Diversity Oriented Clicking for Modular Synthesis

Abstract

Diversity Oriented Clicking (DOC) fuses the reliability of classical click reactions with the versatility of Fluoride Exchange (FEx) chemistry, in which fluorine atoms bound to sulfur or phosphorus serve as highly selective reaction handles. This integrated platform forms covalent bonds rapidly and under mild, biocompatible conditions, broadening the synthetic toolkit for building molecular complexity. By employing DOC workflows, researchers can create structurally diverse compound libraries and explore previously inaccessible regions of chemical space. DOC holds particular promise in drug discovery where it accelerates the profiling of protein function in disease contexts and streamlines the design, synthesis, and high-throughput screening of novel therapeutic candidates.

Introduction

Modern drug discovery programs depend on transformative synthesis technologies that deliver structurally diverse scaffolds^1-3. Access to structural complexity allows researchers to generate novel molecules for biological evaluation, perform rigorous structure–activity relationship analyses^4,5, and systematically optimize the pharmacokinetic (PK), pharmacodynamic (PD), and toxicological properties of candidate molecules^6,7.

Target-Oriented Synthesis (TOS)⁸ has long been a cornerstone of organic chemistry and drug discovery^9,10, functioning both as a showcase of synthetic skill and as a route to bioactive molecules. By design, TOS pursues a single, pre-defined target—often a complex natural product^11-13—via multi-step sequences (Figure 1A). As chemical biology and pharmaceutical research goals have broadened, however, attention is increasingly shifting from the total synthesis of individual molecules to strategies that rapidly create structurally diverse libraries of small molecules.

Figure 1.

A) Approaches for synthetic organic chemistry. B) Exemplary click reactions grouped by stages of development. Formed bonds are bolded. C) Examples of drugs containing functional groups accessible through click reactions. A, B, C = click chemistry groups with orthogonal reactivity. X = O or NH. Y = O-, N-, or C-centered nucleophile.

Diversity-Oriented Synthesis (DOS)^14-17 exemplifies this transition. By combining simple, readily available building blocks, DOS rapidly yields compound libraries that maximize scaffold diversity, incorporating a range of functional groups and varied stereochemical configurations. Generating such compound libraries markedly increases the probability of identifying previously inaccessible bioactive molecules that can be progressed as leads in drug discovery programs.

Building on this foundation, Diversity Oriented Clicking (DOC)¹⁸ has emerged as a complementary method that combines the modularity and near-quantitative efficiency of click chemistry transformations¹⁹ to diversity-centric workflows. At its core, click chemistry¹⁹ embraces a pragmatic tenet —"all searches must be restricted to molecules that are easy to make"²⁰— thereby prioritizing reactions that are high-yielding, modular, and chemoselective. In DOC, multifunctional connective hubs serve as modular nodes for connecting several ‘clickable’ fragments, enabling rapid access to diverse molecular architectures. This approach originated from insights into how organic syntheses were conducted and the potential significant impact on every stage of subsequent drug discovery, development, and manufacturing^21-27.

Click chemistry has broadly progressed through three stages of development (Figure 1B). The first focused on classical "spring-loaded" reactions involving opening strained rings like epoxides and aziridines^19,28-32, cycloaddition chemistries, and specific condensation reactions¹⁹. Next followed the Copper-Catalyzed Azide-Alkyne Cycloaddition (CuAAC)^33,34 reaction, which became a cornerstone reaction due to its efficiency and reliability in forging 1,4-disubstituted 1,2,3-triazole rings from terminal alkynes and azides^35-37. Copper-free variants soon emerged with the Strain-Promoted Azide-Alkyne Cycloaddition (SPAAC)³⁸, which leverages highly strained cycloalkynes to dispense with the cytotoxic Cu(I) catalyst and thereby enhance biocompatibility^39,40.

The click repertoire has since expanded to encompass Fluoride Exchange (FEx) reactions—including Sulfur Fluoride Exchange (SuFEx)⁴¹ and Phosphorus Fluoride Exchange (PFEx)⁴². Whereas cycloaddition-based click processes typically furnish near-planar linkers, FEx chemistry exploits polyvalent hubs^43,44 to form robust heteroatom bonds and assemble three-dimensional molecular architectures whose very linkages can serve as pharmacophores^45-47.

Sulfur(VI) Fluoride Exchange (SuFEx) and Phosphorus(V) Fluoride Exchange (PFEx)

SuFEx click chemistry has been described in another Primer⁴⁴. Briefly, it relies on the exchange of the fluorine atom found in high oxidation sulfur groups (e.g., sulfonyl fluorides (–SO₂F), fluorosulfates (–OSO₂F)) with a range of nucleophiles, forming robust sulfur–oxygen (S─O)^41,48,49, sulfur–nitrogen (S─N)^50-52, or sulfur–carbon (S─C)^53,54 connections (Figure 1B). Several important SuFEx connectors, including sulfuryl fluoride (SO₂F₂)⁴¹, ethenesulfonyl fluoride (ESF)^55,56, and sulfur tetrafluoride (SOF₄)^53,57,58171, have gained prominence in the field. Consistent with the principles of click chemistry, SuFEx reactions are characterized by their rapidity, chemoselectivity, and ability to proceed under mild reaction conditions (i.e., at room temperature in relatively benign solvents). These reactions benefit from being catalyzed in a metal-free environment, enhancing their practicality and appeal in biological sciences. SuFEx has consequentially found wide-reaching applications in synthesizing polymers^49,57,59, bioconjugation^58,60 and drug discovery^26,61-63.

Similarly, PFEx reactions enable the exchange of fluorine atom in phosphorus(V) fluoride groups with nucleophiles⁴², facilitating the formation of stable phosphorus–oxygen (P─O), phosphorus–nitrogen (P─N), or phosphorus–carbon (P─C) bonds (Figure 1B). This extension of click chemistry includes phosphorus-containing linkages, which are naturally occurring and privileged connections^64,65. PFEx also allows for the precise tuning of the physicochemical properties of the phosphorus core by carefully selecting the modules that are clicked in. The introduction of PFEx to the click chemistry toolbox has broadened the collection of clickable hubs, now including alkyl and aryl phosphonyl fluorides^42,66 and cyclotriphosphazenes⁶⁷.

Integrating Click Chemistry with Drug Discovery

Click transformations have revolutionized the preparation of small-molecule therapeutics^26,68, bioconjugates⁶⁸, and chemical probes^36,69. When incorporated in drug discovery pipelines, click reactions can accelerate every phase of lead evolution, enabling rapid library assembly for high-throughput screening^46,62,70,71, systematic derivatization of primary hit molecules to refine potency, selectivity, and Absorption, Distribution, Metabolism, and Excretion and Toxicity (ADMET) profiles^72,73. Additionally, click chemistry can be used to install reporter handles for detailed mechanism of action studies.

Particularly noteworthy are the 1,4-disubstituted 1,2,3-triazoles, generated through CuAAC reactions, which serve as bioisosteres of peptide bonds that provide increased resistance to protease-mediated degradation^74,75. This characteristic has been pivotal in developing drugs such as rufinamide^76,77, an anticonvulsant, and solithromycin⁷⁸, a ketolide antibiotic (Figure 1C). CuAAC has also been an effective tool in enhancing the efficacy of vancomycin dimers in combatting drug-resistant bacterial strains such as methicillin-resistant Staphylococcus aureus (MRSA) and vancomycin-resistant enterococci^79,80.

SuFEx and PFEx click reactions are now setting the stage for the efficient synthesis of sulfur-^81-83 and phosphorus-containing⁸⁴ functional groups commonly found in drug candidates. This evolution in click chemistry promises to continue advancing drug development and therapeutic innovation. While PFEx remains in its infancy, SuFEx chemistry has enabled the in situ creation of potent acetylcholinesterase (AChE) inhibitors⁴⁶, the conversion of known anticancer drugs to more potent aryl fluorosulfate derivatives in one synthetic step⁸⁵, and facilitated the modification of lead molecules in formats suitable for automated systems^62,73,86. SuFEx chemistry has likewise become integral to high-throughput hit-to-lead workflows, markedly accelerating the conversion of screening hits into potent drug candidates^{62,71,73,87,88}.

Electrophilic handles tailored for FEx chemistry—exemplified by sulfonyl fluorides⁸⁹—are now prized for their biocompatibility and finely tuned reactivity toward proteins⁶³. Their selective activation dovetails with a wide range of diversity-building reactions, rendering them indispensable for integrating click processes into platforms such as DOC.

From Click Chemistry to Diversity Oriented Clicking

Building upon the impact of click chemistry, DOC was conceived to achieve broader molecular diversity by merging modular click reactions into a sequence that promotes unique chemical structures. CuAAC^19,90-95 and SuFEx^{20,41,44,61,96-101} click reactions have been extensively reviewed elsewhere; here, we concentrate on integrating FEx click transformations with classical click reactions to construct DOC-based discovery frameworks for hit molecule identification and optimization.

