In vivo structure-activity relationship of dihydromethysticin in reducing NNK-induced lung DNA damage against lung carcinogenesis in A/J mice

Abstract

Lung cancer is the leading cause of cancer-related deaths and chemoprevention should be developed. We recently identified dihydromethysticin (DHM) as a promising candidate to prevent NNK-induced lung tumorigenesis. To probe its mechanisms and facilitate its future translation, we investigated the structure-activity relationship of DHM on NNK-induced DNA damage in A/J mice. Twenty DHM analogs were designed and synthesized. Their activity in reducing NNK-induced DNA damage in the target lung tissues was evaluated. The unnatural enantiomer of DHM was identified to be more potent than the natural enantiomer. The methylenedioxy functional moiety did not tolerate modifications while the other functional groups (the lactone ring and the ethyl linker) accommodated various modifications. Importantly, analogs of high structural similarity to DHM with distinct efficacy in reducing NNK-induced DNA damage have been identified. They will serve as chemical probes to elucidate the mechanisms of DHM in blocking NNK-induced lung carcinogenesis.

Graphical Abstract

The structure-activity relationship of DHM on NNK-induced DNA damage in A/J mice was investigated, which not only identified more potent analogs but also developed in vivo chemical probes for future mechanistic characterization of DHM in preventing lung carcinogenesis.

Introduction

Over two million individuals are diagnosed with lung cancer each year worldwide.^[1] Compared to many other cancers, the clinical outcome of patients with lung cancer is much worse: the five-year overall survival remains to be at or below 15 – 20% even with the recent introduction and significant progress in targeted therapies and immunotherapies. Lung cancer thus causes more deaths than any other cancer, with estimated 1,800,000 deaths each year. Alternative strategies therefore are needed in order to improve lung cancer management. Tobacco use is well established as the leading etiological cause of lung cancer, contributing to 80 – 90% of lung cancer risk. Controlling tobacco exposure is the most effective approach in reducing lung cancer risk and incidence. Due to the addictive nature of nicotine in tobacco products, the success in tobacco control and tobacco cessation, however, has been limited.^[2–5] Globally, over one billion individuals actively use tobacco products and the prevalence is increasing in certain parts of the world. Lung cancer chemoprevention therefore should be considered as an important pillar to reduce lung cancer incidence.

For the practical translation of lung cancer prevention, a chemopreventive agent needs to be safe for long-term human use. Candidates with a historical and safe human exposure are of particular advantage. The preparation of the chemopreventive candidates needs to be economical as well so that the cost will not be prohibitive. In addition, individuals with higher risk of lung cancer need to be identified or at least enriched among smokers or former smokers because not all of them would develop clinical lung cancer during their life span and chemoprevention may not be necessary for such individuals. Lastly, the mechanism(s) of the chemopreventive candidate need to be elucidated in details. Such mechanistic insight will not only help identify high-risk individuals, who are more likely to benefit from the chemopreventive intervention but also offer the opportunity of timely monitoring the chemopreventive efficacy. In our effort to develop solutions towards these challenges, we demonstrated kava as a promising source for lung cancer chemopreventive agents.^[6–8] Kava is a beverage consumed among the South Pacific Islanders to help relax and improve the quality of sleep with a long history of chronic use.^[9–14] Despite its disputable hepatotoxic risk, the WHO and many others have reviewed the reported kava safety data and concluded that kava’s purported hepatotoxic risk is rare if any (< 0.3 cases per one million daily doses), particularly with the right cultivar and the traditional preparation.^[15–20] Kava has also been on the US market as a dietary supplement for decades to improve relaxation.^[21]

We recently identified dihydromethysticin (DHM, Fig. 1) as the lead compound in kava in blocking 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanone (a tobacco specific lung carcinogen, commonly known as NNK)-induced lung tumor formation in A/J mice.^[22] Interestingly, dihydrokavain (DHK, Fig. 1) was completely ineffective even though it is structurally similar to DHM. NNK-induced DNA damage has been established as one root cause of its lung carcinogenicity.^[23] We found that DHM preferentially reduced DNA damage induced by 4-(methylnitrosamino)-1-(3-pyridyl)-1-butanol (the major in vivo metabolite of NNK, commonly known as NNAL) with minimal effects on DNA damage induced by NNK (Fig. 2).^[22] Since the inactive DHK was not able to modulate such DNA adducts, reducing NNAL-induced DNA damage appears to be an essential mechanism for DHM in blocking lung carcinogenesis.^[22,24,25]

Figure 1.

The structures of natural DHM, natural DHK, and kavalactone numbering.

Figure 2.

Proposed mechanism of DHM in differentially reducing NNK/NNAL-induced DNA damage.

Interestingly, our recent time-course mapping data of DHM on NNAL-induced DNA damage and lung adenoma formation suggest that DHM likely inhibited NNK-induced lung carcinogenesis via DNA damage-driven as well as DNA damage-independent mechanisms (Fig. 3).^[24] Specifically, DHM pre-exposure (gavage 16h before NNK) reduced NNK-induced tumor multiplicity by 98% while it only reduced NNK-induced DNA adducts (O⁶-methylguanine, O⁶-mG) by 75%. Furthermore, DHM (gavage 40h before NNK) reduced tumor multiplicity by 52% with NO significant reductions in O⁶-mG while DHM and NNK concurrent exposure reduced tumor multiplicity by only 50%, even though O⁶-mG was reduced by 63%. These data overall demonstrate that the extent of DNA damage reduction by DHM did not perfectly correlate with its suppression of lung tumor formation. Indeed phenethyl isothiocyanate (PEITC) and indole-3-carbinol have been reported to more effectively inhibit NNK-induced DNA damage in A/J mice than DHM while their efficacy in blocking NNK-induced adenoma formation was not as efficient as DHM,^[26,27] further supporting that DNA damage is not the sole mechanism for NNK-induced lung carcinogenesis.

Figure 3.

