Download PDF
Synlett 2024; 35(19): 2201-2206
DOI: 10.1055/a-2239-6965
letter
Isotopic Labeling

Directing Hydrogen Isotope Exchange with Aryl Carboxylic Acids

Richard J. Mudd 
a   Department of Pure and Applied Chemistry, University of Strathclyde, 295 Cathedral Street, Glasgow, Scotland, UK, G1 1XL
,
Marc Reid 
a   Department of Pure and Applied Chemistry, University of Strathclyde, 295 Cathedral Street, Glasgow, Scotland, UK, G1 1XL
,
Laura C. Paterson 
a   Department of Pure and Applied Chemistry, University of Strathclyde, 295 Cathedral Street, Glasgow, Scotland, UK, G1 1XL
,
Jens Atzrodt
b   Sanofi Germany, R&D Operations, Industriepark Höchst, 65926 Frankfurt am Main, Germany
,
Volker Derdau 
c   Sanofi Germany, R&D, Integrated Drug Discovery, Isotope Chemistry, Industriepark Höchst, 65926 Frankfurt am Main, Germany
,
William J. Kerr  ∗
a   Department of Pure and Applied Chemistry, University of Strathclyde, 295 Cathedral Street, Glasgow, Scotland, UK, G1 1XL
› Author Affiliations

The work was funded by postgraduate studentship provision from the University of Strathclyde (R.J.M.) and the Carnegie Trust for the Universities of Scotland (M.R.).
› Further Information
 
 PDF Download Permissions and Reprints All articles of this category


Abstract

A highly effective and selective ortho-directed hydrogen isotope exchange process for aryl carboxylic acids has been achieved by using an iridium(I) N-heterocyclic carbene/phosphine complex under mild and neutral conditions. Good levels of deuterium incorporation have been delivered across a wide array of examples, including a number of biologically active drug compounds.


#

Key words

iridium catalysis - hydrogen isotope exchange - deuteration - carboxylic acids - benzoic acids

Transition-metal-catalysed methods for direct and selective hydrogen isotope exchange (HIE) through C–H activation is a key technology that has been studied over several years and is increasingly adopted within the chemistry community.[1] As part of drug-discovery programmes in particular, medicinal chemists rely heavily on such preparative methods to ensure fast and efficient incorporation of a radiolabel tracer into drug candidate molecules, which then enables various metabolic, stability, and toxicity studies to be performed.[2] Stable isotopic labels, such as 2H (D) and 13C, are also routinely used as internal standards for quantitative bioanalysis and mechanistic studies.[3] In addition to these uses, more recently, there has been an increased interest within the pharma domain in the development of deuterated versions of existing therapeutic molecules.[4] Such analogues can often offer higher systemic exposure and reduced clearance (permitting lower dosage), alongside an improved pharmacokinetic and safety profile, all while maintaining potency. Based on all of this, practically accessible and widely applicable HIE methodologies are continually sought.

As part of our own research endeavours, we have developed a series of highly active iridium-based catalysts capable of delivering heavy isotopes of hydrogen (deuterium or tritium) in a single step to a wide range of organic structures, including many pharmaceutically active drug-type motifs (Scheme [1]).[5]

Zoom Image
Scheme 1 Ir-catalysed directed HIE and proposed application to aryl carboxylic acids.

In terms of their mode of action, we have reported that our iridium catalyst binds to Lewis basic atoms within a directing group, activating a proximal C–H bond and generating a metallocycle (most commonly a five-membered metallocyclic intermediate) that can then undergo C–D or C–T bond formation.[6] To date, we have shown extensive applicability of this Ir-catalysed HIE technology within aromatic compounds, whereby ortho-directed HIE is achieved by using a structurally diverse range of directing groups, including ketones, amides, esters, the nitro unit, sulfones, sulfonamides, and an array of N-heterocycles, all with high levels of isotope incorporation and under mild reaction conditions.[7] Our work has also been extended to include the β-labelling of nonaromatic unsaturated systems,[8] as well as at more-challenging C(sp3) centres.[9]

