Solid-phase synthesis of protein-polymers on reversible immobilization supports

Facile automated biomacromolecule synthesis is at the heart of blending synthetic and biologic worlds. Full access to abiotic/biotic synthetic diversity first occurred when chemistry was developed to grow nucleic acids and peptides from reversibly immobilized precursors. Protein–polymer conjugates, however, have always been synthesized in solution in multi-step, multi-day processes that couple innovative chemistry with challenging purification. Here we report the generation of protein–polymer hybrids synthesized by protein-ATRP on reversible immobilization supports (PARIS). We utilized modified agarose beads to covalently and reversibly couple to proteins in amino-specific reactions. We then modified reversibly immobilized proteins with protein-reactive ATRP initiators and, after ATRP, we released and analyzed the protein polymers. The activity and stability of PARIS-synthesized and solution-synthesized conjugates demonstrated that PARIS was an effective, rapid, and simple method to generate protein–polymer conjugates. Automation of PARIS significantly reduced synthesis/purification timelines, thereby opening a path to changing how to generate protein–polymer conjugates.

DMSO, Sigma Aldrich) at 37 °C. The residual activity was calculated as a ratio of initial rates of the reaction at given incubation time over initial activity of native AChE (0.4 µM) in 100 mM sodium phosphate buffer (pH 7) at time zero. Rates were monitored by recording the increasing in absorption at 412 nm using a UV-VIS spectrometer.
Determination of Michaelis-Menten kinetics of AChE-pCBMA conjugates Acetylthiocholine iodide (0 -100 μL of 10 mM in 100 mM sodium phosphate buffer (pH 7.4)) and 10 µL of DTNB solution (50 mM in DMSO) was mixed with 100 mM sodium phosphate buffer (980 -880 μL, pH 7.4). Conjugate solution (10 μL, 5.5 µM of AChE) was added to the substrate solution. The initial substrate hydrolysis rate was monitored by recording the increase in absorbance at 412 nm using an UV-VIS absorbance spectrometer with a temperature-controlled cell holder at 37 °C. Michaelis-Menten parameters were determined by nonlinear curve fitting of initial rate versus substrate concentration plots using Enzfitter software.
Activity assays of Lyz-pCBMA conjugates Activity of Lyz-pCBMA conjugates was determined by two different substrates. Lyophilized Micrococcus lysodeikticus (Sigma Aldrich) was used to monitor enzymatic catalysis of cell wall lysis. 1 Absorption at 450 nm of suspended M. lysodeikticus (990 µL, 0.2 mg/mL) in 50 mM phosphate buffer (pH 6.0) was measured by UV-VIS spectrometer. 10 µL of Lyz-pCBMA solution (2.8 µM in 50 mM phosphate buffer (pH 6.0)) was added and the change of absorbance at 450 nm at room temperature was monitored.
p-nitrophenyl -glycoside of N-acetylchitooliosaccharide 2 was also used to determine the activity of the conjugates. To the solution of 4-Nitrophenyl β-D-N,N′,N′′-triacetylchitotriose (10 µL of 50 mM in DMSO, Sigma Aldrich) in 50 mM phosphate buffer (980 µL, pH 6.0), conjugate solution (10 µL of 714 µM in 50 mM phosphate buffer, pH 6.0) was added and the absorption was measured at 405 nm using an UV-VIS absorbance spectrometer with a temperature-controlled cell holder at 37 °C. The p-nitrophenol releasing rate was reported as hydrolysis activity of the conjugates.
Binding affinity of HABA to Avi-pCBMA conjugates 4'-hydroxyazobenzene-2-carboxylic acid (HABA, Sigma Aldrich) is a reagent that binds to Avidin and shows spectral changes, thus it is utilized to for determination of Avidin binding affinity. 3 Absorption at 500 nm of 300 µM HABA solution in phosphate buffered saline without calcium or magnesium (986 µL, Lonza) was measured using UV-VIS spectrometer. 16 µL of the conjugates solution (1.25 µM of Avidin in deionized water) was added to the HABA solution and incubated at room temperature for 1 min, and then absorption at 500 nm was measured. Change in absorbance at 500 nm was used to determine bound HABA to the conjugate.
Activity assay of Uox-pCBMA Enzymatic activity of the Uox-pCBMA conjugates was determined by oxidation of uric acid to allantoin. 4 Absorption at 290 nm of 50 µM uric acid in 20 mM sodium borate buffer (pH 8.5, 990 µL) was measured using an UV-VIS absorbance spectrometer with a temperature-controlled cell holder at 37 °C. Conjugate solution (10 μL, 57 µM of Uox in 20 mM borate buffer (pH 8.5)) was added to the substrate solution. The initial reaction velocity was monitored by recording the decrease in absorbance at 290 nm using the UV-VIS absorbance spectrometer at 37 °C. Activity of the conjugates (U/g) was determined from the initial velocity and concentration of Uox.

