Sustainable Flow Synthesis of Encoded Beads for Combinatorial

Publication Date (Web): July 3, 2018. Copyright © 2018 American Chemical Society. *E-mail: [email protected]. Cite this:ACS Comb. Sci. XXXX, XXX, XXX...
1 downloads 0 Views 1MB Size
Subscriber access provided by TUFTS UNIV

Article

Sustainable Flow Synthesis of Encoded Beads for Combinatorial Chemistry and Chemical Biology Hongxia Hu, Sergei Valeryevich Nikitin, Adam Bjørnholdt Berthelsen, Frederik Diness, Sanne Schoffelen, and Morten Meldal ACS Comb. Sci., Just Accepted Manuscript • DOI: 10.1021/acscombsci.8b00052 • Publication Date (Web): 03 Jul 2018 Downloaded from http://pubs.acs.org on July 4, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Combinatorial Science

Sustainable Flow Synthesis of Encoded Beads for Combinatorial Chemistry and Chemical Biology Hongxia Hu, Sergei Nikitin, Adam Bjørnholdt Berthelsen, Frederik Diness, Sanne Schoffelen and Morten Meldal* Center for Evolutionary Chemical Biology, Department of Chemistry University of Copenhagen Universitetsparken 5, 2100 Copenhagen, Denmark E-mail: [email protected]

TOC – for table of contents use only:

ACS Paragon Plus Environment

1

ACS Combinatorial Science 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 21

ABSTRACT: Monosized beads of polar resins were synthesized for combinatorial chemistry and chemical biology by sustainable microchannel flow synthesis. Regular, biocompatible and optically encoded beads could be efficiently prepared on large scale and in high yield. In a preparative flow polymerization instrument, taking advantage of a designed T-connector for droplet formation, quality beads were synthesized with accurate size control using a minimal amount of recirculating silicon oil as suspension medium. Bead-size was controlled through shear imposed by the silicon oil flowrate. This process provided 86 % yield of ~500 µm macrobeads beads within a 20 µm size range with no deformities or vacuoles, ideally suited for combinatorial chemistry and protein binding studies. The simple flow equipment consisted of a syringe pump for monomer and initiator delivery, a T-connector, a gear pump for oil recirculation, a long, heated coil of Teflon tubing and a collector syringe. The method was used for preparation of PEGA1900 beads, optically encoded with fluorescent microparticles. The microparticle matrix (MPM) encoded beads were tested in a MPM-decoder showing excellent recognition in bead decoding.

INTRODUCTION We introduced the biocompatible PEG-based resins for application, not only in synthesis, but also in aqueous protein chemistry and chemical biology. The advantage of the PEG-resins is that they have an open, highly porous structure and show little or no interaction with biomaterials.1-2 PEGA resin has been successfully used in numerous combinatorial screens of affinity3-4 and activity.5-10 The biocompatible nature of PEGA resins has changed the ways in which combinatorial on-bead screening is performed today. This includes on/in-bead assays for enzyme

ACS Paragon Plus Environment

2

Page 3 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Combinatorial Science

substrates11 and inhibitors,12 for GPCR active ligands with cells on beads13 as well as more conventional affinity and binding assays.4 On-bead screening is a powerful technology for identification of active molecules in combinatorial libraries generated by split-mix synthesis as introduced independently by Lam14 and Furka.15 However, screening beaded libraries have frequently been compromised by the lack of biocompatibility of the resins employed,16,

17

particularly in screens performed with larger proteins unable to enter the resin.2, 18 PEGA resins provide excellent biocompatibility and the porosity allows larger proteins to diffuse unhindered throughout the polymer matrix. The PEG based material swells excellently in buffered aqueous solutions and show little or no unspecific interaction with the biomolecules and even cells consider PEGA resin as a stealth material. Furthermore, the PEGA resin has no autofluorescence and can be used in fluorescence based assays at all the usual wavelengths of emission and excitation employed in screening. Apart from aqueous chemistry and biocompatibility the quality and homogeneous size of the polymer beads are crucial properties for the application in on-bead screening of synthetic peptides and oligosaccharides for biological activity.19-20 Beading of PEGA resins is conventionally performed in bulk stirred reactors by inverse suspension radical polymerization of macromonomers under zero gravity conditions with mixtures of mineral oil and carbon tetrachloride,21 which has been characterized as environmentally unsafe. Such inverse suspension polymerization is generally difficult to control and usually a rather heterogeneous distribution of polymer particles is obtained with significant amounts of imperfect beads and bead aggregates due to bead coalescence even under optimal polymerization conditions (Figure 1). Preparation of high-quality beads therefore requires sieving procedures