In vivo data indicating DNA damage-independent mechanisms of DHM in blocking NNK-induced lung tumor formation. Adapted with permission from our earlier work.^[24]

In order to characterize the detailed mechanism for DHM in blocking NNK-induced lung carcinogenesis, we have designed and synthesized twenty DHM analogs with structural modifications on key functional groups and characterized their efficacies on reducing NNK-induced lung DNA damage in A/J mice. The results determined the structural moieties on DHM that are important in reducing DNA damage and identified structure-similar analogs of distinct efficacies. These DHM-based compounds are expected as powerful in vivo chemical probes to help elucidate the molecular basis of DHM in blocking NNK-induced lung carcinogenesis, particularly in defining the contributions of NNK-induced DNA damage-driven vs. DNA damage-independent mechanisms.

Results and Discussion

Synthesis of dihydromethysticin (DHM) and rational design of its analogs

We have previously developed facile and efficient synthetic routes to obtain kavalactones.^[25,28] Such synthetic routes were employed to synthesize racemic DHM and its analogs with slight modifications (Scheme 1). Briefly, aryl aldehydes (1) with various substituents were subjected to a standard Wittig reaction to generate the aryl alkene intermediates (2). The alkenes (2) were oxidized to the corresponding aldehydes (3). Aldehydes (3) were subjected to dianion alkylation with ethyl acetoacetate, resulting in the intermediate aldol products, followed by K₂CO₃ mediated lactonization and methylation using dimethyl sulfate, affording compounds (4). Palladium-catalyzed hydrogenation of (4) led to DHM and its analogs (5). This synthetic route can accommodate various arylaldehydes, leading to analogs 5a – 5j in excellent overall yield (Table 1). These analogs have modifications on the aromatic ring of the kavalactones, because the only difference between DHM and DHK is on the aromatic ring (Fig. 1), which results in significant differences in their efficacy in blocking NNK-induced DNA damage and adenoma formation. The synthetized compounds are racemates. The natural (+)-DHM and unnatural (−)-DHM enantiomers were obtained through chiral resolution of the synthetic racemic DHM.

Scheme 1.

The general synthetic route for racemic DHM and its analogs.

Table 1.

DHM analogs with modifications on the aromatic ring of kavalactones and their activities in reducing NNK-induced O⁶-mG in the target lung tissues in A/J mice.

No.	R1	R2	R3	R4	O6-mG% (mean ± SD)	p-value
(±)-DHM	H	-OCH2O-		H	34 ± 7	****
KA-8 (5a)	H	-OCH2CH2O-		H	126 ± 11	NS
KA-82 (5b)	H	-CH2CH2O-		H	82 ± 17	NS
KA-110 (5c)	H	-OCH2CH2-		H	114 ± 20	NS
KA-108 (5d)		-OCH2O-	H	H	68 ± 3	NS
KA-69 (5e)	H	-OCD2O-		H	75 ± 11	NS
KA-25 (5f)	H	-OCF2O-		H	108 ± 9	NS
KA-10 (5g)	H	OH	OCH3	H	97 ± 23	NS
KA-76 (5h)	H	OCH3	OH	H	136 ± 8	*
KA-81 (5i)	H	OCH3	OCH3	H	116 ± 22	NS
KA-120 (5j)	H	OH	OH	H	93 ± 6	NS

One-way ANOVA was used to compare the values among NNK and treatment groups. Dunnett test was used for comparisons of the quantity between NNK control and treatment groups. p value < 0.05 was considered statistically significant.

^*p < 0.05

^**p < 0.01

^***p < 0.001

^****p < 0.0001.

NS: not significant. n = 2 for 5d.

Employing similar synthetic routes, other analogs were synthesized to explore the linker and the lactone functional groups on DHM (Fig. 4). Briefly, subjecting racemic DHM to hydrogenation generated the saturated analog 5k. The same racemic DHM, upon oxidation with DDQ, resulted in analog 5l. 5k and 5l offer the opportunity to explore the importance of the oxidation state of DHM, which also affects the conformation and flexibility of these compounds. By incorporating different alkylating agents in the final step, analogs 5m – 5n were obtained, which could explore the steric tolerance at the fourth position. Lastly, a methyl walk was employed to generate analogs (5o – 5r), exploring the potential tolerance of modifications at the 3, 5, 6, and 7 positions of the kavalactone system.

Figure 4.

Structures of DHM analogs 5k – 5r.

Effects of DHM analogs on NNK-induced O⁶-mG adduct in the lung tissues in A/J mice

Bioactivation is essential for NNK and NNAL to induce DNA damage and there are no validated biochemical or cell-based assays that faithfully model these in vivo complex processes to evaluate lung cancer chemopreventive agents. We therefore employed the A/J mouse model herein to explore the SARs of DHM analogs. Given the low-throughput nature of this in vivo assay, DHM and its analogs were evaluated on four different days. An NNK control group was included in each day. In order to compare the results from different days, the levels of O⁶-mG in each treatment group were normalized to that of the NNK group of the same day.

Dose-response comparison of DHM enantiomers

We first evaluated the dose response of the natural (+)-DHM and unnatural (−)-DHM to explore the importance of the chirality in reducing NNK-induced DNA damage. The dosages chosen were 0.10, 0.20, and 0.40 mg/mouse, which approximated to a dose of 5, 10 and 20 mg/kg of bodyweight respectively for a mouse of 20 g in bodyweight. Such a dose range was based on our previous data showing that natural (+)-DHM, at a dose of 0.80 mg/mouse, reduced O⁶-mG by 90% under the same treatment regimen.^[24] We rationalized that the selected dose range between 0.10 – 0.40 mg/mouse would offer a good opportunity to detect efficacy differences. Briefly, compounds were administered to A/J mice via gavage one hour before NNK exposure, which was given via intraperitoneal (i.p.) injection. Mice were euthanized four hours later. We focused on quantifying O⁶-mG in the lung tissues of A/J mice initially because it is the most carcinogenic DNA damage type in A/J mice and its abundance is much higher than the 4-(3-pyridyl)-4-oxobut-1-yl (POB) and 4-(3-pyridyl)-4-hydroxobut-1-yl (PHB) adducts.^[24,25] As shown in Fig. 5, dose-response relationships were observed for both compounds. (−)-DHM were much more potent than (+)-DHM in reducing O⁶-mG, suggesting that the unnatural enantiomer may be more effective in lung cancer prevention than the natural (+)-DHM.

