Spatial Separation of Microbeads into Detection Levels by a

Subscriber access provided by Nottingham Trent University

Article

Spatial Separation of Microbeads into Detection Levels by a Bioorthogonal Porous Hydrogel for Size-Selective Analysis and Increased Multiplexicity Anna Herrmann, Stefan Jörg Rödiger, Carsten Schmidt, Peter Schierack, and Uwe Schedler Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.9b01586 • Publication Date (Web): 07 Jun 2019 Downloaded from http://pubs.acs.org on June 7, 2019

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 17 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

Analytical Chemistry

Spatial Separation of Microbeads into Detection Levels by a Bioorthogonal Porous Hydrogel for Size-Selective Analysis and Increased Multiplexicity

Anna Herrmann,1# Stefan Rödiger,2# Carsten Schmidt,2 Peter Schierack,2 Uwe Schedler1,3* 1

Institut für Chemie und Biochemie, Freie Universität Berlin, Takustraße 3, 14195 Berlin, Germany 2

Brandenburgische Technische Universität Cottbus-Senftenberg, Universitätsplatz 1, 01968 Senftenberg, Germany 3

PolyAn GmbH, Rudolf-Baschant-Straße 2, 13086 Berlin, Germany

#

These authors contributed equally.

Correspondence: [email protected]

Abstract Multiplex detection techniques are emerging within the fields of life science research and medical diagnostics where it is mandatory to analyze a great number of molecules. The detection techniques need to be highly efficient but often involve complicated and expensive fabrication procedures. Here, we present the immobilization and geometric separation of fluorescence-labeled microbeads for a multiplex detection in k levels. A compound of differently sized target molecules (DNA, proteins) is channeled into the respective detection levels by making use of a hydrogel as a size selective filter. The immobilized microbeads (10-20 µm) are considerably larger than the pores of the hydrogel network and therefore stay fixed at the well bottom and in higher elevations, respectively. Small biomolecules can diffuse through the pores of the network, whereas medium-sized biomolecules pass slower and large molecules will be excluded. Besides filtering, this method discriminates the used microbeads into k levels and thereby introduces a geometric multiplexity. Additionally, the exclusion of large entities enables the simultaneous detection of two target molecules, which exhibit the same affinity interaction. The hydrogel is formed through the combination of two macromonomers. One component is a homobifunctional polyethylene glycol linker, carrying a strained alkyne (PEG-BCN) and the second component is the azide-functionalized dendritic polyglycerol (dPG-N3). They react via the bioorthogonal strain-promoted azide alkyne cycloaddition (SPAAC). The hydrogel creates a solution-like environment for the diffusion of the investigated biomolecules all the while providing a stable, bioinert, and surface bound network.

1 ACS Paragon Plus Environment

Analytical Chemistry 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 17

Introduction The demand for speed, increased sensitivity, and lower costs is driving miniaturization and multiplex technologies. Multiplex detection techniques are emerging in the fields of clinical diagnostics, point-of-care testing, personalized medicine, and pharmaceutical research, with an important objective to understand interaction processes at the molecular level.1,2 Multiplexity can be achieved by geometrical coding, color coding, magnetic coding, or barcodes, only to name a few.3-6 Assays based on geometric coding are 2-dimensional, i.e., the sensor molecules are located at different positions in the x,y plane and get thereby distinguishable. Platforms, which are most commonly used, are microtiter plates or microarrays. Assays based on color coding introduce different dyes to either the targets or the substrate and hence discriminate several probes in a single testing. For instance, fluorescence microbeads of different sizes and equipped with dyes are distinguished by flow cytometry, microfluidics or with fluorescence microscopy-based detection platforms.7-9 An important goal of our work was the extension of such microparticle assays into the 3rd dimension to achieve a x,y,k-coding. The fluorescence microbeads are therefore not only placed into different cavities on a planar substrate but also into k elevations in z-direction and analyzed by a fluorescence microscope. This increases the degree of multiplexity by the number of elevations in which the fluorescence microbeads are arranged compared to a classical assay. The separation of each level is achieved by embedding the microbeads in a polymeric 3D hydrogel network. Hydrogels are soft materials consisting of natural or synthetic hydrophilic polymers that are crosslinked by either chemical or physical methods. They highly swell in water without dissolving while preserving their 3-dimensional structure. Hence, hydrogels provide a solution like environment throughout the pores for the diffusion of small molecules. These characteristics make hydrogels not only a great candidate for drug delivery,10 as injectable therapeutics11 and tissue engineering,12 but also for biosensor materials.13 Recently, a group of the authors developed the herein used hydrogel based on two components. A highly branched polyglycerol and a linear polyethylene glycol linker which were reacted by the bioorthogonal strain-promoted azide-alkyne cycloaddition (SPAAC) and used as a versatile matrix for biosensors with enhanced sensitivity.14 The size of the pores can be controlled and to some extent adapted to the size of target molecules. The reaction is fast, proceeds at room temperature, and is stable against all common buffers. In this research project we want to utilize the hydrogel to introduce a 3-dimensional coding and exploit the filtering effect of such a porous network to control the pathway of differently sized target molecules. A biomolecule mixture consisting of an IgG-type antibody, a Fab fragment and two oligonucleotides of different lengths was selected because of its differences in size and flexibility. Thus, immunoassays and hybridization assays were combined in one multiplex assay. Antigen-binding fragments (Fab fragments) are the part of an antibody that interacts with the antigen and is therefore much smaller than a whole antibody. They are an important class of functional biomolecules, which are used in research for the prediction of antigen-antibody interaction kinetics, in immunotherapy and bioanalytical applications.15,16 Contrary to wholesize antibodies the modified (e.g., biotinylated) Fab fragments show superior antigen-binding performance. This is mainly due to their smaller size, which leads to a higher density of the antigen binding sites on the surface and correct orientation.17 The interpenetration of the diffusing species and the polymer matrix varies with the flexibility of the diffusing species.18 This effect was studied by using two kinds of biomolecules. The heavy and light chains of both 2 ACS Paragon Plus Environment

