Improvement of Endothelial Cell-Polycaprolactone Interaction through Surface Modification via Aminolysis, Hydrolysis, and a Combined Approach

,


Introduction
Coronary heart disease is the most common type of heart disease and the foremost single cause of mortality and loss of disability adjusted for life years globally [1,2].Coronary bypass surgery is one of the most frequent treatments for coronary disease using vascular grafts to replace damaged blood vessels [3,4].Ideally, autologous artery segments are used for this surgery as they are both compliant and nonthrombogenic [4,5].However, autologous grafts are not always available or suitable to use as grafts due to occlusion or pathological changes, specifcally in patients sufering from peripheral vascular disease [5][6][7].Te large number of patients in need of cardiovascular grafts and the associated requirements thus create a demand for alternative solutions [8].
Current synthetic commercial vascular grafts, especially small-diameter grafts, are associated with a high incidence of failure due to thrombus formation and infection [5,9,10].Vascular graft infections are rare, but they are associated with mortality rates between 26 and 55%, depending on the graft location and surgical procedure [11].Furthermore, Staphylococcus aureus and Escherichia coli are collectively responsible for a signifcant part of all medical device-associated infections [12].Tis results in a growing demand to fabricate small-diameter vascular grafts with a diameter under 6 mm with improved biocompatibility and adequate mechanical properties that do not inherit a risk for bacterial infection [9,13].
Biodegradable materials are preferable because the material degrades over time while cells infltrate the matrix, producing collagen, elastin, and proteoglycans and replacing the degrading material [8].Tus, a fully functional vessel is created, constituted of autologous tissue, smooth muscle cells, and endothelial cells [14].Importantly, the material should degrade over years to ensure a prolonged mechanical support for infltrating cells while keeping the risk of an infammatory response low and to avoid rupture and aneurysms of the vessel [8,14].Furthermore, the graft should not increase the risk for bacterial attachment to avoid vascular graft infections.
Polycaprolactone (PCL) is a biodegradable and biocompatible polyester used in tissue engineering [15,16].It is characterized by a rubbery state at a physiological temperature of 37 °C leading to superior mechanical properties over other polyesters [17,18].In addition, PCL degrades in the physiological environment in vivo and can undergo hydrolytic degradation through bulk or surface degradation mechanisms [15,[19][20][21].Another advantage of PCL scaffolds is the long degradation time of more than 18 months in vivo due to their hydrophobic nature and high level of crystallinity [17,19,22].However, PCL has a poor cell affnity due to its high hydrophobicity and lack of cell-binding sites [16].Tus, to use PCL as a scafold material, its hydrophilicity needs to be increased to allow cellular compatibility with endothelial cells forming a neo-endothelial layer at the surface of the graft.
Te cellular compatibility of grafts is primarily determined by the surface properties, including morphology, topography, chemical structure, and functional groups [23,24].Tese surface properties can be changed by multiple methods including treatment with drugs, peptides, growth factors, or adhesive proteins, or chemical reactions such as hydrolysis and aminolysis [15,[25][26][27][28][29][30][31].Te advantage of chemical surface modifcation is the possibility of combining it with other treatments due to its simple and short procedural steps.For PCL, hydrolysis in an alkaline medium results in the formation of carboxylates and hydroxyl groups on the graft surface [32].Aminolysis is another option to functionalize the graft surface by a fast reaction that creates amide bindings and amino groups on the surface [15,32].
Te aim of the present study was to compare the efect of diferent surface treatments to increase the endothelialization of PCL.Both hydrolysis and aminolysis, as well as a combination thereof, were used to treat PCL flms.Te efects of these modifcations on the hydrophilicity and surface morphology of the PCL flms were investigated and the adhesion and survival of endothelial cells to the PCL flms were studied.Te bacterial adhesion was examined to ensure that the surface treatment does not lead to an increase in bacterial infectivity.Te adequate surface modifcation would bring PCL a step closer to being a promising biomaterial for vascular prosthesis and other applications.

PCL Film Preparation.
A PCL solution was made by dissolving PCL (Sigma-Aldrich, 440744) in tetrahydrofuran (Merck KgaA, 107025) (10% w/v).Four ml of this solution were poured in a fat-bottomed glass box of 70 × 105 × 85 mm (Heinz Herenz, A11520).To ensure a homogeneous layer, the glass box was shaken gently while the PCL flm dried.Te resulting PCL flm was cut into disks (d � 16 mm) using a circular biopsy punch.Te conditions were selected based on existing literature and adapted, if necessary, such that comparison between the diferent treatment methods was possible [15,[33][34][35][36][37].

