Spacer Effects in Sulfo‐ and Sulfabetaine Polymers on Their Resistance against Proteins and Pathogenic Bacteria

The resistance of zwitterionic polymer coatings against the adsorption of proteins and the attachment of pathogenic bacteria is influenced by the precise molecular architecture of the polymers. Two until now rarely studied molecular variables in this context are side chain spacer groups separating the zwitterionic moieties from the polymer backbone and spacer groups separating the cationic and anionic groups within the zwitterionic moiety. Therefore, a set of six poly(sulfobetaine)s and poly(sulfabetaine)s is prepared, in which these spacer groups are systematically varied, incorporating ethylene, propylene, and undecylene side chain spacers, as well as ethylene, propylene, and butylene inter‐charge spacers, and their effects on the antifouling behavior are explored. Hence, the corresponding zwitterionic methacrylates are copolymerized with a photo‐reactive methacrylate bearing a benzophenone moiety. All zwitterionic coatings reveal hydrophilic properties when immersed in water and those with relatively short spacers show effective suppression of non‐specific protein adsorption. Polysulfobetaines outperform the polysulfabetaine ones in terms of resistance against adhesion of bacteria. The overall best fouling protection is observed for the polysulfobetaine bearing a propylene side chain spacer, which coincides with their relatively highest water solubility. The results corroborate previous findings that even apparently minor molecular changes of polyzwitterions can strongly affect their antifouling performance.


Introduction
Nosocomial infections are a major problem associated with hospitalizations, resulting in many casualties each year.In the EU, an average of 5.7% of patients contract such an infection during hospital treatment. [1]According to the US National Institutes of Health, 80% of nosocomial infections involve the formation of biofilms, which play a key role in infection progression. [2]Pathogenic bacteria irreversibly adhere to living tissue or non-living surfaces and produce extracellular polymeric substances that form a structural matrix and facilitate their spreading into biofilms. [3]Biofilm formation is generally considered to be an advantageous survival strategy for the bacteria compared to single planktonic bacteria.6] Implanted synthetic medical devices, such as central venous catheters, stents, urinary catheters, or intubation tubes, are particularly prone to biofouling. [7]Such artificial interfaces may be protected by coatings that kill bacteria due to incorporated biocides, such as antibiotics, [8] peptides, [9] or silver compounds. [10][13][14] If such materials reduce the non-specific adsorption (NSA) of proteins like fibrinogen, chances are increased that also platelet adhesion is reduced. [15,16]Among the most intensely studied materials, poly(ethylene glycol) (PEG)-based polymers or ethyleneglycols with varying chain length were widely used as low-fouling coatings due to their hydrophilic properties. [17,18]21] On top of a certain hydrophilicity, [22][23][24][25][26] a neutral net charge and hydrogen bond acceptors at the interface were identified as key design rules for inert coatings. [17]Zwitterionic materials [27][28][29] fulfill these requirements and are capable to form strong bonds with surrounding water molecules by electrostatically induced hydration. [30,31]Inspired by cell membranes, zwitterionic moieties in form of naturally occurring phosphatidylcholine (PC) groups were first introduced into fouling-resistant polymers. [32][35] Nowadays besides PC-based, [36] also carboxybetaine-(CB), [37,38] sulfobetaine-(SB), [39][40][41] and the recently established sulfabetaine-(SaB) [42][43][44][45] based polymers are intensely studied in this general context. [46,47][50] In addition, the dipole orientation, [44,[51][52][53] backbone composition, [54][55][56][57] or the spacer lengths [58,59] have been reported to affect the hemocompatibility and resistances against the NSA of proteins.Due to the diverse possibilities for structural modification, these polymers possess a high degree of variability and application-specific adaptability. [60]Still, conclusive structure-activity relationships are yet sparse.A direct comparison of polymers of SB-with CB-methacrylate demonstrated a higher blood plasma and alginic acid adhesion on the latter, while at the same time, a