Page 3 of 17 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


Fab fragment and antibody are interconnected by disulfide bonds. Therefore, they exhibit an overall more rigid structure than the oligonucleotide sequences do.19,20

Experimental Section The used chemicals were synthesized in our lab. The labeled biomolecules as well as the labeled microbeads were obtained from commercial sources. The VideoScan technology is a fluorescence microscopy-based technology developed in our laboratory and is described elsewhere.7,21,22 Microbead functionalization and detection. Carboxylated polymethylmethacrylate microbeads (PolyAn, Berlin) were used. They carry two coding fluorescent dyes (Rhodamin 6G and Coumarin 6), which are detected in two channels (em,1 = 489-531 nm, em,2 = 550610 nm) by the VideoScan software. By employing a certain ratio of these dyes and varying the size of the microbeads, a multiplex detection is enabled by distinctly assigning the microbead populations to the respective probe molecules.23 In our new method presented here, we replaced this color/size coding and introduced the distinction of different probes on the microbead by placing them into different detection levels (k1 and k2) in z-direction. This way we accomplished a compound analysis in k elevations (CAkE) without the need of many different microbead populations. Consequently, we used the same microbead population and functionalized the microbeads with an antibody (µB-Ab) and also with a fluorescence labeled oligonucleotide (µBSA-Ofl), using the strong streptavidin-biotin interaction for immobilization. For detection and quantification, the microbeads are incubated in target solution, the microbead corona was scanned in a third channel (em,3 = 662-737 nm), and the fluorescence intensity of the biomolecules is determined. In case of µBSA-Ofl, the maximum fluorescence intensity of the immobilized probe was quenched during incubation with the target sequences (150mer and 19mer, 20 µL, 100 nM), which carried the quencher (BHQ2). For µB-Ab, however, the initial intensity rose during incubation due to an affinity interaction between probe and targets (Ab * and Fab*, 20 µL, 0.8 µg/mL). During incubation the cavities were covered by 30 µL of oil to prevent evaporation of water and to maintain a constant and reproducible concentration. A stock solution of μBSA-Ofl and μB-Ab were stored at 8 °C. Table 1 gives a summary of all substrates and the probes with their respective targets. Scheme 1 shows the functionalization of the microbeads as well as their detection. For more details see Supporting Information.



Page 4 of 17

Table 1 Overview over the used microbead substrates and assignment of probes, targets and specification of the used labels. For more details please see Supporting Information.

Substrate (Abbreviation)

Probe (Abbreviation)

Probe Label

Target (Abbreviation)

Target Label

Ofl

Atto 647N

19mer

µBSA-Ofl

µBSA-Ofl

Ofl

Atto 647N

150mer

BHQ2, intern T132

µB-Ab

Ab

-

Ab*

Alexa Fluor 647

µB-Ab

Ab

-

Fab*

Alexa Fluor 647

BHQ2

Scheme 1 Reaction scheme for the functionalization of carboxylated microbeads by carbodiimide crosslinker (EDC) chemistry with both SA (top) and Ab (bottom). The streptavidin beads are further functionalized with a biotin-labeled oligonucleotide sequence (Ofl). The µBSA-Ofl microbeads were later incubated with either 150mer or 19mer to quench the fluorescence signal; the µB-Ab were incubated with either Ab* or Fab* to obtain a fluorescence signal.


Page 5 of 17 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


Encapsulation of microbeads by dendritic polyglycerol-based hydrogel. The two macromonomers homobifunctional polyethylene glycol-bicyclo[6.1.0]non-4-yne (PEG-BCN) and dendritic polyglycerol azide (dPG-N3) were synthesized as previously described.14,24 Hydrogel samples were prepared by crosslinking PEG-BCN with dPG-N3 via strain-promoted azide-alkyne cycloaddition (SPAAC). This Click reaction works in all common buffers, at room temperature, without the addition of a catalyst, and without further external activation like heat or UV (for the chemical structure see Supporting Information). We worked with two different sizes of PEG-BCN, one with the molecular weight of 6 kDa to yield hydrogel dPG-G6 and secondly a PEG chain of 10 kDa to give dPG-G10. The ratio of the linear PEG-BCN to the multivalent sphere-like crosslinker dPG-N3 (MW = 3 kDa, 20% azide) was kept constant at 3:1. The two components were mixed as aqueous solution and thoroughly vortexed before the hydrogel solution was distributed into the cavities of Nunc™ NucleoLink™ 8-well strips. They already contained a suspension of microbeads in TBS-T buffer (Tris-Buffered Saline with 0.01% Tween 20, pH 7.4) which was mixed with the hydrogel solution. The concentration of the hydrogel in solution was adjusted to 4 wt% so that the gelation was slow enough for the microbeads to sediment before the hydrogel formed and captured the microbeads at the bottom of the cavity. The hydrogel formation proceeded for 2 h before measurements could be performed. However, it was found that the non-covalent adhesion of the hydrogel in the cavity was improved if the gel was left to dry before being reswollen by the addition of PBS, water or TBS-T buffer (30 µL). In the cases of height determination or swelling/deswelling experiments unfunctionalized microbeads were used. Washing (scooping out the solution, adding 50 µL buffer, agitation, scooping out, and removing surface water by tapping on blotting paper, repeat) of the entrapped microbeads is possible but not necessary for classical evaluations and was therefore only performed, when we wanted to ensure that the hydrogel matrix had securely entrapped all the microbeads. In the cases where we claim exclusion of large molecules by the hydrogel matrix, we performed control assays without the hydrogel. Here, the microbeads were suspended in 30 µL of buffer and incubated with the same volume of target solution as in the hydrogel case.