As related to our work with aromatic molecules [i.e., directed isotope incorporation at an aromatic C(sp2) centre], we have also disclosed the first case of an ortho-H/D and -H/T exchange of pharmaceutically relevant aryl tetrazole species.[10] Whilst our methodology until this stage had always employed neutral conditions, in this latter investigation it became apparent that the use of a base was necessary to deliver high levels of incorporation adjacent to the (acidic) tetrazole directing group. These studies represented the first case of selective C–H functionalisation of unprotected tetrazole species, while also revealing evidence to support a base-assisted concerted metalation-deprotonation (CMD)-type catalysis process. Indeed, the proposed CMD mechanism that is active in the case of tetrazole substrates was distinct from the σ-complex-assisted metathesis-type mechanism previously evidenced for our H/D exchange methods with nonacidic directing groups.[6] [7f]

The research described herein follows on from our findings with tetrazoles, as we have endeavoured to develop an effective labelling method for aryl carboxylic acids. Despite the abundance of such functionality within bioactive compounds, a generally applicable and highly efficient iridium-based method for ortho-directed HIE with aromatic carboxylic acid substrates has not yet been established. While individual examples using a variety of metals can be found in the literature,[11] varying success is reported where, in many cases, suboptimal isotope incorporation levels are delivered and/or typically forcing reaction conditions are required. It is presumed that this is due to the relatively weak coordinating ability of the carboxylic acid moiety. Given the chemical similarities between tetrazoles and carboxylic acids,[12] we anticipated that our previously developed base-assisted process might also provide the foundations of a method to facilitate efficient labelling of benzoic acid substrates.

To initiate this study, carboxylic acid 1 was reacted with 5 mol% of iridium(I) N-heterocyclic carbene/phosphine complex 2 at 50 ℃ in MeOH under an atmosphere of deuterium gas and in the presence of caesium carbonate as base (Scheme [2]). It was pleasing to find that these conditions, as applied previously to tetrazole substrates, were indeed applicable in the labelling of aryl carboxylic acids, albeit with only modest levels of exchange. Specifically, 22% deuterium incorporation was observed for substrate 1 after six hours. The first consideration towards improving the reactivity of our model system was to assess the base. Indeed, it was considered that the use of a carbonate as base could well be hindering isotope exchange through competitive binding to the iridium centre relative to the carboxylate functionality of the substrate. For this reason, the noncoordinating organic base diisopropyl(ethyl)amine (DIPEA) was also considered. At the same time, and based on our experience that alcoholic solvents are not the most generally applicable media in related processes,[13] a range of solvents were tested in conjunction with both bases. As shown in Scheme [3], a variety of solvents were applicable in this process, albeit with generally modest incorporations being observed. Except for the experiments carried out in methanol, switching the base from caesium carbonate to DIPEA improved the incorporation in every solvent, including methyl tert-butyl ether (MTBE), which delivered a promising 59% deuterium incorporation level for carboxylic acid substrate 1.

Zoom Image
Scheme 2 Initial labelling of carboxylic acid 1 under unoptimised conditions. BArF = tetrakis[3,5-bis(trifluoromethyl)phenyl]borate.
Zoom Image
Scheme 3 Solvent and base screen. CPME = cyclopentyl methyl ether.

With MTBE and DIPEA providing improved effectiveness as the solvent and base, respectively, the catalyst structure itself was then examined in an attempt to enhance the overall efficiency of the exchange process (Scheme [4]). The range of iridium-based complexes chosen for examination reflected previously active structures within HIE[1] [5] [6] [7] [8] [9] [10] or hydrogenation processes.[14] The neutral chlorocarbene complex 3, an excellent choice of catalyst type for the labelling of primary sulfonamides,[7f] was, unfortunately, completely inactive towards substrate 1 under the developed conditions. Furthermore, the cationic PF6 species 4 and 5 displayed minimal activity. Pleasingly, moving to complexes with a combination of the noncoordinating tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (BArF) counterion and a phosphine/NHC ligand sphere (complexes 2 and 6–8) led to dramatic increases in the levels of deuterium incorporation. In particular, complex 6 delivered an excellent 89% incorporation over the indicated positions, which we believe is attributable to the flexible but bulky tribenzylphosphine ligand conferring an optimal balance of thermal stability and reactivity.

Zoom Image
Scheme 4 Catalyst screen.

