A new organic polymer based on polydopamine, and containing active boronic units, is proposed in this work for the first time. Our investigation de facto begins with the use of green, one-pot chemistry to synthesize a poly-hydroxy-indole bulk modified with phenyl boronic moieties. This synthesis starts recurring to a co-polymerization of dopamine monomer with boronic aromatic species bearing specific functional groups directly addressing dopamine, such as free amines or thiols. Based on a co-polymerization protocol20, 3-aminophenylboronic acid (APBa, Fig. 2a, polymer 1) and other related chemicals (3-mercaptophenylboronic acid, MPBa; phenylboronic acid, PBa; 3-carboxyphenylboronic acid, CPBa; Fig. 2a 2–4, respectively), have been considered for dopamine functionalization during its first oxidative conversion step, using a solution of 1% ammonia in bidistilled water, at room temperature, under stirring, for 24 h (Supplementary Information (SI) section for control reactions). Among all the attempted reactions, only output 1 gave a water soluble polymer. Furthermore, polymer 1 exhibits high fluorescence intensity when excited at 375 nm (emission wavelength, 497 nm) in water (Figure S1). Polymer 1 was first characterized in solid phase, using FT-IR spectroscopy (Fig. 2b) and Raman Spectroscopy (Fig. 2c–e). Together with the lone dopamine polymerization control (Figure S2), a further control reaction has been introduced in this work exploiting a phenylboronic unit with amino-group indirectly linked via a methyl function (Figure S3 and comments). The comparison between the FT-IR spectrum of PDA-APBa, those of PDA monomers and free APBa (Fig. 2b), evidenced the presence of a sharp signal at 1395 cm-1, attributable to the B-O stretching vibration. Moreover, multiple serial signals after 1900 cm-1 underline the presence of the overtones related to the aromatic rings.21 Fig. 2c–e shows Raman spectra of PDA films drop casted on silicon before (PDA) and after co-polymerization with 3-aminophenylboronic acid (PDA-APBa), compared again with the Raman spectrum of the reference 3-aminophenylboronic acid (APBa). Raman spectra in the wide range, 100–1000 cm−1 (Fig. 2d), are dominated by a very intense band at ~485 cm−1 due to C–C–C–C torsion (ΓCCCC) hiding weaker Raman signals. To highlight them, Raman measurements have been also performed in two zoomed ranges, 70–450 cm−1 (Fig. 2c) and 500–1000 cm1 (Fig. 2e). Raman spectra of pure PDA (black lines) show bands corresponding to vibrational frequencies of bonds between C and C, O, H and N atoms, while the Raman spectra of the reference 4-aminophenylboronic acid (green lines) shows bands corresponding to vibrational frequencies of bonds between B, C and O atoms22,23.Figure 2(a) Co-polymerization of R-arene species and dopamine; (b) FT-IR characterization of PDA-APBa in comparison with APBa and PDA; (c–e) Raman spectroscopy characterization of PDA-APBa in comparison with APBa and PDA. Red asterisks in the Raman spectra of PDA-APBa films indicate the formation of B-C/B-O modes, not present in pure PDA. Raman vibrations, (torsion), (bending), (stretching) and (rocking) are also specified.After doping with 3-aminophenylboronic acid (PDA-APBa, red lines), the Raman spectra of PDA change, showing the presence of additional peaks, evidenced by red asterisks, due to the formation of bonds between B, C and O (see Fig. 2c and e), in both measurement ranges. Raman characterization has also been extended to wavenumbers higher than 1000 cm-1 as shown in Figure S4, confirming the efficacy of the PDA doping with 3-aminophenylboronic acid. Going over the investigation and for an insight into the structure, polymer 1 dissolved in D2O was characterized by using 1H-NMR and 13C-NMR. These methods were used to record spectra of PDA-APBa for comparison with those of lone soluble PDA, obtained using pyridine catalyst (section S2)23 and APBa monomer. An important and common group of signals ranging from 6 to 7.6 ppm in the 1H spectrum was assigned to the aromatic protons for both PDA and PDA:APBa. A last set of signals appearing from 8 to 8.6 ppm was assigned to hydroxyl protons of catechol moieties. In the 1H-NMR spectrum belonging to PDA:APBa the presence of multiplets at 7.05 and 7.45 reflected the presence of the APBa moieties24, whose spectrum is magnified in the inset of Figure S5. The 13C-NMR spectrum of the