I) Synthesis and Tandem/Sequential Reactions of Main Group Element (Boron, Aluminum and Silicon) Substituted 1,3-Dienes.

A) Present Project Objectives.  

1) Preparation of 2-Main Group Element-Substituted 1,3-Dienes. We have already prepared some boron, silicon and aluminum-substituted dienes. The methods used to prepare these readily available compounds will be extended to additional diene structure types and they will be used in a variety of tandem reactions.

2) Tandem Diels-Alder/Cross-Coupling Reactions of Main Group-Substituted Dienes. We have already demonstrated that boron and aluminum-substituted dienes can affect this type of tandem reaction and will expand these studies to determine the scope and limitations of this new synthetic methodology.

3) Tandem Transmetallation/Diels-Alder Reactions and Sequential Transmetallation/Diels-Alder/Cross Coupling Reactions of Main Group Substituted-Dienes. We have now demonstrated the feasibility of these types of reactions using aluminum substituted dienes. We will also be studying these reactions using boron-substituted dienes. Based on precedent in the literature and new preliminary results, we should be able to affect catalytic, exo selective, and enantioselective Diels-Alder reactions, a long-standing goal of the Welker group.

4)Use of this Methodology for the Construction of Biologically Important Core Structure Targets. While the bulk of the new chemistry proposed above centers on new diastereo- and enantioselective reaction development, this methodology can access biologically significant core structures in the cis clerodane terpenes. These compounds have biological activities ranging from insect antifeedants to biomedical science applications.

B) Results of Prior NSF Support.

We first became interested in preparing transition-metal-substituted h 1 -1,2-butadienyl ( h 1 -allenic) complexes (3) (for use in 3 + 2 cycloadditions) and 2-transition-metal substituted-1,3 butadienes (4) (for use in 4 + 2 cycloadditions). Our group [1] and the Tada group [2] independently reported different routes to cobaloxime dienyl complexes ten years ago. We thought complexes of the general form (3) and (4) should be available via reactions of transition- metal anions (1) with 1,2-butadienyl electrophiles (2) or hydrometalation of the alkyne portion of an enyne (5).

We used those dienyl complex synthesis strategies to prepare a number of cobaloxime (4, L n M =cobaloxime = (pyridine)(dimethylglyoxime) 2cobalt(III)) substituted 1,3-dienes. [3, 4] The cobalt- substituted dienyl complexes are air-stable orange solids that have high thermal stability and can be prepared on a multigram scale from inexpensive materials. The cobalt dienyl complexes reacted with a variety of dienophiles both regioselectively and stereoselectively (with high exo- diastereoselectivity) to produce air-stable cobalt-substituted cyclohexenes in good yield. [5-9] Exo diastereoselectivity improved as glyoxime ligand set size increased, so we rationalized exo- selectivity based on dienophile-cobaloxime ligand set steric interactions. [10] We also developed a number of methods for cleavage of the cobalt-carbon bonds in the cycloadducts, which would yield organic products as well as a cobalt complex that could be recycled into the synthesis of the starting dienyl complexes. [11]

When our previous proposal ( CHE-0104083) was submitted , we were directing a lot of effort on how best to affect enantioselective as well as exo-selective Diels-Alder reactions using stoichiometric amounts of transition metal substituted dienes. We prepared both enantiomers of new dienyl complexes (7) with optically active salen (salen = N, N’-bis salicylideneaminato) ligand sets and used ligand-centered chirality rather than Lewis acid chirality to control the enantioselective outcomes of Diels-Alder reactions (Current Organic Chemistry, 2001, 5, 89-111; European Journal of Org. Chem, 2001, 12, 2273). [3, 12] The dienyl complexes (7) reacted with a variety of dienophiles to produce cycloadducts of very high enantiomeric purity, and the optically active cobalt salen complex used in the synthesis of the dienyl complexes was recovered in high yield and enantiomeric purity.