Introducing geometric multiplexity by a 3D-coding for compound analysis in k elevations (CAkE). The k levels were prepared layer by layer, incubated with the analyte, and analyzed in levels k1 and k2 with the VideoScan Software by quantifying the fluorescence intensity of the microbead corona. For the first layer k1, the microbeads were suspended in a diluted gel solution and sedimented before the first hydrogel layer dPG-G6 formed on top of the microbead level. The hydrogel formation was left for 1 h at room temperature. Then 20 µL of TBS-T buffer was added and the next detection level of microbeads was disseminated on top. This second level k2 was covered by a second hydrogel layer dPG-G10 which was prepared according to the aforementioned protocol. The protocol was adapted concerning the amounts added and the waiting time to achieve a smooth surface of the underlying hydrogel layer Gn-1 before the next microbead detection level is arranged. This is crucial because the microbead levels are focused automatically by the VideoScan software and, hence, need to be positioned as planarly as possible. If the hydrogel surface is too rough, no focal plane can be found. The prepared multilayers could then be used directly for a CAkE or be stored in buffer as well as in dry condition. For the incubation with target molecules we used 15 µL for each target. To obtain an overall volume of 45 µL for incubation TBS-T buffer was added if necessary. 5 ACS Paragon Plus Environment


Page 6 of 17

Swelling experiments. Two levels (k1, k2), loaded with different microbead populations, were separated by dPG-G6 and dPG-G10, respectively. The microbeads in k1 were focused with the fluorescence microscope at a certain z position. The difference to the z-value of microbeads in level k2 is taken as the distance between k1 and k2. The different populations in k1 and k2 allowed us to be sure that the respective microbeads always stayed in separated levels, even in the collapsed, dry state. To relate the z-distance of the microscope to an actual distance in [µm] we determined a correction factor. Therefore, we took a calibration slide developed in our lab of which we knew the exact thickness in [µm], and which had a cavity to fill with water, TBS-T buffer or with hydrogel. The calibration slide also carried a focusable mark at the top and at the bottom. The z-distance between these marks was determined with the microscope and then the height was divided by this value. We calculated the correction factor for air, water, TBS-T buffer, and the hydrogel dPG-G6 to be 0.99 ± 0.02 µm/a.u., 1.38 ± 0.02 µm/a.u., 1.42 ± 0.01 µm/a.u., and 1.39 ± 0.02 µm/a.u., respectively. The procedure was performed three times for each medium and values are given as mean ± standard deviation. We are aware that this is a rough estimation and only valid in a certain size regime, but the results are in good agreement with general assumptions (~1 µm/a.u. for air and ~1.4 µm/a.u. for liquid media) and exhibited good reproducibility. For the mass swelling ratio hydrogel samples of 150 µL (n = 4) of both hydrogels, dPG-G6 and dPG-G10, were prepared, and the swelling was investigated both in solution (qsol) and on surface adhered hydrogels (qsurf). qsurf was performed in the Nunc™ NucleoLink™ cavities. The pads for qsol were prepared in a casting mold and then transferred into a larger entity for incubation. All samples were initially dried and then rehydrated through the addition of an excess amount of water and kept hydrated for three days. After that, the water from the cavities for qsurf was scooped out and carefully tapped on blotting paper to remove surface water and weighed. The samples incubated in bulk solution were taken out of the water and tapped softly on blotting paper before being weighed. Thereafter, all samples, for qsurf and qsol, were freezedried and the dried mass was determined. The swollen mass (ms) divided by the dry mass (md) yields the mass swelling ratio qsurf and qsol, respectively (cf. Equation 1).25 In both cases, and for each hydrogel dPG-G6 and dPG-G10, 4 samples were prepared, q determined, and the data are presented as mean ± standard deviations of the replicates. The swollen mass was also revised on day 4 and 5, with no further increase in mass. 𝑞𝑠𝑢𝑟𝑓/𝑠𝑜𝑙 =

𝑚𝑠 𝑚𝑑

1

Mesh size determination and rheological measurements. Rheological experiments were performed on a Malvern Instrument Kinexus in parallel plate geometry with an 8 mm upper plate geometry at 25 °C. For the gelation time tg we determined the storage modulus (G’) and loss modulus (G’’) at 1% strain and at 1 Hz frequency and took the crossover of G’ and G’’ as gelation point. Each hydrogel was measured 3 times and both hydrogels were used at the same concentration of 7.7 wt%. For the mesh size determination, we performed a frequency sweep from 0.1-10 Hz at a constant strain of 1%. G’ was determined as the average of at least 5 points were G’ remains constant and the experiments were repeated three times. The hydrogel samples dPG-G6 and dPG-G10 were prepared to yield discs of 8 mm and subsequently swollen to equilibrium overnight in MilliQ water. To prevent solvent evaporation during measurements 6 ACS Paragon Plus Environment