polymer based on CB-acrylate outperformed the one made with the analogous SB-methacrylate.However, the performance of the different polymers must be compared with caution, because several structural parameters such as backbone composition, inter-charge spacer length, and cationic moiety, were changed simultaneously. [61]A more recent comparison found a somewhat increased fouling resistance by polymers of CB-over its analogous SB-methacrylamide, [58,62] while a study on polyaspartamides reported better protection by the SB-over its CB-analogue. [63]Comparing SB-and SaB-systems of analogous structure, the latter have significantly lower solubilities in aqueous media. [42,43,45]Still, SaB systems showed comparable or even better antifouling performance as analogous SB-systems in recent marine fouling experiments, whereas the length of the inter-charge spacer strongly influenced their antifouling performance. [64][67] Concerning studies on CB-systems, such with methylene and ethylene inter-charge spacer chains (CBAA-1 and CBAA-2) showed improved blood plasma resistance, compared to systems with the propyl spacer equivalent (CBAA-3). [59]urther studies confirmed the strong resistance against the NSA of proteins for CBAA-1-and CBAA-2-based systems, but varying degrees of resistance against bacteria attachment were reported.The methylene-bearing CBAA-1-polymer resisted the adhesion of Pseudomonas aeruginosa, while [Mg 2+ ] dependent adhesion was detected on CBAA-2. [68]Still, the reports on inter-charge spacer effects in CB-polymers seem ambiguous. [69]Studies on the effect of the length of spacer group separating the zwitterionic moieties from the polymer backbone ("side chain spacer") on the fouling behavior have been particularly rare up to now. [58,64,70]n this work, we compare the fouling resistance of sulfobetaine and sulfabetaine functionalized polymethacrylates of systematically varied structures, exploring the effects of both the inter-charge and the side chain spacers.A series of zwitterionic polymers were synthesized by free radical copolymerization of the monomers with about 1 mol.% of 2-(4′-benzoylphenoxy)ethyl methacrylate (BPEMA) to form cross-linkable coatings that are fixed by UV-curing. [44]To allow a direct comparison of polysulfobetaines with polysulfabetaines, [64] two de novo synthesized sulfobetaine copolymers were prepared.The distance between the opposite charges was kept similar by adding an additional CH 2 group into the sulfobetaine monomers to compensate for the additional oxygen atom in the analogous sulfabetaine.Furthermore, a third sulfobetaine copolymer was prepared that possesses a marked amphiphilic character due to its hydrophobic undecylene side chain spacer. [71]In addition to their stability in different media, the ability to resist NSA of proteins and the influence of the chain length on the reduction of attachment of gramnegative (Escherichia coli and Pseudomonas fluorescens) and grampositive (Bacillus subtilis) bacteria was systematically investigated.

Results
The methacrylate monomers underlying the photo-crosslinkable copolymers used for the low-fouling coatings are shown in Figure 1.While BMA serves to synthesize the hydrophobic, negative reference coating, BPEMA is the photo-reactive comonomer used to crosslink the polymer chains upon irradiation by near-UV light via the CHic mechanism. [72]As functional building blocks, a set of custom-made zwitterionic monomers was employed with a systematic variation of the length of the aliphatic side chain spacer between the polymer backbone and the zwitterionic moiety, and a variation of the inter-charge spacer between the two oppositely charged ionic groups.In addition, the chemical nature of the anionic group was varied, comparing sulfabetaine containing copolymers against sulfobetaine ones with an adapted inter-charge spacer length to keep the relative distance between the ionic groups equal.The identifiers for the resulting molecules comprise three letters: the first −uppercase− letter indicates the anionic group (S for sulfonate, or Z for sulfate), the second −also uppercase− letter denotes the length of the inter-charge spacer (E for ethyl, P for propyl, B for butyl), while the third −lowercase− letter defines the backbone spacer length (e for ethyl, p for propyl, u for undecyl).To distinguish monomers from polymers, a lowercase "p" is added as prefix to the identifiers of the latter.While some of the copolymers were already