During the current award period, we have also investigated several other aspects of cobaloxime and cobalt salen chemistry. We explored [3 + 2] cycloaddition reactions of cobaloxime allyl complexes (8) and discovered that cobaloxime allyls react with some electron-deficient alkenes in [3+2] cycloaddition chemistry to produce metal-substituted cyclkopentanes. Others, however, such as benzoquinone (9), react via unexpected allylation or Michael addition/retro Claisen tandem reaction pathways to produce alkoxy substituted phenols (11) (Journal of Organometallic Chemistry, 2002, 656, 217. [13] In 2000, in somewhat related cobaloxime alkenyl chemistry, we reported a new method for preparing cobalt-sp 2 carbon bonds that involved a zinc-mediated coupling of alkenyl halides and triflates to (pyr) 2 (dmg) 2 Co. [14] More recently, we discovered that a coupling product prepared using this methodology, 2-cobaloxime-substituted 1-hexene (14), isomerized readily to (E)-2-cobaloxime-2-hexene (15). The rate constant for this somewhat unusual metal alkenyl isomerization was determined, and the 2-cobaloxime 1-hexenyl complex (15) characterized by X-ray crystallography ( Acta Crystallographica , 2003, C59, m193-m195). [15]

We prepared a water-soluble (aquo)cobaloxime dienyl complex ( 16 ) and developed its Diels-Alder chemistry in water as well as organic solvents (Organometallics, 2004 , 23, 2257). To our knowledge, this is the first report of a transition metal- or main group-substituted diene that can tolerate water as a solvent for a Diels-Alder reaction.

We investigated [6+4] as well as [4+2] cycloaddition chemistry of cobaloxime dienyl complexes with tropones and found that tropones that were unsubstituted at the 2 and 7 positions (such as 18 ) reacted with cobaloxime dienes via [6 + 4] cycloaddition ( 19 ), whereas substitutents at those positions ( 21 ) caused a change to [4 + 2] cycloaddition chemistry ( 22 ). Both types of tropone cycloaddition chemistry proceeded

to produce a single diastereomer ( 19 & 22 ) of the observed products. The [6+4} cycloadditions shown are stereocomplementary to Rigby’s reports of metal-mediated higher order cycloadditions [16] , and the [4+2] cycloaddition chemistry shown provides access to the core structures in the tubulin polymerization-inhibiting estratropones. [17]

17	18	19

20	21	22

We knew that the cobaloxime Diels-Alder chemistry we have worked on would be even more useful to the organic community, if the stoichiometric reaction sequences described above could be affected without isolating and purifying dienes like 4, 7, 17, or 20 and their Diels-Alder cycloadducts. In theory, a catalytic cycle for the reaction sequence can be drawn (Figure 1), and our group has reported all of the individual steps in the cycle- (cobalt hydride generation (1); enyne hydrometallation (2); Diels-Alder reactions of cobalt dienyl complexes with dienophiles to generate metal-susbstituted cycloadducts (3); and reductive demetallation of cycloadducts (4) have been reported previously by our group. [3, 4]

As we developed the cycle proposed in Figure 1, we reported a new method for cobaloxime dienyl complex generation and Diels-Alder cycloadditions that did not require isolating and purifying the cobalt dienyl complexes (23 & 24) (24 isomerizes to 23 faster than it participates in [4+2] chemistry, so generating an E/Z mixture does not matter here) (Journal of Organometallic Chemistry, 2003, 681, 120. [18] We also reported a new silane-mediated cobalt-sp 2 carbon bond cleavage that yielded Diels-Alder cycloadducts (26) and cobalt salts that were converted back into dienyl complexes in high yield. Both of these observations are significant advances over our previously reported procedures.[3, 4] While all three steps of the cycle proposed in Figure 1 can be performed without intermediate characterization and purification, in practice, the chemical yields are higher if just the first two steps of the cycle are run in this manner.

Using our optimized protocol, which relied on enyne hydrometallation for dienyl complex preparation, cobaloxime-substituted Diels-Alder cycloadducts (25) were isolated and purified prior to performing the silane-mediated demetallation reaction. We are now directing our efforts toward developing a cycle that uses transmetallation rather than hydrometallation for initial dienyl complex synthesis and will discuss that work more fully beginning in section V below.

C) Preliminary Results on Current NSF Project Objectives.

• Diels-Alder/Cross Coupling of Boron Dienes.

Potassium organotrifluoroborates were first introduced as alternatives to boronic esters and acids in 1995. [52] Since then, many have reported on their utility and advantages-atom economy compared to boronic acids and esters, ease of purification and disposal, their monomeric rather than trimeric nature, their air stability. [53] Given the reported stability and utility of this class of compounds, we recently set out to prepare the first 1,3-dienyl-2-potassium trifluoroborate. We chose to prepare the butadiene initially and used a route that involved preparing the Grignard reagent of chloroprene (36), [54, 55] followed by its quenching with trimethylborate (B(OMe) 3 ) and aqueous KHF 2 . This new boron-substituted dienyl (37) is a white, air stable solid, as hoped, and shows no propensity to dimerize. It has now been prepared on several gram scale (70% isolated and reproducible), characterized by 1 H, 13 C, 11 B, and 19 F NMR, and appears by NOESY to be predominantly in a solution conformation close to s-trans (38).

