International Journal of Molecular Sciences

Article Exploring Diverse-Ring Analogues on Combretastatin A4 (CA-4) Olefin as -Targeting Agents

1, 1, 2 2 3 Ming-Yu Song †, Qiu-Rui He †, Yi-Lin Wang , Hao-Ran Wang , Tian-Cheng Jiang , Jiang-Jiang Tang 1,* and Jin-Ming Gao 1

1 Shaanxi Key Laboratory of Natural Products & Chemical Biology, College of Chemistry & Pharmacy, Northwest A&F University, Yangling 712100, China; [email protected] (M.-Y.S.); [email protected] (Q.-R.H.); [email protected] (J.-M.G.) 2 College of Innovation and Experiment, Northwest A&F University, Yangling 712100, China; [email protected] (Y.-L.W.); [email protected] (H.-R.W.) 3 College of Food Science and Engeering, Northwest A&F University, Yangling 712100, China; [email protected] * Correspondence: [email protected]; Tel.: +86-2987-09-2662 These authors contributed equally to this work. †  Received: 15 February 2020; Accepted: 4 March 2020; Published: 6 March 2020 

Abstract: Combretastatin-4 (CA-4) as a polymerization inhibitor draws extensive attentions. However, due to its weak stability of cis-olefin and poor metabolic stability, structure modifications on cis-configuration are being performed. In this work, we constructed a series of novel CA-4 analogues with linkers on olefin containing diphenylethanone, cis-locked dihydrofuran, α-substituted diphenylethanone, cyclobutane and cyclohexane on its cis-olefin. Cytotoxic activity of all analogues was measured by an SRB assay. Among them, compound 6b, a by-product in the preparation of diphenylethanone analogues, was found to be the most potent cytotoxic agents against HepG2 cells with IC50 values of less than 0.5 µM. The two isomers of 6b induced cellular apoptosis tested by Annexin V-FITC and propidium iodide (PI) double staining, arrested cells in the G2/M phase by PI staining analysis, and disrupted microtubule network by immunohistochemistry study in HepG2 cells. Moreover, 6b-(E) displayed a dose-dependent inhibition effect for tubulin assembly in in vitro tubulin polymerization assay. In addition, molecular docking studies showed that two isomers of 6b could bind efficiently at binding site of tubulin similar to CA-4.

Keywords: combretastatin-4; tubulin polymerization; diphenylethanone; cis-locked dihydrofuran; α-substituted diphenylethanone; cyclobutane; cyclohexane

1. Introduction The natural product combretastatin A-4 (CA-4, Figure1) was first isolated from the South African tree Combretum caffrum in 1989 [1,2] and was serendipitously discovered in 1998 by strongly blocking the polymerization of tubulin with the colchicine binding site [3]. CA-4 and its water-soluble prodrugs such as CA-4P (fosbretabulin and zybrestat) [4] and AVE8062 (ombrabulin) [5] have entered clinical trials for the treatment of anaplastic thyroid and other solid tumors in both USA and Europe [6]. Due to its small molecular weight and simple structure, an impressive number of synthetic research was carried out over the past decades [7–10]. The structure–activity relationships (SARs) have made clearly important features of maintaining the tubulin interaction for CA-4. The presence of 3,4,5-trimethoxy groups at ring-A, 4‘- at ring-B and cis-configuration of olefin are the indispensability for inhibiting tubulin assembly [11]. Nevertheless, the Z-restricted configuration can rapidly convert into 100-fold less

Int. J. Mol. Sci. 2020, 21, 1817; doi:10.3390/ijms21051817 www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2020, 21, 1817 2 of 17 active trans-isomerization under the conditions of suffering from heat, light and protic media [12]. Meanwhile, the poor metabolic stability caused by the olefinic bridge and the weak water solubility of CA-4 resulted in restrictions for its exploitation [13]. Thus, an extensive focus of research was directed towards the rational modifications of the unstable olefinic bond as the main pursued strategy. Various olefinic linkers including silicon [14], butadiene [15], cyclopropane [16], cyclopropyl [17], chiral-lactam [18], furan [19], furazan [20], imidazole [21], 2,3-diarylthiophene [22], isoxazole and terphenyl [23], triazoles [24,25], thiazole [26,27] and 1,2,4-triazolo-4,3-b-pyridazines [28] has unveiled Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 2 of 17 novel agents endowed with cytotoxic activity.