described in previous publications, [64] the sulfobetaine copolymers of SPp, SPu, and SBe were de-novo synthesized by co-polymerization of the underlying sulfobetaine methacrylate with BPEMA, and characterized.Copolymer pSPu which has an unusually long (C11) side chain spacer was included in our study since recent experiments Table 1.Dry thickness and wetting properties of the polymer coatings.The reported values were determined on three sample replicates at three positions per sample.Errors reflect the standard deviation.suggested that amphiphilic properties may be advantageous for low-fouling behavior. [73,74]The polymers were compared to those derived from the sulfabetaine methacrylates N-[(2′methacryloyloxyethyl)-N,N-dimethylammonio] propyl-1-sulfate (ZPe) and 2-[N-(3′-methacryloyloxypropyl)-N,N-dimethylammonio] ethyl-1-sulfate (ZEp). [64]The copolymer pSPe was included in the study as benchmark system with a well-established low-fouling behavior. [39,44,64,75]ey properties of the coatings like thickness and wettability are summarized in Table 1 and Figure S2 and S3, Supporting Information).Within the precision of the measurements, the ellipsometrically determined film thicknesses of the photo-cured coatings decreased rapidly after immersion somewhat, but subsequently remained stable afterward even after 7 d of incubation, having about 85 % of the original dry thickness in MilliQ water and 80 % in PBS (Figure 3A,C).The data shows that the UV-induced crosslinking by C, H-insertion of the benzophenone moiety leads to surfaces that are stable underwater and in PBS, in agreement with previous reports on closely related systems. [43,44]esides the thickness, the wettability of the dry coatings was characterized as property with relevance for their foulingresistance.The hydrophobic pBMA had the highest static water contact angle (WCA) with 87°, while all zwitterionic coatings were much more hydrophilic with WCAs between 24°and 43°.The prominent exception are the films of the amphiphilic polyzwitterion pSPu with 64°.The pronounced hydrophilic  properties are consistent with the findings for other hydrogel coatings made from these or similar zwitterionic polymers. [44,64]o better understand the wettability underwater, additionally the captive bubble contact angles (CBCAs) were determined in MilliQ water and in PBS buffer.All zwitterionic polymers showed contact angles close to 22°in MilliQ water and close to 25°in PBS (Figure 3), which remained constant during the entire incubation period of up to 336 h.Although pSPu had a long alkyl chain, its reorientation in the medium was very fast.Already after 10 s of immersion, hydrophilic wetting comparable to the other zwitterionic polymers were measured.Interestingly, the originally hydrophobic film of pBMA showed a slow and steady contact angle decrease both in MilliQ water and in PBS.After 336 h, the CBCA in MilliQ water reached 52°, while in PBS, an even much lower CBCA of 30°was observed.
NSA of human serum albumin (HSA) and fibrinogen (Fb) was determined by surface plasmon resonance (SPR) spectroscopy.PBS-equilibrated coatings were treated with a 1 mg mL −1 protein solution at a flow rate of 10 μl min −1 for 10 min and afterward re-equilibrated in PBS for 15 min until the SPR signal was again stable.The amount of irreversibly bound protein was determined from the signal difference after and before protein injection (Figure 4).
The largest amount of protein adsorbed non-specifically on the hydrophobic pBMA, which is consistent with previous reports. [73,75]All coatings carrying zwitterionic moieties strongly reduced the NSA of HSA as well as Fb in comparison to pBMA.For HSA, all zwitterionic coatings showed a reduction of the protein adlayer to <8 % compared to the hydrophobic pBMA reference.For Fb, <30 % NSA were detected on pSPu and <2% on the other zwitterionic coatings compared to pBMA.Similarly, upon closer analysis comparing only the HSA results among the zwitterionic polymers (excluding pBMA), the NSA on pSPu is significantly higher than on the other zwitterionic polymers (Figure S10, Supporting Information) which was also found for Fb.Thus, the undecylene side chain spacer in pSPu significantly promoted the HSA and Fb adsorption.In contrast, for both proteins, a statistical analysis revealed no significant differences between sulfoand sulfabetaine polymer analogs and between those with propyl and ethyl spacers, even when both pBMA and pSPu were excluded.
