Stable Immobilization of Enzymes in a Macro- and Mesoporous Silica Monolith.

Horseradish peroxidase isoenzyme C (HRP) and Engyodontium album proteinase K (proK) were immobilized inside macro- and mesoporous silica monoliths. Stable immobilization was achieved through simple noncovalent adsorption of conjugates, which were prepared from a polycationic, water-soluble second generation dendronized polymer (denpol) and the enzymes. Conjugates prepared from three denpols with the same type of repeating unit (r.u.), but different average lengths were compared. It was shown that there is no obvious advantage of using denpols with very long chains. Excellent results were achieved with denpols having on average 750 or 1000 r.u. The enzyme-loaded monoliths were tested as flow reactors. Comparison was made with microscopy glass coverslips onto which the conjugates were immobilized and with glass micropipettes containing adsorbed conjugates. High enzyme loading was achieved using the monoliths. Monoliths containing immobilized denpol-HRP conjugates exhibited good operational stability at 25 °C (for at least several hours), and good storage stability at 4 °C (at least for weeks) was demonstrated. Such HRP-containing monoliths were applied as continuous flow reactors for the quantitative determination of hydrogen peroxide in aqueous solution between 1 μM (34 ng/mL) and 50 μM (1.7 μg/mL). Although many methods for immobilizing enzymes on silica surfaces exist, there are only a few approaches with porous silica materials for the development of flow reactors. The work presented is a promising contribution to this field of research toward bioanalytical and biosynthetic applications.

Introduction

Among the various methodologies which have been developed during the last decades for the immobilization of enzymes on silica surfaces,[1−26] one recently developed method is based on the preparation of dendronized polymer (denpol)–enzyme conjugates.[27−30] The principle of the method is simple. In a first step, enzyme molecules are attached to a dissolved, water-soluble denpol of second generation (abbreviated as de-PG2), which carries in each repeating unit (r.u.) four amino groups; that is, for a de-PG2 molecule consisting of n = 1000 r.u., as many as 4000 amino groups are distributed along the denpol chain. The attachment of the desired enzyme molecules to de-PG2 occurs via covalent, spectrophotometrically quantifiable,[31−33] bis-aryl hydrazone (BAH) bonds (Figure ). In a second step, an aqueous solution of the denpol–BAH–enzyme conjugate is brought in contact with a silica surface. This results in a stable, noncovalent adsorption (immobilization) of the conjugate on this very surface. This is achieved through many (individually weak, possibly mainly electrostatic) interactions between the conjugate and the many hydroxyl groups of the surface. In this way, Engyodontium album proteinase K (proK) was previously immobilized successfully as a de-PG22000-BAH-proK conjugate on microscopy glass coverslips, inside glass micropipettes, or on a microfluidic chip.[29] Similarly, Aspergillus sp. glucose oxidase and horseradish (Armoracia rusticana) peroxidase isoenzyme C (HRP) could be immobilized on microscopy glass coverslips,[27] inside glass micropipettes,[27] and onto mesoporous silica nanoparticles.[28] In the most recent work, a microbial transglutaminase (MTG, recombinantly produced in Escherichia coli with a gene derived from Streptomyces mobaraensis) was immobilized as de-PG2500-BAH-MTG conjugate on 75 μm-sized glass beads,[34] and bovine erythrocytes carbonic anhydrase was successfully immobilized on glass coverslips, inside glass micropipettes, and in porous glass fiber filters.[30]

Figure 1

Illustration of a short section of a de-PG2-BAH-HRP conjugate. Two denpol r.u. are shown, whereby to one of them one HRP molecule is attached via a spectrophotometrically quantifiable, stable BAH linker unit (in red), most likely through Lys232 of HRP.[47] PDB 1HCH. The glycosydic chains present on the surface of HRP are not shown.

In the work presented here, we investigated whether the same type of denpol–BAH–enzyme conjugates can also be immobilized inside silica monoliths which have a hierarchical pore structure consisting of a macroporous silica network, which itself contains smaller macropores as well as mesopores (see below and Figure ). Because this immobilization turned out to be possible, we evaluated the feasibility of de-PG2-BAH-enzyme-loaded silica monoliths for their use as flow reactor systems. Moreover, we made a direct comparison of the immobilization inside the monoliths with the immobilization of the same conjugates on silica glass coverslips and inside silica micropipettes. Furthermore, we investigated the denpol chain length dependency of the amount and stability of the immobilized enzymes. The enzymes used in the present investigations were HRP and proK.

Figure 2

SEM images of the silica monolith MH1 showing the presence of two types of macropores, accessible ones which go through the entire structure (diameter: 20–30 μm) and isolated ones within the silica skeleton (diameter: a few μm). Accessible mesopores (diameters of ∼22 and ∼2.5 nm), which are also present, cannot be seen at the resolution of the micrographs shown.[37] The bottom figure is reprinted from Szymańska et al. (2016)[37] with permission from Elsevier.