Figure 5.

The dose-response results of natural (+)-DHM and unnatural (−)-DHM on O⁶-mG in A/J mouse lung DNA. One-way ANOVA was used to compare the values among NNK and treatment groups. Dunnett test was used for comparisons of the quantities between NNK and treatment groups. p value < 0.05 was considered statistically significant. *: p < 0.05; ****: p < 0.0001.

Structure-activity relationships of DHM analogs

The dose-response results of DHM suggest that the dose of 0.20 mg/mouse should be reasonable to evaluate DHM analogs, allowing room for detection of either efficacy enhancement or attenuation. As shown in Table 1, modifications on the methylenedioxy function group of DHM (5a – 5j) all result in significant or complete loss of activity in reducing NNK-induced DNA damage. Specifically, expanding the methylenedioxy functional group by one methylene carbon (5a) resulted in complete loss of activity. Replacing either of the oxygen atoms in the methylenedioxy functional group with a carbon atom resulted in significant (5b) or complete (5c) loss of activity. Moving the position of the methylenedioxy functional group also significantly reduced the activity (5d). Even replacing the hydrogen atoms of the methylene with the deuterium atoms (5e) resulted in significantly compromised activity, while replacing the hydrogen atoms with the fluorine atoms (5f) resulted in complete loss of activity. Given the similarity in electronic properties and size between fluorine and hydrogen, 5f could be a promising chemical probe to help define the DNA damage-independent mechanism of DHM in preventing lung carcinogenesis. Analogs 5g - 5j were designed to be able to generate similar metabolites of DHM. None of them were active in vivo. In combination with the results from 5e and 5f, the active form of DHM in reducing DNA damage may be a metabolic intermediate, such as carbene, that functions as a mechanism-based inhibitor, which will be investigated in the future.

The reduced and oxidized form of DHM, 5k and 5l respectively, also lost the activity (Table 2), potentially because the lactone ring in 5k may be too flexible while 5l may be too rigid. On the other hand, functional groups on the lactone ring of DHM tolerates modifications. First of all, the 4-methoxy functional group can be modified that an n-propyl analog (5m) has similar activity as (±)-DHM and even the isopropyl analog (5n) had some activity. The four analogs with methyl at the 3, 5, 6 and 7 positions (5o – 5r) all showed comparable activity as (±)-DHM. Future systematic modifications at these positions will be explored in the future. It also should be noted that analogs 5p and 5r evaluated herein could be mixtures of diastereomers since the synthesis is not expected to be stereospecific. Future effort is needed to separate the isomers to determine whether specific stereoisomers may be more potent, particularly for 5p, which has the best activity among the DHM analogs evaluated.

Table 2.

DHM analogs with modifications at the lactone ring of kavalactones and their activities in reducing O⁶-mG in the target lung tissues in A/J mice.

	O6-mG% (mean ± SD)	p-value
(±)-DHM	34 ± 7	****
KA-84 (5k)	97 ± 18	NS
KA-83 (5l)	97 ± 7	NS
KA-106 (5m)	50 ± 11	**
KA-112 (5n)	81 ± 11	NS
KA-66 (5o)	48 ± 6	****
KA-65 (5p)	31 ± 12	****
KA-67 (5q)	48 ± 6	****
KA-109 (5r)	53 ± 14	**

One-way ANOVA was used to compare the values among NNK and treatment groups. Dunnett test was used for comparisons of the quantity between NNK control and treatment groups. p value < 0.05 was considered statistically significant.

^*p < 0.05

^**p < 0.01

^***p < 0.001

^****p < 0.0001

NS: not significant.

Differential effects of DHM analogs on NNK vs. NNAL-induced DNA damage

Lastly, since DHM preferentially reduces DNA damage derived from NNAL over that from NNK (Fig. 2),^[22] the top analogs were evaluated for their effects on DNA damages induced by NNK and NNAL respectively. As shown in Scheme 2, NNK-induced 4-(3-pyridyl)-4-oxobut-1-yl (POB) DNA damage results in the formation of 4-hydroxy-1-(3-pyridyl)-1-butanone (HPB) upon hydrolysis while NNAL-induced 4-(3-pyridyl)-4-hydroxobut-1-yl (PHB) DNA damage results in the formation of a corresponding 1-(3-pyridyl))-1,4-butanediol (Diol) upon hydrolysis.^[29] We therefore determined the effects of DHM analogs on NNK and NNAL-induced DNA damages via indirect quantification of HPB and Diol (Table 3). For both HPB and Diol, (−)-DHM resulted in the most significant reductions followed by (±)-DHM, 5p, and (+)-DHM, which overall parallels with their relative potency in reducing O⁶-mG. As expected, the extent of reductions in Diol by these compounds is much higher than the reductions in HPB. DHM analogs therefore also preferentially reduced NNAL-induced DNA damage in comparison to NNK-induced DNA damage.

Scheme 2.

The formation of HPB and Diol from the POB and PHB DNA adducts respectively.

Table 3.

The effects of active DHM analogs on HPB and Diol.

	HPB% (mean ± SD)	p-value	PBD% (mean ± SD)	p-value
(+)-DHM	112 ± 20	NS	86 ± 8	NS
(±)-DHM	71 ± 3	*	38 ± 7	***
(−)-DHM	46 ± 16	***	29 ± 14	****
5p	98 ± 18	NS	64 ± 18	**

One-way ANOVA was used to compare the values among NNK and treatment groups. Dunnett test was used for comparisons of the quantity between NNK control and treatment groups. p value < 0.05 was considered statistically significant.

^*p < 0.05

^**p < 0.01

^***p < 0.001

^****p < 0.0001

NS: not significant.

