One-Pot Bioisosteric Replacement of Alkyl Carboxylic Acids via Organic Photoredox Catalysis

Abstract

Drug development is often hindered by the pharmacokinetic and physicochemical limitations of lead compounds. Bioisosteric replacement is commonly employed to bolster drug viability; however, methods for accessing drug analogs are often synthetically inefficient. To address this challenge, we disclose strategies to access carboxylic acid bioisosteres via organic photoredox catalysis. A one-pot method enables the direct conversion of carboxylic acids to their most common bioisostere, tetrazoles. Additional functionalization to an amidoxime intermediate affords access to three other bioisosteres: oxathiadiazolones, oxadiazolones, and oxadiazole thiones. HPLC lipophilicity measurements of carboxylic acids and bioisostere derivatives illustrate the synthetic value of these methods for improving lead compound viability.

Graphical Abstract

Hit-to-lead optimization is a crucial component of drug discovery in which identified drug candidates are iteratively refined to lead compounds with high potency and biological viability. These drug development campaigns cost billions, span decades, and often fail to reach patients. Even for those with viable synthetic routes, only 10–20% of clinical trial drugs will ever reach the market.¹ Many potential drugs fall short of success at clinical trial stages, not due to a lack of potency, but to shortcomings in pharmacokinetic (PK) and physicochemical properties such as drug absorption, distribution, metabolism, and excretion.² For these reasons, there is an ongoing demand for synthetic methodologies that can improve the limiting properties of candidate drugs and advance them along the development pipeline in more expeditious timeframes.

While optimizing a drug’s pharmacokinetic properties, bioisosteric replacements are employed, often to great positive effect. Stemming from the concept of isosterism developed in the early 20th century, bioisosterism involves exchanging a functional group within a molecule with an alternative but similar group to enhance PK and physicochemical properties.³ Bioisosteres have been developed for various functionalities prevalent in drug molecules, including phenyl rings, amides, and carboxylic acids.^4–6 Carboxylic acids are of particular interest due to their involvement in hydrogen bonding to facilitate drug–protein interactions and are present in over 450 drugs marketed worldwide, including nonsteroidal anti-inflammatory drugs (NSAIDs), antihypertensives, antibiotics, and statins. Despite facilitating protein binding, carboxylic acids can cause unfavorable solubility, membrane permeability, and off-target reactivity, which can limit their viability in clinical settings.^3,7

Viable alternatives for carboxylic acids include acylsulfonamides, phosphoric acids, squaramides, and tetrazoles. In marketed drugs, the most prominent bioisosteric replacements for carboxylic acids are tetrazoles, which have shown promise in mitigating these issues (Figure 1A).⁸ These carboxylic acid replacements have enabled the discovery of improved drug candidates for hypertension, hepatitis C, and B-cell lymphoma.^9,10 Tetrazoles mimic the two-point hydrogen bonding and acidity of carboxylic acids that facilitate key drug–protein interactions, making them comparable alternatives. Substitution with this moiety has also contributed to improved drug efficacy via charge delocalization and extension of acidic protons further from the molecule’s core. This property assists binding to biological targets and enhances metabolic stability.¹¹ It has also been suggested that the delocalized charge distribution of tetrazoles leads to greater drug lipophilicity in comparison to carboxylic acids, however, this is still an open debate.^9,12

Figure 1.

(A) Tetrazoles as bioisosteres of carboxylic acids. (B) Precedented strategies for acid to subsequent bioisosteres. (C) Decarboxylative cyanation of aliphatic carboxylic acids to radiolabeled nitriles. (D) One-pot synthesis of tetraoles via a decarboxylative cyanation.

Given the prevalence of carboxylic acids in biological contexts, we became interested in a methodology for their direct conversion to tetrazoles. Current methods for the conversion of carboxylic acids to their corresponding tetrazoles typically involve three or more synthetic steps and use highly toxic reagents. In an Alzheimer’s disease treatment campaign to selectively inhibit β-amyloid_1–42 secretion, a variety of bioisosteres of the carboxylic acid NSAID Flurbiprofen were explored to reduce off-target reactivity. The tetrazole-containing analogue required a formal four-step sequence from the carboxylic acid and featured a cycloaddition with an excess of hazardous tin azide. Further experiments afforded the oxadiazolone, another well-precedented carboxylic acid bioisostere, but necessitated five formal steps (Figure 1B).¹³ We envisioned a direct one-pot conversion of carboxylic acids to their tetrazole bioisosteres via organic photoredox catalysis. Generally, there are limited examples of direct late-stage bioisostere interconversion,^14,15 none of which accomplish the overall transformation in this work.