MeO A MeO OH MeO CA-4 X = OH B CA-4P X = OPO Na OMe 3 2 MeO OMe X AVE8062 X = NH-serine HCl OMe OMe isoCA-4

This work Previous work R O O MeO HO R MeO 2 R1 X MeO O MeO MeO * OH OMe OH OMe OTBS OMe MeO OMe OMe OMe O MeO X = C, N Inhibit tubulin polymerization MeO MeO OH R MeO MeO OMe OMe OH OMe OMe

Figure 1. The structures of CA-4 and synthetic prodrugs, isoCA-4 and designed analogues. Figure 1. The structures of CA-4 and synthetic prodrugs, isoCA-4 and designed analogues. Isocombretastatin A-4 (isoCA-4, Figure1), an unexpectedly synthesized isomer of CA-4 found The structure–activity relationships (SARs) have made clearly important features of by Alami [29], exhibited a very strong inhibition on cancer cell growth similar to that of CA-4. maintaining the tubulin interaction for CA-4. The presence of 3,4,5-trimethoxy groups at ring-A, The synthetic method is simple without the control of cis-configuration [30,31]. The Alami group have 4‘-methoxy group at ring-B and cis-configuration of olefin are the indispensability for inhibiting completed plentiful studies on the modification of 1,1-diaryl double bond scaffold using naphthenic [32], tubulin assembly [11]. Nevertheless, the Z-restricted configuration can rapidly convert into 100-fold naphthalene [33], arylchromenes [34], azaisoerianin [35], arylbenzoxepins [36], etc. In our previous less active trans-isomerization under the conditions of suffering from heat, light and protic media works [37], dihydrofuran and dihydroisoxazole analogues of isoCA-4 were prepared through [3+2] [12]. Meanwhile, the poor metabolic stability caused by the olefinic bridge and the weak water reactions and could inhibit tubulin polymerization. solubility of CA-4 resulted in restrictions for its exploitation [13]. Thus, an extensive focus of In our endeavor aiming at finding novel potential rigid analogues of CA-4 and isoCA-4, research was directed towards the rational modifications of the unstable olefinic bond as the main we hypothesized that the replacement of the unstable olefinic bond of CA-4 with cis-locked dihydrofuran, pursued strategy. Various olefinic linkers including silicon [14], butadiene [15], cyclopropane [16], α-substituted, cyclobutane and cyclohexane nucleuses (Figure1) could be a good strategy to unravel cyclopropyl [17], chiral-lactam [18], furan[19], furazan [20], imidazole [21], 2,3-diarylthiophene [22], a novel class of antitubulin agents. Indeed, different number rings are widely used as a molecular isoxazole and terphenyl [23], triazoles [24,25], thiazole [26,27] and 1,2,4-triazolo-4,3-b-pyridazines scaffold in natural products and medicinal chemistry because of their favorable pharmacokinetic [28] has unveiled novel agents endowed with cytotoxic activity. features and their ability to act as a privileged structure when properly decorated. In this work, Isocombretastatin A-4 (isoCA-4, Figure 1), an unexpectedly synthesized isomer of CA-4 found we present the synthesis through the modification of olefin scaffold with diphenylethanone, cis-locked by Alami [29], exhibited a very strong inhibition on cancer cell growth similar to that of CA-4. The dihydrofuran, α-substituted diphenylethanone, cyclobutane and cyclohexane. Subsequently, their synthetic method is simple without the control of cis-configuration [30,31]. The Alami group have cytotoxic activity was evaluated against HepG2 cells, and the most potent cytotoxic compound was completed plentiful studies on the modification of 1,1-diaryl double bond scaffold using naphthenic selected to carry out mechanism studies including apoptosis, cell cycle analysis, immunohistochemistry, [32], naphthalene [33], arylchromenes [34], azaisoerianin [35], arylbenzoxepins [36], etc. In our tubulin polymerization assay and molecular docking. previous works [37], dihydrofuran and dihydroisoxazole analogues of isoCA-4 were prepared through [3+2] reactions and could inhibit tubulin polymerization. In our endeavor aiming at finding novel potential rigid analogues of CA-4 and isoCA-4, we hypothesized that the replacement of the unstable olefinic bond of CA-4 with cis-locked dihydrofuran, α-substituted, cyclobutane and cyclohexane nucleuses (Figure 1) could be a good strategy to unravel a novel class of antitubulin agents. Indeed, different number rings are widely used as a molecular scaffold in natural products and medicinal chemistry because of their favorable pharmacokinetic features and their ability to act as a privileged structure when properly decorated. In this work, we present the synthesis through the modification of olefin scaffold with diphenylethanone, cis-locked dihydrofuran, α-substituted diphenylethanone, cyclobutane and cyclohexane. Subsequently, their cytotoxic activity was evaluated against HepG2 cells, and the most potent cytotoxic compound was selected to carry out mechanism studies including apoptosis, cell cycle analysis, immunohistochemistry, tubulin polymerization assay and molecular docking.