Our early lab animal and pilot human trial data strongly support the potential of kava in reducing lung cancer risk with DHM as an active ingredient. To facilitate its future translation, detailed mechanisms of action are needed. Whereas our prior studies have shown that reducing NNK/NNAL-induced DNA damage is one key mechanism for DHM, DNA damage-independent mechanisms and its contribution, as revealed in our time-course mapping study, remain to be elucidated. The results described above represented work in progress to achieve a better understanding of the mechanisms of DHM through an SAR approach. The distinct SAR results of DHM analogs on DNA damage overall suggest that DHM likely interacts with the molecular target(s) in a narrow and well-defined binding pocket(s) to introduce its in vivo DNA-adduct attenuating activity. Interestingly the unnatural (−)-DHM enantiomer appeared to be more potent than the natural (+)-DHM. More importantly, several analogs, structurally similar as DHM, have been discovered with distinct potency in reducing DNA damage, particularly (−)-DHM and 5f. Although their pharmacokinetics could be potential variables, we have previously characterized that DHM and dihydrokavain have similar pharmacokinetics,^[30] suggesting that pharmacokinetics unlikely accounts for their distinct SARs observed herein. Upon confirming their similar pharmacokinetics in the future, these DHM analogs will be powerful in vivo chemical probes to define DNA damage dependent and independent mechanisms in NNK-induced lung cancer chemoprevention, which is critical to the basic understanding of NNK-induced lung carcinogenesis and DHM-based chemoprevention.

Experimental Section

Caution:

NNK is a potential human carcinogen and should be handled carefully in well-ventilated fume hoods with proper protective clothing.

Chemicals, reagents and animal diets

The AIN-93G powdered diet was purchased from Harlan Teklad (Cambridgeshire, UK) and stored at 4°C. NNK, [CD₃]O⁶-mG, 4-Hydroxy-1-(3-pyridyl)-1-butanone (HPB), D₄-HPB (4-hydroxy-1-(3-pyridyl)-1-butanone [3,3,4,4-D₄]), Diol, and D₄-Diol (1-[3-Pyridyl-d4]-1,4-butanediol) were purchased from Toronto Research Chemicals (Toronto, ON, Canada). O⁶-methylguanine (O⁶-mG) was purchased from Midwest Research Institute (Kansas City, MO).

General procedures of synthesis

All solvents, reagents and starting materials, including anhydrous solvents and chemicals, were purchased from commercial vendors, and used without any further purification or distillation unless otherwise stated. Analytical thin layer chromatography was performed on Whatman silica gel 60 Å with fluorescent indicator (Partisil K6F). Compounds were visualized by UV light and/or stained with potassium permanganate solution followed by heating. Flash column chromatography was performed on Whatman silica gel 60 Å (230–400 mesh). NMR (¹H and ¹³C) spectra were recorded on a Bruker AVANCE NEO 500 MHz NMR spectrometer with N₂-Cryo-Platform for PRODIGY probe and calibrated using an internal reference. Chemical shifts (δ values) and coupling constants (J values) are given in ppm and Hz, respectively, using tetramethylsilane (TMS) as an internal standard. High-resolution mass spectrometer data were acquired. The final structures were characterized by ¹H NMR, ¹³C NMR and HRMS. All compounds synthesized are racemic mixtures with purity analyzed based on HPLC analysis with purity ≥ 95%.

General synthetic procedures. The synthesis for 5a is described herein, which can be applied to the other DHM analogs. Briefly, ethyltriphenylphosphonium bromide (1.0 equiv.) was dissolved in dry THF and stirred at 0 °C under nitrogen atmosphere for 10 minutes. nBuLi (1.1 equiv.) was added dropwise and the reaction was stirred at same temperature for 30 minutes. The corresponding aldehyde (1, 1.0 equiv., dissolved in dry THF) was added slowly and the reaction was brought to room temperature after 30 minutes. Once TLC confirms the consumption of the starting aldehyde, water was added and the organic layer was extracted with ethyl acetate. The organic layer was dried using MgSO₄ and evaporated to dryness to get the crude alkene, which was purified by silica gel column chromatography using increasing percentage of ethyl acetate in hexane as an eluent to get the pure alkene (2). The corresponding alkene (2, 1.0 equiv.) was dissolved in 1,4 dioxane. Selenium dioxide (1.0 equiv.) was added in the reaction mixture and refluxed overnight. Once TLC confirms complete consumption of the alkene, the reaction mixture was passed through celite bed and washed with ethyl acetate. The organic part was evaporated to dryness to get the crude aldehyde, which was purified by silica gel column chromatography using increasing percentage of ethyl acetate in hexane as an eluent to get the pure aldehyde (3). Sodium hydride (2.0 equiv.) was dissolved in dry THF and stirred at 0 °C under nitrogen atmosphere for 10 minutes. Ethyl acetoacetate (1.0 equiv.) was added dropwise and stirred the reaction mixture at the same temperature for 30 minutes. nBuLi (2.0 equiv.) was added dropwise at the same temperature and the reaction was stirred for another 30 minutes. Then the reaction was brought to −78°C and the corresponding aldehyde (3, 1.0 equiv., dissolved in dry THF) was added slowly and the reaction was brought to room temperature after 30 minutes. Once TLC confirms the consumption of the starting aldehyde, saturated solution of NH₄Cl was added in the reaction mixture and the organic layer was extracted with ethyl acetate. The organic layer was dried using MgSO₄ and evaporated to dryness to get the crude reaction mixture, which was used for the next step without further purification. The mixture was dissolved in dry methanol and K₂CO₃ (2.0 equiv.) was added. The reaction mixture was stirred at room temperature under nitrogen atmosphere until TLC confirms the complete consumption of the intermediate. Then methanol was removed and the crude mixture was dissolved in dry acetone and the corresponding dialkyl sulphate (2.0 equiv.) was added. Then the reaction mixture was stirred at room temperature for another 12 hours. Water was then added in the reaction mixture and the organic layer was extracted with ethyl acetate. The organic layer was dried using MgSO₄ and evaporated to dryness to get the crude reaction mixture, which was purified by silica gel column chromatography using increasing percentage of ethyl acetate in hexane as an eluent to get the pure compound 4. Compound 4 was then dissolved in dry THF and 5 % Pd/C (0.1 equiv.) was added. The hydrogenation reaction was performed in presence of 1 atom hydrogen balloon and stirred for 2 hours. After that, the reaction mixture was passed through celite bed and washed with ethyl acetate. The organic part was evaporated to dryness and purified by silica gel column chromatography using increasing percentage of ethyl acetate in hexane as an eluent to get the final compound (5). For 5k, DHM was dissolved in dry THF and 5 % Pd/C (0.3 equiv.) was added in the reaction mixture. The reaction was performed in presence of hydrogen balloon and stirred for 8 hours. After that the reaction mixture was passed through celite bed and washed with ethyl acetate. The organic part was evaporated to dryness to get the pure compound 5k. For 5l, methysticin was dissolved in dry benzene and DDQ (1.5 equiv.) was added in the reaction mixture. The reaction mixture was refluxed for 3 hours. Then the reaction mixture was evaporated to dryness and purified by silica gel column chromatography using increasing percentage of ethyl acetate in hexane as the eluent. For the synthesis of 5m and 5n, intermediate before methylation reaction was dissolved in dry DMF (1.0 equiv.) and consecutively K₂CO₃ (3.3 equiv.) and the corresponding alkyl bromide (3.3 equiv.) was added. The reaction mixture was stirred at 60 °C for 6 hours. After that water was added in the reaction mixture and the organic layer was extracted with ethyl acetate. The organic layer was dried using MgSO₄ and evaporated to dryness to get the crude reaction mixture, which was purified by silica gel column chromatography to obtain the pure compound.