In 2015, our lab published a strategy for the hydrodecarboxylation of aliphatic carboxylic acids using an acridinium photocatalyst.¹⁶ In addition to capturing the resulting alkyl radical with a hydrogen atom, copper-catalyzed cyanation of the subsequent radical was also realized.¹⁷ In 2024, we leveraged organic photoredox and copper cocatalysis to accomplish the decarboxylative [¹¹C] cyanation of carboxylic acids to first generate the corresponding [¹¹C] alkylcyanide and following hydrolysis, the corresponding [¹¹C] carboxylic acids (Figure 1C).¹⁸ We envisioned that the nitriles generated via this strategy could serve as a handle for a thermal [3 + 2] cyclization with an azide source to access tetrazoles in a one-pot method. This approach would represent the first direct carboxylic acid to tetrazole interconversion and feature a linchpin intermediate that could be derivatized to access additional carboxylic acid bioisosteres. Herein, we present the reaction development and scope of a one-pot route to tetrazoles from alkyl carboxylic acids via decarboxylative cyanation and [3 + 2] cycloaddition depicted in Figure 1D. Efficient routes to additional carboxylic acid bioisosteres such as oxathiadiazolones, oxadiazolones, and oxadiazole thiones via an amidoxime intermediate are also disclosed. Finally, HPLC lipophilicity measurements provide evidence for the enhanced membrane permeability of bioisostere drug candidates, illustrating the value of this method for accelerating drug development campaigns.

Results and Discussion

Optimization for the direct conversion of carboxylic acids to tetrazoles was performed using commercially available 4-biphenyl acetic acid.^16,18,19 Our previously reported decarboxylative cyanation conditions were incompatible with this approach as cycloadditions of nitriles and azides require temperatures upward of 100 °C. Thus, a solvent evaluation was conducted (Table 1, entries 1–3). Initial attempts using aromatic solvents with high boiling points revealed that chlorobenzene improved cyanation yield when combined with 2,2,2-trifluoroethanol (TFE) as a cosolvent to ensure the solubility of the copper complex. Increasing the concentration to 0.15 M further improved the result to 91% yield (entry 4). Having identified optimal cyanation conditions, we next evaluated reaction variables for the [3 + 2] cycloaddition with sodium azide and triethylamine hydrochloride with the crude reaction mixture following irradiation.^19–21 Variation of reaction temperature and duration were investigated and revealed that 110 °C for 16 h was essential for reaction completion (entry 8).

Table 1.

Reaction Optimization^a

entry	solvent (M)	temp	time	nitrile yieldb	tetrazole yieldb
1	PhMe:TFE 10:1, (0.10 M)	35 °C	16 h	20%
2	PhCF3:EtOAc 10:1, (0.10 M)	35 °C	16 h	60%
3	PhCl:TFE 10:1, (0.10M)	35 °C	16 h	78%
4	PhCl:TFE 10:1, (0.15M)	35 °C	16 h	91%
5	PhCl:TFE 10:1, (0.30M)	35 °C	16 h	73%
6	PhCl:TFE 10:1, (0.15M)	80 °C	16 h		30%
7	PhCl:TFE 10:1, (0.15M)	100 °C	16 h		48%
8	PhCl:TFE 10:1, (0.15M)	110 °C	16 h		93%
9	PhCl:TFE 10:1, (0.15M)	110 °C	8 h		81%
10	PhCl:TFE 10:1, (0.15M)	110 °C	24 h		86%

^aReactions run on 0.3 mmol scale.

^bIsolated yields.

The identified conditions for tetrazole synthesis were subsequently evaluated in the context of biologically active carboxylic-acid-containing molecules and other small molecules (Figure 2). A variety of primary (1–6) and secondary-substituted (7–14) carboxylic acids were effectively converted to the tetrazole analogue in good to moderate yields. A broad range of functionalities were compatible with this chemistry, including halogens (9 and 10) as well as oxygen and sulfur-containing heterocycles (3 and 12). We were pleased to find that functional groups typically prone to oxidation, such as pyrrole 5 and amines 10 and 14 were tolerated in the oxidative catalytic cycle. Unactivated alkyl substrates, such as 2 and 14, proceeded in moderate yields, further demonstrating the versatility of the reaction conditions. Additionally, we found that tertiary acids 15 and 16 were tolerated, albeit in lower yields, likely due to less reactive tertiary radical intermediates. The sampling of drug molecules selected for this transformation includes an array of functional groups that were amenable to the reaction conditions, demonstrating the potential of this method for late-stage functionalization and rapid conversion to bioisostere-containing analogs.

Figure 2.

Carboxylic acid scope for a one-pot tetrazole synthesis. HPLC lipophilicity measurements are reported for carboxylic acid starting materials (Log P_A) and tetrazole-containing products (Log P_T) at pH 6. Difference between Log P_T and Log P_A are defined as Δ Log P.