We catalog the functional groups and clickable hubs compatible with each click modality—emphasizing FEx transformations—and provide detailed protocols for their synthesis, handling, and practical deployment in the ‘Experimentation’ section. Next, we explore notable applications of FEx and DOC in identifying lead compounds, particularly those with antibacterial and anticancer properties, discussing considerations for efficient biological testing, and underscoring the impact of FEx click reactions in advancing drug development (see ‘Applications’). Finally, we explore some of the limitations currently faced by the field (see ‘Limitations and Optimizations’) and suggest what the future may hold for this nascent approach to chemical synthesis and what role it might play in shaping the future of pharmaceutical research (see ‘Outlook’).

Experimentation

This section outlines the functional groups that can be applied in a DOC discovery strategy, detailing the compatible hubs, reagents, workflows, and safety considerations. We offer guidance and recommendations to ensure reliable and robust results as they integrate the DOC technique into their research endeavors.

Azide-Alkyne Cycloaddition (AAC)

The catalyzed AAC reaction, built upon the pioneering work of the Huisgen 1,3-dipolar cycloaddition¹⁰², necessitates two essential coupling partners: azides (R─N₃) and alkynes (refer to Figure 1B). Alkyne building blocks are commonly available from commercial suppliers or can be readily synthesized using established protocols¹⁰³. In contrast, access to organic azide reagents is more constrained due to their potential toxicity and explosive properties (see ‘Safety Considerations’)¹⁰⁴. Nevertheless, organic azides are easily prepared in one step from alcohols (R─OH) via deoxyazidation¹⁰⁵ or from amines (R─NH₂) using fluorosulfuryl azide (FSO₂N₃)¹⁰⁶, among other protocols^104,107.

CuAAC reactions involve terminal alkynes and a Cu(I) catalyst, which is introduced directly as a Cu(I) salt or generated in situ by reducing a Cu(II) salt^33,34. While ligands are generally not required for CuAAC reactions, they may help prevent the disproportionation of Cu(I) to inactive Cu(0) and Cu(II)¹⁰⁸. Conversely, ruthenium catalysts can achieve AAC reactions involving more sterically demanding internal alkynes (termed the RuAAC reaction)¹⁰⁹, generating 1,5-disubstituted 1,2,3-triazoles that complement the 1,4-disubstituted 1,2,3-triazole linkages formed through CuAAC transformations. Other transition metals like iridium¹¹⁰, nickel¹¹¹, and zinc¹¹² have also been used to facilitate AAC reactions. The development of the biocompatible SPAAC method was a significant advancement³⁸ and allows the cycloaddition of azides and strained alkynes without the need for a Cu(I) catalyst, becoming popular in bioconjugation and chemical biology applications where the cytotoxicity of copper can be problematic^39,40.

SuFEx Hubs

This section will highlight DOC-compatible SuFEx hubs; that is, those containing several distinct and individually manipulatable functional groups with orthogonal reactivity. Examples will include ethenesulfonyl fluoride (ESF), (Z)-1-bromo-2-arylethene-1-sulfonyl fluoride (Z-BASF), 1-bromoethene-1-sulfonyl fluoride (BESF), β-chloro alkenylsulfonyl fluoride (BCASF), fluorosulfuryl isocyanate (FSI), and ethene-1,1-disulfonyl fluoride (EDSF), and methanedisulfonyl fluoride (MDSF). Their structures and associated diversity-generating reaction pathways are detailed in Figure 2.

Figure 2.

SuFEx-compatible hubs and subsequent diversification reactions.

Ethenesulfonyl fluoride (ESF)

Combining unsaturated carbon bonds and sulfur(VI)-containing functional groups yields a collection of versatile hubs particularly suited to DOC strategies. One such example is ESF, a superb Michael acceptor¹¹³ applied in material science¹¹⁴ and medicinal chemistry^20,115-117 before its inclusion as a modular click hub for SuFEx chemistry⁴¹. ESF can be readily derived from 2-chloroethanesulfonyl fluoride in a two-step process on a kilogram scale¹¹⁸. The fluoride-chloride exchange of 2-chloroethenesulfonyl chloride is initially conducted using potassium fluoride under acidic conditions, affording the intermediate 2-chloroethanesulfonyl fluoride¹¹⁸. The chlorine is subsequently eliminated upon exposure of the intermediate to manganese oxide, yielding ESF in near quantitative yields¹¹⁸.

ESF exhibits remarkable orthogonal reactivity at its two reactive sites^56,119. The most well-developed reaction involving ESF at the vinyl site is the Michael addition, where it reacts effectively with primary and secondary amines, anilines, certain nitrogen-containing heterocycles, and thiols^113,119. Other reactions, such as cycloadditions, radical reactions, and transition metal-mediated coupling reactions^120,121, have been documented, all while preserving the sulfonyl fluoride group^56,122. The sulfonyl fluoride functional group is well-placed for SuFEx reactions involving aryl alcohols^41,56, whereas harsher conditions are required for amines^41,56,123. Impressively, the reactivity of the sulfonyl fluoride group remains largely unchanged, regardless of the reactions occurring at the vinyl group.

(Z)-1-Bromo-2-arylethene-1-sulfonyl fluorides (Z-BASF)

Z-BASF is available from the corresponding 2-aryl alkenyl sulfonyl fluorides, themselves accessed from ESF and arenediazonium tetrafluoroborates via a Heck-Matsuda cross-coupling reaction¹²⁴, using lithium bromide in the presence of sodium periodate (NaIO₄, 40 mol%) in acetic acid¹²⁴. This method demonstrates high regio- and stereoselectivity for the (Z)-1-bromo product. Z-BASF can react selectively at its three functional sites under carefully optimized conditions, including palladium-catalyzed coupling reactions, nucleophilic substitution reactions, SuFEx reactions, and selective debromination reduction¹²⁴. When reacted with an excess of benzylamine nucleophile, a four-membered sultam (cyclic sulfonamide) can be generated through a Michael addition followed by an intramolecular SuFEx reaction¹²⁴.

1-Bromoethene-1-sulfonyl fluoride (BESF)

BESF (also called BrESF)^125-127 and its bench-stable precursor, 1,2-dibromoethane-1-sulfonyl fluoride (DESF), have been introduced as robust connectors for SuFEx reactions. DESF is synthesized from ESF through bromination, with subsequent elimination of HBr using tertiary amines, which yields BESF¹²⁵.

BESF participates in regioselective 1,3-dipolar cycloadditions with nitrile oxides, azides, and diazo compounds to produce isoxazole-5-sulfonyl fluorides, 1,2,3-triazole-4-sulfonyl fluorides, and 1H-pyrazole-5-sulfonyl fluorides¹²⁸, respectively^125-127. BESF's reactivity has further been explored in [3+2] cycloadditions with various aminopyridines and related compounds (e.g., N-aminoquinolines and N-aminoisoquinolines), yielding heterocycles in good yields while preserving sulfonyl fluoride functionality¹²⁹. The sulfonyl fluoride group retained in these products can be subjected to SuFEx functionalization, as seen by the calcium bistriflimide-mediated SuFEx between 1H-pyrazole-5-sulfonyl fluorides and amines¹²⁸. The DESF precursor can also react with amines to form sulfonyl fluorides and β-sultams¹²⁵.

β-Chloro alkenylsulfonyl fluoride (BCASF)

BCASF is an alkenyl sulfonyl fluoride hub synthesized by reacting a sulfuryl fluoride radical (SOF₂•), generated through the irradiation of sulfuryl chloride fluoride (FSO₂Cl) with either internal or terminal alkynes under photoredox conditions¹³⁰. This process employs fac-Ir(ppy)₃ (2 mol%) as a catalyst in a solvent mixture of diethyl ether and trifluorotoluene (5:1) illuminated by a blue LED (λ_max = 460 nm)¹³⁰.

BCASF is a versatile DOC hub with a chlorine atom and a sulfonyl fluoride site. The chlorine allows for transformations such as palladium cross-couplings (e.g., Suzuki, Sonogashira), selective reductions, and nucleophilic substitutions with S-, O-, and N-based nucleophiles, all while maintaining the integrity of the sulfonyl fluoride group suitable for subsequent SuFEx chemistry¹³⁰. Moreover, the late-stage modification of peptides and drugs with BCASF highlights its potential for broad applications in biological and medicinal chemistry¹³⁰.

Fluorosulfuryl isocyanate (FSI)

FSI is a bis-electrophilic linker useful for the stepwise conjugation of alcohols and amines¹³¹. FSI is prepared by reacting chlorosulfuryl isocyanate neat with an excess of sodium fluoride at 130°C for 2 hours. The reaction mixture was subjected to micro-distillation at 130 °C, with volatiles collected in a dry ice/acetone-cooled receiving flask. The pure FSI is obtained following a second, short-path distillation under atmospheric pressure (760 mmHg)¹³².

