Part 2: Playing around with CYP inhibition in metabolism studies.

Introduction 

Let’s start off with another analogy. Every car is made in such a way that the average consumer will fit in it, can reach the pedals, and can drive comfortably at a reasonable speed. In enzymology language: the substrate fits well in the active site, the catalytically active amino acids are in close proximity to the substrate and the turnover rate is good. Now picture a cartoon character with a normal body and arms but comically short legs [1]. Since this person that occupies the seat cannot reach the pedals, the car is not moving. In enzymology: the inhibitor occupies the active site but is not converted. Enzyme engineering is like adjusting the seat and the steering wheel into an optimal configuration. The result? Fast catalysis and probably a lot of speeding tickets.

I just explained competitive inhibition in the way it is normally regarded: an un-reactive inhibitor molecule binds to the enzyme active site and prevents the real substrate from binding and undergoing catalysis. How good the inhibitor binds to the active site is expressed as the inhibition constant K_i. Substrate binding is (roughly) expressed as K_m.[2] How severe the effect of the inhibitor will be depends on how good the substrate binds in relation to the inhibitor binding. A low value for K_i only means that this is potentially a very good inhibitor.

Table 1: Two examples that show that the ratio K_i/K_m is a better indication of inhibition effect and not K_i by itself.

Now, inhibition must be seen in a much broader sense.

• Every molecule of substrate X (once it sits in the active site) acts as an inhibitor of the conversion of another molecule of substrate X.
• A molecule of a different substrate Y can act as an inhibitor to substrate X (depends on Ki and Km).
• The same goes for the product of the reaction (which is merely a substrate in the reverse reaction). In some cases, like with transaminase enzymes, the product is a strong inhibitor to the conversion of substrate.

Confused? Don’t be – just remember that an inhibitor is not always a substrate but that every substrate is an inhibitor. Let’s look at how the tool of inhibition can be used in R&D.

CYP phenotyping by inhibition in drug development

In drug development, the metabolic fate of a drug (candidate) is one of the important things you need to know. Most metabolism is done in the liver, by cells that contain microsomes that contain a cocktail of ~6-7 major oxidative enzymes from the cytochrome P450 family (CYP’s). “CYP phenotyping” is the study of elucidating which of the CYPs are involved in drug metabolism.

Figure 1: Design of the CYP phenotyping experiment. The assumed measured rates are listed.

A classical phenotyping experiment is done by two comparative measurements.[3] In experiment #1, the un-inhibited rate is determined. In experiment #2, the rate in the presence of a specific CYP inhibitor. There can now be two cases (Figure 1). In Case #1, there is no sign of inhibition since both rates are 25 and the particular CYP that is inhibited is not involved in metabolism of your candidate. In case #2, it seems there is some inhibition of an involved CYP with a rate down to 15 and we have found the culprit – or one of the culprits.

Why is this phenotyping experiment fundamentally flawed? Well, if we don’t know what the K_m of the candidate is, how can we make such a bold conclusion that there is or isn’t any inhibition? For all you know, the K_m of the candidate is very much lower than the Ki of the inhibitor and if that’s the case we should expect no effect (Table 1).

It is also possible that the metabolism is ‘taken over’ by some of the other CYPs in the microsomes. Theoretically, there can be a different CYP that has a higher K_m for the candidate but that now comes into play. The rate of that enzyme may even be higher, so it could happen that the measured rate now comes out close to 25! This is shown in Figure 2 with some hypothetical numbers.

Figure 2: (Using hypothetical numbers). In blue, the normal case without inhibitor. If we do the experiment at, say, a candidate concentration of 50 µM, CYP3A4 because of its low K_m will bind the candidate much better and do the majority of the metabolism, at a rate of 50 turnovers per second. In green is the case if we inhibit CYP 3A4. CYP2D6 will now be the only enzyme ‘in play’. Since the candidate concentration is much below the K_m, the enzyme does not work at its fastest but the 3-fold higher turnover rate compensates for that. The resulting metabolism rate can thus come close to the rate of CYP 3A4.

So, how do we need to do the CYP phenotyping? The better strategy is to use each CYP individually and test for reactivity. If there is no measurable reaction rate at a physiologically relevant concentration of the candidate, it is safe to conclude that that particular CYP will not be involved in the candidate’s (initial) metabolism. Nowadays, all important CYPs are available from recombinant expression systems in relatively pure form from various commercial sources.