The macro- and mesoporous silica monoliths used throughout the work were of the so-called type MH1, prepared according to a modified “Nakanishi process”[35,36] and characterized as described previously.[37] The key feature of the 4 mm thick MH1 monolith is its uniform bicontinuous, foam-like structure with pore sizes of 20–30 μm, which results in a very low hindrance to liquid flow when used in a flow-through system. A second type of pores with sizes of a few μm exists within the silica skeleton. These pores are inaccessible, that is, isolated from the much larger pores, as clearly seen in the scanning electron microscopy (SEM) images (Figure ). Because pores with diameters larger than 50 nm are classified as “macropores”,[38] MH1 is a macroporous silica. Moreover, because the silica skeleton of MH1 itself also has two types of much smaller pores with average diameters of about 2.5 and 22 nm,[37] classified as “mesopores” (2–50 nm),[38] MH1 is called a “macro- and mesoporous silica”. The accessibility of the mesopores (which are not seen in Figure due to the limited resolution of the SEM analysis) was determined previously by nitrogen adsorption measurements at 77 K and by Hg porosimetry measurements.[37] The experimentally determined total volume of all pores of MH1 is about 4 cm3/g, of which about 1 cm3/g is the total volume of the mesopores, and the remaining 3 cm3/g is the calculated volume of the accessible macropores.[37] From this latter value and assuming spherical macropores with an average diameter of 25 μm, the total surface area limiting the accessible macropores is estimated to be 0.72 m2/g (corresponding to 0.25% of the total surface area, 287 m2/g,[37] of all pores in the monolith), see Supporting Information, part 1. This is the available surface area of the monolith onto which the denpol–enzyme conjugate may adsorb. This adsorption was tested by preparing and testing different de-PG2-BAH-HRP and de-PG2-BAH-proK conjugates.

Experimental Section

Enzymes, Chemicals, and General Methods

Horseradish (A. rusticana) peroxidase isoenzyme C (HRP, EC 1.11.1.7, M ≈ 44 000 g/mol, ε403 = 102 000 M–1 cm–1)[39,40] was purchased from Toyobo Enzymes, Japan [catalogue number PEO-131, grade I, lot 0240160000, RZ (A403/A260 = 3.09)]. E. album proteinase K (proK, EC 3.4.21.64, recombinant from Pichia pastoris, M = 28 900 g/mol, ε280 = 41 000 M–1 cm–1)[41] was from Roche Applied Sciences, Switzerland (catalogue number 03115879001, lot 14321500 and lot 1016630). N-Succinimidyl 6-hydrazinonicotinate acetone hydrazone (S-HyNic) and N-succinimidyl 4-formylbenzoate (S-4FB) were synthesized as described before.[42] For the commercial sources of 4-nitrobenzaldehyde (used for the quantification of HyNic in the HyNic-modified denpols), 2-hydrazinopyridine dihydrochloride (used for the quantification of 4FB in the 4FB-modified enzymes), 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) diammonium salt (ABTS2–(NH4+)2, used as substrate of HRP), and succinyl-l-alanyl-l-alanyl-l-prolyl-l-phenylalanyl-para-nitroanilide (Suc-AAPF-pNA, used as substrate of proK), see Küchler et al. (2015)[27] and (2017).[43] The buffer salts sodium phosphate monobasic (anhydrous, NaH2PO4) and sodium phosphate dibasic (anhydrous, Na2HPO4) were from Sigma-Aldrich, Switzerland; 2-(N-morpholino)ethanesulfonic acid monohydrate (MES) and 3-(N-morpholine)propanesulfonic acid (MOPS) were purchased from AppliChem, Germany. For all experiments, ultrapure water (filtered with a Synergy water preparation apparatus from Millipore) was used; all solutions were stored at 4 °C.

The following buffer solutions were prepared. PB0, 0.01 M phosphate, pH = 7.0; PB0b, 0.01 M phosphate, 0.15 M NaCl, pH = 7.0 (see Section ); PB1, 0.1 M phosphate, 0.15 M NaCl, pH = 7.2; MesB1, 0.1 M MES, 0.15 M NaCl, pH = 4.7; MesB2, 0.1 M MES, pH = 5.0; MopsB1: 0.01 M MOPS, pH = 8.0; MopsB2: 0.01 M MOPS, pH = 7.0.

Because the deprotected dendronized polymer strongly adheres to glass surfaces, all reactions, purification steps by ultrafiltration and quantification reactions were carried out in polypropylene (PP) reaction vessels (Eppendorf tubes from Greiner Bio-One or Sarstedt).[43,44]N,N-dimethylformamide (extra dry over molecular sieve) was obtained from Acros Organics (Switzerland).

Amicon ultrafiltration devices (Ultra-4, Mcutoff = 10 or 100 kDa) were obtained from Merck Millipore or from Sigma-Aldrich. Centrisart I ultrafiltration units (Mcutoff = 300 kDa) were from Sartorius (Switzerland).

UV/vis-spectroscopy was carried out either on a V-670 UV/Vis/NIR spectrophotometer (from JASCO Inc., Japan) equipped with a Peltier-temperature control (EC-717), on a SPECORD S600 diode array spectrophotometer (from Analytik Jena), on a Cary 1E spectrophotometer (from Varian, Australia), or on a Nano-Drop ND 1000 spectrophotometer (from Witec AG, Switzerland).

Round microscopy glass coverslips (diameter 8 mm, thickness 0.16–0.19 mm) were obtained from Science Services (Germany), glass micropipettes (intraMARK, 200 μL) were from Brand (Germany), and borosilicate DURAN NMR tubes economic 178 × 4.95 mm (cat. no. 23 170 0117) were from the Duran group (now DWK Life Sciences).