ACS Paragon Plus Environment

3

ACS Combinatorial Science 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 21

subsequent to polymerization with a significantly reduced yield as a result. We introduced cation ring opening polymerization of SPOCC resins in silicon oil22 in order to perform cation catalyzed ring opening reactions during beading in suspension and this approach could potentially also be a safer alternative for radical polymerization.23 Nano-droplet preparation in microchips24 indicated the useful properties of silicon oil as a suspension medium.25 Recently, porous PAM microparticles were prepared by polymerization in a paraffin suspension using microchannel technology.26 One-bead-one-compound libraries generated by the split-mix approach have frequently been employed for the screening and identification of new protein ligands.14 Active hits may be identified by analytical procedures such as MSMS or an encoding strategy may be used to identify the resin bound material. While analytical methods may work for simple peptide libraries, success in compound analysis rapidly decline with the complexity of the compounds. The use of a chemical or preferably optical encoding strategy may circumvent this problem as previously described. The most relevant methods for optical encoding includes in-situ modification of resin beads during reaction steps or post modification of the beads with the equivalent of a barcode providing each bead with an identity, which is recorded at each chemical modification and during compound activity screening. Microparticle matrix (MPM) encoding of macrobeads with fluorescent microparticles constitute one such barcode approach, the method of which has previously been described.27 During the assembly of a library by this technology, the identity of each bead is recorded during the split process prior to any chemical modification. Subsequently, during the on-bead screening process the identity of beads with active library members is recorded and correlated with each split record to provide the history of compound assembly. The “barcode” is provided a three dimensional matrix of fluorescent microparticles

ACS Paragon Plus Environment

4

Page 5 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Combinatorial Science

immobilized inside the MPM-encoded beads. This matrix can be rapidly determined with 3 orthogonally positioned cameras in front of a carousel handling the beads and computational reconstruction of 3D-space from the three 2D-images. This allow recording of the history of synthesis and screening result for each individual bead.28 The present flow synthesis provides a method to produce very regular MPM-encoded beads (Figure 1c) for combinatorial studies.

Figure 1. In conventional bulk polymerization (a) of biocompatible resins, polymer beads form simultaneously with many imperfect beads and broad size distributions. In contrast, the sustainable flow synthesis (b) provides significantly improved control over shape, size and bead quality. The image (c) illustrates the regular and narrow size distribution (415 ± 9.2 µm, see

ACS Paragon Plus Environment

5

ACS Combinatorial Science 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 21

Figure S10) of the 86% encoded macrobeads obtained through continuous flow synthesis in a microchannel flow synthesizer (Figure 2). RESULTS AND DISCUSSION Initially, synthesis of encoded macrobeads by inverse suspension droplet formation in silicon oil was attempted in a stirred bulk reactor. However, this method invariably provided beaded polymers showing some aggregation or a haze of soft polymer at the bead surface. Only tiny amounts of perfect encoded beads could be obtained from rather large volumes of silicon oil. Furthermore, in preparative synthesis of beads for combinatorial chemistry, where aggregated beads must be completely avoided, the amount of oil used in inverse suspension is 30-50-fold that of the volume of macromonomer solution. This excessive volume of suspension medium is to prevent the droplets that all simultaneously pass through the sticky phase of the polymerization reaction from colliding during this critical point. Based on initial experiments of controlled addition of macromonomer to the stirred reactor, we envisaged a method for “green” preparation of very regular beads, which would circumvent all of the issues encountered during conventional bulk inverse suspension polymerization. Thus evenly sized beads should preferentially be generated one by one in a flow channel and kept separate during the polymerization reaction using a flow in an extended coil of heated tubing, in a suspension medium of recirculating silicon oil. Here we describe a reproducible high yielding synthesis of biocompatible, regular macrobeads (See Figure 1c) with and without MPM-encoding, exceptionally useful for synthesis and biomolecular screening of split/mix libraries,14 for determination of protein binding constants4 and for diagnostics with bead bound compounds.