Is there merit to the CYP phenotyping experiment with microsomes and inhibitors? Well, there is for a couple of reasons:

1. One CYP may do the first oxidation which only then is accepted by a second CYP for further oxidation. You will not see this in the individual experiments.
2. You may be interested in the minor products of the less-important CYPs and by knocking out the major metabolism, you can produce a bigger quantity of the minor metabolites for identification.
3. Any metabolite that is formed, may have an (adverse) effect on the metabolism by the other CYPs. So, overall metabolism is not just the sum of metabolism by each individual CYP. The use of inhibitors can assist in elucidating the clearance pathways.
4. Knowing the effect of inhibition of certain CYPs on the candidate's overall metabolism is important to uncover potential drug-drug interactions. This happens when one drug’s metabolism interferes with another drug’s metabolism, leading to slow clearance and too high drug concentrations. CYP inhibition is actually also the reason why you should never drink grapefruit juice when you are on statin drugs like Lipitor® for lowering cholesterol.[4]

Conclusion

I hope I have shown that enzyme inhibition is an important parameter to study in drug development and that due to the complex nature of the liver microsomal metabolism and all the enzymes involved, there is not one easy method to study this.

As always feedback or questions welcome. I purposely did not distract too much by using all the proper units, etc.

References

[1] The “Asterix and Obelix” comic books by René Goscinny and Albert Uderzo from my childhood (and afterwards, I have to admit) come to mind. All Romans have names ending in –us (I guess to make it sound Latin). Below is legionnair Crismus Bonus; look at the short legs. http://www.asterix.com/

[2] Technically, K_s is the binding affinity of the substrate to the enzyme. However, because of its reactivity, this parameter is very difficult to measure and it is custom to use the Michaelis constant K_m instead.

[3] Edward H Kerns and Li Di in “drug-like properties: Concepts, Structure Design and Methods – from ADME to Toxicity Optimization. Academic Press, 2008. Pg. 341 “Metabolic phenotyping has been performed using the microsomal assay and adding an inhibitor that is specific for an isozyme [CYP family, EdV]. An increase in half-life in the presence of the inhibitor is an indication that the isozyme is important for the metabolism of the compound.”

[4] http://en.wikipedia.org/wiki/Grapefruit_drug_interactions

Part 1 - How many catalysts for ketone reductions do you need?

Introduction

Let’s start off with a simple question. If you already have a bicycle and you want to go faster on it, wouldn’t it be easier to just switch out some gears and maybe get better handle bars instead of finding all the new and improved parts that you need to build an entire bike?

It is exactly like that in current-day biocatalysis: it is much easier to modify an existing enzyme scaffold to enable a reaction to the exact specifications than it is to find a new enzyme in Nature.

Results

In the figure 1 below, I show six examples of products of ketone reduction with mutants of one and the same ketoreductase enzyme (KRED). This particular KRED has been studied for years now and all the relevant issues have been dealt with: manufacturability at scale, inherent enzyme stability, screening data is available, etc. In other words “we have a pretty good bicycle”. To achieve the results that are summarized in the table, protein engineering was applied (‘directed evolution’) to modify the active site for substrate-fit and selectivity, much like in changing out the gears on a bicycle.

Figure 1: One KRED catalyst adapted for six processes.[1-5]

Table 1: Summary process descriptions. The alcohols had an e.e.>99% (or d.e. for penem’s) in all cases and the yields were >95% of theoretical in less than 24 hrs.

Discussion

I hope you will agree that it is pretty impressive that this enzyme’s active site is so flexible that mutagenesis allows it to accept a large molecule like the Montelukast intermediate but also a tiny, almost symmetrical molecule like the Sulopenem intermediate.

Some more conclusions:

• Substrate loading is always higher than 100 g/L (10%), even up to 36% without loss of productivity.
• Temperature is relatively mild (but can be high if required), as is the pH.
• Use of co-solvents is fine, even up to 90% 2-propanol (iPA).