Following this initial optimisation, a number of substrates were selected for study, alongside an investigation of whether a base was actually necessary in our developing system. Indeed, given the presence of a basic site within its own structure, we felt the need to go beyond model substrate 1 to probe the role of the base. Furthermore, the change in pK a across substituted benzoic acids is well recognised in the literature,[15] and any changes in reactivity correlating with this could offer insights into factors influencing the reaction pathway. To this end, six p-substituted carboxylic acids (including model substrate 1), along with unsubstituted benzoic acid (9f), were treated under the HIE conditions using catalyst 6 in the presence or absence of added base (Scheme [5]). Note that a lower 2.5 mol% catalyst loading was also employed in these experiments to permit a more obvious assessment of reactivity differences between substrates.

As noted in Scheme [5], the initial model substrate 1 delivered an impressive deuterium incorporation of 91% in the absence of base, compared to 74% labelling when one equivalent of DIPEA was employed. This trend continued across each substrate whether it was electron-rich, as in 9a and 9b, electron-deficient (9c–e), or neutral (9f), giving a clear indication that the use of base is not required in this HIE process, and is, indeed, detrimental to achieving excellent levels of D incorporation. Furthermore, and as was most apparent under base-free conditions, the varying pK a values of the carboxylic acids has little effect on reaction efficiency.

Zoom Image
Scheme 5 Investigations of the use of a base across substrates.

Following the investigations relating to both the optimal catalyst system and base usage, we next examined the continuous variables of catalyst loading, reaction time, and reaction temperature. To this end, a three-factor, two-level design-of-experiments (DoE) approach was employed in an effort to minimise the number of experiments while maximising the chemical space incorporated into the optimisation (see the Supporting Information for full details). The outputs from this study indicated that the most significant factor in our process is the necessity for a mildly elevated temperature (50 ℃); additionally, a 5 mol% catalyst loading and a reaction time of just two hours were indicated as optimal for efficient labelling.

With this improved understanding of the parameters of our developing HIE system, we then investigated a range of aryl carboxylic acid substrates to fully explore the reaction scope (Scheme [6]).[16] Compounds 1 and 9a–f were reassessed using our newly optimised procedure, with excellent levels of deuterium incorporation achieved in just two hours. To investigate the impact of steric congestion around the labelling site and the directing group, a further range of substituted acids was also examined. Unfortunately, 3-(dimethylamino)benzoic acid (9g) performed poorly, with just 30% D incorporation at the 2-position only. We rationalised this lowered labelling outcome by considering the formation of a less favourable pincer-type complex intermediate between the iridium centre, the amino moiety, and the carboxylic acid, coupled with the strongly electron-donating NMe2 substituent potentially increasing the strength of the aryl C–H bonds at positions ortho and para to this unit, i.e. at the positions where HIE would be expected to occur. The additional amino ligation would also provide a rationale for the labelling regioselection observed with this substrate. A somewhat similar labelling level variance was observed between the expected sites within 3-methoxybenzoic acid (9h); however, this was to a much lesser extent, and high incorporation levels were restored in this case. Similarly, meta-substituted substrates 9i and 9j performed exceedingly well under our protocol with overall incorporation levels of 94 and 90%, respectively. With 9i and 9j, resolution of the two individual ortho-positions was, unfortunately, not possible. Greater degrees of substitution on the ring, as in 9k, were tolerated well, with 94% D incorporation achieved with this substrate. Furthermore, an assessment of ortho-substitution by using species 9l–p exemplified the robustness of our system in the presence of proximal steric encumbrance; with the exception of salicylic acid (9p), all such ortho-substituted aryl carboxylic acids delivered deuterated products with 90% D or above. With regards 9p, one can envisage the formation of a stable chelate between the phenol and acid moieties; nonetheless, only a moderate drop in the incorporation level to 81% was observed. Overall, with the exception 3-(dimethylamino)benzoic acid (9g), the 16 tested benzoic acid derivatives delivered selective isotopic incorporations of 81–95% at the position ortho to the carboxylic acid, and in just two hours, with only a 5 mol% catalyst loading being required.

Zoom Image
Scheme 6 Substrate scope.