Silica Monolith

Cylindrical monoliths of type MH1 with a diameter of 4 mm (Figure ) were prepared from tetraethyl orthosilicate, polyethylene glycol 35 000, water, nitric acid (HNO3), and cetyltrimethylammonium bromide at a molar ratio of 1:0.52:14.25:0.26:0.027, according to the procedure described by Szymańska et al. (2016).[37] In this previous publication, the details about the characterization of the monolith are also provided, including an SEM analysis.[37]

Dendronized Polymers (Denpols)

Throughout the entire work, three deprotected second generation denpols composed of the same r.u. (Figure ) but with different average degrees of polymerization, n, were used. They are abbreviated as de-PG2750, de-PG21000, and de-PG220000, for corresponding values of n = 750, 1000, and 20 000. The syntheses of de-PG21000 and de-PG220000 were carried out by free radical polymerization of the Boc (tert-butyloxycarbonyl)-protected first generation macromonomer (MG1) to yield the protected first generation denpols PG11000 and PG120000. Boc-protected PG1750 was synthesized by reversible addition–fragmentation chain transfer polymerization of MG1. Polymerization was followed by deprotection with trifluoroacetic acid (TFA) and dendronization to yield the second generation Boc-protected denpols PG2750, PG21000, and PG220000. Deprotection by TFA afforded the deprotected denpols (de-PG2750, de-PG21000, and de-PG220000) used in this study (Figure S1).[43] For de-PG2 dissolved in aqueous solution, the molar absorption coefficient at 285 nm, ε285nm, which was used for the spectrophotometric determination of the r.u. concentration is ε285 = 5000 M–1 r.u. cm–1 (Fornera et al., 2011).[44]

Molar mass determinations of the Boc-protected second generation denpols, PG2750, PG21000, and PG220000, were carried out on a Viscotek GPCmax system (from Malvern), equipped with a Viscotek TDA detector array and a 2× D5000 column set using DMF containing 0.1% w/v LiBr as eluent. The denpols were dissolved in the eluent containing an additional 0.1% v/v toluene as flow marker. The results summarized in Table were evaluated by using universal calibration employing narrow-dispersity poly(methyl methacrylate) standards (from Polymer Laboratories).

Table 1

Molar Mass Determinations of the Boc-Protected Second Generation Denpol Samples, PG2, Useda

PG2n sample	Mw (g/mol)	Mn (g/mol)	PDI = Mw/Mn (-)	Mn/Mr.u. (-)	approximate number average degree of polymerization, n
PG2750	1.18 × 106	0.92 × 106	1.3	755	750
PG21000	2.88 × 106	1.23 × 106	2.4	1000	1000
PG220000	8.40 × 107	2.51 × 107	3.4	20 441	20 000

Mw, weight average molar mass; Mn, number average molar mass; Mr.u., molar mass of the r.u. of PG2 (=1225.5 g/mol); PDI, polydispersity index.

Partial Modification of de-PG2 for Obtaining de-PG2-HyNic

The partial modification of de-PG2750, de-PG21000, and de-PG220000 with S-HyNic at pH = 7.2 (PB1)—followed by purification with MesB1(pH = 4.7)—and the quantification of the modification with 4-nitrobenzaldehyde at pH = 5.0 (MesB2) for determining the molar substitution ratio, MSR(HyNic), were carried out in a similar way as described before (using ε390 = 24 000 M–1 cm–1),[43] see Supporting Information (Table S1). The concentration of the denpol r.u. was determined with the trypan blue assay, see Supporting Information (Figure S2). To visualize the modified denpols de-PG2-HyNic, tapping mode atomic force microscopy (AFM) was applied, as described previously,[27] see Figure .

Figure 3

Tapping mode AFM height images of HyNic-modified de-PG2750 (a), de-PG21000 (b), and de-PG220000 (c) on mica. Length of the bar: 200 (a,b) or 400 nm (c).

Partial Modification of HRP and proK for Obtaining HRP-4FB and proK-4FB

The partial modification of HRP and proK (through accessible lysine residues on the surface of the enzyme molecules) with S-4FB in either PB1 (pH = 7.2 for HRP) or MopsB1 (pH = 8.0 for proK)—followed by purification in MesB1 (pH = 4.7)—and the quantification of the modification with 2-hydrazinopyridine at pH = 4.7 (in MesB1) for the determination of the molar substitution ratio MSR(4FB) were carried out as described before (using ε350 = 24 500 M–1 cm–1),[29,42,43] see Supporting Information (Table S2).

Preparation of de-PG2-BAH-HRP and de-PG2-BAH-proK

The denpol–enzyme conjugates de-PG2-BAH-HRP and de-PG2-BAH-proK were prepared by incubating at pH = 4.7 (MesB1) a solution of de-PG2-HyNic with either a solution of HRP-4FB, or a solution of proK-4FB, as described before,[29,42,43] see Supporting Information, part 6. The concentration of BAH bonds formed in the conjugation reaction was determined by using ε354 = 29 000 M–1 cm–1(Küchler et al., 2017).[43]

The denpol–enzyme conjugates prepared in this work are abbreviated in the same way as in our previous work,[27,29,42] whereby the average number of enzyme molecules per average chain length, n, is indicated: de-PG2750-BAH-HRP62, de-PG21000-BAH-HRP70, de-PG220000-BAH-HRP1850, and de-PG2750-BAH-proK115.