The attachment of gram-positive and gram-negative bacteria was analyzed under dynamic conditions on an orbital shaker (Figures 5 and 6).In all cases, the accumulation on the zwitterionic coatings was significantly lower than on the hydrophobic reference pBMA.On a closer view, the relative extent of fouling protection was highest in the case of E. coli, and the least effective in the case of P. fluorescens.Also, the order by which the fouling resistance of the polyzwitterion films increased, is not uniform but specific for the particular bacteria strain investigated.The positive control pSPe showed the lowest relative accumulation densities of attached bacteria in the case of E. coli (ca.2% relative to pBMA) and of P. fluorescens (ca.8% relative to pBMA).Comparably low bacteria densities were seen on pSPp and pSBe with ca. 3 % and 4% in the case of E. coli, respectively, and ca. 13 % and 18 % in the case of P. fluorescens.However, in the case of the gram-positive bacterium B. subtilis, polysulfobetaine pSPp significantly outperformed the positive control pSPe with a relative bacterial density of ca.7 % compared to 24 %.The corresponding sulfabetaine polymers pZEp and pZPe showed significantly higher settlement densities, whereby pZEp with relative densities of ca.24 % (E.coli) and 54% (P.fluorescens) performed significantly weaker than pZPe with relative densities of ca.12% (E.coli) and 41 % (P.fluorescens).The overall weakest protection against settlement by bacteria was shown by films of pSPu that is characterized by a hydrophobic undecylene side chain spacer.While in the E. coli and B. subtilis attachment assays, the settlement densities on pSPu were still roughly comparable to the ones on pZPe and pZEp, films of pSPu exhibited much higher settlement in the P. fluorescens attachment assay reaching a level that approaches the one of the negative control pBMA.

Discussion
A series of zwitterionic polymers containing sulfa-and sulfobetaines with alkyl spacers of different lengths between the charged units and between the zwitterionic group and the polymer backbone were synthesized, spin-coated onto surfaces and crosslinked by UV irradiation.By this method, coatings of uniform thicknesses between 120 and 150 nm were fabricated.All investigated polyzwitterionic coatings, made of polysulfobetaines as well as polysulfabetaines, were stable during immersion in different media.An overview of the wettability and the antifouling behavior of the different coatings is provided in Table 2.During immersion, the zwitterionic polymers showed the ability to rapidly reorganize and presumably expose the zwitterionic moieties at the film-liquid interface, since an immediate contact angle decrease from ≈24-64 °to similar hydrophilicities of ≈22-25 °was observed for both the sulfa-and the sulfobetaines.The strongest difference between WCA and CBCA of ΔΘ∼ 42 °was found for pSPu.This can be explained by the low surface energy hydrophobic moieties accumulating at the surface in the dry state and dominating the initial wettability.During immersion for up to 14 d, the contact angles dropped immediately after immersion and both in MilliQ and PBS, but barely evolved further for the remaining incubation time.This indicates that a stable arrangement was quickly reached after immersion.The observed behavior of the reference surface pSPe is in line with the results of previous work. [73]eference polymer pBMA that does not contain zwitterionic groups showed a different behavior in the long-term CBCA experiments.The CBCA value of pBMA steadily decreases slightly throughout the incubation time in MilliQ water, and even more in PBS.As reported earlier [73] , this could be explained by the hydrophobic butyl-chains reorienting toward the film while the polar connecting ester bonds increasingly orient toward the aqueous interface.This process seems to be even more pronounced in PBS where phosphate groups might coordinate to the ester groups, stabilize this conformation, and reduce the WCA further.The rapid reorientation of zwitterionic groups underwater with a consequently stronger hydration, as well as the slow reorientation of pBMA underwater are in line with previous findings. [73,76,77]he film thickness of all coatings remained very stable during the entire immersion duration of 14 d.The strongest decrease was found for the sulfonates pSPp, pSBe, and pSPu, which lost up to 15 % of their thickness in MilliQ water, which is only slightly beneath most of the error bars of the measurement.The slightly lower stability of the sulfonates compared to the sulfates was in line with the slightly higher solubility of the noncrosslinked polymers in MilliQ water.While the solubility of pBMA and the sulfabetaines was in general very low, pSPe, pSPp, and pSBe seemed to be partially soluble after 1 h as cloudy solutions were obtained (Figure S6, Supporting Information).Also during incubation in PBS buffer, the dry film thickness of all coatings remained very stable for the immersion duration of 14 d.The sulfonates pSPp and pSPu, and to a lesser extend pSBe and pSPe, were again slightly less stable and showed a film thickness decrease by up to 20 %, which is again close to the error bars.The difference between sulfo-and sulfabetaines was similar as observed in the MilliQ experiments as the sulfobetaines were slightly less stable than the sulfabetaines.The solubility of the non-crosslinked sulfobetaine polymers in PBS were again slightly higher than for the sulfabetaines (Figure S7, Supporting Information).
