Full Thermal Switching of Enzymes by Thermoresponsive Poly(2‐oxazoline)‐Based Enzyme Inhibitors

Abstract Controlling the activity of enzymes is an important feature for many processes in medicine, bioanalytics, and biotechnology. So far, it has not been possible to fully switch biocatalysts on and off by thermoresponsive enzyme inhibitors. Herein, we present poly(2‐oxazoline)s with iminodiacetic acid end groups (POx‐IDA) that are lower critical solution temperature (LCST) polymers and thus thermosensitive. They are capable of reversibly inhibiting the activity of horse radish peroxidase and laccase by more than 99 %. Increasing the temperature makes the POx‐IDA precipitate, which leads to 100 % recovery of the enzyme activity. This switching cycle is fully reversible. The LCST of the POx‐IDA can be tuned by varying the polymer composition to generate a wide range of switching windows.


Introduction
Controlling the activity of enzymes is ak ey issue for influencing the biochemistry of life-forms, required in medicine and biotechnology,b ut it is also useful for varying the activity of biosensors. For example, many processes such as measuring metabolites in blood or determining certain oxidizing compounds in water are based on enzymatic assays, whichm easure as ignal increase with time affordingaprecisep reparation time andp unctual measurement. Controlling the activity of the enzyme would greatlyi mprove the reliability of such crucial measurements by switchingt he respective enzyme on only for ac ertain periodo ft ime and switching it off afterwards. In terms of medical treatment it would be of great benefit to reversibly control the in vivo activity of drugs,w hich are very often enzymei nhibitors, by using externalt riggers.
Typical strategies for controlling enzyme activity are shown in Figure 1.
The most commons trategyi nc ontrolling the activity of biocatalystsi st he encapsulationo fe nzymes into light-or thermoresponsivea mphiphilic copolymers, hydrogels,p olymersomes, or nanoparticles ( Figure 1a). [1][2][3][4] Thisway,the accessibility of the substrate is controlled by the switchable aggregation or swelling behavior of the surrounding shell. This works best for isolated enzymes andi su sually not specific and only useful for larger substrates.
Another way to achieve this was pioneered by Staytona nd Hoffmann in 2003, who have managed to control enzyme activity by attaching at hermosensitive copolymer based on N,Ndimethylacrylamide (DMAM) and N-4-phenylazophenylacrylamide (AZAAm) monomers near the active site of the enzymee ndoglucanase 12A (for the concept, see Figure 2b,b elow). This conjugate showed8 2% of its originale nzyme activity and a full activity shuto ff above the phase transition temperature of the UCSTp olymer. [5] Others have successfullyf ollowed this approach by an umber of further polymers attached to enzymes. [6][7][8] Although elegant,c reatings uch conjugates is still quite elaborate.
Alternatively,r esearchers have createdl ow molecular weight enzymei nhibitors, which can be changed in their structure. [9,10] or form aggregates that are not active anymore. [11] This method illustratedi nF igure 1c has the limitation that the change in structure is an equilibrium state not fully in favor of one or the other form andt hus usually not efficient to fully switche nzymea ctivity.
The combination of polymers with enzymei nhibiting functionalg roups is ap ossibility to use the thermally induced phaset ransition of the latter with activity control of the biocatalysts ( Figure 1d). However,t he enzyme inhibiting groups are still accessible after the phase transition resultingi nn of ull switcha bility. [12] We proposet hat this can be circumventedb yu sing enzyme inhibitors attached to the terminal of at hermosensitive polymer ( Figure 1E). Am acromolecule designedt his way should be able to "hide" the inhibitor inside the coil after thermally inducedp hase transition, which should enableafull on and off switching of the enzyme activity.C hang et al. could already show that the antibiotic ciprofloxacin, whichi sa ne nzyme inhibitor,c ould be thermally controlled in its activity by hindering the passage of the polymer through the bacterial cell membrane. [13] So far it wasn ot explored if the thermally inducedp hase transition can truly deactivatea ni nhibitor attachedt ot he end group.