Int.Int. J. J. Mol. Mol. Sci. Sci.2020 2020, 21, 2,1 1817, x FOR PEER REVIEW 33 of of 17 17

2. Results and Discussion 2. Results and Discussion 2.1. Chemistry 2.1. Chemistry 5-bromo-2-methoxyphenol was firstly converted into the key intermediate 4 by four steps in total5-bromo-2-methoxyphenol 65% yield referred to was our firstly previous converted report into [ the37]. key Subsequently intermediate, 44 bywas four oxidized steps in totalwith 65% yield referred to our previous report [37]. Subsequently, 4 was oxidized with rearrangement of rearrangement of olefinic bond through iodobenzene acetate (PhI(OAc)2) to afford the key olefinic bond through iodobenzene acetate (PhI(OAc) ) to afford the key compound 5a. The synthetic compound 5a. The synthetic conditions were optimized2 in Scheme 1. In the synthesis of 5a, NaHSO4 conditions were optimized in Scheme1. In the synthesis of 5a, NaHSO was used as the initial activator was used as the initial activator of PhI(OAc)2. However, the conversion4 of 4 was always very low in ofthe PhI(OAc) presence2. However,or absence the of conversionsolvent as ofshown4 was in always entry 1 very and low 2, and in the the presence yield of or5a absence was only of about solvent 10% as shown in entry 1 and 2, and the yield of 5a was only about 10% when the ratio of 4, PhI(OAc) and when the ratio of 4, PhI(OAc)2 and NaHSO4 was 1:1:1. As the ratio of PhI(OAc)2 and NaHSO2 4 NaHSOcontinue4 wasd to 1:1:1. increase As the, the ratio conversion of PhI(OAc) rate2 and of 4 NaHSO was indeed4 continued greatly to improvedincrease, the when conversion the reaction rate of 4 was indeed greatly improved when the reaction adopted MeCN-H O (5:1) as the solvent in entry 3 adopted MeCN-H2O (5:1) as the solvent in entry 3 and 4. When both2 of PhI(OAc)2 and NaHSO4 andreached 4. When 2 equivalent both of PhI(OAc) (entry 5)2 ,and 4 could NaHSO react4 reachedcompletely, 2 equivalent but the yield (entry of 5), 5a4 wascould still react only completely, about 10% but the yield of 5a was still only about 10% or even lower. After the NaHSO was replaced with H SO or even lower. After the NaHSO4 was replaced with H2SO4 (50% methanol4 solution), the reaction2 4 (50% methanol solution), the reaction was completed with a 61% yield at -20 C, while the ratio of 4, was completed with a 61% yield at -20 °C, while the ratio of 4, PhI(OAc)2, and◦ H2SO4 was 1:1.1:1.1 in PhI(OAc)entry 6. Accidentally2, and H2SO4, awas by-product 1:1.1:1.1 6a in entrywas given 6. Accidentally, under the condition a by-product of entry6a was6 with given 30% under yield,the and condition6b was prepared of entry 6 withfrom 30%6a using yield, andtetrabutyl6b was ammonium prepared from fluoride6a using (TBAF tetrabutyl) to remove ammonium the protection fluoride (TBAF)group toin remove 95% yi theeld. protection To the formation group in of 95% the yield. by-product To the formation 6a, MeOH of may the by-product attack the 6aintermediate, MeOH may of attack the intermediate of PhI(OAc) and H SO in reaction kinetics and through following elimination PhI(OAc)2 and H2SO4 in reaction 2kinetics2 and4 through following elimination reaction to trigger the reactionformation to trigger of 6a. theThe formationpossible reaction of 6a. The mechanism possible reaction was shown mechanism as Scheme was S shown1 in the as Supplementary Scheme S1 in 13 thematerial Supplementary. Furthermore, Material. the Furthermore,HPLC analysis the and HPLC 13C analysisNMR spectrum and C of NMR 6b revealed spectrum that of 6b therevealed ratio of that(Z)/ the(E) ratioisomers of (wasZ)/( Eabout) isomers 1:1. To was explore about biological 1:1. To explore activities biological of different activities isomer ofs di, 6bfferent-(Z) and isomers, 6b-(E) 6b-(wereZ) and separated6b-(E) wereby preparative separated byHPLC preparative. The assignment HPLC. The of assignment the olefin of geometry the olefin of geometry 6b-(Z) was of 6b-(determinedZ) was determined using NOESY using NMR NOESY experiments. NMR experiments. Specifically, Specifically, NOESY NOESYsignals signals of 6b-(Z of) 6b-(wereZ ) readilywere readilyapparent apparent between between the proton the proton of phenyl of phenyl ring and ring the and proton the proton attached attached to the to olefin the olefin in Figure in Figure 2 (2see (seeSupplementary Supplementary data data).).

Entry Activator Equivalent 1 Solvent Conversion (%) Yield (%) of 5a

1 NaHSO4 1: 1: 1 free 30 10 2 NaHSO4 1: 1: 1 MeCN-H2O 2 25 8 3 NaHSO4 1: 1.3: 1.3 MeCN-H2O 88 12 4 NaHSO4 1: 1.5: 1.5 MeCN-H2O 90 10 5 NaHSO4 1: 2: 2 MeCN 100 < 5 6 H2SO4 3 1: 1.1: 1.1 MeOH 100 61

1 Equivalent represents the ratio of 4: PhI(OAc)2: activator; 2 MeCN-H2O is 5:1 (v/v); 3 H2SO4 is 50% methanol solution; 4 6a was obtained under the condition of the Entry 6 with 30% yield. 6b was prepared from 6a in 95% yield.