boronic-functionalized polymer (Figure S6, in dark red) shows chemical shift signals set around 161–163 ppm, corresponding to carbonyl carbon atoms from the quinone present in the polymer structures. Signals at 119, 121 and 125 ppm belong to carbon atoms of aromatic moieties present in the lone APBa molecule and in the proposed co-polymer. Lone PDA was not solved in 13C NMR experiment due to its low solubility. The presence of characteristic DA-APBa dimers (285.0821 Da, Figure S7) in the organic layers deriving from the washing steps of the co-polymer, has been recorded via high resolution LC–MS (Chan, 2019). PDA shows DA-DA dimer set around 293.113 Da. The fragmentation profile of the dimer has been reported in Figure S8.The chemical characterization of the proposed polymer gave enough information about structure, monomer repetition, spectroscopic and optical features and the presence of active boronic species. This new synthesis of boronic-bearing polymer 1 is intended to be exploited to directly address living cells in cell media, keeping it stable in contact with water, waterish media and biological matrices. The addressing of polymer 1 aims to occur on sugar moieties of living cells. This should be done in principle via the esterification reaction of the boronic unit with the vicinal diol functionalities of the saccharides on living cell surface. The two scenarios for obtaining this functionalization are: (i) the co-polymer production directly around living cells, by incubating these biological partners with the monomers, and (ii) the incubation of living cells with the preformed polymer 1. Unfortunately, changing the conditions of the polymerization reaction and, specifically, avoiding the use of ammonia and performing the polymerization of dopamine and APBa in biocompatible media (sea water, Tris:HCl, Tris(hydroxymethyl)aminoethane and Dulbecco’s modified Eagle Medium (DMEM)), the resulting polymer presents low solubility in water and in many organic solvents , with no fluorescence emission. The 1H-NMR spectra of polymer 1, synthesized from dopamine and APBa, in sea water, Tris:HCl and DMEM, are reported in Figure S9. As shown in the latter figure, a lack of peculiar signals from the APBa inner moiety is evident in the spectra.After the polymer structure characterization, the amount of phenylboronic units into PDA-APBa has been quantified by a titration in solution with increasing amounts of Alizarin Red S (ALI, Figure S10)25, a reagent specifically used to form catechol:boronic interaction with phenyl-boronic derivatives. ALI interaction with PDA-APBa leads to fluorescence quenching, directly proportional to the reagent amount. After the titration, 0.504 mg of ALI were consumed to quench 0.1 mg of PDA-APBa, underlining the presence of 14 μmol of -B(OH)2 per mg of the polymer.Resilient, sugar-enriched, living marine microalgae were chosen here to perform the biological experiments with the polymer 1, aiming to find the best combination in terms of cells viability, surface addressing and polymer bulk concentration in waterish stocks. The following microalgae cells, exhibiting a plethora of glycoproteins and membrane-associated polysaccharides, have been exploited here as living platforms: Dunaliella tertiolecta (D) and Chlorella vulgaris (C) as green microalgae, Thalassiosira weissflogii (T) and Navicula pelliculosa (N) as model diatoms. For the biological approaches and investigations, all these cells were incubated with polymer 1 at the concentration of 0.1 mg/mL in sea water for 1 h, 4 h and 24 h (Figures S11–S12). As expected, Dunaliella tertiolecta did not exhibit any blue fluorescence signal due to the low density of glycoproteins on cell surface, while Navicula pelliculosa accumulates the emitting polymer between cell structures, underlying that it links to the extracellular polysaccharide matrix (EPS). In this case N cells do not act as living platforms, because the saccharide units are not uniformly distributed over the cell surface, however the polymer successfully addressed these functionalities in EPS. T cells often internalize the polymer, maybe due to serial processes of passive surface interaction and active cell uptakes. This internalization phenomenon can be also due to the natural degradation of the external T cell membranes, leading to an inner accumulation and a cytoplasmic addressing of the polymer 1. The best