We have begun to explore the tandem reaction chemistry of this dienyl potassium trifluoroborate (37). We have run a few tandem Diels-Alder/cross coupling reactions as shown in Table 1, below. We first heated the boron diene (37) and dienophile (so far using common, monosubstituted ones like ethyl acrylate and methyl vinyl ketone), then adding Pd(OAc) 2 (0.5 mol%), 3 eq K 2 CO 3 , and refluxing in EtOH or MeOH for 2 hours. The sequence appears useful for unsubstituted phenyl halides, phenyl halides substituted by electron donating or withdrawing groups, and heteroaromatic halides. The preference for the para over meta regioisomer in these initial experiments ranges from 3 to 4:1. Even in its limited form, Table 1 demonstrates that 37 can serve as a synthon for a host of 2-substituted-1,3-butadienes. We have not worried about regiochemistry here in these early studies, since in our later work we transmetallate to Rh or Pd and we have already demonstrated that low valent transition metal substituted dienes participate in Diels-Alder reactions with excellent regio- and stereoselectivity. [3]

We have also prepared the tetra n-butylammonium (TBA) salt of the BF 3 substituted diene (41). [56] TBA salts of other trifluoroborates have been shown to improve cross coupling yields considerably, presumably due to their greater organic solvent solubility. The bulkier ammonium salt should also increase organic solvent solubility of this class of dienes and may drive their solution conformation more toward s-cis and increase their Diels-Alder reactivity. The effects of this change on Diels-Alder reactivity and regioselectivity remain to be investigated.

Last, we have also prepared the 2-triethoxysilyl-1,3-butadiene (42). We prepared this compound because related aryl trialkoxysilanes are known to have good stability and readily participate in cross coupling reactions. [57] We have not yet started to investigate reaction chemistry of 42.

Table 1. Tandem Diels-Alder/Cross coupling Reactions of Diene 37.

• Directed Hydrometallation/Cross Coupling of Enynols.

Directed hydroalumination and/or hydroboration of the yne portion of enynols appeared to us to be another potentially easy way of accessing 2-boron or 2-aluminum substituted 1,3-dienes. A previous reviewer had noted that attempted hydroboration of 4-methyl-4-penten-2-ynol (43) followed by EtOAc and aqueous KHF 2 quenches looked troublesome and they were correct. These reactions yielded complex mixtures of products where it was clear by 1 H NMR that alkene and alkyne hydroboration was

occurring. [58] However, directed hydroalumination/cross coupling proved much more promising than the hydroboration approach. We first found that methyl pentenynol (43) could be treated sequentially with LAH followed by Pd(II), EtOAc, and aromatic iodides. Hydroalumination/cross coupling products (45) were isolated in good yield as a single stereoisomer from these 1 pot, sequential reactions.

• Directed Hydroalumination/Diels-Alder/Cross Coupling Reactions of Enynols.

We have also started performing sequential rather than just tandem reactions using these enynols as the starting materials. We have performed these preliminary reactions using a readily available mixture of E & Z hexenynol(46). We didn’t worry about using a E/Z mixture rather than pure E or Z since we were just exploring proof of concept reactions. In these initial reactions, the enynol was treated with LAH followed by PdCl 2 (PPh 3 ) 2 (5%) or NiCl 2 (PPh 3 ) 2 (10%), EtOAc, then methyl methacrylate and chloroiodobenzene. When nickel catalysis was used no Diels-Alder/cross coupling products were observed but reduction (50) and Diels-Alder products (48) were isolated. However, when Pd catalysis was employed, the Diels-Alder/cross coupled product (47) was isolated in good yield along with some chloroiodobenzene. A 4:1 mixture of E/Z stereoisomers in 46 leads to a product which is a 5:1 mixture of stereosiomers. The regiochemistry of the Diels-Alder reaction leading to 47 was inferred from the fact that the product had spontaneously formed a lactone. At this point we haven’t tried to determine the stereochemistry of the major and minor isomers of 47, we just know that the reaction gives two products in 5:1 ratio and that they are lactones with molecular weights (GCMS) corresponding to that of 47 and NMR data consistent with the proposed structure. When the order of addition of the last two reagents was reversed, ie the chloroiodobenzene was added first and heated followed by methyl methacrylate, the products and product ratios were substantially different indicating that the Diels-Alder reaction is competitive with or faster than the cross coupling reaction. This order of relative rates will become important later when we propose sequential reactions which could be enantioselective.