To synthesize compound 5o and 5p, instead of ethyl acetoacetate, corresponding β-keto ester ethyl 2-methylacetoacetate and ethyl propionylacetate respectively were used. For compound 5q and 5r, commercially available carbonyl compound was used as Michael acceptor and subsequent Dieckmann condensation furnished six membered cyclic diketones which underwent alkylation to yield the final compounds.

6-(2-(2,3-dihydrobenzo[b][1,4]dioxin-6-yl)ethyl)-4-methoxy-5,6-dihydro-2H-pyran-2-one (5a)

Yellow solid, yield 36%; ¹H-NMR (CDCl₃, 500 MHz): 6.75 (d, J = 10.0 Hz, 1H), 6.68 (s, 1H), 6.65–6.63 (m, 1H), 5.11 (s, 1H), 4.34–4.33 (m, 1H), 4.22 (s, 4H), 3.72 (s, 3H), 2.74–2.71 (m, 1H), 2.682.63 (m, 1H), 2.50–2.44 (m, 1H), 2.30–2.26 (m, 1H), 2.08–2.02 (m, 1H), 1.87–1.82 (m, 1H); ¹³C-NMR (CDCl₃, 125 MHz): 172.7, 167.3, 143.3, 141.8, 134.0, 121.3, 117.1, 117.0, 90.3, 74.7, 64.3, 64.2, 55.9, 36.3, 32.9, 30.1. HRMS (ESI+) m/z calcd for C₁₆H₁₉O₆ [M+H]⁺, 291.1232, found 291.1222.

6-(2-(2,3-dihydrobenzofuran-5-yl)ethyl)-4-methoxy-5,6-dihydro-2H-pyran-2-one (5b)

White solid, yield 52%; ¹H-NMR (CDCl₃, 500 MHz): 7.02 (s, 1H), 6.91 (d, J = 10 Hz, 1H), 6.69 (d, J = 5 Hz, 1H), 5.13 (s, 1H), 4.52 (t, J = 10 Hz, 2H), 4.37–4.31 (m, 1H), 3.71 (s, 3H), 3.16 (t, J = 10 Hz, 2H), 2.81–2.75 (m, 1H), 2.72–2.66 (m, 1H), 2.52–2.45 (m, 1H), 2.31–2.27 (m, 1H), 2.11–2.03 (m, 1H), 1.90–1.83 (m, 1H); ¹³C-NMR (CDCl₃, 125 MHz): 172.7, 167.3, 158.3, 132.6, 127.7, 127.1, 124.9, 109.0, 90.2, 74.6, 71.0, 55.9, 36.7, 32.9, 30.2, 29.6. HRMS (ESI+) m/z calcd for C₁₆H₁₉O₄ [M+H]⁺, 275.1283, found 275.1272.

6-(2-(2,3-dihydrobenzofuran-6-yl)ethyl)-4-methoxy-5,6-dihydro-2H-pyran-2-one (5c)

White solid, yield 53%; ¹H-NMR (CDCl₃, 500 MHz): 7.09 (d, J = 10 Hz, 1H), 6.69 (d, J = 5 Hz, 1H), 6.63 (s, 1H), 5.13 (s, 1H), 4.54 (t, J = 5 Hz, 2H), 4.37–4.33 (m, 1H), 3.72 (s, 3H), 3.16 (t, J = 5 Hz, 2H), 2.83–2.78 (m, 1H), 2.74–2.69 (m, 1H), 2.51–2.46 (m, 1H), 2.31–2.27 (m, 1H), 2.12–2.07 (m, 1H), 1.91–1.87 (m, 1H); ¹³C-NMR (CDCl₃, 125 MHz): 172.6, 167.2, 160.3, 141.0, 124.7, 120.5, 109.3, 90.3, 74.7, 71.3, 55.9, 36.3, 32.9, 30.9, 29.4. HRMS (ESI+) m/z calcd for C₁₆H₁₉O₄ [M+H]⁺, 275.1283, found 275.1273.

6-(2-(benzo[d][1,3]dioxol-4-yl)ethyl)-4-methoxy-5,6-dihydro-2H-pyran-2-one (5d)

Yellow gummy liquid, yield 49%; ¹H-NMR (CDCl₃, 500 MHz): 6.77–6.74 (m, 1H), 6.71–6.66 (m, 2H), 5.92 (s, 2H), 5.14 (s, 1H), 4.39–4.35 (m, 1H), 3.73 (s, 3H), 2.87–2.81 (m, 1H), 2.77–2.73 (m, 1H), 2.55–2.48 (m, 1H), 2.36–2.31 (m, 1H), 2.13–2.10 (m, 1H), 1.99–1.95 (m, 1H); ¹³C-NMR (CDCl₃, 125 MHz): 172.7, 167.2, 147.1, 145.4, 122.5, 122.1, 121.5, 106.8, 100.5, 90.3, 74.9, 55.9, 34.2, 32.9, 25.0. HRMS (ESI+) m/z calcd for C₁₅H₁₇O₅ [M+H]⁺, 277.1076, found 277.1066.

