Temperature control of the enantioselectivity in the lipasecatalyzed resolutions

Takashi Sakai

Division of Chemistry and Biochemistry, Graduate School of Natural Science and Technology, Okayama University, Okayama, Japan


In the lipase-catalyzed resolution, temperature control of enantioselectivity has been generally accepted for its simplicity and theoretical reliability. Lowering the reaction temperature usually enhances the enantioselectivity. Here, the historical and theoretical backgrounds of the temperature control of enantioselectivity and its applicability to the method are described. Recent literatures for the lipase-catalyzed resolutions to which the "low-temperature method" seems to be promising to enhance the enantioselectivity are also summarized.


Recently, a variety of methods for increasing the enantioselectivity in lipase-catalyzed reaction1ab have been developed, involving modulation of organic media; additive, acyl donor; and immobilization support. Ionic liquid1c d and supercritical CO21e—g recently emerged as efficient reaction media and/or supporting material in enzymatic reactions. Beside these methods, temperature control in the lipase-catalyzed resolution of alcohols is now accepted as a theoretically reliable and generally applicable method, since we demonstrated that a lipase exerts its function even at very low temperatures down to — 80° C in an organic solvent.2a'f In our first attempt for resolution of 3-phenyl-2H-azirine-2-methanol (1), the enantioselectivity was found to be maximal at —40° C in ether as reported in 1997. We recently demonstrated that temperature modulation is generally effective in the lipase-catalyzed resolution of primary and secondary alcohols.2a—h Enzymes have long been believed to be temperature labile and have been used at around ambient temperatures; however, our finding opened the way to the wide range of temperature control of enantioselectivity in enzymatic reactions like chemical asymmetric reaction.

Although unusual temperature-induced rate control of enzymatic reaction was well investigated before 1970,3 the temperature control of enantioselectivity in enzymatic reaction (PLE, — 10°C) was initially reported by Jones and co-workers in 1986.4a Since then, many attempts4b—i were reported, although the reactions were mostly carried out above 0°C. The rational explanations for the temperature effect were then proposed by Overbeeke and Hult5 with lipases and by Phillips6a—f and co-workers with secondary alcohol dehydrogenase and PLE. In 1997, we reported that the temperature effect is observable with regulation below to very low temperature, and practically applicable to the lipase-catalyzed resolution of alcohols.2a Moreover, an immobilized lipase on porous ceramic particles (Toyonite), now commercially available as Amano lipase PS-C II, was found to be highly effective for enhancement of the reaction rate even at low temperatures.2c—h The immobilized lipase was also found to be effective for high-temperature reaction up to 120°C, in which bulky alcohol can be resolved without loss of enantioselectivity,7 maintaining the transition-state structure.8 On the other hand, the low-temperature method can also be applicable to hydrolytic asymmetric protonation of enol acetate.8 Here, the features and current applications of the temperature modulation of enantiose-lectivity are summarized. In addition, reported examples for the lipase-catalyzed resolution of primary alcohols are listed.


We first examined the lipase-catalyzed resolution of azirine-2-methanol 1,2a which we expected to have a versatile synthetic utility.211'9 As expected for primary alcohols,7'10 the enantioselectivity obtained in the transesterification with lipase PS in ether was low (E11 = 17 at best) at room temperature despite considerable efforts such as screening of lipases, solvents, additives, and acylating agents.12-17 As the final choice, we examined the reaction at low temperatures because of its high reaction rate and found that lowering the temperature to —40°C increased the E value of 17 (at 30°C) to a practically acceptable level of 99; however, further lowering the temperature rather decreased the enantioselectivity (Scheme 1).

Scheme 1: "Low-temperature method" in the lipase-catalyzed resolution of 3-phenyl-2H-azirine-2-methanol (1) for enhancement of the enantioselectivity.


Lowering the temperature in the lipase-catalyzed resolution usually enhances the enantioselectivity. The phenomenon does not come from the temperature-induced conformational change of lipase, but it is understandable on the basis of the theory of physical organic chemistry as explained below.56

In general, the enantioselectivity (E value11) in a kinetic-control reaction is determined by the ratio of the rates of two enantiomers and defined by Equation 1:

where kA and kB are the rate constants for a faster-reacting enantiomer (A) and a slower-reacting enantiomer (B). AAG* is the difference in the transition-state energies (AAG* = AGjiast — AGflow) for each reaction process, and R is gas constant. The lipase-catalyzed kinetic resolution is represented in the same way. Equation 1 is changed to Equation 2:

Figure 1 shows the plots for the relation between AAG* and E at three different temperatures: 50, 0, and —50°C. At the lower temperature (—50° C), the curve flattens quickly, and E value of 100 requires AAG* = 2047 cal mol-1, less than those of 2506 (at 0°C) and 2965 cal mol-1 (at 50°C), and thus a small difference in transition-state energy (AAG*) between the enantiomers gives a large effect on the enantioselectivity. Thus, lowering the temperature increases the





7 2500


J 2000

2047 cal mol-1

-it 1500

<1 1000

0 20 40 60 80 100

Figure 1: Relation between AAG and E value.

enantioselectivity if the temperature-induced change of AAG* were small between the temperatures.