It is results like the above that, already in 2007 led Merck researchers to write “Isolated enzymes have clearly supplanted whole cell bioreductions and in most instances chemo-catalytic ketones reductions at Merck.”[6]

Conclusion

Now, the big question that remains is “is this one catalyst, or are these 6 different ones?”. Of course, every enzyme variant is a specific unique chemical entity and as such it is a different catalyst. However, keep in mind that these catalysts are more than 90% identical, so humor me, and agree that you only need one catalyst :-). Let me know what you think!

References

[1] From my webinar hosted together with process chemists from Merck, Oct 2010. 21st Century Biocatalysis: an easy-to-use precision tool for every Process Chemists’ toolbox. http://www.icis.com/Webcast.htm

[2] For a nice general paper on KREDs: Gjalt W Huisman, Jack Liang, and Anke Krebber

Practical chiral alcohol manufacture using ketoreductases
Current Opinion in Chemical Biology, Volume 14, Issue 2, April 2010, Pages 122-129

[3] For Sulopenem: Jack Liang , Emily Mundorff, Rama Voladri, Stephan Jenne, Lynne Gilson, Aaron Conway , Anke Krebber , John Wong, Gjalt Huisman, Susan Truesdell, and James Lalonde

Highly Enantioselective Reduction of a Small Heterocyclic Ketone: Biocatalytic Reduction of Tetrahydrothiophene-3-one to the Corresponding (R)-Alcohol

Org. Process Res. Dev., 2010, 14 (1), pp 188–192

[4] For Montelukast: Jack Liang, James Lalonde, Birthe Borup, Vesna Mitchell, Emily Mundorff, Na Trinh, D. A. Kochrekar, Ramachandran Nair Cherat, and G. Ganesh Pai

Development of a Biocatalytic Process as an Alternative to the (−)-DIP-Cl-Mediated Asymmetric Reduction of a Key Intermediate of Montelukast

Org. Process Res. Dev., 2010, 14 (1), pp 193–198

[5] Patent numbers are shown in the figure.

[6] Jeffrey C. Moore, David J. Pollard, Birgit Kosjek, and Paul N. Devine

Advances in the Enzymatic Reduction of Ketones

Acc. Chem. Res. 2007, 40, 1412–1419

This is from their conclusions:” The substrate range and enantioselectivity for ketone reductions are excellent, providing high ee of either alcohol on the majority of ketone substrates. The enzymes also demonstrate valuable chemoselectivity and diastereoselectivity as described on the para-diketone 10 and several other substrates. The enzymes can be screened and scaled-up as rapidly as their chemocatalytic counterparts, and their cost to use at large scale and the environmental impact of their use is less. They have been used at Merck to economically deliver kilogram quantities of chiral intermediates with excellent yields and ee values. As a result, ketoreductases are the preferred catalyst for ketone reductions at Merck.”.

Purpose of this blog

Dear Reader,

Welcome to the first post of my blog on Industrial Biocatalysis.
In the next few weeks, I plan to regularly post articles on some of the lesser-known facts and aspects of Biocatalysis. The main focus will be on real applications of enzymes.

I tentatively plan to write a series, so come back every now and then!
Topics:

1.      Enzyme flexibility – How many catalysts do you need?
2.      CYP inhibition in metabolism studies – How fundamentally wrong this is!
3.      Lead diversification – How this oversight can cost you money!
4.      Substrate or product instability as limit to enzyme catalysis – How biocatalysis has changed.
5.      Cascade reactions as mimic of whole cell conversions – How to overcome some limitations
6.      Dynamic kinetic resolution of epoxides – How can you have ~30 reactions all lead to ultimately one product?

      Some of the topics can get pretty technical and in those cases, I will provide links to background information. Please provide me with feedback on my posts, but also feel free to suggest topics you want me to write about.

      Cheers,

    

      Erik

     devries.erikjan@gmail.com

      Biocatalytic Asymmetric Synthesis of Chiral Amines from Ketones Applied to Sitagliptin Manufacture. Christopher K. Savile, Jacob M. Janey, Emily C. Mundorff, Jeffrey C. Moore, Sarena Tam, William R. Jarvis, Jeffrey C. Colbeck, Anke Krebber, Fred J. Fleitz, Jos Brands, Paul N. Devine, Gjalt W. Huisman, Gregory J. Hughes. Published in Science 16 July 2010: 305-309.Published online 17 June 2010 [DOI:10.1126/science.1188934]. Free download after free registration.