Immobilization of de-PG2-BAH-enzyme on Silicate Surfaces

Immobilization on Microscopy Glass Coverslips and inside Glass Micropipettes

Details about the immobilization of the conjugates on microscopy glass coverslips [diameter: 8 mm; thickness: 0.16–0.19 mm; total surface area (both sides): 1.0 cm2] and inside glass micropipettes (“200 μL”; full length: 14 cm; inner diameter: 1.6 mm; inner surface area: 7.0 cm2; total inner volume: 280 μL) are given in the Supporting Information (parts 7 and 8). The procedures used were similar to the ones described before.[27,29,43]

Immobilization in Monolith MH1 for Use as Batch Reactors and for Flow Reactor Applications with NMR Tubes as Monolith Holders

The cylindrical monoliths with diameters of 4 mm were cut into pieces of about 5 mm length and then cleaned with ethanol by inserting the monolith pieces into 1 mL ethanol, followed by three times bath sonication with fresh ethanol for 10 min and finally drying with a stream of nitrogen gas, followed by storage in air for 3 days. The conjugates were then immobilized in/on the monolith pieces for two different applications, as batch reactor systems and as flow reactors.

For the batch reactor applications, a monolith piece was put inside a 2 mL PP reaction tube to which 1 mL MesB1 was added for wetting the monolith. After removal of MesB1, a solution of de-PG2–BAH–enzyme (prepared in MesB1 at [enzyme] = 2 μM, as determined from the BAH bond formation, see Supporting Information, part 6) was added to the tube and stored for 2 h at room temperature. After removal of the conjugate solution, the monolith piece was washed three times by immersing in 1 mL fresh buffer solution (PB0 in the case of HRP, MopsB2 for proK). The monolith piece was then stored immersed in 1 mL buffer solution (PB0 in the case of HRP, MopsB2 for proK) at 4 °C until further use.

For the flow reactor applications, the monolith piece was inserted into a borosilicate NMR tube (outer diameter: 4.95 ± 0.05 mm; inner diameter: 4.20 ± 0.05 mm), which was cut to a length of ∼5 mm. The small glass tube containing the monolith piece was then connected to PTFE tubes. After wetting with MesB1, the procedure for the immobilization of the conjugates was the same as described above for the glass micropipettes, see Section .

Enzyme Activity Measurements

The activity of HRP, either dissolved in aqueous solution or immobilized on silicate surfaces, was measured at ∼25 °C with 1.0 mM ABTS2– and 0.2 mM H2O2 at pH = 7.0 (PB0) by taking into account ε414nm (ABTS•– = 36 000 M–1 cm–1),[27,45] see Supporting Information, part 9.1 and 9.2. For measuring the activity of proK, the substrate used was Suc-AAPF-pNA (0.5 mM) at ∼25 °C, pH = 7.0 (MopsB2), using ε410nm (p-nitroaniline) = 8800 M–1 cm–1,[29,46] see Supporting Information, part 9.3 and 9.4.

Hydrogen Peroxide Quantification Experiments with Additionally Synthesized de-PG21000-BAH-HRP71

Preparation of de-PG21000-BAH-HRP71

The preparation of de-PG21000-BAH-HRP71 was carried out in a similar way as described in Sections , 2.5, and 2.6. All details are provided in the Supporting Information, part 11.

Immobilization of de-PG21000-BAH-HRP71 in Monolith MH1 Placed inside LDPE Tubes for Flow Reactor Applications

A 2.5 cm long tube was cut from a disposable, sterile low-density polyethylene (LDPE) transfer pipette (inner diameter ≈ 4 mm, from VWR Switzerland, product number 612-4466, lot no. 120627). A 5 mm long piece of the cylindrical monolith MH1 (d = 4 mm) was inserted into the LDPE tube and placed in its center part, followed by flushing with nitrogen gas to remove monolith debris. The monolith was then washed with deionized water by pumping it through the monolith with the help of a P-1 peristaltic pump (from Pharmacia), followed by drying with nitrogen gas. A solution of de-PG21000-BAH-HRP71 was prepared with 0.01 M sodium phosphate buffer solution (pH = 7.0) containing 0.15 M NaCl (abbreviated as PB0b). The HRP concentration in this solution was set to 0.5 μM, as determined on the basis of activity measurements with ABTS2– (1.0 mM) and H2O2 (0.2 mM) as substrates in PB0b and a calibration curve prepared with native HRP. A volume of 50 μL of this de-PG21000-BAH-HRP71 solution was then pipetted from one side onto the monolith inside the LDPE tube. The entire 50 μL volume (corresponding to the accessible pore volume, see Supporting Information, part 1) was taken up by the monolith through capillary forces, that is, no pumping was necessary and no leakage of the solution from the monolith was observed. After incubation for 1 h at RT, the monolith inside the LDPE tube was washed with PB0b. After washing, the tube was filled with PB0b, closed at both ends with Parafilm and stored at ∼4 °C in the refrigerator until further use. For more details, see Supporting Information, part 13.

Flow-through Activity Measurements of de-PG21000-BAH-HRP71 Immobilized in Monolith MH1 Placed inside LDPE Tubes

The activity measurements of the HRP-loaded monoliths placed inside LDPE tubes were carried out with ABTS2– (1.0 mM) and H2O2 (0.2 mM) in a similar way as described for the monoliths placed inside the borosilicate NMR tube. The only difference was that PB0b was used instead of PB0 and that the flow rate was 200 μL/min instead of 35 μL/min. For more details, see Supporting Information, part 13.

