Compartmentalized cross-linked enzyme nano aggregates (c-CLEnAs) toward pharmaceutical transformations

A new immobilization strategy using compartmentalized nanoreactors is herein reported for two biocatalytic processes: (1) N-acetylneuraminate lyase (NAL) is internalized in NAL-c-CLEnAs and used in a continuous flow aldol condensation of N-acetyl-d-mannosamine with sodium pyruvate to N-acetylneuraminic acid; (2) two hydroxysteroid dehydrogenases (HSDH) 7α- and 7β-HSDH are incorporated in c-CLEnAs and used in a two-step cascade batch synthesis of ursodeoxycholic acid (UDCA). The versatile use of c-CLEnA demonstrates that this immobilization methodology is a valuable addition to the toolbox of synthetic chemists.


S-2
Size Exclusion Chromatography (SEC): For an efficient separation of the stomatocytes from the unencapsulated enzymes, a Shimadzu Prominence SEC system equipped with a Superose™ 6 column and a UV detector (220 nm) was used. The separation was performed using filtered PBS buffer at 0.8 mL min -1 .
Transmission electron microscopy (TEM): TEM images were recorded using a FEI Tecnai 20 (type Sphera) at 200 kV. 5 μL sample was dropped on top of a carbon-coated copper grid (200 mesh, EM science), and the desalted samples were left to dry at room temperature overnight.

NAL-c-CLEnA Activity in Flow:
Multiple NAL-c-CLEnAs (10 mg mL -1 ) containing ca 1.4 mg of NAL were mixed together to reach a final amount of ca 20 mg NAL. The final solution was adjusted to 3 mL via spin filtration (0.1 μm, 3000rpm) for 10 mins. The aqueous suspension was vortexed and loaded into a syringe (1 mL), after which it was coupled directly to the side-inlet of the flow reactor. Using a syringe pump, the solution was slowly added to the reactor (< 0.5 mLmin -1 ), effectively eluting excess aqueous solvent. This procedure was repeated 3 × after which the reactor was sealed before the catalysis experiments. After the indicated time intervals (Table S1)  NAL-c-CLEnA stability test in a membrane in-flow reactor: The Microkros ® reactor was loaded with c-CLEnA enzyme dispersion (containing up to 18 mg NAL) according to the aforementioned procedure. An HPLC pump was loaded with a solution of ManNAc (1, 500 mM) and sodium pyruvate (2, 100 mM) and the pH was adjusted to 7.0 using 1 M aqueous NaOH or HCl. The Microkros reactor was connected directly to the HPLC pump. The reactor was submerged in a 35 °C water bath for 10 minutes before starting the experiment. The flow rate was set to 25 µLmin -1 , and the substrate solution was continuously pumped through the reactor. After the indicated time intervals, the reaction mixture was separately collected for 10 min (3 ×). The samples were concentrated in vacuo and the conversions were determined using 1 H NMR in D 2 O.
Enzymatic assay: the enzymatic activity of 7α-HSDH and 7β-HSDH (on purified enzyme, encapsulated and coencapsulated) was determined at room temperature (25 °C) using 2.0 mM CDCA or UDCA respectively, 1.0 mM NAD + , in 50 mM KPi buffer and 10% methanol (v/v), pH 8.0. The 7α-HSDH activity was measured using the conversion of CDCA into the intermediate 7-oxo-LCA. The activity of 7β-HSDH was measured using the reverse reaction from UDCA to 7-oxo-LCA (as the intermediate compound 7-oxo-LCA is not commercially available). The extinction coefficients of NADH, at 340 nm is 6.220 M −1 ·cm −1 . One unit (U) was defined as the amount of enzyme producing 1 µmol of product per minute at 25 °C and at pH 8.0. Blank measurements were performed in the absence of CDCA, NAD + and enzyme. Results are reported in Table S2.
The supernatant was filtered (syringe filter 0.2 µm) and 10 µL of these preparations were analyzed by HPLC.
Epimerization of CDCA to UDCA with co-encapsulated enzymes (7α/7β-HSDH stomatocytes and 7α/7β-HSDHc-CLEnA ): All conversions were carried out (duplicates) employing different amounts of 7α/7β-HSDH c-CLEnA (300 µL of different dilution, Table S3) on 10 mM CDCA, using 1 or 0.5 mM NAD + . As a general procedure, 1 mL of reaction mixture containing 10% MeOH and 50 mM of KPi buffer, pH 8.0 was incubated at 25 °C on a rotatory wheel. At fixed times of incubation 50 μL of reaction mixture were diluted with 200 µL of mobile phase and centrifuged in order to separate the nanoreactors from the mixture (14000 rpm, 2 min). The supernatant was filtered (syringe filter 0.2 µm) and 10 µL of these preparations were analyzed by HPLC.
Bradford assay: The Bradford method was used to quantify the enzyme loading. Pierce™ Coomassie Plus (Bradford) assay kit was used as described in the protocol of the assay, Table S1 reports the amount of enzyme used in all the experiments.