candidate in terms of cell surface glycoprotein distribution is Chlorella vulgaris, as reported in the literature26, thus reflecting a uniform coating with the emissive polymer 1. In this case the boronic unit provokes a direct condensation reaction with 1,2- or 1,3-diol groups, typical of mono-, di- and polysaccharides both in extracts and living cells. This click reversible reaction requires mild experimental conditions (pH 7–8) with a pH-dependent reversibility, allowing a better, fast control of the attachment/detachment process27. These receptor:ligand-like interactions have been also exploited in the production of surface devices for mammal cell adhesion, attachment and recognition, till the final modulated release28. In this case, the interaction of boronic free units and aromatic catechol (in DOPA moiey) of specific peptides has been utilized, leading to the formation of a dynamic catechol:phenyl boronic cyclic ester functional to the tuning of cell attachment and cell migration29. Our polymer, containing covalently bound phenylboronic units, best performs at a contact concentration of 0.5 mg/mL (Figure S13). As shown in Fig. 3 (Figure S15 for 3D reconstruction), chloroplasts appear huge, round, strongly red emitting, and the polymer signals (blue emission) seem to perfectly profile the cell surfaces. The merge mode evidenced the absence of a co-localization between the two fluorescence signals. Coated living cells have been, then, incubated with FITC-DX as an external source of vicinal diols (Figure S14 for the bidimensional microscopy images), to investigate if the already exploited polymer 1 could attract and catch soluble biomolecules along the free boronic moieties exposed on the engineered cell surface (Fig. 4). An experiment of FITC-DX catching by PDA-APBa coated silica microparticles (diatomaceous earth, Sigma Aldrich, 1 mg) has been carried out, leading to a trapping of 19% of FITC-DX from a solution containing 1 mM of the fluorescently labelled polysaccharide (Figure S16). At pH 7.8, a saturation of 72% has been achieved by loading over 60 μg of polymer. Further evidence of PDA-APBa targeting towards polysaccharides was achieved by saturation percentage curves.Figure 3Confocal microscopy images of living Chlorella cells coated with PDA-APBa. Red and blue colors were arbitrarily assigned for chloroplast, inside cells, and PDA-APBa blue emission around cells, respectively. (Size bar 10 μm).Figure 4Confocal microscopy images of living Chlorella cells coated with PDA-APBa and incubated with FITC-DX. Red and blue colors were arbitrarily assigned for chloroplasts, inside cells, and PDA-APBa around cells. Green signals, present alone and in merge mode, are related to the fluorescent polysaccharide probe. (Size bar 10 μm).These functions were obtained by using different polymer concentrations incubated for 24 h with the same amount of Chlorella cells and comparing the polymer absorbance recorded before and after the addition of the interacting living cells. At pH 7.8, a saturation of 72% was already reached by loading over 60 μg of the polymer (Figure S17). The same workability function defined at pH 8.5 and at pH 6.5 reported a general set of saturation below 20%. The saturation % curves have been built also for 3 other chemical species: Alizarin Red S which targets polysaccharides (Figure S18) (Gu et al., 2010), a T-BTZ-T lipid targetting moiety (Passos et al., 2021) (Figure S19) and a Cy3-NHS(Southwick et al., 1990) typically used to chemically bind to amino-termini of free Lysine of protruding membrane proteins (Figure S20).Similar saturation % values have been obtained between PDA-APBa and Alizarin Red S (84% for ALI), conversely to the other staining agents (saturation for T-BTZ-T 65% at 7 μg, saturation for Cy3-NHS 95% at 7 μg) and this strongly suggests a common interaction of PDA-APBa and Alizarin Red S with vicinal diols of the exposed polysaccharides functionalities. In particular the rapid saturation reached by the lipophilic T-BTZ-T dye during the staining underlines the low amount of exposed lipid units, despite the abundant presence of simple sugar moieties (mainly glucose, rhamnose, mannose, xylose, galactose, arabinose and fucose)30,31 displayed over Chlorella cells in glycoproteins or organized cell wall forms. Moreover, carbohydrate competition technique (displacement) has been involved to calculate the