H., Tani, A., Sumizawa, T., trospinning and emulsion templating. Materials (Basel). 12, 2643. https://doi.org/ Haraguchi, M., Komatsu, M., Uchimiya, H., Ren, X.Q., Motoya, T., Yamada, K., 10.3390/ma12162643. Akiyama, S.I., 2002. Molecular basis for the inhibition of hypoxia-induced apoptosis Aldemir Dikici, B., Reilly, G.C., Claeyssens, F., 2020. Boosting the osteogenic and an- by 2-deoxy-D-ribose. Biochem. Biophys. Res. Commun. 291, 806–812. https://doi. giogenic performance of multiscale porous polycaprolactone scaffolds by in vitro org/10.1006/bbrc.2002.6432. generated extracellular matrix decoration. ACS Appl. Mater. Interfaces 12, Jiraritthamrong, C., Kheolamai, P., U-Pratya, Y., Chayosumrit, M., Supokawej, A., 12510–12524. https://doi.org/10.1021/acsami.9b23100. Manochantr, S., Tantrawatpan, C., Sritanaudomchai, H., Issaragrisil, S., 2012. In vitro Andleeb, A., Dikici, S., Waris, T.S., Bashir, M.M., Akhter, S., Chaudhry, A.A., MacNeil, S., vessel-forming capacity of endothelial progenitor cells in high glucose conditions. Yar, M., 2020. Developing affordable and accessible pro-angiogenic wound dressings; Ann. Hematol. 91, 311–320. https://doi.org/10.1007/s00277-011-1300-6. incorporation of 2 deoxy D-ribose (2dDR) into cotton fibres and wax-coated cotton Koch, S., Tugues, S., Li, X., Gualandi, L., Claesson-Welsh, L., 2011. Signal transduction by fibres. J. Tissue Eng. Regen. Med. in press. https://doi.org/10.1002/term.3072. vascular endothelial growth factor receptors. Biochem. J. 437, 169–183. https://doi. Azam, M., Dikici, S., Roman, S., Mehmood, A., Chaudhry, A.A., Rehman, I., MacNeil, S., org/10.1042/BJ20110301. Yar, M., 2019. Addition of 2-deoxy-d-ribose to clinically used alginate dressings sti- Kovacs, K., Decatur, C., Toro, M., Pham, D.G., Liu, H., Jing, Y., Murray, T.G., Lampidis, mulates angiogenesis and accelerates wound healing in diabetic rats. J. Biomater. T.J., Merchan, J.R., 2016. 2-deoxy-glucose downregulates endothelial AKT and ERK Appl. 34, 463–475. https://doi.org/10.1177/0885328219859991. via interference with N-linked glycosylation, induction of endoplasmic reticulum Brooks, P., Montgomery, A.P., Cheresh, D., 1999. Use of the 10-day-old chick embryo stress, and GSK3β activation. Mol. Cancer Ther. 15, 264–275. https://doi.org/10. model for studying angiogenesis. Integrin Protoc. 129, 257–269. https://doi.org/10. 1158/1535-7163.MCT-14-0315. 1385/1-59259-249-X:257. Lamalice, L., Houle, F., Huot, J., 2006. Phosphorylation of Tyr 1214 within VEGFR-2 Brown, N.S., Jones, A., Fujiyama, C., Harris, A.L., Bicknell, R., 2000. Thymidine phos- triggers the recruitment of Nck and activation of Fyn leading to SAPK2/p38 activa- phorylase induces carcinoma cell oxidative stress and promotes secretion of angio- tion and endothelial cell migration in response to VEGF. J. Biol. Chem. 281, genic factors. Cancer Res. 60, 6298–6302. 