FSI leverages the pronounced electrophilicity of its –NCO group to engage alcohols in near-quantitative formation of fluorosulfuryl carbamates¹³¹. The transformation proceeds smoothly in routine solvents such as dichloromethane or acetonitrile, reaching completion within 5 minutes to 12 hours at 0 °C to ambient temperature—highlighting its operational simplicity and broad functional-group tolerance¹³¹.

FSI likewise engages secondary aliphatic amines in high-yielding additions, and the resulting adducts undergo a subsequent SuFEx step in aqueous K₃PO₄ (80 °C, 2–16 h) to deliver sulfamides¹³¹. Most recently, treatment of FSI with diaryl ketones under inert atmosphere has furnished bench-stable (diphenylmethylene)sulfamoyl fluorides—versatile, bifunctional synthons that enable modular access to aliphatic (β-imino)sulfonyl fluorides and racemic, protected β-amino-sulfonyl fluorides¹³².

Ethene-1,1-disulfonyl fluoride (EDSF) and methanedisulfonyl fluoride (MDSF)

Ethene-1,1-disulfonyl fluoride (EDSF) is a connective hub generated from 1,1-bis(fluorosulfonyl)-2-(pyridin-1-ium-1-yl)ethan-1-ide in situ¹³³. This precursor is itself synthesized by reacting methanedisulfonyl fluoride¹³⁴ (MDSF) with paraformaldehyde in the presence of pyridine in dichloroethane¹³³.

The reaction between EDSF and aryl acetylenes yields 1,1-bissulfonylfluoride-substituted cyclobutenes via a regioselective [2+2] cycloaddition (using H₂SO₄ in acetonitrile, −10°C to room temperature for 1 hour)¹³³. Subsequent diversification of the cyclobutene cores under Cs₂CO₃-facilitated SuFEx conditions successfully affords the mono-sulfonate ester product in high yields¹³³. MDSF can also be reacted with aromatic and alkenyl aldehydes to generate β-arylethenesulfonyl fluorides and unsaturated 1,1-disulfonyl fluorides, respectively¹³⁴.

PFEx Hubs

Several methods have been developed for generating the P(V)─F bonds needed for PFEx, including fluorination of phosphine oxides^135-138 or phosphinic acids^135,137,138, and F─Cl exchange from phosphorous chlorides⁴².

Phosphinic and phosphine fluorides

Phosphine oxides and phosphinic acids react with SO₂F₂ gas and Et₃N in dichloroethane to afford phosphoric fluorides (Figure 3A)¹³⁵. Additionally, phosphinic acids can be converted to the corresponding fluoride derivatives using potassium salts of perfluoroalkyl ether carboxylic acids (PFECA) in dimethyl formamide (DMF)¹³⁸. Sulfone iminium fluoride is another viable reagent for this transformation, reacting with phosphinic acids and phosphine oxides in the presence of Et₃N in acetonitrile¹³⁷. Reacting secondary phosphine oxides with electrophilic fluorine sources like Selectfluor^™ also provides facile access to phosphoric fluorides in good yield under mild conditions¹³⁶.

Figure 3.

Phosphorus-based click chemistry hubs. A) Synthesis of P─F-containing PFEx hubs. B) Polyfluorinated hubs derived from HFP. C) Reported P(V) hubs containing other DOC-compatible functional groups. D) Multi-clickable hub.

Phosphoramidic difluorides and phosphoramidofluoridates

Phosphoramidic difluorides can be prepared by reacting secondary amines with POCl₃ (or PSCl₃ to access the corresponding thiophosphoryl equivalent) in the presence of Et₃N as a base, followed by F─Cl exchange with excess potassium fluoride in acetone⁴². Given their limited stability¹³⁹, phosphoramidic difluorides should be freshly prepared and used promptly for subsequent PFEx reactions with aryl alcohols⁴². The initial PFEx product, a phosphoramidofluoridate, exhibits enhanced stability over its difluoro precursor when prepared using a catalytic system of hexamethyldisilazane (HMDS) and tert-butyl-1,1,3,3-tetramethylguanidine (BTMG) (20 mol%) within 15 minutes⁴².

Phosphoramidofluoridates can also be prepared in situ by sequentially adding aryl alcohols and primary or secondary amines to POCl₃ in the presence of several equivalents of Et₃N. This stepwise decoration of the P(V)-core generates a phosphoramidochloridate that readily undergoes F─Cl exchange facilitated by potassium fluoride and a phase transfer catalyst, n-Bu₄NCl (10 mol%)⁴². The fluorine found in phosphoramidofluoridates undergoes PFEx reactions with aryl alcohol nucleophiles when treated with a catalytic system of HMDS and the stronger organic base 1,5,7-triazabicy-clo[4.4.0]dec-5-ene (TBD, 20 mol%)⁴².

Fluorinated cyclotriphosphazenes

Polyfluorinated organophosphorus compounds, such as hexafluorocyclotriphosphazene (HFP)^140-144, are enticing hubs for PFEx-centered DOC⁴² (Figure 3B). Using a series of PFEx reactions with aryl and alkyl alcohols and amines, the sequential decoration of the HFP core has been demonstrated, exchanging five of the six P─F bonds with various modules⁴². Of note, HFP readily sublimes at room temperature and can be susceptible to hydrolysis; the PFEx-mediated derivatization of the cyclotriphosphazene ring markedly enhances the stability and ease of handling of derivatives^42,67.

The PFEx reaction between HFP and lithium alkynides produces 2-substituted-alkynyl-1-cyclotriphosphazene (SACP) hubs that feature an internal alkyne for cycloaddition chemistries^145-148, similar to the alkyne present in SASF SuFEx hubs⁶⁷. By carefully adjusting the equivalent of delivered alkynide, SACP hubs with multiple internal alkynes can be generated⁶⁷.

The cyclotriphosphazene group of SACP hubs polarizes the internal alkyne and activates it towards 1,3-dipolar cycloaddition chemistries⁶⁷. Reacting SACP hubs with a variety of dipoles produce a diverse array of heterocycles under different conditions, including pyrrolines, pyrrolizines, isoxazoles, 2,3-dihydroisoxazoles, triazoles, and pyrazoles⁶⁷. For instance, 3,4,5-trisubstituted isoxazoles can be synthesized under mild conditions from the corresponding nitrile oxides generated in situ from imidoyl chlorides and Et₃N in toluene upon microwave irradiation. In contrast, the formation of 1,3,4-trisubstituted pyrazoles from sydnones typically requires prolonged heating in toluene. Regiochemcial assignment was achieved using NMR and X-ray crystallography. Additionally, sequential functionalization of each internal alkyne in the bis-alkynyl SACP hub has proven to be successful. In line with DOC principles, the subsequent PFEx reactions with aryl and alkyl alcohols allowed the elaboration of the central P(V)-hub in a controlled manner to generate a library of unique compounds, marking the first integration of PFEx click chemistry within the DOC framework⁶⁷.

Difluorophosphoryl isocyanate

Difluorophosphoryl isocyanateis synthesized by reacting ethyl carbamate with PCl₅, yielding the precursor phosphorisocyanatidic dichloride, a molecule found to be stable for prolonged periods at room temperature following distillation¹⁴⁹ (Figure 3C). Subsequent treatment with antimony trifluoride (SbF₃) produces phosphorisocyanatidic difluoride¹⁴⁹. Alternatively, difluorophosphoryl isocyanate can be obtained by reacting difluorophosphoric anhydride (F₂(O)POP(O)F₂) and potassium cyanate (KOCN)^150,151. The isocyano group found in difluorophosphoryl isocyanate readily reacts with aryl alcohols, sodium salts of alkyl alcohols and amines, forming substituted carbamatophosphorodifluoridates and ureidophosphorodifluoridates, respectively, while preserving two P─F bonds for further FEx derivatization¹⁴⁹.

Difluorophosphoryl azide

Difluorophosphoryl azide (F₂P(O)N₃), bearing two P─F bonds and an azide, can be prepared by condensing neat difluorophosphoryl chloride onto dry NaN₃ and allowing the reaction to proceed at room temperature for 24 hours^150,152. The desired F₂P(O)N₃ can be collected in good yield by distillation through a −20 °C cold trap, followed by condensation in a dry ice/acetone-cooled receiving system^150,152.

Alkynyl phosphonic difluorides

To access alkynyl phosphonic difluorides, alkynylphosphonic acid diethyl esters can be reacted readily with PCl₅ to form the intermediate alkynyl phosphonic dichlorides¹⁵³. These intermediates can then be fluorinated with arsenic (AsF₃) or antimony trifluoride (SbF₃) to yield the target P─F products¹⁵³.