General procedure for polymersome preparation
The polymersomes were self-assembled using a slightly modified variation of a previously reported solvent switch method.
In short, 20.0 mg synthesized PEG 44 -b-PS 200 polymer was dissolved in a 2 mL mixture of THF: dioxane (4:1 v/v), to which 1.0 mL MilliQ was added via a syringe pump with a flow rate of 1.0 mL h -1 , resulting in the formation of a cloudy solution. The assembly was performed inside a 5.0 mL vial which contained a magnetic stirring bar and which was capped with a septum. The cloudy solution was then dialyzed against MilliQ water for 24 h, with the MilliQ frequently refreshed.

General procedure for stomatocyte preparation and enzyme loading
Stomatocyte nanoreactors were prepared using the previously reported solvent addition methodology. 1 300 μL THF:dioxane solution (4:1 v/v) was added via syringe pump at a rate of 300 μL h -1 to 500 μL of the previously prepared polymersome solution (10.0 mg mL -1 ), while continuously stirring. The organic mixture was removed from the polymeric solution using spin filtration (20 mins, 13523 rcf) which was repeated two times using Amicon 3 kDa filters). The polymersomes were re-suspended to their initial concentration by adding MilliQ water. At the end of this process, opened neck stomatocytes were formed which were used for enzyme entrapment.
Next, 1 mL of a 10 mg mL -1 NAL solution in 50 mM sodium phosphate buffer (50 mM pH 7.5), was added to the stomatocytes and mixed vigorously at 7000 rpm for 30 mins. To narrow the neck of the stomatocytes 150 µL THF: dioxane (4 : 1 v/v) at 150 μL h -1 flow rate was added to the solution. To remove the THF, samples were purified using spin filtration (15 mins, 13523 rcf) two times with Amicon 3 kDa filters. To remove nonencapsulated enzymes, stomatocytes were purified from the solution mixture using size exclusion chromatography (SEC). After SEC, the stomatocytes were concentrated again to a final volume of 500 µL (10 mg mL -1 ). The same procedure was used for the 7α-HSDH and 7β-HSDH stomatocytes, using 8 mg mL -1 of 7α-HSDH and 8 mg mL -1 of 7β-HSDH or a solution of [7α-HSDH +7β-HSDH]=8 mg mL -1 with both enzymes. In the case of 7α-HSDH and 7β-HSDH, the samples were re-dispersed in PKi (50mM pH = 8).