Lastly, as new preliminary results in enynol chemistry, we wanted to take a quick look at the possibility of transmetallating from aluminum to zinc prior to Diels-Alder and cross coupling (Alkenyl boranes transmetallate to zinc and addition of 10% chiral amino alcohol ligands enables enantioselective transformations of those reagents, so this also might have some implications for our future enantioselective work. [59] We found that the E/Z hexenynol mixture (46) could be treated with LAH followed by EtOAc and ZnEt 2 prior to heating with methylmethacrylate and hydrolysis to yield 48.

iv) Transmetallation/Enantioselective Diels-Alder/Hydrolysis as a Sequence for Effecting Catalytic Enantioselective and Exo Selective Diels-Alder Reactions.

Figure 1 outlined one approach to catalytic exo and enantioselective Diels-Alder reactions but the demetallation conditions (step 4) required were incompatible with the other 3 steps in the proposed cycle. We thought the metal carbon bond hydrolysis step of Hayashi’s rhodium catalyzed asymmetric 1,4 addition of organoboronic acids to enones may offer a solution. [60-62] We started pursuing this approach (outlined in more detail below), which is based on our proven ability to make stable 2-boron substituted 1,3-dienes and Hayashi’s demetallation conditions.

We have just done some proof of principle reactions here to date. The basic idea is that we want to transmetallate from boron to rhodium, perform a Diels-Alder reaction under conditions where the boron diene is itself unreactive but the rhodium diene (ultimately containing optically active ligand and inducing asymmetry for products formed through exo transition states) is reactive, then hydrolyze the rhodium-carbon bond in that rhodium bonded cycloadduct to generate a metal free cycloadduct and rhodium catalyst.

We took the potassium salt of the BF 3 substituted diene (37) along with either maleimide or N-phenyl maleimide (52a or b) as a dienophile to do this preliminary work. We first determined temperature limits where we should work. We can’t go up to 100 o C (entry 1) because the boron diene itself (37) is reacting at that temperature in the absence of rhodium. We can’t run the reactions at 25 o C for either dienophile (entries 2-5) because in the presence or absence of rhodium we see less than than 5% cycloadduct even after 48h. Rh(COD) 2 OTf as catalyst with N-phenyl maleimide at 50-55 o C looks most promising (compare entries 6 & 8). We certainly haven’t optimized this yet but we isolate cycloadduct (53b) in 62% yield from those conditions and only see a trace of product in the absence of rhodium under these same conditions (entry 7). This proof of principle reaction will be exploited and optimized in the future.

Table 2. Rhodium Catalyzed Transmetallation/Diels-Alder/Hydrolysis Reactions of Maleimides.

[a] Reactions were conducted in sealed tube. [b] heated w/out sealed tube. [c] Determined by 1 H NMR analysis. Catalyst used A = [Rh(COD) 2]BF 4.xH 2O, B= [Rh(COD) 2]OTf, Co-catalyst = S-BINAP

II) Synthesis and Evaluation of New Cancer Chemopreventive Agents.

A. Project Specific Aims. The project’s long term goal is to produce nontoxic cancer chemopreventive agents. A comprehensive cancer treatment strategy will ultimately involve the use of small molecules for both the treatment and prevention of cancer. To date, much more progress has been made in identifying small molecule antitumor agents than small molecule cancer prevention agents. The proposed work helps to close this gap.

Chemoprevention of cancer involves the use of chemical agents either to retard or to block the initiation of carcinogenesis. These agents affect the metabolism of xenobiotic procarcinogens by inducing the enzymes that detoxify potential carcinogens. Typically, phase 1 of xenobiotic metabolism involves oxidative processes, and phase 2, redox or conjugation chemistry. This project will search for chemical agents of low toxicity that elevate phase 2, but not phase1, enzymes as a cancer prevention strategy.

Aim 1 . To prepare and completely characterize new oxathiolene oxides, dithiolene oxides and aminothiolene oxides, using quinone oxidoreductase (NQO1) inducing ability as a guide for the synthesis of new compounds.