Results and Discussion

Preparation and Characterization of the Denpol–BAH–Enzyme Conjugates de-PG2-BAH-HRP and de-PG2-BAH-proK (n = 750, 1000, or 20 000)

During the first phase of the investigation, four different denpol–BAH–enzyme conjugates were synthesized from the denpol de-PG2 (Figure ). Three batches de-PG2 with different number averaged degrees of polymerization, n, and different polydispersity indices (PDI) were prepared and used, abbreviated as de-PG2750, de-PG21000, and de-PG220000, see Table and Section . In a first step, some of the amino groups of the three denpols were modified with N-succinimidyl 6-hydrazinonicotinate acetone hydrazone (S-HyNic) and then purified by ultrafiltration, see Section . The AFM images taken from the HyNic-modified denpols, which were deposited from aqueous solution onto mica and then visualized in the dry state, confirm qualitatively that one of the three batches (de-PG220000) contained rather long chains (Figure ). In a second step, the two enzymes, HRP and proK, were modified with N-succinimidyl 4-formylbenzoate (S-4FB) and then purified by ultrafiltration, see Section . Finally, by simple mixing of two aqueous solutions of HyNic-modified denpol (de-PG2-HyNic) and 4FB-modified enzymes (HRP-4FB or proK-4FB) and subsequent purification by ultrafiltration, the following four denpol–BAH–enzyme conjugates were obtained, see Section : (i) de-PG2750-BAH-HRP62; (ii) de-PG21000-BAH-HRP70; (iii) de-PG220000-BAH-HRP1850; and (iv) de-PG2750-BAH-proK115. The subscript after the enzymes HRP and proK indicates the average number of enzyme molecules per denpol chain. Based on the quantification of the enzyme modification with 4FB [MSR(4FB) < 1.0, Table S2], the enzyme molecules were considered to be bound to the denpol through one single BAH bond.

Activity and Stability of the Denpol–BAH–Enzyme Conjugates in Aqueous Solution

After preparation of the denpol–BAH–enzyme conjugates, their activities were tested. In the case of HRP, the substrates used were ABTS2– (1.0 mM) and H2O2 (0.2 mM), measured at pH = 7.0 (PB0) and 25 °C. For proK, the substrate was Suc-AAPF-pNA (0.5 mM), also measured at pH = 7.0 (MopsB2) and 25 °C.

The stability of pH = 7.0 solutions of denpol-bound HRP stored up to 30 days at 4 °C is the same as the stability of solutions of free HRP and of HRP-4FB at the same HRP concentration of 2 μM; in all cases, the HRP remained highly stable, see Figure . In the case of pH = 7.0 solutions of proK, again stored at 4 °C, there was a significant loss in enzyme activity for all samples, independent of whether proK was bound to de-PG2750 or whether it was free in solution, unmodified (proK) or 4FB-modified (proK-4FB); the drop in activity during 30 days storage was about 45% (Figure ), in agreement with our previous findings for de-PG22000-BAH-proK140.[29]

Figure 4

Stability of native HRP, HRP-4FB, and de-PG2750/1000/20000-BAH-HRP stored at 4 °C, pH = 7.0 (PB0). The HRP and HRP-4FB solutions were stored at a concentration of 2 μM. The concentration of the de-PG2750/1000/20000-BAH-HRP solution corresponded to 2 μM HRP, as determined from the change of the UV/vis absorption caused by the formation of the BAH bond during the conjugation reaction. The enzymatic activity measurements were performed in quartz cuvettes (l = 1.0 cm) at pH = 7.0 (PB0) using a 100× diluted storage solution, 1.0 mM ABTS2- and 0.2 mM H2O2 as substrate.

Figure 5

Stability of native proK, proK-4FB, and de-PG2750-BAH-proK115 stored at 4 °C, pH = 7.0 (MopsB2). The proK and proK-4FB solutions were stored at a concentration of 1 μM. The concentration of proK in the conjugate solutions was 10 μM, as determined from the BAH concentration. The enzymatic activity measurements were performed in quartz cuvettes (l = 1.0 cm) at pH = 7.0 (MopsB2) using a 100× diluted storage solution and 0.5 mM Suc-AAPF-pNA as substrate.

Activity and Stability of Denpol–BAH–Enzyme Conjugates Immobilized on Microscopy Glass Coverslips

The prepared denpol–BAH–enzyme conjugates were immobilized on round-shaped microscopy glass coverslips (see Section ) and the activity of the immobilized enzymes was measured. Before that, the possible release of the immobilized enzymes from the glass surface into the buffer solution in which the coverslips were stored was analyzed by determining the enzyme activity in the buffer solution after repetitive washing cycles. In all cases, no significant amounts of released enzyme could be detected (Figure S3). This indicates that there was either (i) no significant enzyme release from the glass surface under the conditions used, (ii) the released enzymes were inactivated, or (iii) there was no enzyme immobilized at all which could be released. This latter possibility was not the case, as was demonstrated with experiments in which the activity of the immobilized enzymes was measured by first replacing the pH = 7.0 storage buffer solution with a substrate solution, followed by incubation for 2 min at room temperature (see Section ). Afterward, the amount of product formed was quantified spectrophotometrically, and the measurements were repeated after washing the coverslips carrying the immobilized enzymes again with storage buffer (pH = 7.0), see Figure . For all denpol–BAH–enzyme conjugates investigated, immobilization was successful and the activity of the immobilized enzymes remained stable for several washing cycles.