General procedure for the formation of compartmentalized cross-linked enzyme nano aggregates (c-CLEnA) with glutaraldehyde
Having ensured complete removal of free enzyme from the previously prepared stomatocyte nanoreactors, glutaraldehyde (100 μL, at different concentrations varying between 100 mM and 300 mM) was slowly added, at a rate of 100 μL h -1 , to a 500 μL solution of enzyme loaded stomatocytes (10 mg mL -1 ) while stirring. In the case of NAL samples, the cross-linking reaction was quenched with 1mL of sodium phosphate buffer (1M, pH = 7.5) solution, and in the case of 7α-HSDH and 7β-HSDH, the reaction was quenched with 1mL of PKi (50 mM pH = 8). To remove the excess of buffer and cross-linker glutaraldehyde, all resulting c-CLEnAs were concentrated via spin filtration (15 mins, 13523 rcf) two times with Amicon 3 kDa filters and then were re-dispersed in sodium S-5 phosphate buffer (50 mM, pH = 7.4). In the case of 7α-HSDH and 7β-HSDH, the c-CLEnA was re-dispersed in PKi (50mM pH = 8).

General procedure for the formation of compartmentalized cross-linked
enzyme nano aggregates (c-CLEnA) with genipin 500 µL of genipin solution (1wt%-1.6wt%-0.75wt%) was added to 500 µL of stomatocyte sample (10 mg mL -1 ) in an Eppendorf tube. The solutions were kept for 24h at RT under gentle stirring.
After 24h genipin was removed from the solution, by using 10 kDa filters in a centrifuge at 12000 rpm for 15 min, the final concentration was adjusted to 10mg mL -1 with sodium phosphate buffer for the NAL sample or with 50mM PKi buffer (pH 8) for the 7α-HSDH and 7β-HSDH samples.

Quantification of enzyme loading
The Bradford assay was used to quantify the amount of enzyme loaded in the stomatocytes and in the c-CLEnAs.
All the samples were treated with CH 2 Cl 2 to completely remove the polymeric membrane, which would alter the absorbance measured in the test. 150 μL of enzyme loaded stomatocytes were mixed with 500 μL of CH 2 Cl 2 for 30 mins. The final solution was then spin filtered with a centrifugal filter Unit 3 kDa (Millipore) to remove the organic solvent. The fraction collected was adjusted with buffer to the final volume of 150 μL. The measurements were performed in triplicate using 50 μL aliquots. For protein quantification, the Coomassie Plus (Bradford) assay kit was used (Pierce™) according to the manufacturer's instructions. In each cuvette both 1.5 mL of Coomassie reagent and 50 μL of sample were added. Before measuring the absorbance at 595 nm, all samples and the standard solutions were incubated for 5 mins at room temperature and the spectrophotometer was calibrated with a cuvette containing a blank solution.
Using the protein concentrations that were measured, the encapsulation efficiency (E.E. %) was determined by considering the protein concentration in the initial feed solution.
(Top) Schematic setup and conversions to N-acetylneuraminic acid 3 with NAL-c-CLEnA-Conditions: A stock solution of N-acetyl-D-mannosamine, pH = 7 (1) and sodium pyruvate (2) (respectively 500 mM and 100 mM) in water (900 µL) was injected in a sample loop (1 mL) and pumped (at different flow rates) over the microporous hollow fibre reactor (mPES 10 kDa, 1.5 mL) loaded with genipin crosslinked NAL-c-CLEnA ( 0.34 g mL -1 of c-CLEnA, containing 18 mg of NAL) at various temperatures. The product flow was collected for at least 2.0 × t R . The difference in the activity observed could be explained by the effect of the changed ratio between enzyme concentration and cross-linker concentration ([enzyme:genipin]). When the amount of genipin was increased (1.6 wt%) the formation of cross-linked nano-aggregates was expected to occur faster but might have also caused deformation of the enzyme structure that resulted in a partial deactivation in case of 7α-HSDSH-c-CLEnA and in the total deactivation in the case of 7β-HSDSH. When less genipin was employed (0.75 wt%) the cross-linking rate was slower, which could have resulted in inefficient formation of nano-aggregates. In this case, 7β-HSDSH was completely deactivated while 7α-HSDSH-c-CLEnA maintained less than 5% of its native activity S-10