Aim 2 . To test the new compounds’ abilities to induce the phase 2 enzyme NQO1 in Hepa1c1c7 cells and to measure their cytotoxicity in that cell line. Compounds that prove excellent NQO1 inducers of low toxicity will also be screened for GST, Ferritin H, and cytochrome P450 inducing activity.

Aim 3 . To observe the reaction between potent inducers and known nucleophilic thiols; to characterize the byproducts of their reaction; and to analyze cell lysates to determine total intracellular levels of inducing compounds and their metabolites. M ost phase 2 detoxification enzyme inducers are thought to work by reacting with nucleophilic thiols. Therefore, study of these reactions will illuminate which structural elements to include in the preparation of more potent enzyme inducers.

Background and Preliminary Results.

Chemoprevention of cancer involves the use of chemical agents either to retard or to block carcinogenesis.[1-10] These agents typically affect the metabolism of xenobiotic procarcinogens and this metabolism proceeds in several phases. In phase 1, procarcinogens are typically oxidized (cytochrome P-450) or reduced and this change many times increases their chemical reactivity. In phase 2, phase 1 metabolites are typically conjugated to biological nucleophiles or electrophiles, such as glutathione (glutathione S- transferases) or glucoronic acid (UDP-glucuronyl transferases). They can also be reduced in oxidation state or hydrolyzed, and rendered less reactive and hence less carcinogenic. In phase 3, phase 2 products are transported out of the cell. As the roles of phase 1 and phase 2 enzymes in initiating and preventing carcinogenesis become better appreciated, the search for nontoxic chemical agents that elevate phase 2 but not phase 1 enzymes intensifies.

A strategy for protection against carcinogenesis as well as other oxidative and electrophilic cellular damage involves inducing the genes which code for the production of the enzymes involved in phase 2 metabolism. The molecular mechanisms involved in phase 2 gene inductions are becoming better understood, as both the structure/activity relationships of known inducers and the molecular events involved in phase 2 gene induction are clarified.

The relative activities of phase 2 enzyme inducers is now most often initially screened using quinone oxidoreductase (NQO1) as the target enzyme in a murine hepatoma cell line (Hepa 1c1c7). This assay is performed on living cells, and thus offers two advantages: (1) it assesses not only phase 2 enzyme induction, but the contribution of cellular metabolic pathways that may enhance or diminish the efficacy of candidate chemopreventive agents; (2) it enables the simultaneous measurement of both phase 2 enzyme inducer activity and the toxicity of candidate agents. The ability to assess toxicity is critical, since lack of toxicity is of major importance in compounds that ultimately will be administered as preventive agents (i.e. to healthy individuals). This screen relies on a simple spectrophotometric measurement of the amount of a reduced tetrazolium dye (3-(4,5-dimethylthiazo-2-yl)-2,5-diphenyltatrazolium bromide, MTT) present; (the dye’s oxidized form is easily reduced by the hydroquinone product of the quinone oxidoreductase enzyme present (Scheme 2). 1,2 Relative strengths of phase 2 inducers are now typically presented via the concentrations of compound required to double the quinone oxidoreductase activity (CD values). The ready availability of mutant Hepa cell lines deficient in cytochrome P-450 activity and the aryl hydrocarbon receptor also makes it easy to screen compounds and determine if they are phase 1 and/or phase 2 (monofunctional or bifunctional) inducers.

Quinone reductase was originally selected as a target enzyme, because its induction coordinates with that of other phase 2 enzymes, and it is ubiquitously distributed in mammalian tissues. 1,2 However, more recent work indicates that NQO1 induction may have even more cancer prevention and treatment relevance than originally thought. Low NAD(P)H:quinone oxidoreductase activity is associated with genetic susceptibility to benzene-induced toxicity and increased risk of leukemia. 3,4 NQO1 is also important in the reductive activation of antitumor agents, such as mitomycin C, and increased levels of NQO1 have been shown to enhance cytotoxicity of mitomycin C toward human tumors. 5,6 Perhaps most important, NQO1 has been shown to bind, and to stabilize against degradation, the tumor suppressor protein, p53. 7-9