Impact of Agarase incubation for CT-pCBMA releasing from DMA beads To a suspension
of obtained CT-pCBMA beads (100 µL) in 100 mM sodium phosphate (pH 6, 99 µL), agarose solution (1 µL, 1 U/µL) was added and rotated at room temperature for a given time (0 -24 h). 20 mM sodium citrate (400 µL, pH 3) was added and rotated at room temperature for 1 h. The supernatant containing CT-pCBMA conjugate was separated from the beads by centrifugation.
Released active CT concentration in the supernatant was determined by an enzymatic activity assay on the hydrolysis of suc-AAPF-pNA using a standard curve with native CT.
A releasing study of CT-pCBMA was carried out by pre-incubation with Agarase to digest agarose 5 for quantitative recovery of the conjugate from the DMA beads. CT-pCBMA that was previously prepared on the DMA beads, was pre-incubated with Agarase (1 U/ 100 µL beads) in 100 mM sodium phosphate (pH 6) at room temperature for a designated time followed by incubation with 20 mM sodium citrate (pH 3). Released conjugates were quantified by an activity assay of hydrolysis of suc-AAPF-pNA. Agarase pre-incubation increased recovering the conjugate with incubation time (Supplementary Figure 23). We adopted the Agarase preincubation (1 U/100 µL of beads) in 100 mM sodium phosphate (pH 6) at room temperature overnight before adding 20 mM sodium citrate (pH 3) for releasing the conjugate from DMA beads.

Model Dye Binding and Release Studies
At pH 7 and 8, both the N-terminal and lysine mimic dyes quickly bound to the DMA beads, with maximum binding occurring within approximately 20 minutes. At pH 5 and 6, however, the binding rate of the N-terminal mimic, GGCy3, was an order of magnitude higher than that for the lysine mimic. Both GGCy3 and Cy5.5 amine showed increased initial binding rates with increased pH. At each pH 5 to 8 investigated, the initial binding rate of GGCy3 was higher than Cy5.5 amine indicating that the N-terminus α-amino had a higher binding affinity to the DMA beads, becoming even more pronounced at pH 5-6. Binding studies were performed by incubating the dye with the DMA beads at pH 5-8 over 60 minutes, washing with pH 8 phosphate buffer, releasing the dye at pH 3, and measuring fluorescence in the supernatant at various time points Release studies were also performed using the GGCy3 and Cy5.5 amine fluorescent model dyes as a function of pH and time. Model dyes that were previously immobilized on the DMA beads at pH 8 were incubated in pH 3 to 6 releasing citrate buffers for 60 minutes.
Supernatant fluorescence intensities were measured at various time points. Cy5.5 amine was rapidly released from the DMA beads with the maximum dye released within approximately 5 minutes at pH 3, 4, and 6 and within approximately 20 minutes at pH 5. GGCy3 also showed fast release at pH 3 and 4 within approximately 20 minutes, but slower release at pH 5 and 6.
Interestingly, Cy5.5 amine showed an order of magnitude higher initial release rate than GGCy3 at all pH from 3 to 6. Also, the release rate for both model dyes decreased as pH increased

Trypsin Digestion
As observed with CT, tryptic digestion studies of the acetylcholinesterase-initiator complex Boc-Gly-Gly-Cy3, oily compound, 1