ACS Paragon Plus Environment

6

Page 7 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Combinatorial Science

Figure 2. The schematics of the continuous flow resin synthesizer. The macromonomer and radical initiator solutions are continuously fed from a syringe pump to a T-connector. After appropriate mixing, droplets are formed in a second Teflon T-connector with bead size rigorously controlled by the shear imposed by the silicon oil recirculation through the Tconnector at high rate. Beaded resins by sustainable flow-polymerization. In order to perform the flow synthesis of large macrobeads in the range of 300 – 1000 µm, we engineered a flow polymerization system as outlined in Figure 2. The equipment consists of a dual syringe pump connected via a Tconnector-1 and short mixer tubing to a second Teflon T-connector-229 for droplet formation. The T-connector-2 for droplet formation is connected to a long coil of Teflon tubing (ID 1.6

ACS Paragon Plus Environment

7

ACS Combinatorial Science 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 21

mm) immersed into a thermostat glycerin bath maintained at an optimal temperature for the radical polymerization reaction. The tubing was connected through a tight silicon stopper to a SPE-filter syringe for resin collection, which in turn was connected to a small gas-tight reservoir for collection of recirculating silicon oil. The reservoir was connected back to the Teflon Tconnector through a gear pump delivering the required accurate and stable flow and pressure to drive the process.

Scheme 1. Potassium persulfate initiated radical polymerization of PEG macromonomers and acrylamide with bis TBDMS-PEG1500 stabilization of droplets in silicon oil.

ACS Paragon Plus Environment

8

Page 9 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Combinatorial Science

The macromonomer of partially acryloylated bis-aminopropyl-PEG2 was mixed with water in a weight ratio of 1:4, with a small amount of acrylamide and with bis-di-t-butylmethylsilylPEG1500 acting as a droplet stabilizer. The radical persulfate initiator, K2S2O8 was dissolved separately in water and the solutions were mixed immediately prior to droplet formation using constant dosing from a dual syringe pump (Scheme 1). Prior to polymerizations all solutions and equipment were thoroughly purged with nitrogen and the solutions of macromonomer and initiator were subjected to 30 mbar vacuum for 15 min to remove any dissolved gas from the solutions.

Optimization of conditions. The optimal amount of potassium persulfate initiator was determined by assessing the mechanical properties of PEGA-polymers obtained in batch-wise polymerization reactions with a range of ratios of initiator to macromonomer. The strength of the polymer formed was assessed by the polymer breakpoint for forced entry of a metal rod into the polymer. It was found that there was an optimal ratio of initiator to acrylate moieties of 1:20, where the reaction proceeded at a reasonable rate and gave a tough polymer material. At higher ratios of initiator the polymer became soft due to short chain-lengths (see SI-Optimizations). The temperature was optimized in the range of 70 – 110 oC for the reaction in the heated Teflon coil and it was found that 84

o

C was optimal. Higher temperatures gave incomplete

polymerization reactions and at lower temperatures the reaction was too slow for continuous flow reactions.

ACS Paragon Plus Environment

9

ACS Combinatorial Science 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 21

The flow was optimized by variation of both the feeding rate for droplet formation and the rate of recirculation of silicon suspension medium, through the heated coil of Teflon tube and a Teflon T-connector (Figure 2, b-d through-hole ID 1.3 mm). The optimal flow of macromonomer feed and silicon oil recirculation was determined by several polymerizations where the yield, homogeneity, material properties, etc. were recorded for a variety of flow combinations. The combination of a flowrate of macromonomer solution of 4 mL/h and a linear recirculation rate through the T-connector of 250 mm/s was found to give the optimal balance of productivity as well as homogeneity and allowed continuous production over extended periods of time. The catalyst tetramethyl ethylenediamine (TEMED) was added directly to the recirculating silicon oil. This prevented the premature polymerization during the mixing of macromonomer solution with the radical initiator, previously considered a problem that needed control by photoinitiation.30

Size control. It was considered which factors could be in control of the exact bead size during flow synthesis of PEG-based macrobeads. The viscosity of the solution and the surface tension in presence of detergent are obvious inherent controlling factors in the system used for the polymerization. Two factors under external control are the dimension of the flow channels used for droplet formation. By application of a variety of T-connectors-2 with different bores of both the b-d through-hole recirculation channel and the a-c side-ways feeding channel revealed that the optimal diameter of the recirculation channel in the T-connector-2 depended on the diameter of the heated reactor coil. The two diameters should not not differ significantly in order to avoid turbulence during droplet formation. This was investigated using T-connectors with b-d throughhole diameters ranging from 500 – 1300 µm. Turbulence resulted in highly heterogeneous bead

