Lithium‐Catalyzed Thiol Alkylation with Tertiary and Secondary Alcohols: Synthesis of 3‐Sulfanyl‐Oxetanes as Bioisosteres

Abstract 3‐Sulfanyl‐oxetanes are presented as promising novel bioisosteric replacements for thioesters or benzyl sulfides. From oxetan‐3‐ols, a mild and inexpensive Li catalyst enables chemoselective C−OH activation and thiol alkylation. Oxetane sulfides are formed from various thiols providing novel motifs in new chemical space and specifically as bioisosteres for thioesters due to their similar shape and electronic properties. Under the same conditions, various π‐activated secondary and tertiary alcohols are also successful. Derivatization of the oxetane sulfide linker provides further novel oxetane classes and building blocks. Comparisons of key physicochemical properties of the oxetane compounds to selected carbonyl and methylene analogues indicate that these motifs are suitable for incorporation into drug discovery efforts.

Organosulfur functional groups are often present in pharmaceuticalc ompounds, found in aq uarter of the top 200 drugs (branded drugs by US retail sales in 2011). [1,2] Benzylic sulfides, sulfoxides and sulfones are particularly prevalent, such as in AstraZeneca's blockbuster antiulcerantN exium, Figure 1. [3] Thioester-containingc ompounds have also been disclosed, but are often used as ap ro-drug due to limited metabolic stability; [4] thioesters are 100 more reactive to aminen ucleophiles than esters. [5] However,t od ate there are no suitable bioisosteres for these functional groupst oprovide improved stability at the C-center,l imiting this design space for medicinalc hemists.
Oxetanes have emerged as valuablem otifs in medicinal chemistry that can conferi mproved physicochemical and metabolic properties. [6] Oxetanes can act as suitable polar replace-ment groups for gem-dimethyl linkers and as bioisosteres for carbonylf unctionality. [7] In recent years, oxetane isosteres have been presented for amide, [8] ketone [9] and carboxylic acid derivatives. [10] These developments continuet oa ccelerate the exploration of oxetanes in medicinal chemistry. [11, 6a] We were interested in the potential of 3-sulfanyloxetane derivativesa si sosteres for sulfides or thioesters (Figure1). This little exploredc lass of compounds would offer similar features to thioesters, based on the dipole and lone pair positionoft he oxetane, but withoutt he electrophilic center. Indeed, our DFT studies indicated that while the thioester CÀOb ond length is calculated to be 0.8 shorter than the oxetanyl sulfide, their electrostatic mapping is very similar. [12,13] Additionally,t hese motifsm ay offer ap rotected benzylicc enterf or sulfides or oxidized derivatives. Encouragingly,B ernardes recently reported a mono-substituted 3-sulfanyloxetane as part of am odified protein which was stable under incubation with blood plasma and with glutathione. [14] However, synthetic access to these motifs remains limited, [15][16][17][18] particularly towards 3,3-disubstituted ex- amples; Ellman [19] and Sun [20] reported 3-alkyl-3-sulfanyloxetanes through conjugate addition of as ulfide to oxetane-Michael acceptors.
We envisaged an S N 1p rocess for the formation of oxetane sulfides from 3-aryloxetan-3-ols ( Figure 1). We recentlyr eported aL i-catalyzed Friedel-Crafts reaction using oxetanols to form diaryloxetanes, invoking an oxetane carbocation. [9] The catalytic activation of alcohols through CÀOa ctivationh as become an attractive alternative to replace more toxic alkyl halides. [21,22] However, there are only infrequent examples of catalytic thiol alkylation on functionalized substrates, which often require high catalystl oadings with acidic reagents. [23] Furthermore, ring opening of oxetanes by S-nucleophiles under acidic conditions [24] presentsas ignificant chemoselectivity concern. Here we report ah igh yielding Li-catalyzed alkylation of thiols with oxetanold erivatives. Preliminary data suggest that the methods described can be appliedi nt he synthesis of targets with attractive physicochemical properties for drug discovery.
Buildingo no ur prior studies and with the above considerations in mind, we investigated the alkylation of benzylmercaptan with oxetanol 1 (Table 1).