Vinyl phosphonic difluoride

Vinyl-phosphonic difluoride is readily obtained through F─Cl metathesis of the corresponding dichloride¹⁵⁴. Although its bifunctional reactivity remains largely unexplored, the molecule is an attractive phosphorus-based analogue of ethenesulfonyl fluoride (ESF). Strategically chosen bases and nucleophiles should unlock complementary reaction manifolds: PFEx ligation at the P(V)─F centre and Michael-type conjugate addition across the activated vinyl group.

Multi-clickable Hubs

An optimal DOC hub is inherently multi-clickable: it incorporates a core scaffold bearing at least two—and ideally three or more—orthogonal functional handles amenable to successive generations of click reactions. To maximize structural reach, the hub architecture should transcend simple aryl frameworks and accommodate conformationally flexible or three-dimensional motifs. Crucially, it must embed functional groups that participate in FEx chemistry, whose latent yet highly selective P─F or S─F reactivity furnishes programmable entry points for modular ligation while preserving downstream diversification capacity.

Orthogonally reactive hubs that tolerate PFEx, SuFEx and CuAAC in a single workflow have now been realized (Figure 3D)⁴². An exemplar sequence—accelerated SuFEx ligation, followed by PFEx diversification and a concluding CuAAC—delivered the desired product in three steps and 73 % overall yield. More recently, a photoredox azido-fluorosulfonylation has transformed simple alkenes into bi-clickable hubs bearing mutually compatible azide and –SO₂F handles¹⁵⁵. Sequential CuAAC/SuFEx modification of these scaffolds enabled rapid generation of a 1,2,3-triazole-sulfonyl-fluoride library and late-stage attachment of bioactive fragments including estrone, fenbufen and flurbiprofen¹⁵⁵.

Practical Synthesis and Reactivity Considerations

FEx and DOC, like other click reactions, are designed with democratization and accessibility in mind¹⁵⁶, prioritizing simple, reliable, and easily scalable reactions that require only mild conditions. Many of the nucleophiles compatible with FEx reactions are widely available from commercial suppliers, often with thousands or even millions of nucleophile options¹⁵⁷, making click chemistry readily accessible.

Hubs compatible with DOC chemistry often require more complex syntheses or specialized equipment (detailed above). For example, generating the BCASF hub involves photo-irradiation, which may necessitate a photochemistry setup that not every laboratory has access to. Additionally, handling gaseous reagents like SO₂F₂ can be challenging. While solid, non-gaseous alternatives are gaining popularity¹⁵⁸, they are often significantly more expensive and not always widely available. Although several DOC-compatible SuFEx hubs, such as FSI¹³¹ and ESF¹¹⁸, have been synthesized on a kilogram scale, the preparation of SASF¹⁸—arguably one of the most versatile DOC hubs—remains challenging. Its synthesis relies on less accessible reagents and a multi-step route, posing hurdles for scale-up.

Consistent with click chemistry guidelines, hubs engineered for SuFEx or PFEx ligations must exhibit broad physicochemical resilience. Sulfonyl fluorides^159-163 and phosphoramidofluoridates⁴² have been shown to withstand reductive, hydrolytic, and thermolytic challenges without loss of integrity while sulfonyl fluorides also withstand oxidative conditions.

Indeed, in the multi-click sequence depicted in Figure 3D, the P─F bond of a phosphoramidofluoridate proved markedly less labile than the aryl fluorosulfate S─F bond, allowing for precise, successive transformations⁴². Crucially, the intrinsic reactivity of S─F and P─F linkages can be modulated through electronic or steric tuning of the substituents attached to the central heteroatom; additional rate acceleration can arise from ring strain, as exemplified by cyclic phosphoramidofluoridates⁴².

Safety Considerations

While prioritizing safe, minimally toxic compounds is key to click chemistry, several reagents and substrates compatible with FEx and DOC require careful handling. It is important to note that, while all organic chemistry can be dangerous, azides, S─F-containing functional groups, and P─F-containing groups can bring with them significant risk if improperly handled.

To ensure the safe preparation and handling of organic azides, several guidelines should be adhered to^105,165-167. Metal spatulas and halogenated solvents like dichloromethane should be avoided. When working with organic azides, the likely stability should be assessed using the carbon-to-nitrogen (C/N) ratio; stable azides generally have a ratio of ≥3¹⁰⁴. Alternatively, the "Rule of Six" suggests that six carbon (or similar-sized) atoms per energetic functional group help maintain stability¹⁹. Aliphatic azides are generally more stable than olefinic, aromatic, or acyl azides. When handling azides, a reducing solution of triphenylphosphine in water or organic solvent should be kept on hand to quench any spills.

The P─F-containing organophosphorus compounds employed in PFEx^42,167,168 can act as covalent nerve agents^168,169 by targeting serine residues in the active site of acetylcholinesterase (AChE). This enzyme is involved in the termination of synaptic transmission and thus inhibition can cause muscle paralysis, seizures, and, in severe instances, fatal respiratory failure^168,169. Toxicity is not set by P─X bond electrophilicity alone; factors such as bioavailability, metabolic stability, and downstream pharmacodynamic effects all play important roles. Researchers must review the toxicological properties of these compounds and adhere to strict safety protocols¹⁷⁰. Following comprehensive guidelines and safety policies, such as those recommended by the American Chemical Society, is critical for safely conducting FEx chemistry and minimizing potential risks.

Designing New DOC-Compatible Hubs

Clickable hubs are now firmly embedded in the synthetic toolbox^{18,42,43,67,130,171}, yet there is still considerable scope for inventing new scaffolds. Here we set out key design principles to guide the creation of the next generation of hubs. These principles, detailed in Table 1, offer a practical framework for innovation, integrating traditional click chemistry tenets while emphasizing user-friendliness and safety.

Table 1.

Guiding principles for DOC hub development. Table repurposed from ref⁶⁷.

Desirable Property	Suggested Hub Design Element
Orthogonal Reactive Centers	Include multiple, distinct functional groups (e.g., azides, sulfonyl fluorides, phosphoramidofluoridates) that do not cross-react to ensure clean chemical transformations.
Chemoselectivity	S/P─F bonds should be sufficiently electrophilic for substitution but balanced to avoid uncontrolled or side reactions. Ideally, S/P─F bonds should only be activated by the correct catalyst/reagent combinations.
Click Group Stability	Click groups should be stable to various synthetic conditions, allowing early integration in DOC sequences. Ensure hubs can withstand typical laboratory and storage conditions, including common solvents.
Predictable Regioselectivity	Click reactions should generate a single product regioisomer.
Facile Access and Scalability	Hubs should be synthesized from readily accessible materials using scalable methods to promote community adoption and ease of use.
Low Volatility	Prioritize to minimize exposure risks and ensure ease of handling.
Quenchable	Safe quenching of the electrophilic bonds using simple procedures (e.g., soaking overnight in isopropanol/sodium hydroxide).

High-throughput Synthesis Workflow and Experimental Considerations

CuAAC^172,173, thiol-ene^174,175, and SuFEx^72,73,86,88 click reactions have been successfully adapted to high-throughput synthesis, in both 96- and 384-well formats^173,176, as well as in on-chip platforms¹⁷⁷, demonstrating potential in accelerating hit identification and expediting the hit-to-lead process¹⁷³. Testing unpurified reaction mixtures for biological activity in ‘direct-to-biology’ workflows further expedites the discovery process^48,72. The DOC approach, which utilizes multi-clickable hubs, enables the generation of more structurally diverse molecules, thereby increasing the likelihood of identifying hit candidates. This is especially relevant when considering reactions like conjugate additions and cycloadditions^23,24 that can install saturated motifs in molecule scaffolds, often favored in drug development space^178-180.

Transitioning click transformations from bulk flasks to microplate arrays imposes practical constraints: the chosen reactions must reach completion at ambient temperature so that routine plate shakers can be used without specialized heating or cooling. Reaction scales vary significantly by chosen format—studies have reported a reaction scale of 3 μmol per well in a 96-well setup⁷², while a 384-well format operates at 0.25 μmol per well¹⁷⁶. Conducting reactions on such small scale requires accurate delivery of reagents and substrates. This can be achieved by preparing stock solutions in either the reaction solvent or in an alternative solvent that can be effectively removed through evaporation. For smaller well-plates, automated delivery can be useful (e.g., acoustic liquid handlers). Selecting well-plates compatible with the required reaction solvent is also important to prevent plate decomposition or compound adsorption. Preparing plates of substrates ahead of time and storing these below −20 °C can be helpful.