Figure 6

Stability of HRP (a) and proK (b) immobilized on glass coverslips. After enzyme immobilization (in MesB1) and subsequent washing with buffer solution (PB0 for HRP and MopsB2 for proK), the glass coverslips carrying the immobilized enzymes were analyzed at 25 °C for enzymatic activity by inserting the coverslips into a substrate solution (1.0 mM ABTS2–, 0.2 mM H2O2 for HRP, pH = 7.0, PB0; 0.5 mM Suc-AAPF-pNA for proK, pH = 7.0, MopsB2) for 2 min. Product formation was monitored at either 414 nm (a) (for HRP, ABTS•–) or 410 nm (b) (for proK, p-nitroaniline). After removal of the substrate solution and washing with PB0 or MopsB2, the activities were measured again. The procedure was repeated another eight (for HRP) or six times (for proK). Control measurements with coverslips which were treated with free HRP or proK instead of the conjugates are also shown.

In the case of the denpol–BAH–HRP conjugates, the lowest activity was observed for the conjugate prepared from the longest denpol (de-PG220000-BAH-HRP1850), see Figure a. Therefore, using very long denpol chains (20 000 r.u.) did not lead to the immobilization of a higher HRP activity, when compared to the conjugates prepared and purified in the same way as the conjugates prepared from de-PG2750 and de-PG21000 and immobilized in the same way.

For all three conjugates used (including those prepared with de-PG21000 and de-PG2750), the activity of the coverslips carrying the conjugates remained stable, without any significant HRP leakage from the coverslip surface (Figure S3a). No immobilization occurred with solutions of free HRP or with free HRP-4FB.

For the de-PG2750-BAH-proK conjugate analyzed, it is clear that it stably adsorbed on the glass coverslip, while the amount of immobilized free proK or prok-4FB was very low, see Figure b. No significant proK release from the coverslip surface could be detected during the applied washing steps, see Figure S3b.

In summary, all experiments showed that all tested de-PG2-BAH-HRP and de-PG2-BAH-proK conjugates can be immobilized on glass coverslips. Very long denpol chains are not required for a stable immobilization and apparently are not beneficial for achieving high immobilization yields. Denpols with about 1000 r.u. are sufficient. These data can be compared with the results obtained previously with de-PG22000-BAH-proK140 (Küchler et al., 2015)[29] and with de-PG21400-BAH-HRP108 (Küchler et al., 2015).[27] They are consistent with each other and confirm the good reproducibility of the denpol-based immobilization method.

Activity and Stability of Denpol–BAH–Enzyme Conjugates Immobilized inside Glass Micropipettes

After confirming the stable adsorption of the prepared denpol–BAH–enzyme conjugates on microscopy glass coverslips for enzymatic reactions in bulk solutions in which the coverslips were immersed (see Section ), we investigated the immobilization of the same conjugates inside glass micropipettes with an inner diameter of 1.6 mm and a total volume of 280 μL for use under continuous flow conditions (see Section ). The chosen standard conditions we used for the immobilization of the enzymes through simple adsorption of the conjugates for 2 h at RT were 2.0 μM enzyme in MesB1 (pH = 4.7), see Section . After washing with buffer solution (PB0 in the case of HRP, MopsB2 for proK), substrate solutions were pumped through the micropipettes and the conversion of the substrates into products was followed by recording the absorption spectrum of the outflow from the micropipette at intervals of 10 min for 6 h, see Figure . All micropipettes which were treated with the conjugates showed enzyme activity, in qualitative agreement with our previous investigations of de-PG21400-BAH-HRP108 (Küchler et al., 2015)[27] and de-PG22000-BAG-proK140 (Küchler et al., 2015),[29] while those micropipettes which were treated with free enzyme (HRP or proK) or free 4FB-modified enzymes (HRP-4FB or proK-4FB) did not lead to a measureable substrate conversion, as there was no stable adsorption/immobilization of the free enzymes. This latter finding is in agreement with the experiments using the glass coverslips (see Section ). Concerning the data for the denpol–BAH–HRP conjugates, the highest activity (for the same conjugate preparation, purification, and immobilization protocols) was achieved with de-PG2750-BAH-HRP62, followed by de-PG21000-BAH-HRP70, and finally de-PG220000-BAH-HRP1850 (Figure a).

Figure 7

Use of “200 μL” glass micropipettes containing immobilized de-PG2-BAH-HRP (a) or de-PG2-BAH-proK (b) as flow-through reactors. Immobilization was performed at 2.0 μM enzyme concentrations in MesB1 for 2 h at room temperature, see Section . For (a), the substrates were ABTS2– (1.0 mM) and H2O2 (0.2 mM), pH = 7.0 (PB0); for (b), the substrate was Suc-AAPF-pNA (0.5 mM), pH = 7.0 (MopsB2). The flow rate was set to 35 μL/min, and the outflow was analyzed spectrophotometrically. Readings of A414 (a) and A410 (b) are plotted. They indicate formation of ABTS•– (a) and p-nitroaniline (b), respectively. For (a), the concentration of ABTS•– in the outflow was very high, such that the A414 values were no more within the linearity range of the spectrophotometer (JASCO V-670) for the used 1.0 cm path length. Therefore, the outflow solution was diluted 10 times with PB0 and then A414 was determined with a 1.0 cm cuvette. The values given in the figure are the calculated values for undiluted solutions in a 0.1 cm cuvette.

In summary, stable immobilization of de-PG2-BAH-HRP and de-PG2-BAH-proK conjugates in glass micropipettes is possible with denpols consisting of about 750, 1000, or 20 000 r.u. For the HRP conjugates, the use of de-PG220000 for the conjugate preparation resulted in lower immobilized enzyme activity, as compared to the conjugates prepared from de-PG2750 or de-PG21000, in qualitative agreement with the results of the coverslips (Figure a). The reason for this denpol chain length difference is not clear. It may be that for the longer denpol–enzyme conjugates, more unoccupied areas within the micropipettes remained because they could not be covered with long chains completely.