quantity of diol moiety/cell amount. As previously reported, PDA-APBa interaction with ALI leads to a fluorescence quenching, due to the production of a non fluorescent chemical specie. The displacement consists of freeing of the PDA-APBa moiety from ALI, by titrating with increasing amounts of a diol-containing chemical model (D-( +)-glucose). The delivery of the free PDA-APBa polymer causes the increase of the fluorescence of this specie, and a calibration curve can be defined, set as Fluorescence Intensity vs glucose load (Figure S21). Then, a triplicate of ALI:PDA-APBa solution has been addressed by 3 106 in order to displace PDA-APBa polymer from ALI, using the diol moieties present at the living cell surfaces. Then, by interpolating the calibration curve previously defined, 3 106 Chlorella cells contain 0.06 ± 0.01 mg of glucose-like moieties (with at least a diol unit/molecule). For Thalassiosira, Phaeodactylum and Navicula the amount of glucose-like moieties set at 0.23 ± 0.05, 0.15 ± 0.02 and 0.13 ± 0.02 mg, respectively. These values, apparently higher than those recorded for Chlorella, are actually taking account also of the polysaccharides not only bound over cells but also embedding the extracellular spaces. As expected, sugar moieties were not detectable for Dunaliella cells.Flow cytometry has been used here both to test viability cells approaching polymer 1 and parametrizing the percentages of blue emitting polymer-coated cells and green emitting FITC-DX-binding biological systems. The cell viability analysis via flow cytometry results in 67% of living cells dispersed in the polymerization medium before coating and 57% for cells after coating treatment, as reported in histograms in Figure S22. Moreover, IC50 value for PDA-APBa was defined as 0.79 mg/mL (Figure S22). The fluorescent polymer FITC-DX has been used as an external probe biosorbed by the PDA-APBa layer over living cells. The persistence of FITC-DX was assessed during the cell growth and duplication, for several days by flow cytometry (Fig. 5). Bare cells with no polymer coating were exploited as control, while samples consisted of Chlorella cells incubated with PDA-APBa for 24 h. Both living samples were incubated with FITC-DX for 4 h with subsequent washing steps and then analysed. Real-time evidence of the extent of FITC-DX decoration, especially related to the presence of the organic polymer coating, was explicitly presented. The bivariate cytofluorimetric plot was obtained by recording red auto-fluorescence of chloroplasts (emission band 670 ± 15 nm) versus green FITC-DX emission (emission band 525 ± 25 nm). A significant difference is shown between control samples and polymer coated samples. Furthermore, an increase in the mean percentage of functionalized cells was observed from day 0 (90% of functionalized cells), to day 1 (99% of functionalized cells), with a decrease from 87.7 to 85.3% and to 58.8% for days 2, 3. and 6, respectively. Cytometric analysis demonstrated that cells have the ability to artificially inherit an external decoration over time, thanks to the bridging effect offered by the PDA-APBa polymer. These results strongly suggest that the functional layer is inherited over generations. Cytometric analysis of living cells with only PDA-APBa coating was also performed, giving the following percentage positivity to the polymer coating: 35.7% for day 0, 48.2% for day 1, 62.1% for day 2, 52.4% for day 3 and 30.2% for day 6 (Figure S23). This cytometric outcome likely reflects a redistribution of the emitting polymer over the membranes along the first mitotic events (0–2 days), thus leading to a decrease of the overall quenching.Figure 5Artificial inheritance study of FITC-DX biosorption by flow cytometry of control cells (uncoated bare and living Chlorella cells) and PDA-APBa coated living Chlorella cells. A number of 10000 cellular events were recorded for each acquisition time, interpolating the red fluorescence values, related to chloroplast auto-fluorescence of living cells, versus green fluorescence values, related to the FITC emission biosorbed on the living cells.As a last evidence of the biocompatibility of polymer 1 after living cells coating, data of nutrients uptake from living microalgae biofilm, with and without polymer 1 organic coatings, have been registered here. Bare and coated samples were immersed for 6 days in their