ACS Paragon Plus Environment

10

Page 11 of 21

size distributions. The dimension of the perpendicular feeding channel (a-c in Figure 2) of the Tconnector, which provided the macromonomer droplets, was only found to influence the bead diameter marginally (Figure 3). In contrast the recirculation rate of silicon oil was crucial in determination of bead size.31 A practical range of linear flow rates from 130 − 300 mm/s through-hole were employed and provided uniform beads with size-control in the size range of 300 – 900 µm (Figure 3, Table S1). Bead size vs inlet diameter

Bead size vs flow rate

900

900

800

800 Bead size / µm

Bead size / µm

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Combinatorial Science

700 600 500

700 600 500

400

400

300

300 0.3

0.4

0.5

0.6

0.7

0.8

0.9

100

150

200

250

300

Lineart flow rate in mm/s

Inlet diameter in mm

Figure 3. Dependence of the average bead diameter on the diameter of the macromonomer feeding channel (a-c, Figure 2) (left) and recirculation flow rate (right), using a Teflon Tconnector-2 with through-hole diameter (b-d, Figure 2) of 1.3 mm. A diameter of 0.9 mm for the side-ways feeding channel (a-c, Figure 2) was used in flow rate experiments. While the diameter of the feeding channel (a-c, Figure 2) only influenced the bead size marginally, bead size could be readily be controlled by the recirculation flow rate. A through-hole flow rate of, for example, 250 mm/s provided clean formation of regular beads in the range of 400-450 mm.

ACS Paragon Plus Environment

11

ACS Combinatorial Science 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 21

Comparison of bulk suspension polymerization with flow synthesis. Suspension polymerization in a stirred reactor usually requires zero gravity conditions.2 This may be achieved by addition of a less dense organic component to the aqueous phase. However, organic solvents such as ethanol, although not mixable with silicon oil, gives a slight increase in mixing across the droplet interface resulting in a haze of soft polymer surrounding the polymerized beads and, associated with this, a tendency to form bead aggregates. A breakthrough that circumvented the requirement for zero gravity conditions was slow controlled addition of the macromonomer mixture to the silicon oil containing catalyst in a stirred bulk reactor. While larger amounts of macromonomers also aggregated with this procedure it allowed small amounts of macromonomer (12-15 mL) to be polymerized into high-quality encoded beads (Figure S2S4) that were excellent for the decoding process (table S2, Figure S9). However, the exorbitant use of oil required for proper bead quality in the bulk polymerization could be completely circumvented by application of the flow system in Figure 1b. This reproducibly provided quality beads. Furthermore, significantly improved size control could be established by flow synthesis. In the flow synthesis the volume ratio between oil and polymer was < 1.5 rather than ~50 as required in the conventional batch suspension process for production of regular beads without aggregation and deformation. Furthermore, several batches of beads could be obtained with the same oil. The uniform size of the beads prepared by flow synthesis (Figure 2, S6, S7 and S8) by far exceeded that of those obtained even in the best bulk polymerization reactions. MPM-encoding was achieved with 10 µm Tentagel microbeads (Scheme S1). These were first reacted with a 9:1 mixture of acryloyl chloride and fluorescent ATOTA-succinimide32 to introduce a chemically stable fluorophore and provide a reactive probe for immobilization within the polymer network of the macrobead formed during the polymerization reaction. The

ACS Paragon Plus Environment

12

Page 13 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Combinatorial Science

microparticle suspension was homogenized in macromonomer solution and any microparticle aggregates were removed by vortexing, short centrifugation and decantation.

Figure 4. Three sets of orthogonal images of beads, encoded with 5, 7 and 13 microparticles, respectively, obtained by flow synthesis and recorded with an MPM-decoder.

The microparticle concentration in the decanted stock suspension was determined by simple counting of particles in 2 µL of a 100 fold dilution under a fluorescence microscope prior to addition to the macromonomer solution to yield 8-12 microparticles per 500 µm macrobead (Figure 4).