Page 7 of 17 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


we implemented a water trap under the cover of the measuring device. We derived an equation for estimating the mesh size 𝜉 from G’ and the volume swelling ratio by employing equation 2.26,27 For detailed derivation of this equation see the Supporting Information. 3𝐶𝑛 𝑅𝑇𝑑𝑀𝑛 1/3 𝜉 = 𝑞𝑠𝑜𝑙 ∙ 𝑙 ∙ √ ( 1/3 ) 𝑀𝑟 𝐸𝑞 𝑀𝑛 + 2𝑅𝑇𝑑 2

Considering the spatial extent of PEG in comparison to dPG (cf. Figure S3) the main contributor to the pore size is the linear PEG-BCN. Hence, 𝑙 is the average bond length of the polymer (0.147 nm), 𝑀𝑟 is the molecular weight of a repeat unit (44 g/mol), 𝐶𝑛 is the characteristic ratio or rigidity factor (𝐶𝑛 = 4 for PEG),28 𝐸 is Young’s modulus estimated by the elastic shear modulus G’ of the linear viscoelastic regime, 𝑞𝑠𝑜𝑙 is the swelling ratio, 𝑀𝑛 is the number average molecular weight of the primary chains, in our case the functionalized PEG-BCN, 𝑅 is the absolute gas constant, 𝑇 the temperature during the experiments (298 K), and 𝑑 the density of the polymer. Results and Discussion Entrapment of the microbeads. Initially, we made sure that a hydrogel layer that was without holes and rifts really captured all microbeads whose position did not change during incubation, washing, drying, and resuspension procedures. Therefore, the series shown in Figure 1 was taken and the position of the fluorescence microbeads was followed. The top row shows a cavity, where the microbeads were captured by the hydrogel (µB-gel) and the bottom row shows the microbeads suspended in solution (µB-sol). In each row the pictures follow the same well, the pictures (A)-(D) were taken after the stated procedures. Both wells with µB-gel and µB-sol were exposed to the same protocol in parallel. Throughout the whole experiment the µB-gel did not change their position and every single microbead remained at its position even after four rounds of washing. However, the µB-sol aggregated upon drying, rearranged after rehydration, and all microbeads were lost after a single round of washing. With this experiment we demonstrated that the hydrogel fixes the microbeads at their initial position and its non-covalent interaction with the well surface is strong enough to sustain mechanical and physical stresses. Additionally, we did not lose a single µB, which means that the hydrogel layer is covering the whole level. The microbeads do not fall through the meshes, which assures a complete segregation from following levels. Because of this tight packing, we hypothesized that the hydrogel components could be too densely associated to the microbead surface and might have disturbed the hybridization or other affinity interactions at the microbead surface during detection. Therefore, we diluted a target solution of a 19mer and compared the detection limit of µB-Ofl in solution with the detection limit of µB-Ofl covered with a hydrogel. We found that the detection was not particularly influenced by the hydrogel (see Supporting Information).



Page 8 of 17

Figure 1 Fluorescence pictures of the position of microbeads at the well bottom taken at different time points during preparation. The top row shows the microbeads entrapped by the hydrogel (µB-gel) whereas the bottom row follows the microbeads suspended in solution (µB-sol): (A)Arrangement of the microbeads (A) initially at the well bottom after preparation, (B) after the well has been left to dry at 37 °C, (C) after the addition of buffer to rehydrate the system and finally (D) after washing. The diameter of the well bottom is 3 mm.

Size selective detection of biomolecules Before a mixture of several biomolecule classes (DNA, protein) was applied, the feasibility to exclude large molecules while the small ones could diffuse was tested in two approaches. We used a fluorescence-labeled Fab fragment (Fab*, 50 kDa) as a medium-sized and an IgG type antibody (Ab*, 150 kDa) as the large diffusing species. Due to their structure, stabilized by hydrophobic interactions as well as hydrogen bonds, these biomolecules exhibit a rigid structural framework and are less deformable.18,29 In a second approach, the more flexible system of two quencher-labeled target oligonucleotide sequences of 19 base pairs (19mer, 6 kDa) and 150 base pairs (150mer, 46 kDa) were chosen to see the possibilities and limits of the size selectivity. 19mer/150mer. Here, the µBSA-Ofl were used in all cases and covered with the hydrogel (G) or just suspended in solution (sol). The relative fluorescence intensity of the probe oligonucleotide sequence Ofl is displayed after 2 h of incubation in Figure 2 A. Values were normalized to the initial fluorescence intensity which diminishes by hybridization. The fluorescence of the dye-labeled probe sequence was quenched by the quencher-labeled target sequences. The detection performed in solution showed no significant difference in the intensity between the 150mer and the 19mer (sol_150 and sol_19). However, if a hydrogel layer covered the µBSA-Ofl, the fluorescence intensities after incubation exhibited a significant difference for the 150mer and 19mer (cf. G_150 and G_19). This means the two oligonucleotide sequences can be discriminated in the same cavity if the hydrogel is introduced. However, for the flexible oligonucleotides this is not an infinite exclusion. If the incubation is not interrupted, the flexible 150mer manages to penetrate the network. See Figure S4 in the Supporting Information for the kinetic plot and for a schematic presentation of the experiment.


Page 9 of 17 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


A

B

Figure 2. (A) shows the fluorescence intensity of µBSA-Ofl after 2 h incubation. (B) shows the fluorescence intensities of µB-Ab after 24 h and 6 days. The filled cyan and red columns give the intensity after incubation with Fab* and Ab*, respectively. As a positive control the µB-Ab were incubated with Fab* (open cyan column) and Ab* (open red column) in solution. Mean values are presented, and error bars reflect the standard deviation (n = 3).