Results and Discussion
We have previously reported on poly(2-oxazoline)s with an iminodiacetic acid (IDA) end group (POx-IDA) as an entropic, noncompetitive inhibitor for horse radish peroxidase( HRP), which is widely used in many biological assays,i ncluding ELISA,g lucose sensors, hydrogen peroxide sensors. [14][15][16][17] These polymers were also found to be competitive inhibitors for laccase, which can be found in bioremediation, [18] chemical synthesis, [19] wine stabilization, and biosensing. [20] Te mperature control of the activity of these enzymes could improvet he biosensors, but might also allow to specifically interact with production cascades to controlt he selective synthesis of fine chemicals. [21,22] Diluting the polymer/enzyme mixtures alwaysr esultsi nf ulla ctivity of the enzyme, indicating that the inhibition is fully reversible. Interestingly,P OX-IDAs are dead-endi nhibitors for both enzymes, that is, they can fully inhibitt he enzyme activity.T his offers the potential of fully switching enzyme activity.P oly(2oxazoline)s have been found to be excellent lower critical solution temperature (LCST)p olymers, which can be adjusted in their cloud-point temperature by copolymerization of different 2-alkyl-oxazolines and by varying the end groups. [23][24][25] In order to render POx-IDA into at hermoswitchable enzyme inhibitor, an umber of different POxw ith LCST behavior was synthesized by polymerizing various2 -alkyl-2-oxazolines as homopolymers and statistical copolymers that are known as LCST polymers from literature. [25] The polymers where then terminated with dimethyl-IDA (IDD) and subsequentlyt reated with aqueous NaOH to achieve an IDA end group. All analytical data of these polymers including 1 HNMR spectra and SEC results are given in Table S1 in the Supporting Information.
The thermoresponsivea ctivity control was performed photometrically at temperatures below and above the cloud-point temperature (T cp ). Several substrates for HRP andl accase are known. ApplyingL CST polymers in ac omplex environment such as an enzymea ssay or ac hemical synthesis is always critical, because various factors such as cosolutes, buffer salts or any other chemical additive, for example, enzymes ubstrates, can interfere with T cp . [26,27] In order to determine if the polymers are affected by the contents of the respective enzyme assay solution, the cloud points were measured in assays for HRP and laccase activity using different substrates in the concentrationsapplied in the assay ( Figure 2).  As seen in Figure 2t he T cp of the investigated polymer of 17 8Ci nw ater downshifts by 2-3 Ki nt he chaotropic acetate buffer.W hen adding ABTS (5 mm)t ot his buffer T cp is increased by 5K.I nc ontrast, addition of the substrate guaiacol to the buffer affords ad ecrease in T cp by 5-6 K. This shows ag eneral problem of thermoresponsive polymers, which are greatly influenced by their surroundings and therefore become unreliable when used in changing natural environments. Ta ble 1s ummarizes the cloud-point temperatures of the different POx-IDAs in their respective enzymea ssay.
As seen in Ta ble 1, the transitionr ange of the POx-IDA can be varied over ab road range of temperatures of T cp on of 4-40 8C( full enzymei nhibition below these temperatures) and of T cp off of 10-48 8C( full enzymea ctivity above these temperatures). The transition range of the investigated POx-IDAs is rather broad compared the literature known systems, which is due to the relativelyb road dispersity (Table S1) andt he IDA end group. The statistical distribution of the different monomers in the copolymers leads to af urther broadening of the transition. This will prevent ad iscrete switching but offers the possibility to gradually control( "dim") the activity in ab road range with varying on and off regions ranging from below 4t o 48 8C. Table 1. Cloud-point temperatures T CP of the synthesizedP Ox with an IDA end groupa t2 0mgmL À1 in aqueousa cetateb uffer pH 5.0. The experimental protocol of the synthesis, the analyticald ata of the polymers, detailso nt he T CP measurements, the respectivep hase transition curves ( Figures S1-S5), and the enzyme activity assay compositions are given in the SupportingI nformation.  The activity of laccase and HRP was measured in the presence of varying amountso ft he respective POx-IDA. The objective was to find ac oncentration where the enzyme activity is completely switchedO FF (more than 99 %i nhibition) at ac ertain temperature and fully switched ON at ah igher temperature (more than 99 %a ctivity). The activity wasm easured at concentrations of up to 8mm POx-IDA below the respective onset temperature T cp on .T his concentration was found to be the highest value that allowed complete recovering of the enzyme activity upon heatinga bove T cp off for all POx-IDA.T he enzyme activities are alwaysc ompared to the native enzyme activity at the respective temperature.