ACS Paragon Plus Environment

13

ACS Combinatorial Science 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 21

MPM-encoded beads obtained by either controlled macromonomer addition to a stirred reactor or by flow synthesis both presented extremely confident bead identification according to a test decoding experiment (Table S2). A population of the encoded beads was passed through the decoder twice and the images were converted to 3-D-matrices of microparticle coordinates. The resulting coordinate matrices were compared and 76 % of the beads were confidently correlated with very few mistakes as indicated by the overlaid images in Figure S9. By repeating the recording of the identity of the hit beads 76 % of the 24% beads that were not identified in the first attempt is identified. Thus by 5 consecutive image recordings of each bead in a hit population, the confidence of recognition of regular beads in split recordings for each hit - split comparison is > 99.9 % ( 100 % * (1-0.24^5), assuming all beads have the same probability of recognition. In our hands the decoding result provides a very significant improvement over that of single bead mass spectrometry analysis.

Conclusions. Solid supports with a biocompatible nature are increasingly used in screening and in studies of biomolecular interactions for chemical biology. The synthesis of high-quality biocompatible polymer beads is therefore of interest to this community. We succeeded in designing an environment-friendly continuous flow process that can easily be established in any laboratory for production of uniform PEG-based polymer resins. The process is based on suspension in silicon oil and readily covered the daily use of resin even for a larger laboratory facility. The silicon oil may be reused and is collected for recycling in a process creating no waste. By performing droplet formation in a T-connector with a fast recirculating phase of silicon oil containing catalyst it was possible to generate a very high yield of mono-sized macrobeads using a minimal amount of suspension oil. The beads could furthermore be optically encoded. The high quality of optically encoded beads allowed for very high fidelity in the

ACS Paragon Plus Environment

14

Page 15 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Combinatorial Science

identification of hits isolated from combinatorial one-bead-one-compound (OBOC) split/mix libraries. The preparation of resins by flow polymerization can be performed with stringent control of bead size, important for resins used in protein binding studies and diagnostics. The process is sustainable because minimal amounts of recirculating non-toxic solvents are used in a near zero gravity radical suspension flow polymerization providing a high yield of mono-sized beads, requiring minimal bead-sorting and allowing recycling of suspension medium. Furthermore, all the materials used in the process are all readily available at low costs.

EXPERIMENTAL SECTION

Flow synthesis of encoded PEGA1900 resin. The polymerization was performed in the equipment outlined in Figure 2. Partially acryloylated bis-2-aminoprop-1-yl-polyethyleneglycol ((Acr0.9-NH-CH(CH3)CH2-)2-PEG1900 4.4 g, ~2 mmol), acrylamide (400 mg, 5.6 mmol) and the detergent (TBDMS)2-PEG1500 (30 mg, ~0.017 mmol) were dissolved in water (16 mL). Another identical macromonomer solution (5 mL) was used to suspend ATOTA-labelled Tentagel microbeads (10 µm, 10 % labelled, 5000 beads/mL). This suspension (1.2 mL, ~ 6.000.000 microbeads) was added to the first macromonomer solution. Silicon oil (100 mL) and tetramethylethylenediamine (TEMED, 0.4 mL, 2.9 mmol), was added to the recirculation reservoir. Potassium persulfate (120 mg, 0.44 mmol) was dissolved in water (6 mL). The three solutions were bubbled vigorously with nitrogen for 20 min and the receiver filter syringe was filled with nitrogen. The two aqueous solutions were degassed at 30 mbar for 15 min with agitation to remove any dissolved nitrogen. The temperature of the heating bath was set at 84 oC and the silicon oil was recirculated through the through-hole Teflon T-connector-2 (blocked at

ACS Paragon Plus Environment

15

ACS Combinatorial Science 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 21