Fab*/Ab*. Here, the µB-Ab were used in all cases, covered with the hydrogel and incubated with 20 µL of fluorescence-labeled biomolecule solution. The fluorescence intensity is depicted after 24 h and after 6d in Figure 2 B and over time for both hydrogels dPG-G6 and dPG-G10 in Figure S4 (Supporting Information). The Fab* slowly but steadily penetrated both gels dPG-G6 and dPG-G10. The diffusion of Fab* in dPG-G10 was slightly faster than in dPG-G6 (Supporting Information), but, more importantly the Ab* was excluded in both cases. To investigate if this held true for longer diffusion times or if the Ab* would have eventually penetrated the network, we performed the assay for up to 6 days. µB-Ab were incubated both in solution (open columns) and covered by dPG-G6 with Fab* and Ab*(filled columns), respectively. The intensities found for Fab* (cyan) showed that we got a detectable signal. However, the intensities shown in red for the Ab* assure that the antibody would not penetrate the hydrogel even after 6 days. This enabled the detection of a Fab fragment next to an antibody which have the same affinity interaction with the same substrate in the same well. This was achieved by applying a hydrogel that selectively lets medium-sized molecules (Fab*) diffuse while the larger ones (Ab*) were excluded. Additionally, it proved that the hydrogel remained 9 ACS Paragon Plus Environment


Page 10 of 17

stable and was not degraded, had no ruptures, and was tightly packed into the cavity, even with longer incubation times. See Figure S4 in the Supporting Information for a schematic presentation of the experiment.

Table 2 Presentation of the mesh sizes of dPG-G6 and dPG-G10, estimated by equation 2 next to the radii of gyration for the used target molecules. Molecular weights are presented in the bottom row and were calculated for the oligonucleotides and for Fab/Ab were given by the manufacturer.

dPG-G6

dPG-G10

19mer

150mer

Fab

Ab

Mesh sizea

5.8 nm

8.4 nm

-

-

-

-

G’b

3490 Pa

397 Pa

Radius of gyration [nm]

-

-

1.96

3.836

3.4830

5.530

Molecular weight

-

-

6 kDa

46 kDa

50 kDa

150 kDa

a Determined by Equation 2. b Determined by oscillatory rheology in the linear viscoelastic regime.

Compound analysis in k = 2 elevations (CAkE2) We present the multiplex detection of Ab*, Fab* and 19mer in a single well by coding the microbeads into k elevations. With the system presented here, coding of the targets by labelling with different dyes was not possible because the fluorescence detection of the three biomolecules was achieved in the same channel of the microscope. Additionally, two molecules with the same affinity interaction, such as Ab* and Fab*, were discriminated in one well by our system which is not possible by other common coding techniques. We blocked the larger Ab* to penetrate the two detection levels and separated the detection levels of µBSA-Ofl and µB-Ab into k1 and k2. Thus, we were able to perform a compound analysis in k elevations (CAkE) as is depicted in Figure 3. We used the same microbead population for all probes, which meant that the multiplexity was only introduced by separating the two detection levels (µBSA-Ofl in k1 and µB-Ab in k2) and covering them with a layer of hydrogel.


Page 11 of 17 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


Figure 3 Representation of the two-layered detection system to detect a small 19mer in the bottom level k 1, a medium-sized Fab* in k2, and Ab* should be excluded from all levels. Therefore, k 2 is covered by another hydrogel layer.

The differently sized target molecules were (1) a short oligonucleotide sequence (19mer) that carried HBQ2 and was detected by quenching the fluorescence of µBSA-Ofl in k1, (2) a medium sized fluorescence-labeled Fab fragment (Fab*) which should diffuse through the top layer and be detected on µB-Ab in k2 and (3) a fluorescence-labeled antibody (Ab*) that was too big to penetrate any of the layers. To prove that our system can detect the right molecule in the right level, we unambiguously added several combinations of the target molecules and showed that the results remained conclusive. Firstly, each target molecule was individually applied into a cavity containing the CAkE system. Secondly, every possible combination of two (19mer/Fab*, 19mer/Ab* and Fab*/Ab*) was used before finally the whole mixture was added. Each line in Figure 4 corresponds to one cavity and follows the fluorescence intensity over time. For better clarity we separated the data into two graphs, hence, Figure 4 A shows the development for µB-Ab in k2 and Figure 4 B for µBSA-Ofl in k1. In Figure 4 B the green lines represent all cavities in which the 19mer was added, solely or combined with Fab*, Ab* or altogether. The orange/red lines show the cavities in which no 19mer was added. The individual results are assigned in the legend. The fluorescence intensity dropped whenever the 19mer was added, also in any combination with other molecules. Additionally, a random oligonucleotide sequence with a fluorescence label (Or*) was added which should have led to an increase in intensity if there would have been any unspecific interaction. But the control line in grey exhibited no increased fluorescence intensity. In Figure 4 A we chose a cyan tone for all cavities where Fab* was added in any combination, and a red tone for all cavities where there was no Fab *. Here, the fluorescence intensity always rose whenever Fab* was added and did not change whenever Ab* was added. The addition of Fab*/Ab* led to the same intensity, which was another indication for the successful exclusion of Ab*. However, the positive control (grey line in Figure 4 B) showed that in solution, with no hydrogel layer present, the detection proceeded in a fluorescence increase. The absence of fluorescence for Ab* is therefore only assigned to the filtering effect of the hydrogel layer and not to the inability of binding. We performed several CAkEs with different combinations of the two shown hydrogels and various thicknesses thereof. These parameters were optimized, and the results are presented. To discuss a less crowded picture we chose to display only one experiment, triplicates can be found in the Supporting Information.