Surface
After treatment, the PCL flms were washed twice with distilled water and left to air dry overnight before further analysis.

Contact Angle Measurement.
To analyze the water contact angle, a Drop and Surface Analyser OCA50 (DataPhysics Instruments GmbH) was used.Te PCL flm samples were fxed on the support plate using tape.A droplet of water was placed on top of the samples and pictures were taken of the droplet and surface.Image analysis was done using edge detection to identify the liquid/gas interface.Te Pendent Drop software was utilized to measure the contact angle from the shape of the drop according to the Young-Laplace equation.

Mass Loss Measurement.
To analyze whether the hydrolysis reaction afected the mass of PCL flms, the flms were weighed before and after treatment using an analytical balance (Kern, ALJ 120-4).Te PCL flms were dried at 37 °C overnight and weighed (W0).Tey were then processed for surface modifcation as described above.Afterwards, the To seed HUVECs on the PCL flms, a custom in vitro model was used (designed by Dr. Vogel [38]).Each PCL disk was placed in the center of a well using a 6-well plate.A plexiglass inlet with an opening of 6 mm diameter was put on top of the flm, and the system was immobilized by placing the well plate in a metal frame (Figure 1).After immobilization, the setup was UV sterilized for 2 h.HUVECs were seeded at a concentration of 145,000 cells/ cm 2 onto the PCL flms and kept at 37 °C for 24 or 72 h as noted.Te medium was replaced each day.As a positive control, HUVECs were seeded at the same concentration in a 1% gelatin (Sigma-Aldrich, G2500) coated 96-well plate.

Endothelial Cell Staining by DAPI and Phalloidin.
After culture, PCL flms seeded with HUVECs were washed with Dulbecco's phosphate-bufered saline (DPBS, pH 7.4; Biowest, L0615), fxed in 4% PFA for 15 minutes at RT, then washed 3 additional times with DPBS at 4 °C and permeabilized with phosphate-bufered saline with Tween for 20 minutes.Te flms were blocked with Tris-NaCl-blocking bufer for 1 h at RT and incubated with DAPI (1/1000) and phalloidin staining (1/ 1000, Abcam, ab176757) for 1 h at RT on a shaker.Afterwards, the flms were washed in DPBS and mounted with Prolong Gold (TermoFisher Scientifc, P36930).Images for quantifcation were taken using an AxioVert 200M microscope (Zeiss).For quantifcation, 10 images per sample were taken, and the number of cells was calculated using Fiji-imaging software [39].Images to visualize the cell morphology were taken using an Axio Imager Z1 (Zeiss).
2.9.Bacterial Adhesion.Te adhesion of S. aureus ATCC 8325-4 and E. coli 25922 to treated PCL flms was investigated.Bacteria were stored in liquid medium supplemented with 15% (vol/vol) glycerol (TermoFisher Scientifc, 17904) at −80 °C.Prior to the experiment, single colonies of the respective bacteria were picked up for overnight culture.S. aureus ATCC 8325-4 was grown in 10 ml Tryptic Soy Broth (Sigma-Aldrich, 22092) and E. coli ATCC 25922 in 10 ml Luria-Bertani Broth (Sigma-Aldrich, L2542) overnight at 37 °C.After two washing steps with DPBS, the concentration was assessed spectrophotometrically at 600 nm (UV-vis) using a BioPhotometer (Eppendorf ).PCL flms were incubated in 2 ml of 1 × 10 7 bacteria/ ml S. aureus and E. coli suspension for 1 h under shaking at 200 rpm at 37 °C.After being washed twice, the flms were transferred to 1 ml of PBS and the adhered bacteria were detached from the flms by using the ultrasound cleaner (VWR) for 20 minutes.Serial dilutions were made from the initial inoculation dose and each solution containing the detached bacteria and the respective PCL samples.Bacterial solutions were plated on an agar plate and incubated overnight at 37 °C to count colony-forming units (CFUs).Results are expressed as a percentage of adhesion of the total amount of bacteria incubated to the respective sample.

Statistical Analysis.
Results are presented as mean-± standard error of the mean (SEM).Each data point represents the average value of one flm.Statistical analysis was performed using GraphPad 9 Prism.Variance was tested with the F-test (p < 0.05) and normality was tested with the Shapiro-Wilk test (p < 0.05).Biological outliers were picked up based on Grubb's test (p < 0.05).Te data were analyzed using a one-way ANOVA with Dunnett's post hoc test or Tukey's post hoc test.If the data were not normally distributed, the Kruskal-Wallis test with Dunn's post hoc test was used.