Cruciferous vegetables, particularly those of the Brassica genus, contain a number of unusual organosulfur compounds that are excellent phase 2 inducers. 10 A compound containing the dithole-thione nucleus was reported in cabbage in the late 1950s, 11 but a 1991 follow-up study failed to find it. 12 During the 1970s and 1980s, oltipraz (4-methyl-5-(2-pyrazinyl)-3H-1,2-dithiole-3-thione) (1) was being thoroughly investigated as an antischistosomal agent. 13,14 Oltipraz was extremely effective against schistosomiasis and also proved to be an excellent glutathione S-transferase and UDP-glucuronosyl transferase inducer. However, reports of paresthesia and fingertip pain following oltipraz exposure, side effects which were exacerbated by exposure to sunlight, led to the discontinuation of schistosomiasis trials of this compound. The connection between these unusual organosulfur compounds and cancer prevention was nevertheless established by the mid-1980s and led to the identification of a number of other naturally occurring sulfur compounds, such as isothiocyanates (2) and disulfides (3) that function as phase 2 inducers. 10,15-19

An obvious chemoprevention strategy driving this search for dietary phase 2 inducers is to identify nontoxic chemoprotective agents that are already present in the human diet. This strategy may work well in parts of the world where crucifers are widely consumed, but in the United States, broccoli, cabbage, cauliflower, brussel sprouts, kale, etc., are unpopular. 20 Thus, in areas where dietary consumption of cancer chemopreventives is low and/or for populations with significantly elevated cancer risk, searching for synthetic, nontoxic phase 2 inducers has high priority.

1	2	3

The Welker group has synthesized unusual organosulfur compounds particularly oxathiolene- and dithiolene oxides for over 15 years. 21-24 We recognized that these compounds were structurally similar to the five-member rings in dithiol thiones and that the thiolene oxides could participate in both Michael additions and S N2’ reactions with soft nucleophiles, which are chemical characteristics of many of the early phase 2 inducers (Scheme 3). 25,26 The oxathiolene oxide nucleus is significantly easier to access (the SO 2 used in the cycloaddition is commercially available) than the dithiolene oxide nucleus (the S 2O used in the cycloaddition is generated in situ from a molecule we must synthesize 27), so we initially looked at four oxathiolene oxides (that we had quantities of on hand from our methodology studies) as phase 2 inducers. 28 No effort was made to perform an exhaustive structure/activity relationship study; we only wanted an idea of their inducing ability and toxicity, and if they proved to be good phase 2 inducers, were they monofunctional or bifunctional inducers?

Scheme 3.

5	6	7	8

All four of these compounds (5-8) were initially screened for NQO1 inducing ability in the murine hepatoma cell line, Hepa 1c1c7. The CD values shown are concentrations of compound required to double NQO1 levels after 24h of cell treatment with the compounds shown. Oltipraz had a CD value of 6 in our hands in this same cell line, so none of these initial compounds was better, but 8 was close (CD of 32 +10) and had low toxicity (IC 50 >200 mM). Compound 8 also has a calculated log P of 2.21, so it was also predicted to be less lipophilic than Oltipraz (log P = 2.79); lipophilicity has been associated with some of the Oltipraz side effects. 14,29

All four of these compounds were then shown to elevate mRNA levels of GST, NQO1, and ferritin H and L in a normal murine liver cell line, BNLCL.2. 28 Compound 8 was again superior and induced an 8.2 fold increase in GST a mRNA and a 18.7 fold increase in NQO1 mRNA at the maximum concentration tested (160 mM). Ferritin H and L mRNA were induced by 1.9 and 1.8 fold, respectively, by this same dose. To verify that induction at the mRNA level was translated into increases at the protein level, compound 8’s ability to increase total GST activity was also evaluated in this cell line. Increased GST levels were seen after 24h of treatment and sustained up to 72h thereafter. The elevated GST protein and enzyme activity seen was also shown to be due to preferential elevation of GST- a rather than the m or p class GST isozymes. Similarly, ferritin H protein levels (in addition to the mRNA levels) were verified to be elevated by 24h treatment with compound 8 (100 mM) in this same cell line. Lastly, compound 8 was shown to be a monofunctional inducer (no cytochrome P450 1A1 induction) and an aromatic hydrocarbon (Ah) receptor independent inducer in Hepa 1c1c7 and mutant Hepa 1c1c7 cell lines, respectively. All NQO1 and ferritin H and L induction screens were performed in the laboratories of Dr Suzy Torti (WFUSM) and the GST induction screens were performed in the laboratories of Dr. Alan Townsend (WFUSM).