The progress of the reactions can be monitored by various methods⁷², though liquid chromatography-mass spectrometry (LC-MS) analysis is more routine. Changes in UV peak area and comparison with previously run samples can be used to calculate reaction conversion and quantify product amount. To enable a direct-to-biology workflow, whereby unpurified reaction wells can be tested directly in biological assays (e.g., evaluated for activity against cancer cells or bacteria), volatile reagents, byproducts, and solvents should be prioritized^48,72. Additionally, FEx reactions that can be performed in biologically relevant media, such as phosphate-buffered saline ideally suited for this purpose^58,62.

Initial biological testing may involve a single high dose to identify any hit molecules, followed by a more detailed 8- or 12-point dose-response curve to determine compound potency. Each hit molecule must be prepared on a larger scale and purified and evaluated in analytically pure form to confirm activity. Suitable controls, such as the fragments used to prepare the hit molecules, should be evaluated for activity as well to ensure biological activity is solely attributed to the product molecule. Given the high number of compounds generated per run, and if possible, robotic high-throughput viability assays can be incredibly valuable.

Applications

This section examines applications of FEx click chemistry and DOC discovery platforms, emphasizing their potential to drive therapeutic innovation, covalent drug modulation, high-throughput screening in medicinal chemistry, and enabling late-stage functionalization in drug synthesis.

Structural Diversity of FEx Hubs and Libraries

SuFEx and PFEx hubs typically have a molecular weight of around 100 Da, which is not a barrier for medicinal chemists in keeping the total molecular weight of hit compounds below or near 500 Da — a consideration for adhering to Lipinski's rule of five¹⁸¹.

Some FEx moieties, such as fluorosulfate (compound 2) and sulfamoyl fluoride (compound 3) or phosphoramidofluoridate (compound 9) and phosphoramidate (compound 10), offer similar shapes and chemical structures with aligned physicochemical properties, such as lipophilicity and polar surface area. By selectively replacing the fluoride group in FEx hubs with nucleophiles, the physicochemical properties underpinning the associated absorption, distribution, metabolism, excretion, and toxicity (ADMET) profiles of the overall molecule can be finetuned while potentially maintaining on-target potency. For example, methyl surrogates of sulfuramidimidoyl fluoride (compound 6) and sulfamide (compound 7) have an almost ten-fold difference in their octanol-water partition coefficient values, which will be useful to fine-tune lipophilicity of the molecules.

Broadening structural diversity in screening libraries increases the chance of engaging complementary protein surfaces—via hydrogen bonds, polar contacts and lipophilic interactions—and thereby improves the probability of identifyingits¹⁷.

FEx chemistry and the associated hubs also allow variation in the three-dimensional structure of their molecular shapes to be realized, as illustrated by the electrostatic potential surface of FEx moieties 1–10 (Figure 4C). The differences in surface characteristics highlight the broad range of electrostatic configurations that can be accessed while retaining reactive fluoride groups for future modification. For instance, different electrostatic potential regions on the molecular surface can enhance specific interactions with protein active sites, improving binding affinity and selectivity¹⁸².

Figure 4.

The diversity and drug-likeness of FEx chemistry generated moieties. A) Physicochemical properties of the fragments and products of FEx chemistry. B) Exact molecular weight vs. SLogP plot, C) Electrostatic potential surface of FEx moieties, and D) PMI plot of published FEx libraries. Library information is from Zheng et al., 2019⁶³ (purple), Smedley et al., 2020¹⁸ (red), Cheng et al., 2022¹⁸³ (green), and Sun et al., 2023⁴² (orange). SLogP = simulated octanol-water partition coefficient. PMI = principal moment of inertia. TPSA = topological polar surface area. I₂/I₃ and I₁/I₃ = normalized principal moment of inertia ratios.

The small molecule libraries amenable to or generated through FEx chemistry also span diverse chemical space. We analyzed the chemical diversity of some of the previously reported FEx-compatible compound libraries, including agnostic SuFEx⁶³ and PFEx⁴² libraries and DOC libraries^18,183. The principal moments of inertia (PMI) analysis¹⁸⁴ of these libraries show that these molecules possess diverse shapes, including three-dimensional, sphere-like structures (Figure 4D). The coverage of these libraries extends beyond the chemical space typically explored by conventional small molecule libraries, which are often limited to disc and rod-like molecules¹⁸⁵. Furthermore, the three-dimensional, non-planar connectivity accessed through FEx click reactions (contrasted to the usually planar linkers generated through AAC click reactions) creates valuable linkages for building drug-like libraries that extend beyond the chemical space covered by conventional small molecule screening libraries. The linkages formed through click reactions have themselves been identified as invaluable pharmacophores, as exemplified in the next section, and can have profound effects on compound PK/PD profiles and in vivo stability.

Consequently, FEx hubs and the associated click chemistries can facilitate the design of chemical libraries with broader interaction coverage with biological targets. Altogether, the diverse properties of possible FEx products enable the synthesis of bioactive molecules that meet desired target binding affinity, drug-like properties, and ADMET profiles against various biological targets.

Discovery and Development of FEx-enabled Bioactive Molecules

The diverse chemical space accessible through FEx chemistry enables the discovery and development of bioactive molecules against a variety of biological targets. Over 150 clinically used drugs contain sulfur(VI)¹⁹¹, such as sulfonamide antibiotics, and phosphorus(V)⁸⁴, including the antiviral (−)-remdesivir and the anticancer agent ifosfamide, demonstrating the importance of these functional groups. Hence, perhaps it is unsurprising that FEx-enabled compound libraries have generated several novel classes of bioactive molecules (key examples are shown in Table 2). Aligned with the typical phases of early drug discovery, examples are divided into two main categories: 1) Screening of FEx-derived molecular libraries (including both agnostic⁶³ and DOC^18,183 libraries) and 2) high-throughput medicinal chemistry (HTMC) enabled by FEx chemistry^{46,62,71,72,88,190}.

Table 2.

Examples of FEx/DOC-derived bioactive molecules.

Reference	Library Size	Biological Endpoint	Example Hit Structure	Potency
Bioactive Molecules Identified from Agnostic Compound Libraries
Wolan and Sharpless 201963	132	hNE (biochemical)		IC50 = 0.24 μM
Moses and Sharpless 2022183	88	hNE (biochemical)		IC50 = 2.1 μM
Rakesh and Qin 2019186	82	Staphylococcus aureus (bacterial)		MIC 0.81 μg/mL
Dong, Zhang and Sharpless 2021187	41	Staphylococcus aureus 43300 (bacterial)		MIC 1.1 μg/mL
Moses 202018	151	Staphylococcus aureus USA300 (bacterial)		MIC 0.625 μg/mL
Murthy 2019188	24	kanamycin-resistant Escherichia coli, MRSA (bacterial)		MIC99 0.66 μg/mL (E. coli) 5.28 μg/mL (MRSA)
Xu and Zhao 2024189	32	Rhizoctonia solani (fungi)		EC50 = 1.5 μg/mL
Wu 201985	40a	HT-29 cell (cellular)		EC50 = 70 nM
Bioactive Molecules from FEx-enabled High-throughput Medicinal Chemistry (HTMC)
Sharpless and Wolan 202062	460a	SpeB (biochemical)		Kd = 18 nM
Erb 2021190	576 (biochemical)a then 43 (cellular)a	ENL-YEATS		EC50 = 63 nM (cellular)
Wilson and Wolan 202471	1,200a,b	Influenza H1 (cellular)		EC50 = 0.7-19 nM (H1strains)
Erb 202488	3,180a	CRBN/GSPT1 (cellular)		EC50 = 55 nM
Moses 202472	288c	PDAC organoid proliferation (hM1A line)		IC50 = 6.8 nM
Liu, Qin and Tang 202346	102a	AChE (biochemical)		IC50 = 94 nM

^aUnpurified reaction mixtures used for screening.

^bTotal number including amide coupling products.

^cUsed after removal of solvent and byproduct.

SpeB = streptococcal pyrogenic exotoxin B. ENL = eleven-nineteen leukemia. YEATS = Yaf9, ENL, AF9, Taf14, and Sas5. CRBN = cereblon. GSPT1 = G1 to S phase transition 1. PDAC = pancreatic ductal adenocarcinoma. AChE = acetylcholinesterase. IC₅₀ = half maximal inhibitory concentration. MIC = minimum inhibitory concentration. EC₅₀ = half maximal effective concentration. K_d = dissociation constant.