Surface Characterization of Treated PCL
3.1.1.SEM Analysis.Te surface modifcation of the PCL flms was confrmed by morphology analysis using SEM.Te PCL surface changed remarkably in all conditions (Figure 2).However, there was no rupture of the PCL flms after treatment, and the surface was homogenously afected.

Evaluation of Hydrophilicity.
To study the efect of the surface treatments on the hydrophilicity of the PCL, the contact angles were measured.Te wettability was signifcantly increased on PCL flms modifed by hydrolysis compared to control PCL flms treated with water (Figure 3).Tere was a signifcant increase in hydrophilicity on the flms modifed with KOH (35.638 °± 1.012 °, p � 0.001) and NaOH (44.724 °± 1.843 °, p � 0.001).However, there was no signifcant diference when comparing the two hydrolysis treatments.Te combined treatments of NaOH + HMD (63.759 °± 0.180 °) and KOH + HMD (62.628 °± 1.820 °) as well as the HMD treatment alone (66.783 °± 0.847 °) showed no signifcant decrease in the contact angle compared to PCL flms treated with water (75.923°± 2.407 °).In addition, there was no signifcant diference in the contact angle between flms treated with HMD and flms treated with HMD in combination with NaOH or KOH.Tus, the results showed that hydrolysis treatments, here demonstrated by KOH and NaOH, were more efcient in increasing the hydrophilicity compared to treatment by aminolysis or the combination of aminolysis and hydrolysis.

Mass Loss.
To study whether the hydrolysis surface modifcation led to any mass loss, the weight of PCL flms was monitored before and after the surface modifcation process (Figure 4).A slight loss in mass was observed in both treatment conditions (NaOH −3.5% ± 0.8%; KOH −5.9% ± 1.4%).However, these changes were not signifcantly diferent from the water-treated control flms (−1.5% ± 1.4%).

Ninhydrin Assay.
To exclude the possibility that aminolysis failed to afect the wettability, alone or in combination with hydrolysis, we performed a ninhydrin assay on the PCL flms to confrm the formation of amide bindings.Amino groups were present on all treated conditions, confrming the efciency in the formation of amide bindings (Figure 5).A signifcantly higher number of amino groups was detected on PCL flms treated with KOH + HMD (70.232 ± 0.521, p � 0.001) and with HMD (64.033 ± 1.931, p � 0.001) alone than on PCL flms treated with the combination of NaOH and HMD (35.184 ± 1.835).

Endothelialization.
To study the efect of surface treatments on cell adhesion to PCL, HUVECs were seeded on chemically treated PCL flms.Analysis of the DAPI staining showed that there was a signifcant increase in cell adhesion to flms treated with NaOH (48000 ± 4600 cells/cm 2 , p � 0.014) after 24 h (Figure 6).All other treatments displayed no  signifcant diference with the control condition, where PCL was treated with water (22000 ± 900 cells/cm 2 ).HUVECs seeded on 1% gelatin were used as a positive control (67000 ± 4000 cells/cm 2 , p � 0.001).
A phalloidin staining showed that HUVECs grown on PCL flms had a normal cytoskeleton structure and cell morphology after 72 h.Tere was no diference in cytoskeleton structure and thus cell morphology between HUVECs seeded on water-treated PCL flms and on flms with a surface treatment (Figure 8).

Bacterial Adhesion.
To investigate the efect of chemical surface treatments on bacterial adhesion, treated PCL flms were incubated with S. aureus and E. coli.CFU counting after bacterial detachment revealed that S. aureus inherently shows a higher adhesion to PCL compared to E. coli (Figure 9).However, both tested bacteria displayed no signifcant diference in adhesion for surface treatment conditions compared to water-treated flms.