Activity and Stability of Denpol–BAH–Enzyme Conjugates in Silica Monoliths

In a first series of experiments, the possible immobilization of the denpol–BAH–enzyme conjugates inside MH1 monoliths was investigated for two different applications: as batch reactor and as flow reactor (see Section ). The two systems are illustrated in Figure . For “loading” the cylindrical monoliths with enzyme, a 5 mm long piece (diameter: 4 mm) of cleaned monolith was first placed inside a 2 mL PP reaction tube to which a denpol–BAH–enzyme solution (in MesB1, 2 μM enzyme) was added (batch reactor). In the case of the flow reactor using the borosilicate NMR tube as a monolith holder, the same conjugate solution was added to the monolith after it was placed inside the tube (cut to the same length as the monolith). For both systems, the solution containing non-adsorbed conjugate was removed after 2 h at RT by washing with PB0 (see Section ).

Figure 8

Schematic drawings of the use of MH1 monoliths as batch reactor (a) and as flow reactor with either a borosilicate NMR tube or an LDPE tube as monolith holder (b).

For the batch system (Figure a), the enzymatic activity and stability of the immobilized enzymes were determined by using different de-PG2-BAH-HRP conjugates and ABTS2– (1.0 mM) and H2O2 (0.2 mM) as substrate solution into which the enzyme-loaded monoliths were immersed. After 1 min incubation at RT, the formation of reaction product (ABTS•–) was analyzed spectrophotometrically, whereby the formed product mainly stayed inside the monolith (dark green color) and had to be forced out by accelerating its release into the outer solution by gentle shaking of the reaction tube. As shown in Figure , for all conjugates (including the one prepared from de-PG220000), enzyme activity in the batch reactor could be detected. The monoliths containing immobilized HRP were used repeatedly with intermittent washing steps with buffer solution (PB0). Moreover, also in the case of the monolith loading with free HRP and with free 4FB-modified HRP (HRP-4FB), enzymatic activity was clearly observed. This indicates that both, the free enzymes and denpol-bound enzymes, adsorbed on and/or inside the monolith, and that in the case of free enzymes the washing steps applied did not result in a (complete) release of the enzymes from the monolith. This is in clear contrast to the case of the coverslips and the micropipettes, see above. We conclude that the porous monolith structure retains free enzymes, at least to some extent.

Figure 9

Activity of HRP immobilized in MH1 monoliths (MesB1) at 25 °C, analyzed with ABTS2– (1.0 mM) and H2O2 (0.2 mM) in a batch reactor system at pH = 7.0 (PB0), see Section . The formation of ABTS•– after 1 min is expressed as A414, measured after shaking the reaction tube to release the formed products from the monoliths.

In a next step, monoliths loaded with enzymes—HRP or proK—were analyzed as flow reactors. Both ends of the tubes were connected to PTFE tubing with short pieces of silicone tubing. After short prewashing, substrate solutions were then pumped through the monolith-filled tubes and the solutions eluting from the tubes were analyzed spectrophotometrically for product formation (Figure ). All monoliths into which the conjugates had been immobilized showed stable product outflow values during an operation time of 700 min. If solutions of free HRP or free HRP-4FB were used, the outflow values for the formed product (ABTS•–) initially were relatively high, but started to drop to very low values (Figure a). This shows that free enzymes are not very stably retained. In the case of proK, the monoliths which were loaded with free enzymes showed low outflows of product (p-nitroaniline) throughout the entire operation time (Figure b).

Figure 10

Activity of HRP (a) and proK (b) immobilized in MH1 monolith flow reactors (placed inside small pieces of an NMR tube, see Section and Figure ) at 25 °C. For (a), a pH = 7.0 solution containing 1.0 mM ABTS2– and 0.2 mM H2O2 was pumped through the enzyme-containing monolith (PB0). For (b), the pH = 7.0 solution contained Suc-AAPF-pNA at 0.5 mM (MopsB2). In both cases, the flow rate was 35 μL/min. The outflow from the monolith was analyzed spectrophotometrically. Readings of A414 [(a) ABTS•–] or A410 [(b) p-nitroaniline] are plotted.