From the transition-state theory the reaction rate is represented by Equation 3, where NA is Avogadro number and h is Plank's constant:

Equation 3 is transformed to Equation 5 by combination with Equation 4:

AH * / 1 \ RT AS* ln k = —— I ~ I + ln — + — (5)

The relation between ln k and 1/T for the reaction in Scheme 1 is shown in Fig. 2, where AH* and AS* for faster (A)- and slower (B)-enantiomers are tentatively estimated to be AH * = 4 kcal mol-1, AS* = 4.7 cal deg-1 mol-1 and AH * = 7 kcal mol-1, AS* = 9 cal deg-1 mol-1, respectively, on the basis of the observed data of AAH* (-3 kcal mol-1) and AAS* (-4.3 cal deg-1 mol-1) which are calculated below (Figs 4 and 5). Figure 2 shows that lowering the temperature (1/T) decreases both rates (ln kA and ln kB) by following nearly straight lines, respectively, because the contribution of the second term, ln(RT)/(NAh), is negligible. Therefore, lowering the temperature increases the difference in the reaction

Figure 2: Relation between ln k and 1/T for faster-reacting and slower-reacting enantiomers.

Figure 2: Relation between ln k and 1/T for faster-reacting and slower-reacting enantiomers.

rates for both enantiomers. The temperature effect on the enantioselectivity is thus primarily governed by the difference of AH * in the first term of Equation 5. On the other hand, Fig. 2 shows that raising the temperature decreases the enantioselectivity, giving racemic product at 425°C. The temperature is called racemic temperature (Tr).6f

Handling of the experimental data for the temperature effect and their theoretical consideration are described as follows. Temperature effect cannot be accurately discussed by use of Equation 2 because AAG* itself is temperature dependent (Equation 6), where AAH* = Af - AHs*low and AAS* = AS** - AS*^:

Combination of Equations 2 and 6 gives the Eyring equation 7:

Figure 3 shows the plot between temperature (T) and observed enantioselectivity (E value) for the reaction in Scheme 1. Lowering the temperature suddenly increases the E value until —40°C. The temperature effect can also be represented by the relation between ln E and 1/T as shown in Fig. 4, where ln E is increased linearly from 30 to —40°C to reach the best E value of 99. In contrast, further lowering the temperature rather decreases the E value by following another line. Therefore, the plot consists of two lines intersecting at —40°C, and this temperature is called "inversion temperature" (Tinv).18-20 The linear correlation suggests that the conformation of lipase is maintained in a course of the observed temperature range. At Tinv (—40°C), the transition-state structure may be changed for some reasons: a temperature-induced structural change of enzyme and/or a solute-solvent

120 100 80

Uj 60 40 20

120 100 80

Uj 60 40 20

Figure 3: Relation between observed E value and 1/T for the lipase-catalyzed reaction of (1) (Scheme 1).

Figure 3: Relation between observed E value and 1/T for the lipase-catalyzed reaction of (1) (Scheme 1).

Figure 4: Relation between ln E and 1/T for the lipase-catalyzed reaction of (1)

Figure 4: Relation between ln E and 1/T for the lipase-catalyzed reaction of (1)

cluster change,18-20 and so on. The phenomena of Tinv are interesting and further study is required to reveal the reason.

The temperature effect is theoretically broadened until 1/T = 0 as shown in Fig. 5, and the line is extrapolated to give the value of AAS* (ln E = AASVR). From the slope of the line, the value of AAH* is calculated (AAH* = -3.0 kcal mol-1, AAS* = -4.3 cal deg-1 mol-1). By this line, raising the temperature decreases the E value to reach the racemic temperature (Tr); however, further raising the temperature begins to increase the E value again toward opposite R/S

Figure 5: Broad range of relation between ln E and 1/T for the lipase-catalyzed reaction of (1) (Scheme 1).

Figure 5: Broad range of relation between ln E and 1/T for the lipase-catalyzed reaction of (1) (Scheme 1).

selectivity. Tr can be calculated to be 425 °C by Equation 8, which is derived from Equation 6 (AAG* = 0). Usually the observable temperature effect is enthalpy-driven like this. On the other hand, in an entropy-driven reaction having a large AAS* (line a) or a small AAH (line b), Tr is observable in a range of lower temperatures. If the Tr is observed at 30 °C, E value is increased by both lowering and raising the temperature from 30 °C, but to opposite R/S enantiomers.


Example of the entropy-driven reaction is rare for enzymatic reaction. One example in which Tr appeared at around room temperature in the enantioselective oxidation of secondary alcohols with secondary alcohol dehydrogenase (SADH) from a thermophilic bacterium, Thermoanaerobactor ethanolicus is shown in Fig. 6.6f The reaction of 2-butanol gave an oxidation product together with a remaining optically active alcohol, showing a linear temperature effect (log E6f — 1/T) having Tr at 26 °C. Thus, raising or lowering the temperature from 26 °C increases the R or S selectivity of the remaining alcohol, respectively. However,

Was this article helpful?

0 -1

Post a comment