6-(2-(benzo[d][1,3]dioxol-5-yl-2,2-d₂)ethyl)-4-methoxy-5,6-dihydro-2H-pyran-2-one (5e)

White solid, yield 64%; ¹H-NMR (CDCl₃, 500 MHz): 6.72 (d, J = 5 Hz, 1H), 6.67 (s, 1H), 6.64 (d, J = 5 Hz, 1H), 5.12 (s, 1H), 4.36–4.31 (m, 1H), 3.72 (s, 3H), 2.80–2.75 (m, 1H), 2.71–2.66 (m, 1H), 2.51–2.46 (m, 1H), 2.30–2.26 (m, 1H), 2.10–2.04 (m, 1H), 1.89–1.83 (m, 1H); ¹³C-NMR (CDCl₃, 125 MHz): 172.6, 167.2, 147.6, 145.8, 134.5, 121.1, 108.7, 108.2, 90.2, 74.6, 55.9, 36.4, 32.9, 30.6. HRMS (ESI+) m/z calcd for C₁₅H₁₅D₂O₅ [M+H]⁺, 279.1202, found 279.1191.

6-(2-(2,2-difluorobenzo[d][1,3]dioxol-5-yl)ethyl)-4-methoxy-5,6-dihydro-2H-pyran-2-one (5f)

Yellow solid, yield 57%; ¹H-NMR (CDCl₃, 500 MHz): 6.96–6.95 (m, 1H), 6.91–6.89 (m, 2H), 5.14 (s, 1H), 4.36–4.30 (m, 1H), 3.73 (s, 3H), 2.91–2.86 (m, 1H), 2.80–2.74 (m, 1H), 2.54–2.48 (m, 1H), 2.31–2.27 (m, 1H), 2.12–2.04 (m, 1H), 1.91–1.90 (m, 1H); ¹³C-NMR (CDCl₃, 125 MHz): 172.6, 167.0, 143.8, 142.1, 136.9, 131.5, 123.3, 109.5, 109.2, 90.3, 74.3, 56.0, 36.4, 33.0, 30.8. HRMS (ESI+) m/z calcd for C₁₅H₁₅F₂O₅ [M+H]⁺, 313.0888, found 313.0877.

6-(3-hydroxy-4-methoxyphenethyl)-4-methoxy-5,6-dihydro-2H-pyran-2-one (5g)

Pale yellow solid, yield 41%; ¹H-NMR (CDCl₃, 500 MHz): 6.77–6.75 (m, 2H), 6.68–6.66 (m, 1H), 5.67 (s, 1H), 5.12 (s, 1H), 4.36–4.31 (m, 1H), 3.85 (s, 3H), 3.72 (s, 3H), 2.79–2.71 (m, 1H), 2.69–2.62 (m, 1H), 2.51–2.45 (m, 1H), 2.31–2.26 (m, 1H), 2.11–2.04 (m, 1H), 1.91–1.88 (m, 1H); ¹³C-NMR (CDCl₃, 125 MHz): 172.8, 167.4, 145.6, 145.0, 134.0, 119.9, 114.5, 110.7, 90.3, 74.8, 56.0, 54.8, 36.3, 33.0, 30.3. HRMS (ESI+) m/z calcd for C₁₅H₁₉O₅ [M+H]⁺, 279.1232, found 279.1224.

6-(4-hydroxy-3-methoxyphenethyl)-4-methoxy-5,6-dihydro-2H-pyran-2-one (5h)

White solid, yield 48%; ¹H-NMR (CDCl₃, 500 MHz): 6.83 (d, J = 5 Hz, 1H), 6.71 (s, 1H), 6.68 (d, J = 10 Hz, 1H), 5.57 (bs, 1H), 5.13 (s, 1H), 4.37–4.31 (m, 1H), 3.86 (s, 3H), 3.72 (s, 3H), 2.83–2.77 (m, 1H), 2.74–2.68 (m, 1H), 2.53–2.47 (m, 1H), 2.31–2.27 (m, 1H), 2.11–2.06 (m, 1H), 1.92–1.85 (m, 1H); ¹³C-NMR (CDCl₃, 125 MHz): 171.8, 166.4, 145.5, 142.9, 131.6, 119.9, 119.9, 113.3, 110.1, 89.3, 73.7, 55.0, 54.9, 35.5, 32.0, 29.6. HRMS (ESI+) m/z calcd for C₁₅H₁₉O₅ [M+H]⁺, 279.1232, found 279.1223.

6-(3,4-dimethoxyphenethyl)-4-methoxy-5,6-dihydro-2H-pyran-2-one (5i)

White solid, yield 61%; ¹H-NMR (CDCl₃, 500 MHz): 6.74–6.72 (m, 1H), 6.68–6.67 (m, 2H), 5.07 (s, 1H), 4.31–4.26 (m, 1H), 3.80 (s, 3H), 3.79 (s, 3H), 3.66 (s, 3H), 2.77–2.73 (m, 1H), 2.70–2.65 (m, 1H), 2.47–2.42 (m, 1H), 2.25–2.21 (m, 1H), 2.06–2.02 (m, 1H), 1.86–1.83 (m, 1H); ¹³C-NMR (CDCl₃, 125 MHz): 171.7, 166.3, 147.9, 146.3, 132.3, 119.2, 110.7, 110.2, 89.3, 73.7, 55.0, 54.9, 54.8, 35.5, 32.0, 29.5. HRMS (ESI+) m/z calcd for C₁₆H₂₁O₅ [M+H]⁺, 293.1389, found 293.1378.