With a library of synthesized tetrazole bioisosteres, we sought to determine the biological properties of this substitution as an alternative to carboxylic acid functionality. A common metric for evaluating the promise of a drug candidate in the discovery stages is lipophilicity. This is often quantified by the log of the octanol/water partition coefficient (Log P), which describes preference for lipid-rich or water-rich environments and influences the ability of a drug compound to cross cell membranes and engage in protein binding.^22,23 The traditional shake flask method may be used to determine Log P, but it is both time-consuming and material-intensive. Computational-derived Log P values (cLogP) also suffer from a high degree of inaccuracy.²⁴ A more straightforward method for Log P determination utilizes HPLC.²⁵ This robust method was employed to determine HPLC Log P values for both the starting carboxylic acid (Log P_A) and the corresponding tetrazoles (Log P_T) at pH 6 (Figure 2). Under these conditions, tetrazoles and other carboxylic acid bioisosteres are deprotonated as they would be at physiological pH, allowing this method to predict the behavior of these compounds in the body.¹¹ Of the substrates examined, 11 showed increased lipophilicity of the tetrazole versus the starting carboxylic acid, expressed as Δ Log P. While optimal lipophilicity is case-dependent, it is a known predictor of pharmaceutical success due to its impact on the successful transportation and reception of drug molecules in vivo. Thus, understanding and developing methods to modulate this property is essential for successful medicinal chemistry drug design.²³ To our knowledge, this is the largest collection compiled to date of HPLC-derived Log P data comparing carboxylic acids directly to their tetrazole analogs.

In addition to accessing tetrazoles, we envisioned expansion of our method to access additional carboxylic acid bioisosteres such as oxadiazolones, oxadiazole thiones, and oxathiadiazolones, all of which have precedence in drug development and are accessible via an amidoxime intermediate.^3,26,27 Thus, we sought to establish photoredox conditions that would permit various direct transformations to complex heterocycles. Initial investigations revealed that direct conversion to the amidoxime could be achieved in a one-pot fashion using methanol as the solvent in the decarboxylative cyanation and reducing the TMSCN to two equivalents. After irradiation, subjection of the crude reaction mixture to excess aqueous hydroxyl amine and heating to 65 °C afforded the amidoxime intermediate. At this stage, an aqueous workup was performed and conditions for conversion to alternative bioisosteres were established.

Evaluation of the reaction conditions revealed that treatment of the amidoxime reaction mixture with dimethyl carbonate and sodium hydroxide afforded efficient conversion to the oxadiazolone bioisostere. In an alternative reaction, thionyl chloride and pyridine aided conversion to the oxathiadiazolone. Due to the crude nature of the amidoxime, excess electrophile and excess base were necessary for both transformations. Commercial nonsteroidal anti-inflammatory drug (NSAID) substrates Flurbiprofen and Fenoprofen were subjected to these reaction conditions to demonstrate the utility of this direct method for diversifying potential drug candidates (Figure 3). Both heterocycles were accessed in modest yields. Additional cyclic bioisosteres such as oxadiazole thiones may be accessed via this method, albeit in reduced yields (see SI for further details). In addition to generating a library of carboxylic acid bioisostere derivatives, we applied our tetrazole HPLC methodology to evaluate the lipophilicity of these new compounds (Log P_o) and gain insight into the impact of each of these substitutions on biological activity. HPLC Log P determinations showed that the oxathiadiazolones had a lipophilicity comparable to that of the carboxylic acid, while the oxadiazolones have a higher lipophilicity compared to that of the carboxylic acid.

Figure 3.

Synthetic routes to alternative bioisostere analogs for NSAIDs Flurbiprofen and Fenoprofen.

Conclusion

We have demonstrated the use of organic photoredox catalysis for the direct conversion of carboxylic acids to the corresponding tetrazoles and other pertinent bioisosteres. The one-pot conversion of carboxylic acids to the corresponding tetrazoles was successfully applied to a variety of drug molecules possessing a range of functionality in good yields by using mild conditions. More efficient routes toward oxathiadiazolone and oxadiazolone bioisosteres were also disclosed. From one established hit, three analogs with distinct PK and physicochemical properties can be synthesized, ideally providing biomedical researchers with additional tools for drug candidate development. In addition to establishing a robust method and accessing new chemical space, Log P experiments reveal the unequivocally enhanced lipophilicity of tetrazoles and demonstrate the utility of these moieties as viable alternative functional groups to carboxylic acids in pharmaceuticals.

Supplementary Material

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.orglett.5c04049.

Experimental procedures and supporting ¹H and ¹³C NMR spectra (PDF)

Data Availability Statement

The data underlying this study are available in the published article and its Supporting Information.