SchemeScheme 1. 1.Synthesis Synthesis of compounds of compounds5a, 5b5a,and 5b 6aand, 6b 6a. Synthetic, 6b. Synthetic conditions conditions optimization optimization of 5a. i) of TBAF, 5a. i) THF,TBAF, 30 min.THF, 30 min.

Int.Int. J. J. Mol. Mol. Sci. Sci.2020 2020, ,21 21,, 1817 x FOR PEER REVIEW 44 of of 17 17

Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 4 of 17

Figure 2. 6b-(Z) 6b-(E) 6b-(Z) Figure 2. 6b-(Z) and 6band-(E) after afterHPLC HPLC preparation. preparation. Selected Selected NOE NOEcorrelations correlations of 6b of-(Z). .

With 5a in hands, dihydrofuran analogues were prepared by [3 + 2] cycloaddition in Scheme2. WithFigure 5a in 2.hands, 6b-(Z) anddihydrofuran 6b-(E) after HPLC analogues preparation. were Selected prepared NOE bycorrelations [3 + 2] cycloadditionof 6b-(Z). in Scheme 2. A free-radical reaction between 5a and conjugated diolefinic compounds was catalyzed by a A free-radical reaction between 5a and conjugated diolefinic compounds was catalyzed by a single-electronWith oxidant5a in hands, Cu(OAc) dihydrofuran2, and followinganalogues were by TBAF prepared to remove by [3 + the2] cycloaddition TBS group ofin Scheme the phenolic 2. single-electron oxidant Cu(OAc)2, and following by TBAF to remove the TBS group of the phenolic hydroxylA free to provide-radical reactionthe dihydrofuran between 5a7a andand conjugated7b. In this free-radicaldiolefinic compounds reaction, Mn(OAc)was catalyzed3 as the by initial a hydroxyl to provide the dihydrofuran 7a and 7b. In this free-radical reaction, Mn(OAc)3 as the oxidantsingle gave-electron extremely oxidant low Cu(OAc) yield, perhaps2, and following the single-electron by TBAF to remove oxidation the TBS energy group ofof Mnthe phenolic (III) (1.5 V) initial oxidant gave extremely low yield, perhaps the single-electron oxidation energy of Mn (III) could stillhydroxyl be too to high provide for this the reaction.dihydrofuran Thus, 7a Cu and (II) 7b with. In one-electron this free-radical oxidation reaction energy, Mn(OAc) of only3 as the 0.16 V (1.5 V)initial could oxidant still begave too extremely high for lowthis yield, reaction. perhaps Thus, the Cu single (II)- electronwith one oxidation-electron energy oxidation of Mn energy (III) of was employed for this reaction and the conditions were optimized by using Cu(OAc)2 as the catalyst only 0.16(1.5 V) V couldwas employed still be too forhigh this for reactionthis reaction. and Thus,the conditions Cu (II) with were one- electronoptimized oxidation by using energy Cu(OAc) of 2 and the acetophenone as the conjugated olefin. It was found that when the ratio of 5a, Cu(OAc)2 and as theonly catalyst 0.16 Vand was the employed acetophenone for this asreaction the conjugated and the conditions olefin. I twere was optimized found that by when using theCu(OAc) ratio 2of 5a, acetophenone was 1:3:3, 5a could react completely in 12 h. However, a large amount of by-products Cu(OAc)as the2 and catalyst acetophenone and the acetophenone was 1:3:3, as 5a the could conjugated react completely olefin. It was in found 12 h. that However, when the a ratiolarge of amount 5a, would appear after 4 h. Therefore, the optimal reaction time should be 4 h, while the conversion rate of of by-Cu(OAc)products2 and would acetophenone appear after was 41:3:3, h. Therefore, 5a could react the optimalcompletely reaction in 12 h. time However, should a larbege 4 h,amount while the 5a was about 70%, and the yield of 7a was 65%. Using this condition, dihydrofuran compound 7b was conversionof by-products rate of would 5a was appear about after 47 0%,h. Therefore, and the the yield optimal of reaction7a was time 65%. should Using be 4 h, this while condition, the also synthesized using isoprene, and the two steps yield was 45%. dihydrofuranconversion compound rate of 5a 7b was was about also synthesized70%, and the using yield isopr of ene,7a was and 65%. the twoUsing steps this yield condition, was 45% . However, under the above condition, an addition reaction between 5a and pentenone occurred to However,dihydrofuran under compound the above 7b was condition, also synthesized an addition using reaction isoprene, between and the two 5a andsteps pentenone yield was 45% occurred. give compoundHowever,8a with under a conversionthe above condition, yield of 74%an addition (Scheme reaction2). The between reason that5a and8a pentenonedid not cyclize occurred into a to give compound 8a with a conversion yield of 74% (Scheme 2). The reason that 8a did not cyclize dihydrofuranto give compound ring might 8a bewith the a electron-withdrawingconversion yield of 74% (Scheme effect of 2 the). The carbonyl reason that group. 8a did8b notwas cyclize afforded into a dihydrofuran ring might be the electron-withdrawing effect of the carbonyl group. 8b was by deprotectioninto a dihydrofuran of the TBS ring group might similar be the electron to the above-withdrawing method. effect Next, of the the carbonyl structure group of 1,5-dione. 8b was of afforded by deprotection of the TBS group similar to the above method. Next, the structure of 8a gaveafforded us a new by idea deprotection to synthesize of the cyclobutaneTBS group similar and cyclohexane to the above analogues method. Next of CA-4, the structurethrough Aldol of 1,5-dione1,5-dione of 8 aof gave 8a gave us aus new a new idea idea to to synthesize synthesize cyclobutane and and cyclohexane cyclohexane analogues analogues of C Aof- 4C A-4 condensation. For example, cyclobutane analogue 9 was prepared under the conditions of LiHMDS throughthrough Aldol Aldol condensation. condensation. For For example, example, cyclobutane cyclobutane analogue analogue 9 9was was prepared prepared under under the the and TsCl at 78 C in a 40% yield (Scheme2). conditionconditions −of LiHMDSs ◦of LiHMDS and and TsCl TsCl at at−78 −78 ° C°C in in a a 40% 40% yield (Scheme(Scheme 2) 2). .