the sideways macromonomer entry) at a local linear flowrate of 220 mm/s (gear pump set at 8, Figure S5). The degassed solutions were placed in syringes (22 and 6 mL, respectively, avoiding bubbles) and these were aligned in a syringe pump set at a flowrate of 3 or 4 mL/h. The mixing T-Connector-1 was attached to the syringes and the lines were allowed to fill completely prior to connection to the Teflon T-connector of the recirculation loop. The silicon oil recirculation was halted and flowrate of the gear pump was set to 0. The feed from the mixer line was connected without turning off the syringe pump and the setting of the gear pump increased stepwise to 8 over 20 sec to avoid excessive pressure build up. The flow polymerization was allowed to proceed for ~7 h until the syringes of the syringe pump were empty. The gear pump was halted; the feeding line to the T-connector-2 was disconnected and replaced with a stopper. Recirculation was continued at pump setting 8 for 30 min to cure the beads. The receiver filter was emptied from silicon oil and was transferred to a vacuum line where the beaded product was subjected to extensive suction followed by extensive washing with water while gently stirring with a plastic spatula. The resin was passed through a stack of sieves with a mesh of 1000-200 µm in 100 µm steps using a flow of water and the majority of beads (86 % yield) resided in the 400 µm sieve. Fifteen images of these beads were recorded and used to determine the average size and deviation of average for 300 beads (415 ± 9.2 µm, see Figure S10). The remaining 14 % of polymer had formed larger particles with a typical conical shape, probably due to occasional sticking and consequently imperfect forward movement of the syringe pistons. A representative aliquot of these beads was transferred to a microscope and imaged as presented in Figure 1. Similar beads were prepared without addition of Tentagel microbeads for protein binding studies. The beads were stored at 4 oC while solvated in water. The beads were mechanically

ACS Paragon Plus Environment

16

Page 17 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Combinatorial Science

quite stable, but they should not be frozen, lyophilised from water or subjected to ultrasound, which, as expected, all result in resin fragmentation.

The procedure for decoding of encoded beads is described in the experimental section of the supporting information and decoding data are presented in Table S2 and Figure S9.

ASSOCIATED CONTENT Table of Contents of Supporting Information Experimental procedures

S2

Materials and Instrumentation Synthesis of Bis-acrylamidopropyl Polyethylene Glycol (1900), (Acr)2PEG1900 Encoded beads by radical inverse suspension polymerization in silicon oil Optimizations Determination of functional loading in beads Decoding Scheme S1

The chemistry of the encoding process

S 8

Figure S1

Encoded beads obtained by bulk inverse radical polymerization

S 8

Figure S2

Encoded beads from controlled addition to a bulk reactor

S 9

Figure S3

Distribution of microparticles from controlled addition

S 10

Figure S4

Product from controlled addition to a stirred reactor

S 11

Figure S5

Determination of linear flowrates generated by gear pump

S 12

Table S1

Tabulated data associated with Figure 3

S 13

Figure S6

Flowsynthesis of encoded beads at 220 mm/s

S 14

Figure S7

Flowsynthesis of encoded beads at 140 mm/s

S 15

ACS Paragon Plus Environment

17

ACS Combinatorial Science 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 21

Figure S8

Image of beads presented in figure S6

S 16

Table S2

Data from the decoding of 323 encoded beads

S 17

Figure S9

Overlaid images from fitting of image sets for 323 beads

S 29

Figure S10

Measurement of bead size and s.d.o.m.

S 40

ACS Paragon Plus Environment

18

Page 19 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Combinatorial Science

AUTHOR INFORMATION Corresponding Author. Professor MortenMeldal, CECB, Department of Chemistry, University of Copenhagen, Universitetsparken 5, Room B304, 2100 Copenhagen, Denmark. Email : [email protected] Author Contributions. The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. Funding Sources. University of Copenhagen – CECB-Lighthouse Grant 2013 (Sergei Nikitin, Frederik Diness, Hongxia Hu, Morten Meldal). Free Research Council of Denmark, Sanne Schoffelen, Sino Danish Center – PhD support (Hongxia Hu, 1/3)

ACKNOWLEDGMENT We are grateful for support from the University of Copenhagen, DFF and Sino Danish Center. We would also like to thank Johannes Thomsen for scientific discussions and assistance.

ABBREVIATIONS Acr-, acryloyl; ATOTA-succinimide, N,N‘,N‘‘-pentaamethyl-N-(5-((2,5-dioxopyrrolidin-1yl)oxy)-5-oxopent-1-yltriaminotrioxatriangulenium; CECB, Center for Evolutionary Chemical Biology; ID, inner diameter; MPM, microparticle matrix; MSMS, tandem mas spectrometry; OBOC, one-bead-one-compound, PEG, polyethylene glycol; PEGA, polyethylene glycol polyacrylamide copolymer; TBDMS, t-butyldimethylsilyl; SPE, Solid Phase Extraction; SPOCC-resin, super-permeable organic combinatorial chemistry resin; TEMED, tetramethylethylenediamine

ACS Paragon Plus Environment

19

ACS Combinatorial Science 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 21