Even though most of the trends are very close to the error bars, we noted a slightly stronger thickness decrease in PBS as compared to MilliQ water.This trend agrees with the lower sol- ubility of the zwitterionic non-crosslinked polymers in MilliQ water than in PBS buffer (Figures S6 and S7, Supporting Information).The low solubility of the zwitterionic polymers in aqueous media [42,43,78] and also the increased solubility after salt addition [27,42] are in line with previous publications.Concluding, minor differences in the stability experiments were observed between polysulfobetaine and polysulfabetaine-containing coatings, but overall, the coatings were stable even for extended immersion durations of 14 d in MilliQ water and in PBS buffer.
All zwitterionic coatings strongly suppressed the NSA of the model proteins HSA and Fb.The differences observed between polysulfobetaines and polysulfabetaines analogs were statistically not significant, even if pBMA and pSPu were not considered.This is because, on all coatings, the amounts of irreversibly bound protein were extremely small.This agrees with previous studies where the amount of irreversibly bound BSA was also so small that different polyzwitterionic coatings became virtually indistinguishable. [64]Only Fb, which is much larger (340 kDa) than HSA (66.5 kDa), [79,80] showed statistically significantly higher NSA on the zwitterionic pSPu coatings compared to all otherzwitterionic coatings.In the case of HSA, a significantly higher NSA is also seen on pSPu, when the zwitterionic coatings are considered independently without the control pBMA (Figure S10, Supporting Information).This was surprising as the wettability of pSPu underwater was comparable to the ones of the other zwitterionic coatings (cf.Table 2).The long alkyl side chain spacer between the zwitterionic group and the backbone seems to favor the NSA of Fb and HSA to a certain extent, possibly due to a quick reorientation enabling access to the aliphatic side group and thus van-der Waals interactions with the hydrophobic domains in the proteins.A fast reorientation of hydrophobic moieties in hydrophilic zwitterionic coatings has been documented in the literature. [77,81,82]Also amphiphilic polybetaines showed an enhanced NSA of Fb when a lipophilic alkyl chain is incorporated, while perfluorocarbon chains were more lipophobic and showed a higher resistance against protein fouling. [55,83]Also the dipole moment of the functional groups seems to be relevant for the attachment. [59,84]he outcome of the dynamic bacterial accumulation experiments was more differentiated than of the protein adsorption experiments.All zwitterionic coatings reduced the attachment of all three tested bacteria strains statistically significantly compared to the pBMA references.The polysulfobetaines pSBe and pSPp showed less fouling than the polysulfabetaines.Besides charge neutrality, one of the key design rules for fouling-resistant surfaces is a hydrophilic character. [17,85]It has previously been shown that polysulfobetaines have a higher net affinity to water than polysulfabetaines, as reflected by the considerably higher upper consolute boundaries and clearing points of the latter in aqueous solution. [43,86,87]This is in agreement with our observation that the solubility of polysulfobetaines in both water and PBS is higher than that of the structurally similar polysulfabetaines (Figures S6 and S7, Supporting Information).The high hydrophilicity of zwitterionic coatings goes along with their high protein resistance.It is possible that the hydration shell formed on the surface is sufficient to shield surfaces from the comparatively small proteins. [24]While proteins are macromolecules with an interface dominated by hydrophilic groups, bacteria are living organisms with a dynamic and complex structure.The resistance of surfaces to bacteria seems to rely on several factors beyond hydrophilicity, [31] and thus, is not easily explained.Also, differences were observed in the attachment behavior between the two gram-negative bacteria E. coli and P. fluorescens.Normalized to pBMA, the zwitterionic coatings showed higher relative densities of P. fluorescens with 7-54% or even 97% for pSPu, while in comparison, the relative densities of E. coli ranged from 4% to 24% for all zwitterionic coatings.In both cases, the polysulfobetaines were more efficient in repelling the gram-negative bacteria than their polysulfabetaine analogues (cf. Figure 5).Comparing the polysulfabetaines among each other, pZPe showed higher bacterial resistance than pZEp.A similar trend was observed in a previous study, where the NSA of lysozyme was higher on pZPe than o pZEp. [64]In particular the relative accumulation density of P. fluorescens on pSPu is with 97 % very high compared to the other zwitterionic coatings (Figure 5B).P. fluorescens produces the peptidolipid biosurfactant viscosin, which might be of advantage for the attachment on hydrophobic surfaces. [88,89]Along these lines, the lipophilic alkyl chain of pSPu might favor its interaction with P. fluorescence, en-abling their adhesion on the coatings.Thus, hydrophobic interactions between pSPu coatings and P. fluorescens may be responsible for the relatively high accumulation densities.