OMe O OMe O O MeO O MeO OTBS MeO MeO OTBS i), ii) MeO MeO i), ii) MeO OMe 5a MeO OMe OH OMe 5a OMe O OMe OH 7a (65%) i) O i), OMe ii) 7a (65%) i) OMe i), O ii) O MeO OMe MeO O * OR O MeOMeO MeO * OR MeO OMe OMe OH MeO O MeO OMe OMe8a (R = TBS) (74%) OMe ii) O 7b (45%) OH 8a (R8b (R= TBS)= H) (95%)(74%) OMe 7b (45%) ii) O 8b (R = H) (95%) HO MeO O iii) 8a MeO HO MeO OMe OTBS iii) 8a OMe MeO 9 (40%) OMe OTBS Scheme 2. Synthesis of 7a, 7b, 8a, 8b and 9. Condition: i) Cu(OAc) , DCE,OMe 110 C, 4 h; ii) TBAF, THF, Scheme 2. Synthesis of 7a, 7b, 8a, 8b and 9. Condition: i) Cu(OAc)22, DCE, 110 °C,◦ 4 h; ii) TBAF, THF, 0.5 h; iii) LiHMDS, TsCl, 78 C. 9 (40%) 0.5 h; iii) LiHMDS, TsCl,− ◦−78 °C.

Scheme 2. Synthesis of 7a, 7b, 8a, 8b and 9. Condition: i) Cu(OAc)2, DCE, 110 °C, 4 h; ii) TBAF, THF, 0.5 h; iii) LiHMDS, TsCl, −78 °C.

Int. J. Mol. Sci. 2020, 21, 1817 5 of 17 Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 5 of 17

ForFor synthesis synthesis of of cyclohexane cyclohexane analogues, analogues, the the Aldol Aldol condensation condensation of 8a ofwas 8a was not smoothnot smooth due todue the to notthe very not stablevery stable TBS protecting TBS protecting group undergroup a under strong a base strong condition. base condition On the one. On hand, the one strong hand, alkalinity strong canalkalinity promote can the promote reaction the to proceed reaction forward. to proceed However, forward. on However the other, hand,on the it other would hand, cause it thewould removal cause ofthe TBS removal under theof TBS action under of too the strong action bases, of too then strong reduce bases, the numberthen reduce of substrate the number and push of substrate the reverse and reaction.push the Therefore, reverse reaction. it is very Therefore, important it tois choosevery important a base with to choose moderate a base alkalinity. with moderate alkalinity. TheThe condition condition optimization optimization of of Aldol Aldol condensation condensation is is shown shown in in Scheme Scheme3. 3. At At first, first the, the reaction reaction waswas performed performed with with KOH KOH in methanol in methanol under under reflux, reflux,8a was 8a completelywas completely converted convert quickly,ed quickly, but there but wasthere no was expected no expected product product10a or 8b 10ain or entry 8b in 1. entry Then, 1 three. Then, inorganic three inorganic bases: KOH, bases: NaOH KOH, andNaOH LiOH and (entryLiOH 2-4) (entry were 2- selected4) were andselected methanol and methanol was employed was employ as the solvented as the at 0 solvent◦C. After at the0 °C. three After bases the were three added,bases the were reaction added, occurred the reaction immediately, occurred and immediately,10a was obtained and 10 ina thewas beginning obtained period. in the However, beginning 10period min later,. However, the reactions 10 min oflater, KOH the andreaction NaOHs of beganKOH and to generate NaOH began8b. After to generate another 8 20b. A min,fter allanother10a and20 min8a were, all turned10a and into 8a 8bwere. However, turned into the 8 reactionb. However, of LiOH the (entry reaction 4) ofstarted LiOH to (entry generate 4) started8b after to 30generate min, while 8b after the conversion 30 min, while yield the of conversion10a is 65%. yield Under of the10a action is 65%. of Under K2CO 3the(1 eq,action entry of 5),K2CO8a could3 (1 eq, rapidlyentry 5 generate), 8a could a mixturerapidly ofgenerate10a, 8b aand mixture a little of deprotected10a, 8b and producta little deprotected of 10a at room product temperature. of 10a at Yet,room when temperature. the temperature Yet, when fell to the0 ◦C, temperature8a was completely fell to unreactive0 °C, 8a was even completely increasing unreactive the equivalent even ofincreasing K2CO3 to the 10 eq equivalent (entry 6). of NaOMe K2CO3 alsoto 10 could eq (entry not make 6). NaOMe8a react inalso solvent could of not benzene make in8a entryreact 8.in Insolvent organic of bases, benzene 1,8-diazabicyclo[5.4.0]undec-7-ene in entry 8. In organic bases, (DBU) 1,8-d couldiazabicyclo[5.4.0]undec make 8a to turn into-7-8benewithin (DBU) 30 could min atmake room 8 temperature,a to turn intoand 8b within lithium 30 diisopropylamide min at room temperature, (LDA) could and give lithium a 10% diisopropylamide yield of 10a at 78 (LDA)◦C, − incould entry give 9 and a 10% 10, respectively.yield of 10a at To -78 sum °C up,, in entry LiOH, 9 0and◦C, 10, and respectively 30 min were. To the sum optimized up, LiOH, conditions 0 °C, and with30 min a conversion were the yield optimized of 65% for conditions10a. After with10a was a conversion deprotected yield by TBAF, of 65% a cyclohexane-based for 10a. After 10 CA-4a was analoguedeprotected10b bywas TBAF, obtained a cyclohexane (Scheme3).-based CA-4 analogue 10b was obtained (Scheme 3).