REFERENCES 1. Meldal, M. Pega - a Flow Stable Polyethylene-Glycol Dimethyl Acrylamide Copolymer for Solid-Phase Synthesis. Tetrahedron Lett. 1992, 33 (21), 3077-3080. 2. Auzanneau, F. I.; Meldal, M.; Bock, K. Synthesis, characterization and biocompatibility of PEGA resins. J. Pept. Sci. 1995, 1 (1), 31-44. 3. Meldal, M.; Wu, B.; Diness, F.; Michael, R.; Hagel, G. Metabolically stable cellular adhesion to inert surfaces. Chembiochem 2011, 12 (16), 2463-70. 4. Li, M.; Hoeck, C.; Schoffelen, S.; Gotfredsen, C. H.; Meldal, M. Specific Electrostatic Molecular Recognition in Water. Chem. Eur. J. 2016, 22 (21), 7206-14. 5. Renil, M.; Ferreras, M.; Delaisse, J. M.; Foged, N. T.; Meldal, M. PEGA supports for combinatorial peptide synthesis and solid-phase enzymatic library assays. J. Pept. Sci. 1998, 4 (3), 195-210. 6. Meldal, M.; Svendsen, I. Direct Visualization of Enzyme-Inhibitors Using a Portion Mixing Inhibitor Library Containing a Quenched Fluorogenic Peptide Substrate .1. Inhibitors for Subtilisin Carlsberg. J Chem. Soc. Perk. T. 1 1995, (12), 1591-1596. 7. Meldal, M.; Auzanneau, F. I.; Hindsgaul, O.; Palcic, M. M. A Pega Resin for Use in the Solid-Phase Chemical-Enzymatic Synthesis of Glycopeptides. J. Chem. Soc. Chem. Comm. 1994, (16), 1849-1850. 8. Leon, S.; Quarrell, R.; Lowe, G., Evaluation of resins for on-bead screening: a study of papain and chymotrypsin specificity using PEGA-bound combinatorial peptide libraries. Bioorg. Med. Chem. Lett. 1998, 8 (21), 2997-3002. 9. Comellas, G.; Kaczmarska, Z.; Tarrago, T.; Teixido, M.; Giralt, E. Exploration of the one-bead one-compound methodology for the design of prolyl oligopeptidase substrates. PLoS One 2009, 4 (7), e6222. 10. Qvortrup, K.; Petersen, R. G.; Dohn, A. O.; Moller, K. B.; Nielsen, T. E. SolventControlled Chemoselectivity in the Photolytic Release of Hydroxamic Acids and Carboxamides from Solid Support. Org. Lett. 2017, 19 (12), 3263-3266. 11. St Hilaire, P. M.; Willert, M.; Juliano, M. A.; Juliano, L.; Meldal, M. Fluorescencequenched solid phase combinatorial libraries in the characterization of cysteine protease substrate specificity. J. Comb. Chem. 1999, 1 (6), 509-23. 12. Buchardt, J.; Schiodt, C. B.; Krog-Jensen, C.; Delaisse, J. M.; Foged, N. T.; Meldal, M. Solid phase combinatorial library of phosphinic peptides for discovery of matrix metalloproteinase inhibitors. J. Comb. Chem. 2000, 2 (6), 624-38. 13. Meldal, M.; Nielsen, T. E.; Hagel, G.; Kaznelson, D. W.; Diness, F.; Thastrup, O. Identification of compounds modifying a cellular response. 2005. Patent: PCT/DK2005/000348, WO/2005/116643. 14. Lam, K. S.; Salmon, S. E.; Hersh, E. M.; Hruby, V. J.; Kazmierski, W. M.; Knapp, R. J. A new type of synthetic peptide library for identifying ligand-binding activity. Nature 1991, 354 (6348), 82-4. 15. Furka, A.; Sebestyen, F.; Asgedom, M.; Dibo, G. General method for rapid synthesis of multicomponent peptide mixtures. Int. J. Pept. Protein Res. 1991, 37 (6), 487-93. 16. Lam, K. S.; Lebl, M.; Krchnak, V. The "One-Bead-One-Compound" Combinatorial Library Method. Chem. Rev. 1997, 97 (2), 411-448. 17. Kang, H.; Jeong, S.; Koh, Y.; Geun Cha, M.; Yang, J.-K.; Kyeong, S.; Kim, J.; Kwak, S.Y.; Chang, H.-J.; Lee, H.; Jeong, C.; Kim, J.-H.; Jun, B.-H.; Kim, Y.-K.; Hong Jeong, D.; Lee,