When comparing the accumulation of gram-negative and gram-positive bacteria, it is noted that pSPe films show an intermediate performance against settlement by B. subtilis, while for the gram-negative bacteria, they were always amidst the best performing coatings (cf.Table 2).Films of pSPp suffered the lowest accumulation of the gram-positive bacteria.This indicates that not only the type of anion but also the inter-charge spacer between the oppositely charged groups and the side chain spacer between backbone and zwitterionic moiety are important molecular parameters to control fouling.As the performance of pSPp was also very good withstanding the attachment of both gramnegative species (statistically indistinguishable from pSPe), pSPp shows the best combined performance against the three gramnegative and gram-positive species studied.Hypothesizing about the reasons, compared to pSPe, pSPp has a higher degree of steric freedom due to longer propyl side chain spacer between the betaine group and the backbone.Also, the distance of the propylene inter-charge spacer between the oppositely charged moieties seems to be an optimum, as pSBe with the larger butylene spacer performed worse against gram-positive species.Possibly, the slightly longer connection in pSPp supports the rearrangement into a favorable geometry for an association of the zwitterionic groups to reach not only overall, but also local charge neutrality, which might be favorable for the inertness to NSA.In a recent study on a series of sulfabetaines against marine organisms, pZPe outperformed pZEp [64] and showed also equally good, or even better resistance than pSPe.However, the coatings have only conditionally analogous lengths of the inter-charge spacers, since in sulfabetaines, the distance between the charged moieties is somewhat larger due to the additional oxygen atom.Here, a direct systematic comparison of polysulfabetaine and polysulfobetainecontaining coatings was addressed by considering the different charge distances also for the polysulfobetaines.The pSPp arrangement is identified as best performing chemical structure for the investigated gram-negative and gram-positive bacteria (cf.Table 2).While it is already becoming clear that sulfobetainecontaining polymers can be optimized by adjustment of their molecular architecture and that pSPp seems to be an excellent candidate for further analysis, further experiments are necessary to understand the underlying mechanisms and surface-related conformational rearrangements.

Conclusion
A series of zwitterionic coatings for a comparison of SBs and recently established SaBs with varying chain length was synthesized to characterize the fouling resistance against proteins and gram-negative and -positive bacteria.For a better comparability, the relative inter-charge distance of SB and SaBs was adjusted by introducing an additional CH 2 group into the sulfobetaines.The assays revealed that in general, all the polyzwitterions investigated show high and comparably effective protection against fouling by the model proteins human serum albumin and fibrinogen, except for a hydrophobically modified variant, pSPu, that proved to be somewhat less resistant.While all tested polyzwitterions provided also good protection against several model bacteria −apart from the mentioned hydrophobically modified variant−, SB-based coatings showed a higher bacterial resistance than their SaB analogs.The extension of the inter-charge chain length from propyl to butyl had only a negligible impact on the fouling resistance, while the extension of the backbone-to-betaine spacer length from ethyl to propyl enhanced the resistance toward B. subtilis.Although the incorporation of an undecyl backbone-tobetaine chain in pSPu resulted in hydrophobic properties under air conditions but hydrophilic underwater properties, the resistance against fouling was diminished, nevertheless.The known ability of zwitterions to rearrange and migrate to the interface [73] is consequently not compromised by a long alkyl chain, but in the case of the fouling proteins and the bacterium P. fluorescens, increased hydrophobic interactions seem to be operative and therefore, the fouling resistance of pSPu was attenuated.Across the tested gram-negative and gram-positive bacteria, the influence of the anion type in the zwitterionic coatings seems to dominate the fouling behavior, while the inter-charge and backbone-to-betaine spacer lengths can be used to optimize the low fouling properties.