enriched semi-artificial microalgae medium, to enable living cells catch nutrients from the solutions. Then microalgae were purified, mixed with reference chemical species (see Methods), and casted/dried over plexiglass round chambers. The living biofilm were analyzed by total reflection X-ray fluorescence spectrometry (TXRF) for detecting nutrients concentration. Except for K and Mn, TXRF analysis showed an overall increase of both macro- and micronutrients concentration in coated Chlorella samples (Table and histograms reporting TXRF analysis results.in Fig. S24). Such an increase was particularly noteworthy for Si, Cl, Ca and Fe, and Br, less evident but present also for S, as it can be deduced from the TXRF sum-spectra comparison in Fig. 6. Among these, Ca almost doubled its concentration compared to the bare sample while Fe concentration even increased eightfold. These data suggest the possible exploitation of the organic artificial coatings to enrich living cells with useful chemical species for the molecular assembly of photosynthetic and respiratory chain, together with the likely activation and support of metabolically active enzymes present in living cells. Figure 6Comparison between the TXRF spectra of a bare sample (black) and polymer 1-coated sample (blue). The K-lines of the elements of interest are labelled, including those of the IS (V).As final remarks, this new dopamine polymer containing phenylboronic moieties presents some specific features, starting from the possibility of easily modifying (via one-pot reaction) the monomers in aqueous buffer, under mild conditions, and the attachment of this polymer to the interface of the living cells is performed directly in water, with no harmful effects on biological systems. As a general comment, cheap and self-activating dopamine here becomes a powerful tool to produce polymeric materials carrying reactive and biologically interesting functional moieties, such as boronic species. It is so necessary to underline how the production of this supramolecular bridge occurs through a simple oxidative polymerization reaction, using oxygen as a catalyst, without consuming expensive reagents and toxic solvents, conversely to what is reported in the literature, where organic boronic-based small molecules find a peculiar place in the multidisciplinary context of functional biomaterials32. In details, rhodamine-based bis-boronic acids are produced via a reductive borylation with sodium boron hydride, in non-aqueous solvents, starting from diamino rhodamine dye33. Other cyclic boronic species proposed for targeting sugars are produced in dioxane or acetonitrile, with aliphatic amines or toxic metal catalysts34. Bis-boronic cross-linkers are synthesized via multistep routes, in halogenated solvents with aliphatic amines and acid activated-chlorides, with a certain ecological impact35. Furthermore, it is well known that PDA can be functionalized to gain the desired properties36 exploiting Aza-Michael reactions, using small addressing molecules37, such as thiol- or amine-terminated small molecules on pre-formed PDA, or in combination with co-polymerization of dopamine with other melanin precursors38. Literature proposes fully clickable polydopamine via complicated methods involving the use of dopamine bearing azide terminals to covalently couple with alkyne species39. Our technology, also under the perspective of the possibility to decorate polydopamine, is simple, one-pot and it yields a highly reactive self-adherent xeno-functional group. Finally, as a main intent, our outcomes underline that the PDA-APBa polymer displays boronic units over living cells, which are still reactive over surfaces, and capable of interacting with specific soluble biomolecules. When the modification is applied to the cell surface, this ensures living cells an original xeno-chemical composition. This polymer technology appears ideal to tether the minimal possible amount of chemicals directly on the cell surface, with the lowest synthetic effort, aiming to realize the desired molecular recognition. Our work paves the way to produce living materials, starting from molecular to nanoscale till microscale level, consisting of engineered living cells with potentiated properties of recognition and integration of external stimuli, holding the potential for broad biological, biomedical or agritechnical applications.