Based on the successful immobilization of the different de-PG2-BAH-HRP and de-PG2-BAH-proK conjugates in MH1 monolith flow reactors, the reproducibility of this finding was investigated in a second phase of the work. For this, an additional de-PG21000-BAH-HRP conjugate was synthesized (de-PG21000-BAH-HRP71, see Section and Supporting Information, part 11, Figures S4 and S5). This conjugate was then immobilized in the monolith, whereby alternative monolith holder tubes were applied. Flow reactors were prepared by using tubes made from LDPE instead of borosilicate (Section , Figures S6 and S7). Furthermore, the monoliths were incubated at pH = 7.0 (instead of 4.7) with lower amounts of conjugates (0.5 μM HRP instead of 2 μM, as determined by activity measurements using a calibration curve with free HRP, see Figure S8) and a shorter incubation time (1 h instead of 2 h). The monolith was washed extensively by using a peristaltic pump so that weakly adsorbed enzymes were washed out. Furthermore, the flow rate was increased substantially (from 35 to 200 μL/min). With these modifications, the immobilization was again successful. The results confirmed the most important finding of this work. The type of denpol–enzyme conjugates which we have investigated in the past couple of years[27−29] can be immobilized stably inside MH1 monoliths for flow-through reactor applications. When the monolith was loaded with a solution of free HRP instead of a solution of de-PG21000-BAH-HRP71, the activity of the monolith was very low and vanished nearly completely after repeated use and storage at 4 °C (Figure ). However, a MH1 monolith carrying de-PG21000-BAH-HRP71 showed constant activity when repeatedly used at room temperature after storage at 4 °C (Figure ). Furthermore, by varying the concentration of H2O2 in a pH = 7.0 solution pumped through another monolith of the same type—at fixed [ABTS2–] = 1.0 mM—the monolith outflow of ABTS•– was shown to linearly depend on the H2O2 concentration in the range of [H2O2] = 1–50 μM (Figure ). In these experiments, 1 mM ABTS2– solutions of increasing H2O2 concentrations (“analyte solutions”) were pumped through the monolith containing HRP conjugate. Between the measurements, buffer solution was pumped through the monolith for a fixed time period (Figure a,b). With this set-up (Figure S9), the outflow signal (ΔA414, originating from ABTS•–) depended linearly on the H2O2 concentration between 1 and 50 μM, corresponding to 34 ng/mL H2O2 to 1.7 μg/mL H2O2 present in the analyte solution (Figure c). In a follow-up work, we will apply this HRP-containing spectrophotometric flow reactor system for the quantitative determination of hydrogen peroxide in biological samples.

Figure 11

Activity of HRP immobilized in a MH1 monolith flow reactor (placed inside an LDPE tube, see Section and Figure ). The monolith was loaded with de-PG21000-BAH-HRP71. After prewashing with buffer solution (0.01 M phosphate, 0.15 M NaCl, pH = 7.0, PB0b), a substrate solution consisting of 1.0 mM ABTS2– and 0.2 mM H2O2 (pH = 7.0, PB0b) was pumped through the monolith at 25 °C at a flow rate of 200 μL/min. A414 (l = 0.1 cm) of the outflowing solution was measured. After a run time of 180 min, the monolith was washed with PB0b and then stored at 4 °C for 2 weeks. Afterward, a substrate solution of the same composition was again passed through the monolith at room temperature, followed by washing with PB0b and storage for another 2 weeks at 4 °C. Control measurements were carried out with a monolith which was loaded with free HRP (red data points). The solid lines indicate the average values to guide the eye.

Figure 12

Use of a MH1 monolith flow reactor containing immobilized HRP for the quantification of H2O2 in aqueous solution. The monolith was loaded with de-PG21000-BAH-HRP71 (placed inside an LDPE tube, see Section and Figure ). After prewashing with pH = 7.0 buffer solution (PB0b), substrate solutions consisting of 1.0 mM ABTS2– and varying amounts of H2O2 (pH = 7.0, PB0b) were pumped through the monolith at 25 °C. (a) Concentrations of H2O2 varied between 10 and 200 μM (flow rate: 200 μL/min for 10 min) and A414 (l = 0.1 cm) of the outflow solution was measured. Before exchanging the substrate solution, the monolith was washed with PB0b for 5 min. (b) Same type of measurements as described in (a), but varying the concentration of H2O2 between 0 and 10 μM and measuring A414 (l = 1.0 cm). (c) Correlation between H2O2 concentration in the substrate solution and the reaction product formed (ABTS•–), as expressed by A414 (values taken from (a,b); the data from (b) are scaled for a path length of l = 0.1 cm).

Conclusions

Previous investigations have shown that the monolith MH1—a “macro- and mesoporous silica”—has excellent properties for the covalent immobilization of enzymes.[10,16,37] We have shown that the immobilization of HRP and proK in this monolith is easily possible with the help of the polycationic, second generation dendronized polymer de-PG2. The enzyme molecules are first bound covalently to de-PG2, followed by stable noncovalent adsorption of the obtained conjugates. Experiments about the immobilization on flat silicate surfaces (glass coverslips and inside glass micropipettes) with de-PG2 of three different average chain lengths showed that the use of very long chains (about 20 000 r.u.) is of no advantage. In fact, the immobilized enzyme activity was lower compared to the same denpol with on average about 1000 r.u. if comparison is made on the basis of identical amounts of enzyme used for the conjugate preparation and subsequent immobilization (Figure ). In other words, the enzyme immobilization efficiency under the conditions used is lower for the conjugates prepared with the very long denpol. This is an important finding for future studies of the application and further examination of this denpol-based immobilization method. Using de-PG2 with about 1000 r.u. appears to be an ideal choice. It allows highly reproducible de-PG2-BAH-enzyme conjugate preparation and immobilization in silica monoliths MH1. As a preliminary proof-of-principle, the use of a de-PG2-BAH-HRP-loaded monolith for the quantitative determination of H2O2 was demonstrated (Figure ). Compared to the previously developed micropipette system,[27,29,30] the monolith MH1 allows a higher enzyme loading due to the very large internal surface area onto which the denpol–enzyme conjugates can be adsorbed. This can easily be seen qualitatively by comparing the product formation in the flow-though reaction from de-PG2-BAH-HRP-loaded micropipettes (Figure ) and from de-PG2-BAH-HRP-loaded monoliths (Figure a). For the monoliths, product formation (expressed as A414) was always the same or higher (for all conjugates) than for the micropipettes, despite the substantially shorter resting time of the substrate within the monoliths as compared to the micropipettes. One of our next attempts is toward the development of MH1-flow reactors for enzymatic cascade reactions.