Silane and Thiol Functionalization of the Substrates: Glass objective slides (Marienfeld) were used as substrates for the contact angle stability investigation and bioassays, and silicon wafers (Siegert Wafer) were used for thickness stability investigation.Prior to use, the substrates were washed with ethanol and functionalized with 5% APTMS/acetone solution (v/v), and rinsed with acetone and ethanol according to previously published protocols. [91,92]The silane layer served as adhesion-promoter for the subsequent application of the polymer films.SPR gold chips (D263, Schott, 5 nm titanium, 60 nm gold [93] ) (used for protein adsorption investigations) were cleaned by irradiation for 1 h with ozone-forming UV light and rinsed with ethanol.The SPR chips were then incubated for 24 h in 1 mM of 11-amino-1-undecanthiol (AUDT) in 3% triethyl amine/ethanol solution (v/v) according to previously published protocols to form aminoterminated self-assembled monolayers (SAM). [94]reparation of the Polymer Coatings: Solutions of 0.5% (w/w) and 1% (w/w) polymer in 2,2,2-trifluoroethanol (TFE) were prepared, and applied to the amino-functionalized surfaces by spin-coating (WS-650MZ-23NPP/Lite, Laurell Technologies Corporation) according to previously published protocols (10 s at 200 rpm followed by 30 s at 3000 rpm). [44]For crosslinking, the thin films were exposed to UV radiation (Dr.Hönle AG, UVA Cube 100 (100 W) with Strahler UV 150 F (150 W) iron-doped mercury vapor lamp, radiate-physical data: 200-400 nm, sample-lamp distance: 17.3 cm) for 30 min under air conditions.The coatings were either applied on silicon slides (100 μl 1% (w/w) polymer solution) (20 × 20 mm) for stability investigation, on Marienfeld glass objective slides (500 μl 1% (w/w) polymer solution) for the bioaccumulation experiments, or on AUDT functionalized gold-coated SPR chips (35 μl 0.5% (w/w) polymer solution) for protein resistance tests.
Contact Angle Goniometry: The sessile water contact angle (WCA) was measured using a custom-built contact angle goniometer with a CCD camera.Sessile drops of tri-distilled water were deposited on the surfaces.The recorded images were evaluated by analyzing the droplet shape, according to established protocols. [44]The captive bubble contact angles (CBCA) were determined in a custom-built container with windows (approx.100 × 100 × 100 mm) following previously published protocols. [76]Surfaces were mounted face down in the liquid-filled (either MilliQ or 1xPBS pH 7.4) container, and an air bubble was deposited on the coating from below.Images were recorded with a CCD camera and the contact angle between the aqueous phase and the air bubble was evaluated manually with the software ImageJ. [95]pectroscopic Ellipsometry: The dry layer thicknesses d were determined by spectroscopic ellipsometry (M-2000 ellipsometer, J. A. Wollam Co. Inc., Lincoln).Measurements were recorded at three different angles of incidence (65°, 70°, and 75°).For the APTMS and AUDT functionalized surfaces (d ≤ 60 nm), the surfaces were modeled as a single transparent layer with a wavelength-dependent refractive index (Cauchy model: A = 1.45,B = 0.01, C = 0).For the polymer coatings (d ≥ 60 nm), the additional layer was modeled as a single absorbing layer according to previously established protocols. [44]tability Assessment: The stability of the polymer films was investigated by incubation at 60 rpm on a linear shaker in MilliQ and in 1xPBS (0.137 m NaCl, 0.0027 m KCl, and 0.0119 m phosphates) pH 7.4 for up to 14 d.Before incubation and after selected immersion durations (10 min, 2 h, 24 h, 48 h, 7 d, and 14 d), the static contact angles and the layer thickness of the polymer coatings were determined by either contact angle goniometry or spectroscopic ellipsometry.In the case of PBS incubation, the samples were briefly rinsed with MilliQ to remove salt residues prior to the measurements.