Page 12 of 17

Figure 4 Compound analysis in k = 2 elevations (CAkE2). The fluorescence intensity was followed over time for the µBSA-Ofl in k1 (bottom) and the µB-Ab detection in k2 (top). Each line follows one well; the key at the bottom assigns the color to the added target or combination of targets. The controls in grey, Or*_sol and Ab*_sol, were not performed with the CAkE system, but instead in a separate cavity with no hydrogel and the respective microbeads suspended in the target solution.

Reversible swelling and hydrogel characterization In this experiment two levels of microbeads of different populations (color- and size-encoded) were separated by a layer of each hydrogel dPG-G6 and dPG-G10. Over an overall timeframe of 3 weeks, these systems were subjected to several drying/swelling cycles (cf. Figure 5A). One cycle comprises the transition from a fully swollen to a completely dry, collapsed state and then reswelling induced by the addition of water. As shown in Figure 5B both hydrogels were fully capable to perform a reversible swelling over 6 cycles. As we distinguished the two levels by two different microbead populations, we assured that it was possible to dry and reswell the system while the hydrogel was not disrupted, and the microbeads did not migrate through the meshes but stayed separated in their respective level. The decrease from the initial height until it became stable after cycle 3 was probably due to reorganization of the dangling ends of PEGBCN, which would find each other upon concentration. This resulted in a higher crosslinking density and therefore reduced the swelling capacity. To make a CAkE become an easy and universally applicable multiplex detection platform, it should need as little expert knowledge and components as possible. These results show that our system could be manufactured by a trained person and then be distributed and stored under dried condition until it is needed. The end user will only have to add water for activation of the system. Additionally, we could show that the height between two microbead levels is linearly proportional to the employed hydrogel amount in the investigated range (see Supporting Information, Figure S6). That means that a desirable height can easily be adjusted within that range. Finally, Table 3 shows a comparison of the swelling ratios, G’, and gelation time tgel for dPG-G10 and dPG-G6. We determined the swelling ratio according to equation 1, both on the surface (qsurf) and in solution (qsol). Unsurprisingly, dPG-G10 had a higher swelling ratio than dPG-G6 by a factor of 1.6 for qsurf and a factor of 1.8 for qsol, respectively. The impact of the longer PEG chain on the swelling ratio can be rationalized by looking at the 3D models of the polymers dPG-N3 and PEG-BCN in the Supporting Information. The PEG has a linear structure and hence, mainly contributes to the swelling capability, contrary to the highly branched dPG-N3, which has a more confined 12 ACS Paragon Plus Environment

Page 13 of 17 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


sphere like expansion. We calculated the factor that the longer PEG 10,000 chain in dPG-G10 contributes to the hydrogel in comparison to PEG 6000 in dPG-G6 and obtained a factor of 1.6. This is in line with the experimentally found values (1.6 and 1.8 for qsurf and qsol, respectively). Because the hydrogel was adhered to the surface the swelling is restricted to only one dimension whereas a three-dimensional expansion was possible in solution. That is the reason why qsol was larger than qsurf for each hydrogel. Lastly, we determined the gelation time by oscillatory rheology. The crossover of G’ and G’’ was determined for both hydrogels and gelation times are given as average ± standard deviation of three independent measurements in Table 3. Slower gelation for the longer PEG chain in dPG-G10 can be explained by the lower mobility for longer PEG chains and hence, worse accessibility of the reactive groups (BCN). This effect is also known in literature.31 Consequently, a customized CAkE system with regards to the demands of the experiment can be designed. On one hand, the layer thickness can be adjusted, which has an influence on the diffusion time and, hence, on the separation efficiency. The pore size of the hydrogel can, to some extent, be tuned by changing the length of the PEG linker. Finally, the hydrogel exhibited chemical and thermal stability. Each hydrogel component was stored as aqueous ready-to-use solution at 8 °C. From synthesis until writing this manuscript the storage time was 24 months without detectable changes of the hydrogel properties. Thermal stability was investigated by heating the microbeads covered in the hydrogel up to 90 °C. From the kinetics of the rehybridization, which was much slower for µBSA-Ofl covered by the hydrogel than it was in solution, we concluded that the hydrogel remained stable and intact (see Supporting Information).

Figure 5 For both gels, dPG-G6 and dPG-G10, a series of swelling and drying was performed. (A) is a schematic overview of the experimental setup. (B) shows 6 drying/swelling cycles for dPG-G6 and dPGG10. The grey columns represent the thickness of the swollen hydrogel, whereas the blank columns show the thickness in the collapsed state. The experiment was performed in duplicates. In the collapsed state the error bars are very small and overlap with the frame of the columns.



Page 14 of 17

Table 3 Presentation of qsurf, qsol, the elastic modulus, and gelation time for both used hydrogels dPG-G6 and dPG-G10.

Surface Swelling ratio (n=3)

Solution Swelling ratio (n=4)