34009–34020. https://doi.org/10.1074/jbc.M603928200. Cao, R., Eriksson, A., Kubo, H., Alitalo, K., Cao, Y., Thyberg, J., 2004. Comparative Langer, R., 2000. Tissue engineering. Mol. Ther. 1, 12–15. https://doi.org/10.1006/ evaluation of FGF-2-, VEGF-A-, and VEGF-C-induced angiogenesis lymphangiogen- mthe.1999.0003. esis, vascular fenestrations, and permeability. Circ. Res. 94, 664–670. https://doi. Madonna, R., Giovannelli, G., Confalone, P., Renna, F.V., Geng, Y.J., De Caterina, R., org/10.1161/01.RES.0000118600.91698.BB. 2016. High glucose-induced hyperosmolarity contributes to COX-2 expression and Cardone, M.H., Roy, N., Stennicke, H.R., Salvesen, G.S., Franke, T.F., Stanbridge, E., angiogenesis: implications for diabetic retinopathy. Cardiovasc. Diabetol. 15, 18. Frisch, S., Reed, J.C., 1998. Regulation of cell death protease caspase-9 by phos- https://doi.org/10.1186/s12933-016-0342-4. phorylation. Science 282, 1318–1321. https://doi.org/10.1126/science.282.5392. Mangir, N., Dikici, S., Claeyssens, F., MacNeil, S., 2019. Using ex-ovo chick chor- 1318. ioallantoic membrane (CAM) assay to evaluate the biocompatibility and angiogenic Cheng, S.Y., Nagane, M., Huang, H.S., Cavenee, W.K., 1997. Intracerebral tumor-asso- response to biomaterials. ACS Biomater. Sci. Eng. 5, 3190–3200. https://doi.org/10. ciated hemorrhage caused by overexpression of the vascular endothelial growth 1021/acsbiomaterials.9b00172. factor isoforms VEGF121 and VEGF165 but not VEGF189. Proc. Natl. Acad. Sci. U. S. Matsushita, S., Nitanda, T., Furukawa, T., Sumizawa, T., Tani, A., Nishimoto, K., Akiba, A. 94, 12081–12087. https://doi.org/10.1073/pnas.94.22.12081. S., Miyadera, K., Fukushima, M., Yamada, Y., Yoshida, H., Kanzaki, T., Akiyama, S.I.,

10 S. Dikici, et al. Microvascular Research 131 (2020) 104035

1999. The effect of a thymidine phosphorylase inhibitor on angiogenesis and apop- vascularised tissues in vitro. Eur. Cells Mater. 15, 27–40. doi:vol015a03 [pii] ET - tosis in tumors. Cancer Res. 59, 1911–1916. 2008/02/22. Meadows, K.N., Bryant, P., Pumiglia, K., 2001. Vascular endothelial growth factor in- Roma-Rodrigues, C., Heuer-Jungemann, A., Fernandes, A.R., Kanaras, A.G., Baptista, duction of the angiogenic phenotype requires Ras activation. J. Biol. Chem. 276, P.V., 2016. Peptide-coated gold nanoparticles for modulation of angiogenesis in vivo. 49289–49298. https://doi.org/10.1074/jbc.M108069200. Int. J. Nanomedicine 11, 2633–2639. https://doi.org/10.2147/IJN.S108661. Merchan, J.R., Kovács, K., Railsback, J.W., Kurtoglu, M., Jing, Y., Piña, Y., Gao, N., Sengupta, S., Sellers, L.A., Matheson, H.B., Fan, T.P.D., 2003. Thymidine phosphorylase Murray, T.G., Lehrman, M.A., Lampidis, T.J., 2010. Antiangiogenic activity of 2- induces angiogenesis in vivo and in vitro: an evaluation of possible mechanisms. Br. J. deoxy-D-glucose. PLoS One 5, e13699. https://doi.org/10.1371/journal.pone. Pharmacol. 139, 219–231. https://doi.org/10.1038/sj.bjp.0705216. 