We optimized the reaction to form oxetane sulfide 2a,a nd to minimize ring-opened side product 3 in which all alkyl CÀO bonds had reacted. The optimized 'standard" conditions used 2equivalentso fb enzylmercaptan and aL ic atalyst in chloroform at 40 8Cf or 25 min. The use of the inexpensive and easily handled salt Li(NTf 2 )( 11 mol %) with Bu 4 NPF 6 (5.5 mol %) as an additive gave a6 7% isolated yield of oxetane sulfide 2a and minimal tri-S-benzylated product 3.C atalystss uch as FeCl 3 , Ca(NTf 2 ) 2 or Ga(OTf) 3 gave lower yields compared to the stan-dard conditions (entries 2-4). Bi(OTf) 3 and TsOH gave no productive reaction (entries 5-6). Using toluenea ss olventw as also successful with slightly reducedy ield and selectivity (entry 7). No reaction occurred in the absence of catalyst or additive (entries 8-9). Reducing the reaction time to 10 ming ave no conversion, due to an activation period for the reaction, believed to involves olubilizing the lithium species, likely the key role of the Bu 4 NPF 6 additive (entry 10). Increasing the reaction time and/or the equivalents of nucleophile resultedi na ni ncreased yield of ring opened 3 (entry 11). Indeed, al onger reaction time of 6hand 6equivalents of benzylthiol,g ave 3 as a single product in ar emarkable 97 %i solated yield (entry 12). [25] This ring openingr eactivity highlights the high initial selectivity achievedw ith the Li catalysti nf orming the putative carbocationic intermediate.
Next, we explored functionalization of the oxetane sulfides. Deprotection of the TIPS group of oxetane sulfide 8,f ollowed by formation of the triflate 14 occurred in high yields, providing ab uildingb lock for furtherr eaction (Scheme 2A). Biaryl 15 was formed by Suzuki-Miyaura cross-coupling in 95 %y ield. [26] The high yield demonstrated the excellent stability of the oxetane sulfide unit to the reactionc onditions. Oxidation of the oxetane sulfides with mCPBA formed selectively 3-sulfinyloxetanes 16 c,d,f or 3-sulfonyloxetanes 17 c,d,f as new oxetane structuralclasses (Scheme 2B).
In the context of drug discovery, the distribution coefficient of ac ompound strongly affects how effectively the drug can reach its intended target, as well as efficacy and pharmacokinetic properties. Hence, LogD is often used by medicinal chemists in pre-clinical drug discoveryt oc onsider the druglikenesso fa ni ntended target molecule. To understand the effect of oxetane sulfides on this keyp arameter,L ogD was measuredf or compounds 21, 2d,a nd 17 d.R eplacing the thioester functionality with the oxetane sulfide had the positive effect of lowering the LogD by approximately 1Log unit ( Figure 2). Furthermore, oxidizing the sulfide to the sulfone further dramatically decreased the lipophilicity of the compound, furnishing 17 d in very favorable propertys pace. In addition, the clearance and cell permeabilityo f20 and 22 were also explored.W hile oxetane 20 displayed slightly lower cell permeability than the corresponding methane analogue 22,t he clearance profile by human liver microsomes of both substrates was determined to be the same, indicating that inclusion of an oxetane in this substrate is not am etabolic liability.The combination of this data, in addition to the improved LogD suggests that previous observations of oxetanes lending improved drug property space extends to these novel oxetane sulfides.
In summary,3 -sulfanyloxetanes offer promising bioisosteric replacement groups for thioesters, and new design elements for medicinalc hemistry.T he first Li-catalyzed thiol alkylation with alcohol substrates is described, suitable for both oxetanol derivatives, andm ore generally with p-activated secondary and tertiary alcohols. Remarkably,c omplete chemoselectivity can be achieved for the activation of the CÀOH group of oxe-  tanols, over competing oxetane ring opening. The use of the mild and inexpensive Li catalyst was crucial and careful control of the reaction conditions gave high yields of the oxetane sulfides. The oxetane sulfides were compatible with palladium catalyzed cross-couplinga nd were converted to sulfoxide, sulfone, and thiol derivatives; themselves providing new classes of oxetane containing compounds. Measurement of key physicochemical properties:L ogD,c learance and cell permeability, indicated that oxetane sulfides and sulfones are attractive for medicinal chemistry applications.