Non-Specific Adsorption of Proteins: The protein resistance of the polymer coatings was investigated using SPR spectroscopy (SR7000DC from Reichert Technologies Life Sciences) according to established protocols at a constant temperature of 37 °C. [44]Protein solutions of 1 mg mL −1 of human serum albumin (HSA) and 1 mg mL −1 fibrinogen (Fb) in 0.5xPBS pH 7.4 were freshly prepared before each measurement.0.5xPBS at pH 7.4 was rinsed over the surface at a flow rate of 10 μL min −1 until a stable baseline was obtained.The respective protein solutions were injected for 10 min, before the flow was switched back to 0.5xPBS.The signal difference before and 15 min after the protein injection phase was used to determine the amount of irreversibly bound protein.
Bioassays: Dynamic bacterial attachment was analyzed using the gram-negative freshwater bacteria Escherichia coli K12 (E.coli, DSM 498) and Pseudomonas fluorescens (P.fluorescens, DSM 50090), and the grampositive bacterium Bacillus subtilis (B.subtilis, DSM 402).All were received as freeze-dried cultures from the German Collection of Microorganisms and Cell Cultures (DSMZ, Braunschweig, Germany).Bacteria were stored at −70 °C according to recommended protocols, [96] thawed, spread, and grown either on nutrition agar plates (4.6 g agar in 0.2 L MilliQ) for E. coli, or on Medium 1 agar plates (5 g peptone + 3 g meat extract per 1 L

Figure 1 .
Figure 1.Chemical structures of the sulfabetaine (ZPe, ZEp) and sulfobetaine (SPe, SBe, SPp, SPu) monomers used as building blocks, of the photoreactive comonomer (BPEMA), and of the hydrophobic monomer BMA used for the negative reference.

2 .
Stability of the film integrity was tested in MilliQ water and phosphate-buffered saline (PBS) (Figure 3, further information in Tables

Figure 3 .
Figure 3. A) Thickness changes after different immersion durations in MilliQ water for up to seven days.B) Sessile (grey background) and captive bubble (blue background) contact angles of/in MilliQ water.C) Thickness change after different immersion durations in 1xPBS (pH 7.4).D) Sessile (grey background) and captive bubble (blue background) contact angles of MilliQ water/in 1xPBS (pH 7.4) of the zwitterionic polymer and hydrophobic pBMA coatings after different immersion durations.Data points are the average of at least nine measurements and error bars represent the standard error.

Figure 4 .
Figure 4. SPR analysis of the NSnon-specific protein absorption from (left) 1 mg mL −1 HSA and (right) 1 mg mL −1 Fb solutions onto the coatings of zwitterionic polymers pSPe, pZPe, pZEp, pSBe, PSPp, pSPu, and of the hydrophobic pBMA reference.Reported values are the average of three measurements.Error bars represent the corresponding standard error.Samples sharing the same letter do not exhibit significant differences, whereas distinct letters (a-e) indicate significant difference (p < 0.05).

Figure 5 .
Figure 5. Dynamic attachment of gram-negative bacteria on the zwitterionic coatings and the pBMA references.Relative density of gram-negative bacteria A) E. coli and B) P. fluorescens on the coatings was assessed after dynamic incubation with 7.5 × 10 6 cells mL −1 at 60 rpm over 45 min.Reported values are the average of three measurements.Error bars represent the corresponding standard error.Samples sharing the same letter do not exhibit significant differences, whereas distinct letters (a-e) indicate significant difference (p < 0.05).

Figure 6 .
Figure 6.Dynamic attachment of the gram-positive bacterium B. subtilis on the zwitterionic coatings and the pBMA references.Relative density was determined on the coatings after dynamic incubation with 7.5 × 10 6 bacteria mL −1 at 60 rpm over 45 min.Reported values are the average of three measurements.Error bars represent the corresponding standard error.Samples sharing the same letter do not exhibit significant differences, whereas distinct letters (a-e) indicate significant difference (p < 0.05).Comparisons between the different samples can lead to multiple designations.

Table 2 .
Wettability, NSA of proteins, and density of attached bacteria of/on the zwitterionic polymer coatings.The reported NSA values and bacteria densities were normalized to the adhesion on pBMA.For easier comparison color code according to percentage intervals (Red 100-75%, Black 50-75%, light green 25-50%, dark green <25%) has been used for the respective values of the zwitterionic polymers.