0013699. Shigematsu, S., Yamauchi, K., Nakajima, K., Iijima, S., Aizawa, T., Hashizume, K., 1999. Mizumachi, H., Ijima, H., 2013. Measuring stability of vascular endothelial growth factor IGF-1 regulates migration and angiogenesis of human endothelial cells. Endocr. J. 46, using an immobilization technique. Adv. Biomed. Eng. 2, 130–136. doi:10.14326/ 59–62. https://doi.org/10.1507/endocrj.46.suppl_s59. abe.2.130. Soker, S., Takashima, S., Miao, H.Q., Neufeld, G., Klagsbrun, M., 1998. Neuropilin-1 is Nakajima, Y., Gotanda, T., Uchimiya, H., Furukawa, T., Haraguchi, M., Ikeda, R., expressed by endothelial and tumor cells as an isoform- specific receptor for vascular Sumizawa, T., Yoshida, H., Akiyama, S.I., 2004. Inhibition of metastasis of tumor endothelial growth factor. Cell 92, 735–745. https://doi.org/10.1016/S0092- cells overexpressing thymidine phosphorylase by 2-deoxy-L-ribose. Cancer Res. 64, 8674(00)81402-6. 1794–1801. https://doi.org/10.1158/0008-5472.CAN-03-2597. Tagg, S.L.C., Foster, P.A., Leese, M.P., Potter, B.V.L., Reed, M.J., Purohit, A., Newman, S. Nakajima, Y., Haraguchi, M., Furukawa, T., Yamamoto, M., Nakanishi, H., Tatematsu, M., P., 2008. 2-Methoxyoestradiol-3,17-O,O-bis-sulphamate and 2-deoxy-D-glucose in Akiyama, S.I., 2006. 2-Deoxy-L-ribose inhibits the invasion of thymidine phosphor- combination: A potential treatment for breast and prostate cancer. Br. J. Cancer 99, ylase- overexpressing tumors by suppressing matrix metalloproteinase-9. Int. J. 1842–1848. doi:https://doi.org/10.1038/sj.bjc.6604752. Cancer 119, 1710–1716. https://doi.org/10.1002/ijc.22014. Teixeira, A.S., Andrade, S.P., 1999. Glucose-induced inhibition of angiogenesis in the rat Nakajima, Y., Madhyastha, R., Maruyama, M., 2009. 2-deoxy-D-ribose, a downstream sponge granuloma is prevented by aminoguanidine. Life Sci. 64, 655–662. https:// mediator of thymidine phosphorylase. Regulates Tumor Angiogenesis and doi.org/10.1016/S0024-3205(98)00607-9. Progression. Anticancer. Agents Med. Chem. 9, 239–245. https://doi.org/10.2174/ Uchimiya, H., Furukawa, T., Okamoto, M., Nakajima, Y., Matsushita, S., Ikeda, R., 187152009787313846. Gotanda, T., Haraguchi, M., Sumizawa, T., Ono, M., Kuwano, M., Kanzaki, T., Obi, N., Toda, H., 2015. Human umbilical vein endothelial cells migration in matrigel by Akiyama, S.I., 2002. Suppression of thymidine phosphorylase-mediated angiogenesis the concentration gradient of vascular endothelial growth factor. J. Biotechnol. and tumor growth by 2-deoxy-L-ribose. Cancer Res. 62, 2834–2839. Biomater. 5, 210. https://doi.org/10.4172/2155-952x.1000210. Vara, D., Watt, J.M., Fortunato, T.M., Mellor, H., Burgess, M., Wicks, K., Mace, K., Oka, N., Soeda, A., Inagaki, A., Onodera, M., Maruyama, H., Hara, A., Kunisada, T., Mori, Reeksting, S., Lubben, A., Wheeler-Jones, C.P.D., Pula, G., 2018. Direct activation of H., Iwama, T., 2007. VEGF promotes tumorigenesis and angiogenesis of human NADPH oxidase 2 by 2-deoxyribose-1-phosphate triggers nuclear factor kappa B- glioblastoma stem cells. Biochem. Biophys. Res. Commun. 360, 553–559. https://doi. dependent angiogenesis. Antioxid. Redox Signal. 28, 110–130. https://doi.org/10. org/10.1016/j.bbrc.2007.06.094. 