Discussion
As autologous grafts are often unavailable for coronary artery bypass surgery, there is a high demand for alternative solutions such as biocompatible, artifcial grafts [3,5].PCL, being a biodegradable polyester, is a possible candidate for the fabrication of cardiovascular grafts after optimizing the cell compatibility and ensuring no increase in bacterial infectivity.To increase the endothelial cell afnity, the surface of PCL flms was modifed by hydrolysis, aminolysis, or a combination of both.Current results show that NaOH treatment leads to the best surface modifcation to increase endothelial cell adhesion and survival on PCL flms, which cannot be further increased by dual treatment.
Research on the efect of surface treatment of PCL with NaOH has been done in tissue engineering approaches in multiple felds including cardiovascular research, orthopedic research, and neurology as PCL can also be used to make bone grafts or artifcial conduits for nerve repair [15,40,41].In our approach, NaOH and KOH were used to introduce -COOH groups on the polymer surface by hydrolysis and -NH 2 groups via aminolysis by HMD treatment.Te efect of the combination of both hydrolysis conditions with aminolysis was also studied, as we wanted to investigate the combined efect of -COOH and -NH 2 groups.
All conditions were found to afect the surface of PCL.SEM analysis showed an intact surface without signs of degradation, while clear modifcations were visible.Surface modifcation with NaOH is known to increase the groves on the surface, creating more space for cells to adhere [40].Te optimal water contact angle for maximum cell adhesion has been found to be in the region of 45-70 °C or in the range of 30-60 °C [17].Contact angles that are too low lead to an increase in water uptake, which decreases the protein adsorption needed for cell attachment [17].On the other hand, a very high contact angle causes low cell-conductive behavior and protein denaturation [42].Our contact angle measurements are in the optimal range for all conditions, except the negative control.In addition, the measurements showed that the hydrolysis reactions signifcantly increased the hydrophilicity, but aminolysis reactions did not.Importantly, hydrolysis and aminolysis have diferent kinetics [15,43,44].In brief, NaOH and KOH drive the hydrolysis of a surface by nucleophilic attacks of hydroxide ions on the carbonyl carbon, generating carboxylic acid groups, while the aminolysis reaction causes bulk degradation [44].Interestingly, the combination of hydrolysis and aminolysis showed no diference in contact angle when compared with aminolysis alone.Te weight monitoring of PCL flms revealed that there was no signifcant mass loss with either NaOH or KOH treatment.Bosworth et al. previously reported that a 10 M NaOH treatment for 24 h caused between 32% and 52% mass loss [34].We allowed the hydrolysis reaction to proceed for only 1h, which was likely short enough to not cause any signifcant mass loss.Tus, the degradation caused by a short 1 hour treatment is negligible while still increasing the hydrophilicity of the PCL flms.Furthermore, the ninhydrin assay revealed signifcantly less amino groups on the surface of NaOH + HMD-treated PCL than on PCL treated with HMD + KOH or HMD alone.One explanation is that the Na + cation is more Lewis acidic than the K+ cation.Lewis acids can activate a substrate towards nucleophilic attacks, which is how the hydrolysis reaction on PCL works [45].Tus, NaOH could be generating a bigger surface modifcation than KOH, creating more carboxylic acid groups whereon the aminolysis reaction cannot occur as easily at the temperatures used [46].Te extent of the difference observed in the ninhydrin assay is hard to reconcile with an efect purely due to Lewis acidity.Alternatively, it is possible that on a molecular level, more water is left on the NaOH-treated flms, resulting in more hydrolysis.
Te endothelialization of biodegradable vascular grafts is needed to ensure that a fully functional vessel can be created, out of autologous tissue, before the degradation of PCL is completed.Te current study confrms that NaOH treatment is the most efective surface modifcation to increase vascular endothelial cell adhesion to PCL flms.Unexpectedly, NaOH treatment and KOH treatment showed diferentially efects, with NaOH allowing for faster coverage with endothelial cells.After 24 h, NaOH was the only treatment which gave a signifcant increase in endothelial cell adhesion when compared to PCL flms treated with water.Tis agrees with the research performed by Serrano et al., who demonstrated that NaOH surface treatment increased the cell adhesion of endothelial cells to PCL after 24 h [41].In addition, Zhu et al. determined that aminolysis, investigated by measuring the adhesion of endothelial cells to PCL treated with 1,6-hexanediamine/2-propanol, does not increase the cell attachment ratio over untreated PCL but improves the cell morphology [33].We therefore investigated whether the combination of hydrolysis and aminolysis gave improved results.However, we found that NaOH without aminolysis increased the cell adhesion the most.In addition, after 72 h, NaOH was also the only treatment which gave a signifcant increase in endothelial  10 Journal of Tissue Engineering and Regenerative Medicine cell survival.However, there is also a clear increase, although not signifcant, in cell survival on PCL flms treated with KOH and HMD + NaOH.If the efect of KOH on hydrolysis is indeed less pronounced, then longer times may be needed to see an efect on the endothelialization of KOH treated PCL flms.Te same principle