Temperature Time Yield (%) Entry Base Equivalent Solvent (°C) (min) 10a 8b 1 KOH 10 MeOH 90 30 - - 2 KOH 10 MeOH 0 30 - > 95 3 NaOH 10 MeOH 0 30 - > 95 4 LiOH 10 MeOH 0 30 65 - 5 K2CO3 1 MeOH rt 30 20 15 6 K2CO3 10 MeOH 0 60 No reaction 7 Na2CO3 10 MeOH 0 60 No reaction 8 NaOMe 10 Benzene rt 45 No reaction 9 DBU 5 CH3CN rt 30 - > 95 10 LDA 2 THF -78 60 10 -

SchemeScheme 3. 3.Synthetic Synthetic conditions conditions optimization optimization of of compound compound10a 10a. .

InIn order order to enhanceto enhance the thepolarity polarity and water and water solubility solubility of 10b , of carbonyl 10b, carbonyl group of group10a was of reduced10a was byreduce sodiumd by borohydride sodium borohydride (NaBH4) to (NaBH a HO group4) to a (SchemeHO group4).. AfterAfter deprotectiondeprotection of of the the TBS TBS group group by by TBAF,TBAF,11 11was was obtained obtained in in two two steps steps with with a a 66% 66% yield. yield. Moreover, Moreover, LiHMDS LiHMDS was was employed employed to to remove remove thethe hydroxyl hydroxyl hydrogen hydrogen of 10aof ,10a and, thenand thenp-toluenesulfonyl p-toluenesulfonyl chloride chloride (TsCl) and(TsCl) the and DMAP-Et the DMAP3N system-Et3N weresystem used were to eliminate used to eliminate the sulfonyl the group.sulfonyl Finally, group. theFinally, TBS groupthe TBS was group deprotected was deprotected by TBAF by to TBAF give to give another cyclohexane analogue 12. The total yield of the three steps was 28%. It was noted that 12 was unable to react under the action of NaBH4 to afford 11.

Int. J. Mol. Sci. 2020, 21, 1817 6 of 17

anotherInt. J. Mol. cyclohexane Sci. 2020, 21, x analogue FOR PEER 12REVIEW(Scheme 4). The total yield of the three steps was 28%. It was noted6 of 17 that 12 was unable to react under the action of NaBH4 to afford 11.