Screening of FEx molecular libraries

Serine hydrolases, including human neutrophile elastase (hNE)^63,183, are well known to be inhibited by sulfonyl fluorides like phenylmethylsulfonyl fluoride¹⁹². Human neutrophil elastase (hNE) provides a convenient test-bed for new FEx motifs: it is clinically relevant and can be monitored by straightforward, highly robust fluorogenic assays, enabling rapid assessment of both covalent reactivity and bioactivity. Several S─F-containing molecules have been found to have selective activity against hNE, including benzene-1,2-disulfonyl fluoride (IC₅₀ = 3.3 μM) and 2-(fluorosulfonyl)phenyl fluorosulfate (IC₅₀ = 0.24 μM)⁶³. Several SASF-derived alkenyl sulfonyl fluorides generated through a DOC approach containing 1,2,3-triazole and carboxylic acid groups were also found to have micromolar activity against the enzyme (IC₅₀ range = 2.13–8.80 μM)¹⁸³. These inhibitors exhibit "dormant" reactivity, meaning they remain inactive until they encounter the right protein target with the precise electrostatic and geometric environment to facilitate the replacement of the S─F bond with a covalent link to the protein. Importantly, this covalent modification only occurs in properly folded and active proteins, as denatured proteins cannot be modified by the same inhibitors^63,183, potentially reducing the risk of off-target effects and associated toxicity.

Another phenotypic screen applicable in rapidly probing the activity of compound libraries is bacterial viability assays. Evaluating compound libraries has revealed molecules with antibacterial effects against Gram-negative (e.g., E. coli)¹⁸⁸ and Gram-positive (e.g., S. aureus)^186,187 bacteria. For example, a "double-click" DOC strategy using FSI with carboxylic acids followed by SuFEx reactions conducted on 96-well microplates revealed several compounds exhibiting antimicrobial activities against S. aureus and MRSA, with minimum inhibitory concentration (MIC) values as low as 6.25 μg/mL¹⁹³; the MIC of a molecule is the lowest concentration that inhibits visible growth of a microorganism after a defined incubation period. Diversifying SASFs in a DOC synthesis platform generated a compound library where 11% of the candidates showed inhibitory activity against MRSA (USA300)¹⁸. The lead compound contained a pyrazole core and intact sulfonyl fluoride and had a MIC value of 0.625 μg/mL¹⁸. An evaluation of a library of aromatic sulfonyl fluorides revealed nitro sulfonyl fluorides to be broad-spectrum antibiotics, identifying a new pharmacophore for potentially addressing drug-resistant pathogens¹⁸⁸. The mechanism of action underlying these compounds potentially involves the formation of reactive intermediates, oxathiazoles, generated from nitroarenes by nitroreductases, that lead to bacterial killing¹⁸⁸. Considering the structural diversity of the hit molecules, including both sulfonyl fluorides and fluorosulfates, and the difference in the bacterial species inhibited, it is possible that the molecules discussed operate by unexplored antibacterial mechanisms.

Given the diverse chemical space and broad range of physicochemical properties found in DOC-derived compound libraries, agnostic FEx libraries hold promise as a rich, untapped source for bioactive molecules. In molecules that retain electrophilic groups containing S─F and P─F bonds, the potential for covalent binding with nucleophilic amino acids (e.g., tyrosine, lysine, serine) in target protein(s) is also likely. Modifying hit molecules to include functional groups like biotin that enable target identification through chemical proteomic analysis. These features of FEx compound libraries may allow researchers to target proteins currently considered "undruggable," analogous to the successful discovery of covalent inhibitors of the KRAS^G12C oncoprotein through selective cysteine targeting^194,195.

After an initial hit or lead is secured, the operational robustness of FEx click chemistry allows swift elaboration into analogue series. The resulting libraries can be screened iteratively to isolate compounds with progressively improved potency and selectivity (Figure 5A)^58,62.

Figure 5:

Applying FEx to drug discovery. A) Schematic representing a click chemistry-enabled high-throughput medicinal chemistry workflow. B) Non-competitive and C) competitive ABPP probe design strategy with FEx warhead moieties. LC-MS = liquid chromatography-mass spectrometry.

Direct-to-biology FEx platforms that couple amines with iminosulfur oxydifluorides via SuFEx chemistry have produced hit compounds whose potencies exceed those of their progenitor leads by 100- to 1,000-fold^{62,71,88,171,190}. The ensuing sulfamide and sulfuramidimidoyl fluoride analogs are generally non-cytotoxic at micromolar concentrations, attesting to the medicinal chemistry-friendliness of SuFEx linkers. Should a particular FEx-derived scaffold display an unfavorable toxicological profile, the modular nature of the chemistry allows the linkage to be swapped for an alternative motif without perturbing the core pharmacophore.

Other HTMC applications have included a variety of biological targets, suggesting the universal applicability of the approach. Examples include bacterial virulence factors⁶², epigenetic readers¹⁹⁰, viral surface proteins⁷¹, molecular glues⁸⁸, and various cancers⁷².

While most studies use biochemical assays for the screening of molecular libraries, fluorogenic assays⁶², Homogeneous Time-Resolved Fluorescence¹⁹⁰, fluorescence polarization⁷¹, and cell-based assays⁷² are also applicable for library screening, such as the cellular thermal shift assay to discover cell active ENL inhibitors¹⁹⁰ and cell viability assay to discover G1-to-S phase transition 1 (GSPT1) protein degraders⁸⁸. Notably, the medicinal chemistry cycle of library generation, screening, and biological evaluation can be iterated until optimal molecules are developed, akin to a typical medicinal chemistry campaign. A combination of biochemical and cellular assays facilitates the development of potent, cell-active molecules from biochemical screening hit molecules, as demonstrated by the development of potent cell-active ENL inhibitors¹⁹⁰. SuFEx chemistry has also been utilized to develop PROTACs against ENL¹⁹⁰ and other targets¹⁹⁶.

Activity-based Protein Profiling and Inverse Drug Discovery

Activity-Based Protein Profiling (ABPP) and Inverse Drug Discovery (IDD) are two strategies that utilize FEx probes to identify covalently modified protein conjugates within complex proteomes. ABPP focuses on employing active-site-directed small molecules¹⁹⁷, while IDD uses protein-activated electrophiles¹⁹⁸ to profile protein function. Interestingly, only 4% of the human proteome is "ligandable", or able to interact with a small molecule with high affinity, by the current collection of FDA-approved small molecules¹⁹⁹. Thus, an overarching goal within the chemical biology scientific community is to target every protein in the human proteome with a chemical probe.

ABPP and IDD have emerged as platforms to discover unique ligands for specific proteins and enable covalent modification of proteins implicated in human pathology. These covalent ligand approaches, together with chemically induced targeted degradation technologies such as PROTACs²⁰⁰, are fundamentally transforming the landscape of drug discovery^201-203. The inherent characteristics of FEx moieties as hydrolytically stable, latent electrophiles with chemoselective reactivity make them well suited for ABPP and IDD strategies^203,204.

Accordingly, P─F- and S─F-bearing hubs have been woven into ABPP and IDD platforms across a range of cell systems and proteomic contexts. The studies collated in Table 3, together with the enrichment workflows coupled to mass spectrometry-based proteomics depicted in Figures 5B and 5C, reveal that these probes covalently tag multiple nucleophilic residues—including serine, threonine and histidine^{44,85,198,204,205}.

Table 3:

Probes integrating FEx-compatible functional groups designed to profile various proteins at specific residues.

Reference	Cell Line/ Tissue	Cell Mode	Enrichment Method	FEx Functional Group (Number of Molecules)	Example Probe Structure	Protein Identified
Cravatt 1999206	Rat tissues	Lysate	Direct (Biotin tag)	Fluorophosphate (1)		FAAH
van der Hoorn and Weerapana 2012204	HeLa (spike)	Lysate	Direct (Alkyne tag)	Sulfonyl fluoride (1)		Chymotrypsin
van der Hoorn 2013207	Mouse liver and pancreas	Lysate	Direct (Alkyne tag)	Sulfonyl fluoride (1)		GST
Sharpless and Kelly 2016208	HeLa	Lysate	Direct (Alkyne tag)	Fluorosulfate (3)		CRABP2 FABP
Taunton 2017209	Jurkat	Intact cell	Direct (Alkyne tag)	Sulfonyl fluoride (3)		133 kinases
Sharpless, Wilson and Kelly 2018203	HEK293T	Lysate	Competition	Fluorosulfate (6)		GSTO1
Sharpless and Kelly 2020210	HEK293T	Lysate	Direct (Alkyne tag)	Sulfuramidimidoyl fluorides (16)		PARP1
Abbasov and Cravatt 2021211	MDA-MB-231	Lysate	Competition (Lysine-directed)	Sulfonyl fluoride (20)		CPOX
Burley and Bush 2023212	Jurkat	Intact cells	Direct (Alkyne tag)	Fluorosulfate (1)		CDK
				Sulfonyl fluoride (6)		CDK
				Sulfamoyl fluoride (1)		PIKK
Taunton 2023213	Jurkat	Intact cells	Direct (Alkyne tag)	Sulfonyl fluoride (20, 10x Enantiomers)		APMAP

FAAH: fatty acid amide hydrolase. GST: glutathione transferases. FABP: fatty acid binding protein. CRABP: cellular retinoic acid binding protein. GSTO1: glutathione S-transferase omega 1. PARP1: poly ADP-ribose polymerase 1. CPOX: coproporphyrinogen oxidase. CDK: cyclin dependent kinase. PIKK: phosphatidylinositol 3 kinase related protein kinase. APMAP: adipocyte plasma membrane-associated protein.