6-(3,4-dihydroxyphenethyl)-4-methoxy-5,6-dihydro-2H-pyran-2-one (5j)

White solid, yield 32%; ¹H-NMR (CD₃OD, 500 MHz): 6.68 (d, J = 5 Hz, 1H), 6.65 (s, 1H), 6.53 (d, J = 10 Hz, 1H), 5.17 (s, 1H), 4.40–4.35 (m, 1H), 3.78 (s, 3H), 2.72–2.67 (m, 1H), 2.62–2.57 (m, 1H), 2.55–2.52 (m, 1H), 2.42–2.39 (m, 1H), 2.05–1.99 (m, 1H), 1.93–1.87 (m, 1H); ¹³C-NMR (CD₃OD, 125 MHz): 174.8, 169.1, 144.8, 143.1, 132.4, 119.2, 115.1, 114.9, 88.8, 75.6, 55.5, 36.2, 32.3, 29.9. HRMS (ESI+) m/z calcd for C₁₄H₁₇O₅ [M+H]⁺, 265.1076, found 265.1066.

6-(2-(benzo[d][1,3]dioxol-5-yl)ethyl)-4-methoxytetrahydro-2H-pyran-2-one (5k)

White solid, yield 47%; ¹H-NMR (CDCl₃, 500 MHz): 6.72 (d, J = 5 Hz, 1H), 6.67 (s, 1H), 6.63 (d, J = 10 Hz, 1H), 5.91 (s, 2H), 4.16–4.10 (m, 1H), 3.75–3.69 (m, 1H), 3.33 (s, 3H), 2.85–2.80 (m, 1H), 2.78–2.74 (m, 1H), 2.70–2.64 (m, 1H), 2.54–2.46 (m, 1H), 2.27–2.23 (m, 1H), 2.03–1.97 (m, 1H), 1.88–1.81 (m, 1H), 1.58–1.51 (m, 1H); ¹³C-NMR (CDCl₃, 125 MHz): 170.4, 147.7, 145.8, 134.6, 121.2, 108.8, 108.2, 100.8, 75.7, 72.3, 56.0, 37.4, 36.2, 34.8, 30.7. HRMS (ESI+) m/z calcd for C₁₅H₁₉O₅ [M+H]⁺, 279.1232, found 279.1223.

(E)-6-(2-(benzo[d][1,3]dioxol-5-yl)vinyl)-4-methoxy-2H-pyran-2-one (5l)

Yellow solid, yield 55%; ¹H-NMR (CDCl₃, 500 MHz): 7.42 (d, J = 15 Hz, 1H), 7.01 (s, 1H), 6.97 (d, J = 10 Hz, 1H), 6.81 (d, J = 10 Hz, 1H), 6.41 (d, J = 15 Hz, 1H), 6.00 (s, 2H), 5.90 (s, 1H), 5.48 (s, 1H), 3.82 (s, 3H); ¹³C-NMR (CDCl₃, 125 MHz): 171.1, 164.0, 158.8, 148.9, 148.3, 135.5, 129.7, 123.5, 116.8, 108.6, 105.9, 101.4, 100.7, 88.5, 55.9. HRMS (ESI+) m/z calcd for C₁₅H₁₃O₅ [M+H]⁺, 273.0763, found 273.0755.

6-(2-(benzo[d][1,3]dioxol-5-yl)ethyl)-4-propoxy-5,6-dihydro-2H-pyran-2-one (5m)

White solid, yield 52%; ¹H-NMR (CDCl₃, 500 MHz): 6.73 (d, J = 5 Hz, 1H), 6.69 (s, 1H), 6.65 (d, J = 10 Hz, 1H), 5.92 (s, 2H), 5.10 (s, 1H), 4.36–4.33 (m, 1H), 3.82 (t, J = 5 Hz, 2H), 2.81–2.77 (m, 1H), 2.73–2.68 (m, 1H), 2.53–2.47 (m, 1H), 2.31–2.28 (m, 1H), 2.09–2.06 (m, 1H), 1.88–1.85 (m, 1H), 1.78–1.72 (m, 2H), 0.98 (t, J = 10 Hz, 3H); ¹³C-NMR (CDCl₃, 125 MHz): 172.0, 167.5, 147.6, 145.8, 134.6, 121.3, 108.8, 108.2, 100.8, 90.5, 74.6, 70.5, 36.6, 33.2, 30.7, 21.8, 10.3. HRMS (ESI+) m/z calcd for C₁₇H₂₁O₅ [M+H]⁺, 305.1389, found 305.1379.

6-(2-(benzo[d][1,3]dioxol-5-yl)ethyl)-4-isopropoxy-5,6-dihydro-2H-pyran-2-one (5n)

White solid, yield 58%; ¹H-NMR (CDCl₃, 500 MHz): 6.73 (d, J = 5 Hz, 1H), 6.69 (s, 1H), 6.65 (d, J = 10 Hz, 1H), 5.92 (s, 2H), 5.09 (s, 1H), 4.46–4.42 (m, 1H), 4.36–4.32 (m, 1H), 2.82–2.77 (m, 1H), 2.73–2.68 (m, 1H), 2.49–2.44 (m, 1H), 2.26–2.22 (m, 1H), 2.12–2.05 (m, 1H), 1.88–1.84 (m, 1H), 1.32–1.29 (m, 6H); ¹³C-NMR (CDCl₃, 125 MHz): 170.7, 167.7, 147.6, 145.8, 134.6, 121.2, 108.8, 108.2, 100.8, 90.6, 74.4, 71.7, 36.6, 33.6, 30.6, 21.4, 21.2. HRMS (ESI+) m/z calcd for C₁₇H₂₁O₅ [M+H]⁺, 305.1389, found 305.1377.

6-(2-(benzo[d][1,3]dioxol-5-yl)ethyl)-4-methoxy-3-methyl-5,6-dihydro-2H-pyran-2-one (5o)

Pale yellow solid, yield 56%; ¹H-NMR (CDCl₃, 500 MHz): 6.73 (d, J = 10 Hz, 1H), 6.69 (s, 1H), 6.65 (d, J = 10 Hz, 1H), 5.92 (s, 2H), 4.29–4.24 (m, 1H), 3.75 (s, 3H), 2.83–2.78 (m, 1H), 2.74–2.69 (m, 1H), 2.52–2.46 (m, 2H), 2.11–2.07 (m, 1H), 1.90–1.85 (m, 1H), 1.77 (s, 3H); ¹³C-NMR (CDCl₃, 125 MHz): 168.5, 165.2, 147.6, 145.8, 134.6, 121.2, 108.8, 108.2, 103.5, 100.8, 73.3, 55.3, 36.7, 30.7, 29.2, 8.8. HRMS (ESI+) m/z calcd for C₁₆H₁₉O₅ [M+H]⁺, 291.1232, found 291.1225.