1089/ars.2016.6869. Olsson, A.K., Dimberg, A., Kreuger, J., Claesson-Welsh, L., 2006. VEGF receptor signalling Vernon, R.B., Sage, E.H., 1999. A novel, quantitative model for study of endothelial cell - in control of vascular function. Nat. Rev. Mol. Cell Biol. 7, 359–371. https://doi. migration and sprout formation within three-dimensional collagen matrices. org/10.1038/nrm1911. Microvasc. Res. 57, 118–133. https://doi.org/10.1006/mvre.1998.2122. Ortega, I., Dew, L., Kelly, A.G., Chong, C.K., MacNeil, S., Claeyssens, F., 2015. Fabrication Wang, S., Li, X., Parra, M., Verdin, E., Bassel-Duby, R., Olson, E.N., 2008. Control of of biodegradable synthetic perfusable vascular networks via a combination of elec- endothelial cell proliferation and migration by VEGF signaling to histone deacetylase trospinning and robocasting. Biomater. Sci. 3, 592–596. https://doi.org/10.1039/ 7. Proc. Natl. Acad. Sci. U. S. A. 105, 7738–7743. https://doi.org/10.1073/pnas. c4bm00418c. 0802857105. Parsons, J.T., 2003. Focal adhesion kinase: the first ten years. J. Cell Sci. 116, 1409–1416. Yancopoulos, G.D., Davis, S., Gale, N.W., Rudge, J.S., Wiegand, S.J., Holash, J., 2000. https://doi.org/10.1242/jcs.00373. Vascular-specific growth factors and blood vessel formation. Nature 407, 242–248. Patterson, J., Mura, C., 2013. Rapid colorimetric assays to qualitatively distinguish RNA https://doi.org/10.1038/35025215. and DNA in biomolecular samples. J. Vis. Exp. 72, e50225. Yar, M., Shahzadi, L., Mehmood, A., Raheem, M.I., Román, S., Chaudhry, A.A., ur Poldervaart, M.T., Gremmels, H., Van Deventer, K., Fledderus, J.O., Öner, F.C., Verhaar, Rehman, I., Ian Douglas, C.W., MacNeil, S., 2017. Deoxy-sugar releasing biode- M.C., Dhert, W.J.A., Alblas, J., 2014. Prolonged presence of VEGF promotes vascu- gradable hydrogels promote angiogenesis and stimulate wound healing. Mater. larization in 3D bioprinted scaffolds with defined architecture. J. Control. Release Today Commun. 13, 295–305. https://doi.org/10.1016/j.mtcomm.2017.10.015. 184, 58–66. https://doi.org/10.1016/j.jconrel.2014.04.007. Yu, P., Yu, D.M., Qi, J.C., Wang, J., Zhang, Q.M., Zhang, J.Y., Tang, Y.Z., Xing, Q.L., Li, Ribatti, D., Nico, B., Vacca, A., Presta, M., 2006. The gelatin sponge-chorioallantoic M.Z., 2006. High D-glucose alters PI3K and Akt signaling and leads to endothelial cell membrane assay. Nat. Protoc. 1, 85–91. https://doi.org/10.1038/nprot.2006.13. migration, proliferation and angiogenesis dysfunction. Natl. Med. J. China 86, Richardson, T.P., Peters, M.C., Ennett-Shepard, A.B., Mooney, D.J., 2001. Polymeric 3425–3430. system for dual growth factor delivery. Nat. Biotechnol. 19, 1029–1034. https://doi. Zachary, I., Gliki, G., 2001. Signaling transduction mechanisms mediating biological org/10.1038/nbt1101-1029. actions of the vascular endothelial growth factor family. Cardiovasc. Res. 49, Rivron, N.C., Liu, J., Rouwkema, J., De Boer, J., Van Blitterswijk, C.A., 2008. Engineering 568–581. https://doi.org/10.1016/S0008-6363(00)00268-6.