G’∞

tgel, 7.7 wt%

n=3

n=3

dPG-G6

12.4 ± 1.1

19.6 ± 1.0

3444 ± 134

137 ± 8 s

dPG-G10

19.8 ± 1.3

35.0 ± 2.9

746 ± 44

Hydrogel

364 ± 71 s

Conclusions We showed that target molecules could be detected in k levels whereby each molecule size was detectable in one elevation. The levels were separated and covered by hydrogel layers. The stacking of detection levels increased the multiplexity by the number of stacks. Furthermore, we showed that a bigger biomolecule such as an antibody could completely be excluded, which was useful for the simultaneous detection of molecules that exhibited the same affinity interaction. This size exclusion was more discrete for rigid structures like Fab/Ab as it was for the flexible oligonucleotide sequences 19mer/150mer. This means not only the size but also the deformability were crucial points to consider when a CAkE system is performed. For example, Fab* and 150mer have roughly the same radius of gyration but the flexible 150mer moved considerably faster through the network. In view of a multiplex coding by CAkE, it is important to know that small and flexible biomolecules diffuse faster and should favorably be arranged in the lower levels. However, a complete size selection is possible for more rigid structures like Fab*/Ab*, and the size finally determines if the molecule manages to penetrate or not. We conclusively showed that the Ab* is detectable in solution but will be excluded throughout the whole timeframe of 6 d, if the microbeads are covered by the hydrogel. Another interesting feature of our presented method was the comparatively easy use of the hydrogel. The noncovalent interaction with the cavity surface was strong enough for commonly used substrates such as glass slides, 96-well plates, and the Nunc™ NucleoLink™ stripes without the need of a previous functionalization (data not shown). Additionally, the PEG- and dPG-based hydrogel is antifouling and therefore provided an inert environment for a variety of biomolecules. The two hydrogel components, once they are synthesized, can be stored for at least 2 years as readyto-use aqueous solutions. That was the investigated time between the first and final experiments for this manuscript. This alleviates the fabrication and, hence, enables the dendritic-based hydrogel to become a universally applicable product. Furthermore, the reversible swelling, the quite reproducible height adjustment, and the adaptable pore size are versatile features of this hydrogel-based detection platform. We consider our developed hydrogel as an adaptable tool to improve and complement other methods, such as we combined it with the color and size coded microbeads.


Page 15 of 17 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


Acknowledgements The authors are grateful to Jörg Nitschke and Alexander Böhm for operating the VideoScan software. The authors acknowledge the BMBF, program KMU-innovative: Biotechnology - Biochance for funding.

Supporting Information. Further methods and characterization of the system are presented. Additional information regarding the performance and the reproducibility are shown, the related information is indicated at the respective position in the text.

References (1) Grossmann, K.; Roggenbuck, D.; Schroder, C.; Conrad, K.; Schierack, P.; Sack, U. Multiplex Assessment of Non-Organ-Specific Autoantibodies with a Novel Microbead-Based Immunoassay Cytometry A 2011, 79, 118-125. (2) Koch, W. H. Technology Platforms for Pharmacogenomic Diagnostic Assays Nat. Rev. Drug Discov. 2004, 3, 749-761. (3) Bong, K. W.; Chapin, S. C.; Doyle, P. S. Magnetic Barcoded Hydrogel Microparticles for Multiplexed Detection Langmuir 2010, 26, 8008-8014. (4) Uchugonova, A.; Hoffman Robert, M. In Multiphoton Microscopy and Fluorescence Lifetime Imaging Applications in Biology and Medicine, 2018. (5) Appleyard, D. C.; Chapin, S. C.; Srinivas, R. L.; Doyle, P. S. Bar-Coded Hydrogel Microparticles for Protein Detection: Synthesis, Assay and Scanning Nat. Protoc. 2011, 6, 1761-1774. (6) Choi, N. W.; Kim, J.; Chapin, S. C.; Duong, T.; Donohue, E.; Pandey, P.; Broom, W.; Hill, W. A.; Doyle, P. S. Multiplexed Detection of Mrna Using Porosity-Tuned Hydrogel Microparticles Anal. Chem. 2012, 84, 9370-9378. (7) Liebsch, C.; Rödiger, S.; Bohm, A.; Nitschke, J.; Weinreich, J.; Fruth, A.; Roggenbuck, D.; Lehmann, W.; Schedler, U.; Juretzek, T.; Schierack, P. Solid-Phase Microbead Array for Multiplex O-Serotyping of Escherichia Coli Microchimica Acta 2017, 184, 1405-1415. (8) Du, H.; Hu, X.; Duan, H.; Yu, L.; Qu, F.; Huang, Q.; Zheng, W.; Xie, H.; Peng, J.; Tuo, R.; Yu, D.; Lin, Y.; Li, W.; Zheng, Y.; Fang, X.; Zou, Y.; Wang, H.; Wang, M.; Weiss, P. S.; Yang, Y.; Wang, C. Principles of Inter-Amino-Acid Recognition Revealed by Binding Energies between Homogeneous Oligopeptides ACS Cent. Sci. 2019, 5, 97-108. (9) Dalili, A.; Samiei, E.; Hoorfar, M. A Review of Sorting, Separation and Isolation of Cells and Microbeads for Biomedical Applications: Microfluidic Approaches Analyst 2018, 144, 87113. (10) Koetting, M. C.; Guido, J. F.; Gupta, M.; Zhang, A.; Peppas, N. A. Ph-Responsive and Enzymatically-Responsive Hydrogel Microparticles for the Oral Delivery of Therapeutic Proteins: Effects of Protein Size, Crosslinking Density, and Hydrogel Degradation on Protein Delivery J. Controlled Release 2016, 221, 18-25. (11) Pertici, V.; Pin-Barre, C.; Rivera, C.; Pellegrino, C.; Laurin, J.; Gigmes, D.; Trimaille, T. Degradable and Injectable Hydrogel for Drug Delivery in Soft Tissues Biomacromolecules 2018, 20, 149-163. (12) Saunders, L.; Ma, P. X. Self-Healing Supramolecular Hydrogels for Tissue Engineering Applications Macromol. Biosci. 2019, 19, e1800313. 15 ACS Paragon Plus Environment