In an attempt to push the substrate scope further, we explored the labelling of 3-nicotinic acid (9q) and its N-oxide (9r). However, no deuterium incorporation was obtained with these substrates.[17] A small number of phenylacetic acid derivatives were also tested, given that this substrate class is found in many common active pharmaceutical ingredients. As seen with compound 10 (Scheme [6]), our system was not applicable to this set of substrates, likely due to the required formation of a more-challenging six-membered metallocyclic intermediate, cf. a five-membered metallocyclic intermediate with the benzoic acid compounds. Details of the full set of tested phenylacetic acid substrates can be found in the Supporting Information.

Finally, to further exemplify the utility of the devised HIE protocol, we examined a number of biologically active drug compounds (Scheme [7]). Pleasingly, probenecid (9s),[18] which is traditionally used in the treatment of gout, delivered high and selective incorporation at the site ortho to the carboxylic acid moiety, with negligible labelling adjacent to the tertiary sulfonamide. Furthermore, the common drug aspirin (9t)[19] delivered an excellent 97% D incorporation exclusively ortho to the acid functionality, and mefenamic acid (9u),[20] an anthranilic acid derivative used as an effective nonsteroidal antiinflammatory drug, was labelled to a notable 85% D, despite the potential for the secondary amine motif to hinder catalysis.

Zoom Image
Scheme 7 HIE on active pharmaceuticals.

In conclusion, we have developed a highly effective hydrogen isotope exchange process using the ubiquitous carboxylic acid functionality to direct ortho-labelling. Furthermore, we have illustrated the utility of this procedure across a wide array of benzoic acid substrates, including multifunctional drugs. A combination of conventional reaction optimisation and a design-of-experiments approach was employed to establish the developed protocol and, ultimately, the highly active iridium(I) NHC/phosphine complex 6 was shown to facilitate the process under mild and accessible conditions without the need for a base. Excellent incorporations of up to 97% D over 20 substrates were achieved providing a general HIE method that has the potential to find appreciable levels of utility with pharmaceutical stakeholders, and across the wider preparative chemistry community. Work within our laboratories is ongoing to further explore the capacity of emerging catalyst systems with extended substrates of potential therapeutic importance.


#

Conflict of Interest

J.A. and V.D. are employees of Sanofi and may hold shares or stock options in the company. All other authors declare no conflict of interest.

Supporting Information

    Supporting information for this article is available online at https://doi-org.accesdistant.sorbonne-universite.fr/10.1055/a-2239-6965.
  • Supporting Information