Integrating FEx-containing warheads in chemical probe design expands the range of ligandable human proteins^{42,44,201,206,209,214}. Several examples now demonstrate how SuFExable probes have inherently unique and latent reactivity²⁰¹. Probes can be introduced to intact cells or cellular lysates to study the function of the target proteins. A carefully designed probe, which utilizes a FEx functional group as a warhead, can be applied to proteome samples without additional ligand or inhibitors to survey samples for all probe-reactive, functional proteins (Figure 5B). Incorporating a FEx group without additional ligands or inhibitors is called a non-competitive process. This strategy helps the exploration of a broad range of protein functions using a relatively straightforward workflow without the need for additional molecules^215,216. However, some probes exhibit promiscuity, complicating the validation of their specificity using non-competitive labeling^215-217.

To overcome such challenges, probe labeling can alternatively be conducted with a known ligand or inhibitor that acts as a competitor of the FEx probe, creating a competitive process (Figure 5C). In this strategy, a known inhibitor is pre-incubated with the proteome to profile target proteins at the specific binding site, thereby reducing subsequent probe labeling as the competitive ligand has already occupied the common active residues^216,217. This powerful strategy has been extensively used to identify novel small-molecule inhibitors, discover new target proteins, analyze the mechanism of probe labeling, and validate both new protein-ligand interactions and the specificity of previously discovered compounds^216,218-220. However, the limitation of this approach is that it requires an additional workflow involving the synthesis and concentration optimization of the competitor inhibitor, and ultimately, the probe specificity is evaluated only relative to the interaction between the competitor and the target protein^215,217.

For example, a known lysine ligand²²¹ can be pre-incubated with proteome samples and chased with a lysine-reactive FEx probe to profile lysine residues that compete for binding both the ligand and probe. Reporter groups, such as biotin azide, can be subsequently attached to the alkyne moieties of the FEx probe-bound proteins via click chemistry. The profiled proteins are enriched using a biotin-streptavidin interaction for subsequent analysis (i.e., mass spectrometry or gel-based visualization).

Chemoproteomic profiling that deployed 34 distinct aminophilic probes—including SuFEx-enabled small molecules—mapped the lysinome of two human cancer cell lines, identifying covalent modification of 818 lysine residues across 581 proteins²¹¹. Of note, several proteins were specifically labeled by a SuFEx-containing small molecule out of the pool of 34 probes (e.g., K404 on coproporphyrinogen-III oxidase) or were more selectively profiled by SuFEx probe(s) (e.g., active-site lysine of protein kinase)²¹¹. These findings, and others^85,203, demonstrate that employing SuFEx chemistry within ABPP designs could significantly broaden the druggable proteome.

In IDD workflows, aryl-fluorosulfate SuFEx reagents applied to HeLa cells selectively tagged tyrosine residues on an intracellular lipid-binding protein^198,208. This seminal IDD study was followed by developing a new SuFEx molecule that integrates fluorosulfate with quinolinic acid or imidazoles as different aryl groups²⁰³. HEK293T cell lysates were exposed to these probes, identifying and validating 11 human proteins with tyrosine or lysine active site residues via mass spectrometry-based proteomics²⁰³. Among this collection of isolated proteins, hydroxysteroid dehydrogenase-like protein 2 was found to bind a small-molecule probe for the first time²⁰³.

The accumulated knowledge acquired from applying SuFEx-based probes in probing the proteome has laid the foundation for establishing P─F-containing probes as warheads. As with SuFEx-based probes, the incoming nucleophile of the target protein residue exchanges with the PFEx probe's fluorine and yields a covalent adduct via a P─N or P─O linkage^42,222. PFEx chemistry has been employed in a biotinylated ABPP approach and successfully labeled multiple active serine hydrolases from cellular and tissue samples²⁰⁶. Additionally, incubation of 14 different proteins with a PFEx-containing probe under in vitro conditions led to the labeling of eight via nucleophilic tyrosine residues, suggesting that the application of PFEx is not limited to serine residues²²³. Given the selective reactivity of FEx moieties, proteome-wide profiling with SuFEx or PFEx molecules may provide an opportunity to achieve comprehensive proteome coverage.

The application of FEx in probe design is accelerating, yet discovering new covalent linkages on diverse proteins requires innovative approaches. Research employing genetically encoded fluorosulfate-l-tyrosine (the first latent bioreactive unnatural amino acid (UAA)) into proteins within E. coli and mammalian cells effectively profile lysine, histidine, and tyrosine residues using proximity-induced labeling with SuFEx chemistry^224,225. A novel workflow using PFEx has been recently developed⁶⁶; P─F-containing UAAs, specifically phosphoramidofluoridate tyrosine and phosphoramidofluoridate lysine , have been incorporated into proteins through genetic encode expansion⁶⁶. The two UAAs were found to react with histidine, tyrosine, lysine, and cysteine residues that were in close proximity⁶⁶. In live cell studies, cytotoxicity remains a big challenge, and in this work, they initially evaluated the cytotoxicity of their molecule using E.coli and HEK-293T cells, which resulted in no obvious toxicity up to 4 mM and 2 mM, respectively⁶⁶. This advancement not only facilitates covalent protein engineering using Nature’s preferred phosphate connectors but also lays the groundwork for PFEx’s wider application in biological and biomedical research, which can deepen our understanding of biological processes and open new therapeutic possibilities.

The use of genetically encoded FEx chemistry does not require external reagents; the biorthogonal chemical moiety is hydrolytically stable, exhibits broad reactivity across various amino acids, and is yet to demonstrate obvious toxicity^66,224,225. To our knowledge, the genetic manipulation approach with FEx has yet to apply to the ABPP or IDD strategies. Looking forward, combining this genetically encoded FEx approach with the ABPP or IDD strategy will help to characterize a broad range of ligandable proteins. Additionally, the synthesis of FEx chemistry-enabled probe libraries combined with advancements in high-throughput proteomic analysis promises to expand the range of FEx ligands profiled in a proteome-wide manner.

Lastly, while the application of ABPP and IDD has expanded to analyze various complex biological systems, but there is a dearth of literature regarding the cytotoxicity of ABPs, which limits their application in live cells or in vivo studies. Moving forward, this area requires robust studies to thoroughly characterize the toxicity of probes. All these efforts contribute to the development of drugs from covalent ligands.

Reproducibility and Data Deposition

Integrating FEx chemistry into DOC processes aims to efficiently synthesize structurally diverse compounds with improved certainty in reaction outcomes. Despite this, meticulous attention to characterization remains indispensable. As with conventional chemical workflows, researchers are expected to report detailed experimental procedures and characterization data as part of the supplementary information accompanying peer-reviewed publications.

Any compounds prepared through a DOC workflow must be accompanied by comprehensive experimental evidence to confirm their molecular structure and guarantee consistency. For previously reported compounds, ¹H and ¹³C NMR analyses, along with appropriate citations, are generally sufficient. In contrast, novel compounds require comprehensive characterization, including ¹H, ¹³C, and heteroatom NMR (e.g., ¹⁹F, ³¹P), infrared spectroscopy (IR), melting or boiling points, and high-resolution mass spectrometry (HRMS). NMR analyses should encompass both one-dimensional (1D) and two-dimensional (2D) experiments (e.g., COSY, HMBC, HSQC, NOESY). While NMR figures are typically provided in supplementary materials, researchers are increasingly encouraged to submit primary data files in JCAMP-DX (.dx, .jdx) format. For chiral molecules, optical rotation measurements offer additional clarity. Furthermore, X-ray crystallographic data stored in repositories such as the Cambridge Crystallographic Data Centre (CCDC) provide unequivocal structural confirmation; it is highly recommended to obtain at least one crystal structure per chemical family synthesized.

We emphasize the importance of obtaining essential characterization data to ensure compound identity and purity before conducting in vitro assays. For high-throughput experiments, complete characterization of all compounds may not be practical before initial screening. Instead, researchers can rely on LC-MS data to confirm product formation and assess conversion rates. Hit compounds identified during screening should be resynthesized to validate reproducibility. At the conclusion of each project, comprehensive characterization of lead compounds is mandatory.