can be applied on PCL modifed with HMD + NaOH, suggesting that the efect of NaOH is diluted by the HMD treatment.None of our measurements of SEM and hydrophilicity, however, indicated a diference between NaOH and KOH.Te endothelial cell layer was less confuent on the PCL flms treated with HMD and KOH + HMD after 72 h but the endothelial cells were growing on all treated surfaces, suggesting that endothelial cells follow their native growth pattern when seeded on PCL flms independent of which surface modifcation is performed.Tissue engineering attempts to promote ideal conditions for cell adhesion to surfaces have to pay attention to the specifcities of various cell types and the required sites for cell attachment.As such, it was important to look specifcally at endothelial cell adhesion.Previous studies have demonstrated that NaOH treatment increases the cell adhesion to PCL of diferent cell types, such as Schwann cells, smooth muscle cells, endothelial cells, and osteosarcoma cells [15,40,41].Contrasting fndings reveal that the potential for cell adhesion difers depending on the cell type.de Luca et al. described that HMD, KOH, and NaOH all increased the cell adhesion of Schwann cells to PCL to the same level [15].On the other hand, Moura et al. found that NaOH does not increase the cell adhesion of bone marrow mesenchymal stem/stromal cells to PCL [47].In other settings, aminolysis treatment has been proven to be favorable.Lee et al. reported that NH 2 groups were the most efective at increasing cell adhesion of K100 erythroleukemia cells to self-assembled monolayers of alkylsilanes.Tey attributed this to the interaction of NH + 3 and negatively charged glycosaminoglycans of the cell membrane [43].However, diferent cell types have diferent requirements for their ideal matrix.Each cell type reacts diferently to diferent surface characteristics and to diferent polymers.Tis highlights the value of our study to investigate cell adhesion specifcally for cardiovascular application.
Tere was no signifcant diference between the negative control and the treated PCL flms when looking at bacterial adhesion.Tese outcomes suggest that the tested treatments do not attract bacterial attachment and thereby do not increase the risk for infection after surface treatment.A higher adhesion of S. aureus compared to E. coli could be explained by the multiple surface molecules of S. aureus, also known as MSCRAMM, which mediate binding to diferent biological components such as fbrinogen and collagen.Albeit these results show that the surface modifcation of PCL does not increase the bacterial infectivity, extra antibacterial factors can be added to the vascular graft to ensure an antiinfectious efect of the material.Tis can be done by adding drug releasing systems, for which PCL is an excellent candidate due to its high drug permeability.Tere has been promising research conducted on combining drug releasing systems with PCL using antibiotics and nitric oxide as well as other compounds [32,[48][49][50][51][52].Te combination of drug releasing systems or nitric oxide with the NaOH surface modifcation can possibly increase the endothelization of the graft in a next step while, in addition, decreasing the bacterial infectivity.Tereby, the complications after vascular graft implantation could be decreased, also reducing the mortality rate in patients.
Our research strengthens the preference for NaOH to modify the surface of PCL while simultaneously showing that other hydrolysis reactions, aminolysis reactions, and the combination thereof are less appropriate candidates to increase the endothelialization of PCL used for cardiovascular approaches.While there are other ways of increasing the endothelialization of vascular grafts, such as modifying the polymer graft with zwitterionic polymers or silk fbroin or coating with gelatin, the relative simple method of surface modifcation tested in this paper opens the possibility of combining the surface modifcation with other treatments [29,53,54].Tese drug or gene treatments can then further improve the endothelialization or decrease the bacterial infectivity and platelet activation.Chemical modifcations are also possibly stable for longer times.For example, gelatin coating degrades within 14 days in the body, while the reendothelization takes much longer [55,56].In addition, the simple nature of the treatment by NaOH might allow a quicker path through the regulatory approval process, allow longer shelf life, or, at least, result in shorter production times than other possible surface treatments.Hence, NaOH treatment shows great potential as a catalysator of endothelial cell adhesion and survival on PCL.Further research is needed to test NaOH-surface-modifed grafts under shear stress, both in vitro and later in vivo.In a next step, the association of a drug delivery system to the graft can also be explored.Tus, NaOH treatment causes promising strides in the optimalization of PCL as a biodegradable graft material for cardiovascular and other applications.

Figure 1 :
Figure 1: In vitro model to seed cells on PCL flms.Custom in vitro model to seed HUVECs on PCL flms using plexiglass inlets, 6-well plates, and metal frames to immobilize PCL flms.Plexiglass inlets viewed from the top (a) and bottom (b); metal frame during (c) and after (d) mounting; HUVECs seeding on the surface of sealed PCL flms (e).

15 Figure 4 :Figure 5 :Figure 6 :
Figure 4: Loss in mass of PCL flms modifed by hydrolysis and water.Weighing of PCL flms treated with water, NaOH, and KOH, n � 5. Data are presented as mean ± SEM.Statistical signifcance was tested with Kruskal-Wallis test.