6-(2-(benzo[d][1,3]dioxol-5-yl)ethyl)-4-methoxy-5-methyl-5,6-dihydro-2H-pyran-2-one (5p)

Yellow solid, yield 63%; ¹H-NMR (CDCl₃, 500 MHz): 6.73 (d, J = 5 Hz, 1H), 6.69 (s, 1H), 6.65 (d, J = 10 Hz, 1H), 5.91 (s, 2H), 5.115.06 (m, 1H), 4.32–4.30 (m, 1H), 3.72 (s, 3H), 2.83–2.78 (m, 1H), 2.69–2.62 (m, 1H), 2.49–2.46 (m, 1H), 2.27–2.09 (m, 1H), 2.001.70 (m, 1H), 1.16–1.12 (m, 3H); ¹³C-NMR (CDCl₃, 125 MHz): 178.5, 167.3, 147.6, 145.8, 134.6, 121.2, 108.2, 100.8, 89.2, 80.3, 77.3, 56.1, 36.3, 33.0, 31.0, 10.8. HRMS (ESI+) m/z calcd for C₁₆H₁₉O₅ [M+H]⁺, 291.1232, found 291.1224.

6-(2-(benzo[d][1,3]dioxol-5-yl)ethyl)-4-methoxy-6-methyl-5,6-dihydro-2H-pyran-2-one (5q)

White solid, yield 68%; ¹H-NMR (CDCl₃, 500 MHz): 6.72 (d, J = 10 Hz, 1H), 6.66 (s, 1H), 6.62 (d, J = 10 Hz, 1H), 5.91 (s, 2H), 5.17 (s, 1H), 3.74 (s, 3H), 2.72–2.57 (m, 3H), 2.37–2.33 (m, 1H), 2.02–1.91 (m, 2H), 1.47 (s, 3H); ¹³C-NMR (CDCl₃, 125 MHz): 171.3, 166.5, 147.6, 145.7, 135.1, 120.9, 108.7, 108.2, 100.8, 89.8, 79.6, 55.9, 42.7, 37.4, 29.6, 24.9. HRMS (ESI+) m/z calcd for C₁₆H₁₉O₅ [M+H]⁺, 291.1232, found 291.1225.

6-(1-(benzo[d][1,3]dioxol-5-yl)propan-2-yl)-4-methoxy-5,6-dihydro-2H-pyran-2-one (5r)

Yellow solid, yield 63%; ¹H-NMR (CDCl₃, 500 MHz): 6.72 (d, J = 5 Hz, 1H), 6.66–6.65 (m, 1H), 6.63–6.60 (m, 1H), 5.91 (s, 2H), 5.13 (d, J = 10 Hz, 1H), 4.29–4.17 (m, 1H), 3.72 (s, 3H), 2.88–2.82 (m, 1H), 2.67–2.51 (m, 1H), 2.47–2.41 (m, 1H), 2.28–2.25 (m, 1H), 2.14–2.10 (m, 1H), 0.99–0.90 (m, 3H); ¹³C-NMR (CDCl₃, 125 MHz): 173.0, 167.4, 147.6, 145.9, 133.7, 133.1, 122.1, 109.5, 108.1, 100.8, 90.3, 78.7, 56.0, 38.0, 29.9, 14.0. HRMS (ESI+) m/z calcd for C₁₆H₁₉O₅ [M+H]⁺, 291.1232, found 291.1292.

The general procedures for the in vivo study of DHM analogs in A/J mice with NNK

Female A/J mice (5–6 weeks of age) were purchased from the Jackson Laboratory (Bar Harbor, ME) and maintained in the specific pathogen-free facilities, according to animal welfare protocols approved by Institutional Animal Care and Use Committee at the University of Florida. After 1-week acclimation, mice were weighed and randomized to different groups (n=3) and switched to AIN-93G powdered diet with the date being defined as Day 1. On Day 7, mice in the respective groups were given a single dose of DHM or its analogs in 15% solutol (200 μL) via oral gavage followed by a single i.p. injection of NNK in saline (100 μL, 100 mg/kg of body weight) one hour later. Mice in the negative control group were given 15% solutol and saline respectively. Mice were euthanized four hours after NNK exposure. The lung tissues were harvested, snap-frozen in liquid N₂ and stored at −80°C until DNA isolation.

Isolation of DNA and quantification of O⁶-mG from the lung tissues of A/J mice by liquid chromatography-electrospray ionization/tandem mass spectrometry

DNA was isolated from 25 mg lung tissue of each individual mouse, following Puregene DNA isolation protocol from Qiagen Corp. O⁶-mG was quantified by targeted UPLC/MS-MS analysis on a Dionex Ultimate 3000 RS and a Q Exactive Hybrid Quadrupole Orbitrap Mass Spectrometer, with parallel reaction monitoring (PRM), following our previously reported protocols ^[24].

Quantification of HPB and Diol in DNA from the lung tissues of A/J mice

D₄-HPB (20 pg) and D₄-Diol (10 pg) were added into each DNA sample as the internal standards. HCl was added to a final concentration of 0.8 N in about 500 μL mixture, and the mixture was heated in 95°C for 5 days. The mixture was neutralized to pH7, and cleaned by Strata X SPE cartridges. SPE cartridges were pre-conditioned with 1 mL MeOH and 1 mL water. After loading samples, the cartridges were washed with 1 mL water and 1 mL 5% MeOH in water. Samples were then eluted with 1 mL MeOH, speedvac to dryness and resuspended in 30 μL 10 mM ammonium acetate for LC-MS/MS analysis

Statistical analysis

The one-way analysis of variance (ANOVA) was used to compare differences among groups represented by the mean values ± SD in GraphPad followed by Dunnett test. Significance was indicated as follows: * p < 0.05, ** p < 0.01, *** p < 0.001, **** p < 0.0001.