ACS Paragon Plus Environment

20

Page 21 of 21 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Combinatorial Science

Y.-S. Direct identification of on-bead peptides using surface-enhanced Raman spectroscopic barcoding system for high-throughput bioanalysis. Sci. Rep. 2015, 5, 10144. 18. Camperi, S. A.; Marani, M. M.; Iannucci, N. B.; Cote, S.; Albericio, F.; Cascone, O. An efficient strategy for the preparation of one-bead-one-peptide libraries on a new biocompatible solid support. Tetrahedron Lett. 2005, 46 (9), 1561-1564. 19. Collot, M.; Eller, S.; Weishaupt, M.; Seeberger, P. H. Glycosylation efficiencies on different solid supports using a hydrogenolysis-labile linker. Beilstein J. Org. Chem. 2013, 9, 97105. 20. Garcia-Martin, F.; Quintanar-Audelo, M.; Garcia-Ramos, Y.; Cruz, L. J.; Gravel, C.; Furic, R.; Cote, S.; Tulla-Puche, J.; Albericio, F. ChemMatrix, a poly(ethylene glycol)-based support for the solid-phase synthesis of complex peptides. J. Comb. Chem. 2006, 8 (2), 213-20. 21. Arshady, R.; Ledwith, A. Suspension polymerisation and its application to the preparation of polymer supports. React. Polym. 1983, 1, 159-174. 22. Grotli, M.; Rademan, J.; Groth, T.; Lubell, W. D.; Miranda, L. P.; Meldal, M. Surfactant Mediated Cationic and Anionic Suspension Polymerization of PEG-Based Resins in Silicon Oil: Beaded SPOCC 1500 and POEPOP 1500. J. Comb. Chem. 2001, 3 (1), 28-33. 23. Wang, X.; Ding, X.; Zheng, Z.; Hu, X.; Cheng, X.; Peng, Y. Magnetic molecularly imprinted polymer particles synthesized by suspension polymerization in silicone oil. Macromol. Rapid Comm. 2006, 27 (14), 1180-1184. 24. Park, J. I.; Saffari, A.; Kumar, S.; Günther, A.; Kumacheva, E. Microfluidic synthesis of polymer and inorganic particulate materials. Annu. Rev. Mater. Res. 2010, 40, 415-443. 25. Kim, J. W.; Utada, A. S.; Fernandez-Nieves, A.; Hu, Z.; Weitz, D. A. Fabrication of monodisperse gel shells and functional microgels in microfluidic devices. Angew. Chem. Int. Ed. Engl. 2007, 46 (11), 1819-22. 26. Guo, S.; Yao, T.; Wang, C.; Zeng, C.; Zhang, L. Preparation of monodispersed porous polyacrylamide microspheres via phase separation in microchannels. React. Funct. Polym. 2015, 91, 77-84. 27. Meldal, M., Christensen, S. F., Micro-Particle Matrix encoding of beads. Angew. Chem. Int. Ed. Engl. 2010, 49, 3473-3476. 28. Rasmussen, J. E.; Schiodt, C. B.; Christensen, S. F.; Norskov-Lauritsen, L.; Meldal, M.; St Hilaire, P. M.; Jensen, K. J. Small-molecule affinity ligands for protein purification: combined computational enrichment and automated in-line screening of an optically encoded library. Angew. Chem. Int. Ed. Engl. 2010, 49 (20), 3477-80. 29. Serra, C. A.; Chang, Z. Microfluidic‐assisted synthesis of polymer particles. Chem. Eng. Technol. 2008, 31 (8), 1099-1115. 30. Tong, D.; Yesiloz, G.; Ren, C. L.; Madhuranthakam, C. M. R. Controlled Synthesis of Poly (acrylamide-co-sodium acrylate) Copolymer Hydrogel Microparticles in a Droplet Microfluidic Device for Enhanced Properties. Ind. Eng. Chem. Res. 2017, 56 (51), 14972-14979. 31. Nisisako, T.; Torii, T.; Higuchi, T. Novel microreactors for functional polymer beads. Chem. Eng. J. 2004, 101 (1-3), 23-29. 32. Laursen, B. W.; Sorensen, T. J. Synthesis of super stable triangulenium dye. J. Org. Chem. 2009, 74 (8), 3183-5.

ACS Paragon Plus Environment

21