For covalent inhibitors, reporting dissociation constants (K_D) alongside IC₅₀ is required due to the distinct nature of covalent interactions. These inhibitors may be irreversible or reversible, each with unique kinetic characteristics. Irreversible inhibitors are characterized by k_inact, the rate of covalent bond formation, and K_i, the inhibitor concentration at half the maximum rate of covalent bond formation^226,227. Reversible covalent inhibitors additionally have a dissociation rate constant, k_off²²⁸. IC₅₀ values, while useful for comparing potencies, must be clearly defined as they depend on assay conditions and time. Proper characterization of these kinetic parameters can significantly impact the identification, optimization, and selectivity of covalent inhibitors, ensuring strong, specific interactions and optimal pharmacokinetic profiles.

Limitations and Optimizations

Reagent Accessibility

A central aspiration of click chemistry—and, by extension, both FEx and DOC—is to democratize molecular construction¹⁵⁶. Realizing this goal demands reagents that are inexpensive, widely available and operationally simple. In practice, many DOC-compatible building blocks remain non-commercial, expensive, or synthetically laborious. Thionyl tetrafluoride (SOF₄)^53,171, for instance, is the preferred gaseous precursor to iminosulfur oxyfluorides yet is unobtainable in many jurisdictions; its solid surrogate, 2-methyl-imidazole-1-(N-tert-octyl)sulfonimidoyl fluoride (ImSF)²²⁹, can be prepared only through a multi-step route that relies on xenon difluoride (XeF₂)—an exotic and costly fluorinating agent.

Access to versatile SASF hubs, while highly valuable and accessible via a one-step transformation from alkynes, requires handling hazardous sulfur dioxide (SO₂) gas¹⁸. Additionally, the synthesis of specifically substituted aryl alkynes often demands multi-step routes, typically beginning from commercially available aldehydes and requiring at least three steps. To help promote the broader adoption of DOC and FEx technologies, it is essential to address these limitations by developing more scalable, cost-effective routes^118,131 and collaborative efforts between academia and industry could play a vital role in streamlining production pipelines and expanding reagent libraries.

Many catalog reagents can be prepared far more cost effectively in-house. For example, benzene-1,3-disulfonyl fluoride (BDSF) is obtained in two straightforward steps from inexpensive benzene-1,3-disulfonic acid disodium salt¹⁰⁵, while ESF is readily generated from 2-chloroethanesulfonyl chloride¹¹⁸, offering substantial cost savings over commercial supply.

Hub Generation

Constructing highly functionalized DOC hubs can be non-trivial: multistep sequences are sometimes needed, with tight control of regioselectivity and reaction conditions. Key intermediates may be scarce, moisture-sensitive, or hazardous, complicating handling and storage. These hurdles can slow the supply of high-quality hubs and remain a major bottleneck to broader DOC adoption. Additionally, while computational methods can be powerful tools for predicting the physicochemical properties of hubs (and resulting product libraries), care must be taken when interpreting the predicted data. For example, when considering predicted lipophilicity, prior research indicates that some SuFEx-containing molecules exhibit unexpected behavior in thin-layer chromatography (TLC) analysis, indicating differences between predicted and experimental LogP values²⁰. Measuring experimental LogP values of FEx-compatible molecules will be an important step forward.

Biological Evaluation

Despite the diverse molecular structures in DOC-generated libraries, previous FEx screening campaigns have been limited to just a few biological assays. Future functional screening efforts should be as diverse as possible to match the structural diversity in the compound libraries. They will be crucial for further discoveries of molecules with diverse biological functions. Most of the previous screenings have involved FEx libraries that were prepared in-house. The generation of such an in-house FEx library is one of the limiting factors for this approach due to the cost associated with the library's synthesis, purification, and quality control. The commercialization of these libraries will accelerate the screening efforts and thus hit compound identification. We expect the expanded FEx library will soon become available for routine screening campaigns, like other covalent fragment libraries such as cysteine targeting molecules²³⁰.

Outlook

DOC facilitates the rapid generation of structurally diverse molecules with novel properties, opening up new pathways for drug discovery^{18,183,231,232}. This approach, particularly in combination with FEx chemistries^41,42, has the potential to unlock previously unexplored chemical space, providing access to unique compounds that could serve as innovative lead structures in pharmaceutical development and accelerate the identification of new therapeutic options^233,234.

However, several challenges must be addressed to fully unlock the potential of these methodologies in the coming decade. Foremost is the need for continued innovation in the design and synthesis of novel DOC hubs that can efficiently and selectively participate in SuFEx, PFEx and new FEx reactions, which will significantly broaden the scope of accessible chemical space. Additionally, the development of improved catalysts that operate under mild, metal-free, and biocompatible conditions will be invaluable for advancing pharmaceutical applications.

SuFEx chemistry is a foundational element of DOC discovery platforms, and future initiatives should broaden the use of PFEx-based hubs. This expansion will likely include the HTMC application of diverse FEx chemistries, such as DOC and PFEx reactions. Exploring covalent ligands within this framework presents a promising avenue. The development of covalent FEx-enabled HTMC could significantly enhance the optimization and effectiveness of covalent lead molecules and therapeutic candidates. Additionally, the pharmacokinetic and pharmacological impacts of incorporating various FEx moieties remain underexplored. This is highlighted by the unexpectedly high aqueous and metabolic stability of sulfuramidimidoyl fluoride molecules^62,171,210. Prioritizing research into the metabolic stability and in vivo efficacy of FEx-containing covalent bioactive molecules is crucial, particularly for advancing these innovations toward clinical application in human patients.

Asymmetric PFEx chemistry offers considerable promise, but precise stereocontrol remains a formidable hurdle. Developing reliable strategies—whether through chiral catalysts, ligands or auxiliaries—to install and preserve stereogenic phosphorus centers will unlock new classes of PFEx hubs and, in turn, drug candidates with superior therapeutic potential.

Given the robustness of the discussed chemistries, adaption of DOC to automated, machine learning-driven setups that support real-time feedback, process optimization, and rapid iterative cycles in library synthesis and screening, is a logical progression. Enabling technologies, such as continuous flow chemistry, will help facilitate this evolution. In fact, continuous flow chemistry setups have already been used to generate SuFEx electrophile^235,236, facilitate SuFEx ligations, and drive multistep cycloaddition chemistries²³⁷. Integrating these reactions in sequence would yield a powerful DOC discovery engine. Although the need for specialized infrastructure may initially limit the adoption of automated DOC, advances in HTMC—by accelerating hit-to-lead development and enabling direct biological testing at lower cost⁷¹—will help drive broader implementation^62,73,86.

Finally, the development of new FEx chemistries beyond SuFEx and PFEx will help expand the capabilities of DOC. Exploring these new chemistries can introduce more variety into molecular design and synthesis, enhancing the scope of chemical libraries available for research. This extension is an important step in broadening the toolkit of DOC, opening potential new paths for drug discovery and providing fresh approaches to tackling complex biological challenges.

The next decade holds substantial promise for DOC and FEx chemistry in advancing drug discovery and development. Addressing the challenges of reaction efficiency, selectivity, and stereocontrol will be pivotal. Researchers can unlock new chemical spaces and therapeutic possibilities by focusing on developing new reagents, catalysts, and methodologies and integrating these chemistries with orthogonal techniques. The continued evolution of SuFEx and PFEx reactions will lead to innovative applications such as antibody-drug conjugates and theranostic agents, ultimately contributing to more effective and targeted treatments. This outlook underscores the importance of strategic research efforts to harness the full potential of DOC and FEx chemistry in the years to come.

GLOSSARY

Structure-activity relationship

The correlation between a molecule’s structural features (e.g., shape, polarity, functional groups) and its biological activity.

Chemoselective

A chemical reaction that selectively reacts one specific type of functional group in a molecule, while leaving other functional groups unchanged.

Multifunctional connective hub

A molecule containing several distinct chemical groups that can be reacted independently and selectively.

Clickable

A molecule that is compatible with click chemistry reactions.

Polyvalent

In the context of synthetic organic chemistry, a polyvalent molecule contains multiple functional groups or reactive centers.

Pharmacophore

The structural features of a molecule that is essential for its biological activity, especially those that enable interactions with a biological target such as a protein, enzyme, or receptor.

Clicked

When two chemicals are connected together via a click chemistry reaction such as CuAAC or SuFEx.

Hit molecule

A compound identified during the early stages of drug discovery or chemical screening that shows desired biological activity, typically against a specific target.

Bioisostere

A functional group that can replace another in a molecule without significantly altering its biological activity. Often bioisosteres are chosen to offer enhanced stability or bioavailability.

Biocompatibility

Functional groups that are tolerant to and reactive in biologically relevant systems as opposed to organic solvents in the absence of water.

Orthogonal reactivity

Functional groups that react selectively under specific conditions while remaining inert in other instances.

1,3-Dipolar cycloaddition

A reaction where a dipole (e.g., azide) and a dipolarophile (e.g., internal alkyne) combine to form a cyclic adduct.

IC₅₀ (half maximal inhibitory concentration)

The concentration of a bioactive molecule required to inhibit a specific biological or biochemical function by 50%.