Page 16 of 17

(13) Jung, I. Y.; Kim, J. S.; Choi, B. R.; Lee, K.; Lee, H. Hydrogel Based Biosensors for in Vitro Diagnostics of Biochemicals, Proteins, and Genes Adv. Healthcare Mater. 2017, 6. (14) Herrmann, A.; Kaufmann, L.; Dey, P.; Haag, R.; Schedler, U. Bioorthogonal in Situ Hydrogels Based on Polyether Polyols for New Biosensor Materials with High Sensitivity ACS Appl. Mater. Interfaces 2018, 10, 11382-11390. (15) Chatellier, J.; Rauffer-Bruyere, N.; Van Regenmortel, M. H.; Altschuh, D.; Weiss, E. Comparative Interaction Kinetics of Two Recombinant Fabs and of the Corresponding Antibodies Directed to the Coat Protein of Tobacco Mosaic Virus J. Mol. Recognit. 1996, 9, 39-51. (16) Lyu, Z.; Kang, L.; Buuh, Z. Y.; Jiang, D.; McGuth, J. C.; Du, J.; Wissler, H. L.; Cai, W.; Wang, R. E. A Switchable Site-Specific Antibody Conjugate ACS Chem. Biol. 2018, 13, 958964. (17) Spitali, M. In Process Scale Purification of Antibodies, 2017, pp 537-558. (18) Cheng, Y.; Prud'homme, R. K.; Thomas, J. L. Diffusion of Mesoscopic Probes in Aqueous Polymer Solutions Measured by Fluorescence Recovery after Photobleaching Macromolecules 2002, 35, 8111-8121. (19) Foldes-Papp, Z.; Baumann, G.; Birch-Hirschfeld, E.; Eickhoff, H.; Greulich, K. O.; Kleinschmidt, A. K.; Seliger, H. The Analysis of Oligonucleotide Preparations by Fractal Measures Biopolymers 1998, 45, 361-379. (20) Arosio, P.; Rima, S.; Lattuada, M.; Morbidelli, M. Population Balance Modeling of Antibodies Aggregation Kinetics J. Phys. Chem. B 2012, 116, 7066-7075. (21) Frommel, U.; Lehmann, W.; Rödiger, S.; Bohm, A.; Nitschke, J.; Weinreich, J.; Gross, J.; Roggenbuck, D.; Zinke, O.; Ansorge, H.; Vogel, S.; Klemm, P.; Wex, T.; Schroder, C.; Wieler, L. H.; Schierack, P. Adhesion of Human and Animal Escherichia Coli Strains in Association with Their Virulence-Associated Genes and Phylogenetic Origins Appl. Environ. Microbiol. 2013, 79, 5814-5829. (22) Rödiger, S.; Liefold, M.; Ruhe, M.; Reinwald, M.; Beck, E.; Deckert, P. M. Quantification of DNA Double-Strand Breaks in Peripheral Blood Mononuclear Cells from Healthy Donors Exposed to Bendamustine by an Automated Γh2ax Assay—an Exploratory Study Journal of Laboratory and Precision Medicine 2018, 3. (23) Rödiger, S.; Ruhland, M.; Schmidt, C.; Schroder, C.; Grossmann, K.; Bohm, A.; Nitschke, J.; Berger, I.; Schimke, I.; Schierack, P. Fluorescence Dye Adsorption Assay to Quantify Carboxyl Groups on the Surface of Poly(Methyl Methacrylate) Microbeads Anal. Chem. 2011, 83, 3379-3385. (24) Dey, P.; Schneider, T.; Chiappisi, L.; Gradzielski, M.; Schulze-Tanzil, G.; Haag, R. Mimicking of Chondrocyte Microenvironment Using in Situ Forming Dendritic Polyglycerol Sulfate-Based Synthetic Polyanionic Hydrogels Macromol. Biosci. 2016, 16, 580-590. (25) Salinas, C. N.; Anseth, K. S. Mixed Mode Thiol−Acrylate Photopolymerizations for the Synthesis of Peg−Peptide Hydrogels Macromolecules 2008, 41, 6019-6026. (26) Canal, T.; Peppas, N. A. Correlation between Mesh Size and Equilibrium Degree of Swelling of Polymeric Networks J. Biomed. Mater. Res. 1989, 23, 1183-1193. (27) Peppas, N. A.; Moynihan, H. J.; Lucht, L. M. The Structure of Highly Crosslinked Poly(2Hydroxyethyl Methacrylate) Hydrogels J. Biomed. Mater. Res. 1985, 19, 397-411. (28) Merrill, E. W.; Dennison, K. A.; Sung, C. Partitioning and Diffusion of Solutes in Hydrogels of Poly(Ethylene Oxide) Biomaterials 1993, 14, 1117-1126. (29) Stingaciu, L. R.; Ivanova, O.; Ohl, M.; Biehl, R.; Richter, D. Fast Antibody Fragment Motion: Flexible Linkers Act as Entropic Spring Sci. Rep. 2016, 6, 22148. (30) Pilz, I.; Kratky, O.; Licht, A.; Sela, M. Shape and Volume of Fragments Fab' and (Fab')2 of Anti-Poly(D-Alanyl) Antibodies in the Presence and Absence of Tetra-D-Alanine as Determined by Small-Angle X-Ray Scattering Biochemistry 2002, 14, 1326-1333. 16 ACS Paragon Plus Environment

Page 17 of 17 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


(31) Ghobril, C.; Rodriguez, E. K.; Nazarian, A.; Grinstaff, M. W. Recent Advances in Dendritic Macromonomers for Hydrogel Formation and Their Medical Applications Biomacromolecules 2016, 17, 1235-1252.

For Table of Contents Only


Spatial Separation of Microbeads into Detection Levels by a

Recommend Documents