Having verified that compounds containing the oxathiolene oxide nucleus could be nontoxic, monofunctional phase 2 inducers at both the mRNA and protein levels, we next evaluated the oxathiolene oxide and dithiolene oxide nucleus as phase 2 inducers in more detail. We developed a nontransition-metal based synthetic route to the oxathiolene oxides which proved convenient and general and prepared about 30 new oxathiolene oxides which were screened as NQO1 inducers. NQO1 screens were again performed in the laboratory of our WFUSM collaborator, Dr. Suzy Torti. The results of this work were published earlier this year. 30 Our most active compounds 20 with R 3 = heteroaromatic or heteroatom substituted aromatic and R2 = vinyl had CD values below 10 mM but IC 50s in the 50-100 mM range. We suspect the toxicity arises from the oxidizable exocyclic vinyl group but we continue structure activity relationships in the search for small molecule cancer chemopreventives.

Preliminary Structure Activity Relationship Data for Oxathiolene Oxides.

20

Preliminary conclusions/hypotheses are:

1) 5 = heteroatom substituted or containing aromatics look promising;

2) 4 = vinyl is good for inducing activity but bad for toxicity;

3) 4 = methyl is good for toxicity but hurts by a factor of 3 on inducing activity;

4) Sulfur oxidation state may have little bearing on activity; and

5) Substitution at the 3 position has little effect on activity.

Recent Publications.

56) [6 + 4] and [4 + 2] Cycloaddition Reactions of Cobaloxime 1,3-Dienyl Complexes and Tropones. Ramakrishna R. Pidaparthi, Mark E. Welker,* and Cynthia S. Day, Organometallics, 2006, 25, web released as an ASAP article 1/5/06.

55) Preparation of 2-BF 3-Substituted 1,3-Dienes and Their Diels-Alder/Cross Coupling Reactions. Subhasis De and Mark E. Welker,* Organic Letters, 2005, 7, 2481-2484.

(54) Oxathiolene Oxide Synthesis Via Chelation Controlled Addition of Organometallic Reagents to Alkynols Followed by Addition of Sulfur Electrophiles and Evaluation of Oxathiolene Oxides as Anticarcinogenic Enzyme Inducers. Marion A. Franks, Edward A. Schrader, E. Christine Pietsch, Daniel R. Pennella, Suzy V. Torti,* and Mark E. Welker,* Bioorganic and Medicinal Chem., 2005, 13, 2221-2233..

(53) Preparation of an (Aquo)Cobaloxime 1,3-Dienyl Complex and Its Diels-Alder Reactions in Water and Organic Solvents. Carmen J. Tucker, Mark E. Welker,* Cynthia S. Day, and Marcus W. Wright, Organometallics , 2004 , 23, 2257-2262.

(52) Simple Preparation of Cobaloxime Dienyl Complexes and Their Exo Selective Diels-Alder Cycloadducts. Progress Toward Transition Metal Mediated Diels-Alder Reactions Which are Catalytic in Metal Dienyl Complex. Kerry A. Pickin, Jennifer M. Kindy, Cynthia S. Day, and Mark E. Welker, J. Organometal. Chem.2003 , 681, 120-133.

(51) Preparation of Substituted Transition-Metal h 1 -Propargyl Complexes and their 3 + 2 Cycloaddition Reactions with Sulfur Dioxide and Disulfur Monoxide. Transition-Metal-Carbon Bond Cleaving Reactions of the Cycloadducts Which Yield Oxathiolene Oxides and Dithiolene Oxides. Elizabeth . E. Scott, Erling. T. Donnelly, and Mark E. Welker, J. Organometal. Chem ., 2003, 673, 67-76.

(50) A Cobaloxime Substituted Terminal Alkene Which Rapidly Isomerizes to an Internal Alkenyl Cobaloxime Complex. K.A.P. Pickin, C.S. Day, M.W. Wright, M.E. Welker, Acta Crystallographica, 2003, C59, m193-m195.

(49) Preparation and Characterization of Trispyrazolylborate Molybdenum Dicarbonyl ( h 3 -Dienyl complexes. D.R. Lantero, A.G. Glenn, C.S. Day, M.E. Welker, Organometallics, 2003 , 22, 1998-2000.

(48) Novel Phase II Enzyme Inducers: Oxathiolene Oxides. E.C. Pietsch, A. Hurley, E. Scott, B.P. Duckworth, M.E. Welker, S. Leone-Kabler, A. Townsend, F. Torti, S. Torti, Biochemical Pharmacology, 2003, 65, 1261-1269.