  • References and Notes

  • 5 For a recent review detailing contributions to the field of iridium-catalysed hydrogen isotope exchange, see: Kerr WJ, Knox GJ, Paterson LC. J. Labelled Compd. Radiopharm. 2020; 63: 281
    CrossrefPubMedSearch in Google Scholar
  • 6 Brown JA, Cochrane AR, Irvine S, Kerr WJ, Mondal B, Parkinson JA, Paterson LC, Reid M, Tuttle T, Andersson S, Nilsson GN. Adv. Synth. Catal. 2014; 356: 3551
    CrossrefSearch in Google Scholar
  • 8 Kerr WJ, Mudd RJ, Paterson LC, Brown JA. Chem. Eur. J. 2014; 20: 14604
    CrossrefPubMedSearch in Google Scholar
  • 9 Kerr WJ, Mudd RJ, Reid M, Atzrodt J, Derdau V. ACS Catal. 2018; 8: 10895
    CrossrefSearch in Google Scholar
  • 10 Kerr WJ, Lindsay DM, Reid M, Atzrodt J, Derdau V, Rojahn P, Weck R. Chem. Commun. 2016; 52: 6669
    CrossrefPubMedSearch in Google Scholar
  • 13 Kennedy AR, Kerr WJ, Moir R, Reid M. Org. Biomol. Chem. 2014; 12: 7927
    CrossrefPubMedSearch in Google Scholar
  • 15 Hansch C, Leo A, Taft RW. Chem. Rev. 1991; 91: 165
    CrossrefSearch in Google Scholar
  • 16 4-Methoxy(2,6-2H2)benzoic Acid (D9a); Typical Procedure Exchange reactions were carried out on a Heidolph Synthesis 1 Liquid 16 device. The device was evacuated and filled with argon, and the water condenser was turned on. A carousel tube was charged with substrate 9a (13.1 mg, 0.086 mmol) and iridium catalyst 6 (7.6 mg, 0.0043 mmol). MTBE (1 mL) was added, rinsing the inner walls of the tube. The tube was then sealed at the screw cap, with the gas inlet left open under argon. The flask was subjected to two cycles of evacuation and refilling of deuterium from a balloon. The gas inlet tube was then closed, creating a sealed atmosphere of deuterium. The carousel shaking motion was initiated (750 rpm) and the temperature was set to 50 ℃. After starting the shaking motion and temperature controller of the device, the timer was started, and a rapid red/orange to clear/yellow colour change was observed. The reaction mixture was stirred for 2 h, then excess deuterium was removed and replaced with air. The solution was then diluted with Et2O (2 mL), basified with 2 M aq NaOH (2 mL), and separated. The aqueous layer was washed with Et2O (2 × 2 mL), acidified to pH 1 with 2 M aq HCl (~3 mL), and extracted with CH2Cl2 (2 × 2 mL). The CH2Cl2 extracts were dried (Na2SO4), filtered, and concentrated in vacuo. The level of incorporation was determined by 1H NMR spectroscopic analysis, with the integrals of the anticipated labelling positions measured against a peak corresponding to a position where labelling was not expected. The percentage deuteration was calculated by using the following equation: %Deuteration = 100 – [(residual integral/no. of labelling sites) × 100]. D incorporation; Run 1: 89%; Run 2: 90%; Average: 90%. 1H NMR (300 MHz, DMSO): δ = 12.68 (br s, 1 H, O–H), 7.93–7.84 (m, 2 H, Ar-H), 7.05-6.97 (m, 2 H, Ar-H), 3.81 (s, 3 H, O-CH3 ). Incorporation expected at δ = 7.93–7.84. Determined against integral at δ = 3.81.
  • 17 It is predicted that the pyridine and pyridine N- oxide functionalities would outcompete carboxylic acid binding; see: Timofeeva DS, Lindsay DM, Kerr WJ, Nelson DJ. Catal. Sci. Technol. 2021; 11: 5498
    CrossrefSearch in Google Scholar
  • 18 García-Rodríguez C, Mujica P, Illanes-González J, López A, Vargas C, Sáez JC, González-Jamett A, Ardiles ÁO. Biomedicines 2023; 11: 1516
    CrossrefPubMedSearch in Google Scholar
  • 19 Vane JR, Botting RM. Thromb. Res. 2003; 110: 255
    CrossrefPubMedSearch in Google Scholar
  • 20 Srivastava R, Mishra MK, Patel AK, Singh A, Kushwaha K. GSC Biol. Pharm. Sci. 2019; 7: 52
    CrossrefSearch in Google Scholar

Corresponding Author

William J. Kerr
Department of Pure and Applied Chemistry
University of Strathclyde, 295 Cathedral Street, Glasgow
Scotland, UK, G1 1XL   

