Small but mighty: the impact of tertiary alcohols in drug design

Small but mighty: the impact of tertiary alcohols in drug design

By Julia Shanu-Wilson 

Creating a drug with perfect drug-like properties is a tough challenge. An often-overlooked group that can significantly influence drug-like properties is a tertiary alcohol (3^o ROH). A review paper just out digs into the literature, analysing MMPs to assess the effect of incorporating a 3^o ROH motif on potency and ADMET properties of drug compounds [1].

Installation of a hydroxy group can decrease lipophilicity, increase solubility and decrease hERG and CYP inhibition, although some primary and secondary alcohols may be more susceptible to further oxidation and glucuronidation. However, by introducing a hydroxyl group as a tertiary alcohol, the possibility of creating a metabolic soft spot for further oxidation or glucuronidation is significantly reduced through “shielding” by the geminal alkyl groups, which provide steric hindrance.

In their analysis the authors examine several properties, illustrated with numerous interesting examples:

• On-target potency
• Permeability and efflux
• Metabolism
• Bioavailability and exposure
• Off-target activity, particularly hERG and CYP inhibition, and PXR activation. However, it is noted that 1^o ROHs are usually better than 3^o ROH at preventing hERG inhibition due to the more exposed OH group preventing binding in the hERG pocket. 

Improvement in metabolic stability

Metabolism of some primary and secondary alcohols may result in further metabolism; to a carboxylic acid for 1^o ROHs via alcohol dehydrogenase and aldehyde dehydrogenase/aldehyde oxidase; and to a ketone for 2^o ROHs. For 2^o ROHs this can be further complicated due to reversibility of the reaction and potential racemisation of the stereocentre. However, these reactions should be viewed as possibilities, not certainties.

Figure 1: Peficitinib

In contrast, drugs containing 3^o ROHs are generally excreted unchanged at this moiety, potentially with metabolism directed elsewhere on the molecule. An interesting example looks at the JAK inhibitor peficitinib which contains an 3^o ROH in the adamantane moiety, the hydroxyl being responsible for a dramatic improvement in metabolic stability (Figure 1). Whilst the hydroxyl group was initially thought to be blocking metabolism, investigations revealed that it was the low cLogP achieved that was important in maintaining metabolic stability.

Inspiration from MetID studies

The paper also highlights that MetID studies can serendipitously lead to improvements in drug design.

Figure 2: Linsitinib (OSI-906), currently being developed by Sling Therapeutics as a treatment for thyroid eye disease through IGF-1R inhibition

Linsitinib is an oral small molecule currently in a Phase 2b trial for thyroid eye disease. Its predecessor suffered from high metabolic stability at the cyclobutyl group. During MetID studies, a mixture of cyclobutanols were isolated, as well as a more minor and transient onwards oxidised cyclobutanone. The moiety was subsequently modified to incorporate both the hydroxyl and a methyl group to a create a 3^o ROH motif at this metabolically vulnerable position (Figure 2). This resulted in an ~8-fold lower clearance, and better potency.

Biocatalysis as a tool to introduce tertiary hydroxyl groups

Biocatalysis is one of the approaches that has been explored to oxidise unactivated C-H bonds to create tertiary alcohols in drug compounds. At Hypha we have observed PolyCYPs enzymes [2] that can target these oxidations (Figure 3), including an active metabolite that was one of three subsequently patented as potentially useful treatments for skin conditions such as acne, atopic dermatitis, actinic keratosis and psoriasis [3].

Figure 3: Examples of (published) tertiary alcohols of drug compounds made by PolyCYPs enzymes

 Introduction of two tertiary hydroxyls: a double whammy

Last year we highlighted a very interesting piece of work by scientists at Bayer, involving incorporation of hydroxy groups at two different tert-butyl groups in the same molecule [5]. Dual hydroxylation was found to improve both selectivity and the PK profile of these IRAK4 inhibitors (Figure 5).

Figure 5: Incorporation of two hydroxyl substituents

Introduction of a hydroxyl group to compound 30 resulted in an equally potent compound 31 but with “protection” from oxidative metabolism resulting in a more promising PK profile. In parallel, replacement of the methoxy group at the 6-position of the indazole moiety with a 2-hydroxy-propan-2-yl to form compound 33 also afforded some protection from oxidative metabolism. Although compound 33 was slightly less potent, it was found to have better selectivity, however still suffered from relatively high clearance.

Combining both hydroxyl substitutions in BAY1830839 (IRAK4= 3 nM, FLT3= 167 nM) resulted in a metabolically stable clinical candidate with high oral bioavailability, retaining high potency with high selectivity.  A further clinical candidate was created by derivatising the 2-position through introduction of a sulfone moiety to make BAY1834845 (IRAK4= 8 nM, FLT3= 243 nM). Both compounds were reported to have advanced to clinical trials.

Lasting lessons from terfenadine: hERG inhibition

Not all primary oxidations should be viewed negatively. Let’s not forget the absolute classic example of the active metabolite of terfenadine leading to a better drug. It was a result of Raymond Woosley’s research on metabolism of terfenadine [4] that resulted in the blockbuster antihistamine drug fexofenadine (Allegra, Allevia). Terfenadine was withdrawn from the market after causing cardiac arrhythmias, later shown to result from potent hERG inhibition. Terfenadine is oxidised by CYP3A4 creating a primary alcohol which is further metabolised to the active carboxylic acid metabolite. It is the carboxylic acid metabolite of terfenadine, fexofenadine, that gave us a drug with no cardiotoxicity, or drowsiness due to reduced CNS exposure, and minimal metabolism (Figure 4). Based on Woosley’s studies, the FDA and other regulatory agencies then published guidelines requiring testing of new drugs for their potential to cause heart arrhythmias.

It is also interesting to note that besides a carboxylic acid moiety, fexofenadine also contains both a 2^o and a 3^o hydroxyl group and is metabolically very stable.

Figure 4: Terfenadine and its major active metabolite fexofenadine

To summarise

The examples shared nicely illustrate the value of inclusion of a tertiary hydroxyl in drug design, offering the possibility of greater permeability, lower efflux, reduced chance of phase II conjugation and a reduction in off-target inhibition. Awareness of the potential for improved properties could also stimulate pursuit of greater learnings from MetID studies where oxidation at specific positions may provide unforeseen benefits!

Glossary of terms used

cLogP: Calculated LogP, a method used to estimate the octanol-water partition coefficient (logP) of a chemical compound and a measure of the lipophilicity of a chemical compound

CYPs: Cytochrome P450 enzymes

hERG: Human ether-à-go-go related gene. Encodes Kv11.1 protein, a potassium ion channel crucial for cardiac repolarisation.

IGF-1R: Insulin-like growth factor 1 receptor

IRAK4: interleukin-1 receptor-associated kinase 4

FLT3: FMS-like tyrosine kinase 3

MetID: Metabolite identification

MMP: Matched molecule pairs

PXR: Pregnane X receptor (involved in CYP induction)

References