All copoly(2-oxazoline)s P(EtOx-stat-BuOx)-IDA fully inhibit both enzymes at 8mm below T cp on and the activity is fully recovereda bove the respective T cp off .I nc ontrast,P (PropOx 55 )-IDA fully inhibits HRP,b ut are is not capable of fully inhibit laccase at 8mm (77 %i nhibition). The copolymer P(PropOx 14 -stat-iProOx 25 )-IDA inhibits only laccase by more than 97 %aconcentrationo f8m m,w hile it does not affect the activity of HRP at this concentration. This shows that not only the IDA group influences the inhibition, but also that changingt he nature of the polymer backbonea llows the creation of enzymes elective inhibitors.
Having established that POx-IDAs are efficient thermal enzyme switches, it was now addressed whether this process is reversible. To this end the polymer P(EtOx 15 -stat-BuOx 15 )-IDA and laccase or HRP,r espectively,w ere added in ac uvette containing the respective activity assay and the absorbance was measured at 7 8C( below T set off )f or 3min. Then the cuvette was quickly heatedt o3 78C( above T set on )u sing aw ater bath, was kept there for 2.5 min, and was then cooled to 7 8Cu sing an ice bath and put back into the spectrophotometer.T his cycle was repeated three times. As seen in Figure 3, the activity of the enzymes can indeed be reversiblys witched ON and OFF with temperature. The switched ON laccase activity is always similar as provenb yt he equidistanta bsorbance increase. The activity of HRP seemst ob el ower with each activation step. This is due to the fact that HRP gets quicklyd eactivated in the activity assay as evident from the activity curve shown in Figure 3d.T hus, the switching of the activity is fully reversible for both enzymes and the general activity of the respective enzyme is not affected by the switching.
While the change of temperature by 30 Ks lows the reaction rate of the laccase catalyzedo xidation by only af actor of 5 (due to the relatively low activation energy,F igure S11), the switchable inhibitor fully stops the reaction, which is af actor of at least 20 times more than the temperature effect.F igure 3e shows photographs of the reaction mixtures of the first cycle that clearly illustrate the difference between inhibited and not inhibited reactionmixture.
Besides switching enzyme activity on and off, the POx-IDA are also capable of "dimming" enzymea ctivity.T his is demonstrated on the example of P(EtOx 26 -stat-BuOx 14 )-IDA that has a particularly broad transition range (T cp = 20 8C, switching range 13-25 8Cat8mm in the laccase assay). Figure 4shows the relative activity of laccase in the presence of 8mm P(EtOx 26 -stat-BuOx 14 )-IDA at different temperatures compared with the re-spectivet hermal phase transition diagram of the polymer and the degree of inhibition at different concentrations at 10 8C. The resultsc learly prove that the laccase activity can be tuned by changing the temperature. For example, whenthe temperature wasi ncreased from 10 8C( activity < 1%)t o1 98C, the relative activity raises to 55 %. This corresponds to the inhibition of the enzyme by 2.5 mm the same polymer at 10 8C. In other words, some 70 %ofthe inhibitor is precipitated and thus inactive at 19 8C. Given the phase transition curve at the bottom of Figure4,t his is al ikely scenario and supports the precipitation of the polymer as switching mechanism. The laccase activity is recovered to 100 %b yf urtheri ncreasing the temperature to 30 8C.
The precipitation of the inhibitord oes not only lead to activation of the enzyme, it also offers the possibility to remove it. This was tested on the example of P(EtOx 26 -stat-BuOx 14 )-IDA, which was added to al accase solution( final concentration of 8mm)a nd cooled to 7 8C, whichr esults in full inhibition of the enzyme. Then the temperature was increased to 37 8Ca nd the precipitate was filtered off using ap olypropylene syringe filter. The resulting filtrate showed 92 %o ft he activity of the respective laccase control solution,i ndicating that the switchable inhibitor can easily be removed in contrast to systems, where the enzyme is covalentlym odifiedw ith thermoresponsive polymer.