Publication History

Received: 24 November 2023

Accepted after revision: 05 January 2024

Accepted Manuscript online:
05 January 2024

Article published online:
05 February 2024

© 2024. Thieme. All rights reserved

Georg Thieme Verlag KG
Rüdigerstraße 14, 70469 Stuttgart, Germany

  • References and Notes

  • 5 For a recent review detailing contributions to the field of iridium-catalysed hydrogen isotope exchange, see: Kerr WJ, Knox GJ, Paterson LC. J. Labelled Compd. Radiopharm. 2020; 63: 281
    CrossrefPubMedSearch in Google Scholar
  • 6 Brown JA, Cochrane AR, Irvine S, Kerr WJ, Mondal B, Parkinson JA, Paterson LC, Reid M, Tuttle T, Andersson S, Nilsson GN. Adv. Synth. Catal. 2014; 356: 3551
    CrossrefSearch in Google Scholar
  • 8 Kerr WJ, Mudd RJ, Paterson LC, Brown JA. Chem. Eur. J. 2014; 20: 14604
    CrossrefPubMedSearch in Google Scholar
  • 9 Kerr WJ, Mudd RJ, Reid M, Atzrodt J, Derdau V. ACS Catal. 2018; 8: 10895
    CrossrefSearch in Google Scholar
  • 10 Kerr WJ, Lindsay DM, Reid M, Atzrodt J, Derdau V, Rojahn P, Weck R. Chem. Commun. 2016; 52: 6669
    CrossrefPubMedSearch in Google Scholar
  • 13 Kennedy AR, Kerr WJ, Moir R, Reid M. Org. Biomol. Chem. 2014; 12: 7927
    CrossrefPubMedSearch in Google Scholar
  • 15 Hansch C, Leo A, Taft RW. Chem. Rev. 1991; 91: 165
    CrossrefSearch in Google Scholar
  • 16 4-Methoxy(2,6-2H2)benzoic Acid (D9a); Typical Procedure Exchange reactions were carried out on a Heidolph Synthesis 1 Liquid 16 device. The device was evacuated and filled with argon, and the water condenser was turned on. A carousel tube was charged with substrate 9a (13.1 mg, 0.086 mmol) and iridium catalyst 6 (7.6 mg, 0.0043 mmol). MTBE (1 mL) was added, rinsing the inner walls of the tube. The tube was then sealed at the screw cap, with the gas inlet left open under argon. The flask was subjected to two cycles of evacuation and refilling of deuterium from a balloon. The gas inlet tube was then closed, creating a sealed atmosphere of deuterium. The carousel shaking motion was initiated (750 rpm) and the temperature was set to 50 ℃. After starting the shaking motion and temperature controller of the device, the timer was started, and a rapid red/orange to clear/yellow colour change was observed. The reaction mixture was stirred for 2 h, then excess deuterium was removed and replaced with air. The solution was then diluted with Et2O (2 mL), basified with 2 M aq NaOH (2 mL), and separated. The aqueous layer was washed with Et2O (2 × 2 mL), acidified to pH 1 with 2 M aq HCl (~3 mL), and extracted with CH2Cl2 (2 × 2 mL). The CH2Cl2 extracts were dried (Na2SO4), filtered, and concentrated in vacuo. The level of incorporation was determined by 1H NMR spectroscopic analysis, with the integrals of the anticipated labelling positions measured against a peak corresponding to a position where labelling was not expected. The percentage deuteration was calculated by using the following equation: %Deuteration = 100 – [(residual integral/no. of labelling sites) × 100]. D incorporation; Run 1: 89%; Run 2: 90%; Average: 90%. 1H NMR (300 MHz, DMSO): δ = 12.68 (br s, 1 H, O–H), 7.93–7.84 (m, 2 H, Ar-H), 7.05-6.97 (m, 2 H, Ar-H), 3.81 (s, 3 H, O-CH3 ). Incorporation expected at δ = 7.93–7.84. Determined against integral at δ = 3.81.
  • 17 It is predicted that the pyridine and pyridine N- oxide functionalities would outcompete carboxylic acid binding; see: Timofeeva DS, Lindsay DM, Kerr WJ, Nelson DJ. Catal. Sci. Technol. 2021; 11: 5498
    CrossrefSearch in Google Scholar
  • 18 García-Rodríguez C, Mujica P, Illanes-González J, López A, Vargas C, Sáez JC, González-Jamett A, Ardiles ÁO. Biomedicines 2023; 11: 1516
    CrossrefPubMedSearch in Google Scholar
  • 19 Vane JR, Botting RM. Thromb. Res. 2003; 110: 255
    CrossrefPubMedSearch in Google Scholar
  • 20 Srivastava R, Mishra MK, Patel AK, Singh A, Kushwaha K. GSC Biol. Pharm. Sci. 2019; 7: 52
    CrossrefSearch in Google Scholar

   Permissions and Reprints All articles of this category
Zoom Image
Scheme 1 Ir-catalysed directed HIE and proposed application to aryl carboxylic acids.
Zoom Image
Scheme 2 Initial labelling of carboxylic acid 1 under unoptimised conditions. BArF = tetrakis[3,5-bis(trifluoromethyl)phenyl]borate.
Zoom Image
Scheme 3 Solvent and base screen. CPME = cyclopentyl methyl ether.
Zoom Image
Scheme 4 Catalyst screen.
Zoom Image
Scheme 5 Investigations of the use of a base across substrates.
Zoom Image
Scheme 6 Substrate scope.
Zoom Image
Scheme 7 HIE on active pharmaceuticals.