Scheme 4. Synthesis of 11 and 12. Condition: (a) i) NaBH4, DCE, 0 ◦C, 1 h; ii) TBAF, THF, 30 min and Scheme 4. Synthesis of 11 and 12. Condition: (a) i) NaBH4, DCE, 0 °C, 1 h; ii) TBAF, THF, 30 min and (b) i) LiHMDS, TsCl, -78 ◦C; ii) Et3N, DMAP, DMF, 80 ◦C and iii) TBAF, THF, 30 min. (b) i) LiHMDS, TsCl, -78 °C; ii) Et3N, DMAP, DMF, 80 °C and iii) TBAF, THF, 30 min. 2.2. Biological Activity 2.2. Biological Activity New synthesized CA-4 analogues were in vitro screened for anticancer activity against HepG-2 (humanNew liver synthesized cancer cell CA line)-4 analogues with CA-4 were and isoin CA-4vitro screened as reference for anticancer drugs using activity the sulforhodamine against HepG-2 (human liver cancer cell line) with CA-4 and isoCA-4 as reference drugs using the sulforhodamine B B (SRB) assay [38]. The IC50 values (the concentration to cause 50% inhibition of cell viability) were summarized(SRB) assay in[38 Table]. The1. ItIC could50 values be seen(the thatconcentration these CA-4 to analogues cause 50% diphenylethanone inhibition of cell viability5b, cis-locked) were dihydrofuransummarized in analogues Table 1. It7a could 7b, α be-substituted seen that these diphenylethanones CA-4 analogues8a d,iphenylethanone8b, cyclobutane 5 analogueb, cis-locked9, dihydrofuran analogues 7a 7b, α-substituted diphenylethanones 8a, 8b, cyclobutane analogue 9, cyclohexane analogues 10a, 10b and 11 showed weak cytotoxic activity with IC50 values over 10 µM. cyclohexane analogues 10a, 10b and 11 showed weak cytotoxic activity with IC50 values over 10 μM. Another cyclohexane analogue 12 exhibited a good cytotoxic activity with IC50 of 3.67 µM, but with far Another cyclohexane analogue 12 exhibited a good cytotoxic activity with IC50 of 3.67 μM, but with less than IC50 of CA-4 and isoCA-4 (0.013 µM and 0.017 µM, respectively). These results suggested thatfar introductions less than IC50 of of ethanone, CA-4 and four iso orCA five-4 or(0.013 six-membered μM and 0.017 rings μM on the, respectively).cis-double bond These of CA-4results weresuggested not a good that introduction synthetic strategys of ethanone for optimizing, four or anticancer five or six- activity.membered However, rings on it the was cis worth-double to notebond of CA-4 were not a good synthetic strategy for optimizing anticancer activity. However, it was that isoCA-4 analogue 6b displayed strongest activity with IC50 of < 0.5 µM among all compounds. worth to note that isoCA-4 analogue 6b displayed strongest activity with IC50 of < 0.5 μM among all Notably, the IC50 results of 6b-(Z) with IC50 of 0.30 µM and 6b-(E) with IC50 of 0.23 µM on HepG2 cellscompound indicateds. Notably, no obvious the gap IC50 in results cytotoxic of 6b activity-(Z) with between IC50 ofcis–trans 0.30 μMisomers. and 6b-(E) with IC50 of 0.23 μM on HepG2 cells indicated no obvious gap in cytotoxic activity between cis–trans isomers.

Table 1. Cytotoxic activities (IC50) of compounds against HepG2 cells. Table 1. Cytotoxic activities (IC50) of compounds against HepG2 cells. 1 Compd IC50 (µM) Compd IC50 (µM) Compd IC50 (µM) 1 Compd IC50 (µM) 5b > 10 9 > 10 6b-(Z) 5b 0.30 > 0.0510 910a > 10 > 10 ± 6b-(E6b) -(Z) 0.230.30 ±0.07 0.05 10a10b > 10 > 10 ± 7a 6b-(E) 0.23> 10 ± 0.07 10b11 > 10 > 10 7b > 10 12 3.67 0.16 7a > 10 11 > 10 ± 8a > 10 CA-4 0.013 0.003 7b > 10 12 3.67± 0.16± 8b > 10 isoCA-4 0.017 0.002 ± 1 8a > 10 CA-4 0.013 ± 0.003 The IC50 values represent the concentration that causes 50% inhibition of cell viability measured by SRB assay. Cells were treated with the compounds8b for 48> h. 10 All dataiso (meanCA-4 SD)0.0 are17 the ± average0.002 of three determinations. ± HepG-2:1 The IC human50 values liver represent cancer cells. the concentration that causes 50% inhibition of cell viability measured by SRB assay. Cells were treated with the compounds for 48 h. All data (mean ± SD) are the average of Apoptosis induction is characteristic of many known anticancer agents. To determine whether the three determinations. HepG-2: human liver cancer cells. inhibition of 6b-(Z) and 6b-(E) on cell growth was related to cell apoptosis, a flow cytometric analysis by stainingApoptosi cellss withinduction Annexin is characteristic V-FITC and of propidium many known iodide anticancer (PI) was agents performed. To determine on HepG-2 whether cells. Thethe resultinhibition showed of 6b that-(Z treatment) and 6b-(E with) on 0.05cell µgrowthM CA-4 was for related 48 h induced to cell apoptosis, 26.5% and a 13.4% flow early cytometric and lateanalysis apoptotic by staining cells, respectively cells with (Figure Annexin3). SimilarV-FITC apoptosis and propidium induced iodide by CA-4 (PI) haswas been performed reported on previouslyHepG-2 cells [3].. The After result treatment showed with that compoundstreatment with6b-( 0.05Z) and μM 6b-(CA-E4) forat 48 0.3 hµ inducedM, the early 26.5% and and late 13.4% apoptosisearly and rates late wereapoptotic significantly cells, respectively increased from (Figure 5.0% 3 and). Similar 3.5% (for apoptosis control group)induced to by 19.2% CA- and4 has 16.1% been (forreported6b-(Z) previouslygroup), and [3 24.1%]. After and treatment 19.4% (forwith6b-( compoundE) group),s 6b respectively.-(Z) and 6b- The(E) at results 0.3 μM indicated, the early that and bothlate analoguesapoptosis could rates causewere signi cellularficantly apoptosis. increased from 5.0% and 3.5% (for control group) to 19.2% and 16.1% (for 6b-(Z) group), and 24.1% and 19.4% (for 6b-(E) group), respectively. The results indicated that both analogues could cause cellular apoptosis.

Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 7 of 17 Int. J. Mol. Sci. 2020, 21, 1817 7 of 17 Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 7 of 17

FigureFigure 3. 3.Induction Induction of apoptosis of apoptosis by 6b-( byZ 6b), 6b-(-(Z)E, )6band-(E CA-4) and at CA the-4 indicated at the indicated concentrations concentrations on HepG-2 on cells.FigureHepG After -2 3. cells. treatmentInduction After fortreatment of 48 apoptosis h, the for cells 48 by h,were 6b the-( Z stainedcells), 6b w-(ere withE) stainedand PI and CA with- Annexin4 at PI the and V-FITC indicated Annexin and concentrationsV viable,-FITC and apoptotic viable, on andHepGapoptotic necrotic-2 cells. and cells After necrotic were treatment analyzed cells were for by 48 analyzed flow h, the cytometry. cells by flowwere Cells cytometry.stained in lower with Cells left PI quadrantsand in lower Annexin left are Vquadrants alive,-FITC in and the are viable, lower alive, rightapoptoticin the quadrants lower and right necrotic are quadrants in early cells apoptosis,were are inanalyzed early in theapoptosis, by upper flow cytometry. rightin the quadrants upper Cells right in are quadrantslower in late left apoptosis quadrantsare in late and areapopto in aliv thesise, upperinand the in leftlower the quadrants upperright quadrants left are quadrants necrotic. are in Percentage are early necrotic. apoptosis, ofthe Percentage total in the signal upper of withinthe right total the quadrants signal quadrant within are is indicated.in the late quadrant apopto Eachsis is experimentandindicated. in the wasEach upper performed experiment left quadrants in was triplicate. performedare necrotic. in Percentagetriplicate. of the total signal within the quadrant is indicated. Each experiment was performed in triplicate. AA striking striking character character of of anticancer anticancer activity activity of CA-4of CA and-4 andisoCA-4 isoCA is-4 to is induce to induce cell cyclecell cycle arrest arrest at the at G2the/M AG2/M phase striking phase [3,39 character]. [3 To,39 confirm]. Toof aconfirmnticancer whether whether activity the biological theof CA biological- mechanism4 and iso mechanismCA of-4 theseis to ofinduce two these isomers cell two cycle was isomers arrest caused was at bythecaused cell G2/M cycle by phasecell accumulation cycl [3,39e accumulation]. To at confirm a G2/M at phase,whether a G2/M the thephase, cell biological cycle the distribution cell mechanism cycle distribution of HepG2 of these cells of two He was pG2 isomers evaluated cells waswas withcausedevaluated PI stainingby cellwith cyclby PI flowestaining accumulation cytometry by flow [at40 cytometry ].a G2/M As shown phase, [40] in. As Figurethe shown cell4, cycle following in Fig distributionure treatment4, following of withHe treatmentpG26b-( cellsZ) and waswith 6b-(evaluated6bE-()Zfor) and 24 with h6b at- (PIE concentrations) stainingfor 24 h byat flowconcentrations of 0.3 cytometryµM and of 1.0[40] 0.3µ.M, AsμM theshown and percentages 1.0in FigμMure, the of 4 cells, percentagesfollowing in the G2treatment of/M cells phase inwith atthe di6bG2/Mfferent-(Z) andphase concentrations 6b at-( Edifferent) for 24 were hconcentrations at 18.5% concentrations and 89.8% were forof 18.5 0.36b-(% μMZ and) and and 89.8 12.8% 1.0% forμM and ,6b the 88.5%-(Z percentages) and for 12.86b-(%E ) ,ofand respectively, cells 88.5 in% the for asG2/M6b compared-(E )phase, respectively, toat 12.2%different as in compared controlconcentrations cells, to 12.2 suggesting were% in control18.5 that% and cells, these 89.8 suggesting two% for isomers 6b -that(Z)caused andthese 12.8 two an%obvious isomersand 88.5 G2caused% /Mfor arrest6ban- (obviousE with), respectively, a similarG2/M arrest mechanism as compared with a tosimilar to that 12.2 ofmechanism% CA-4. in control to cells, that ofsuggesting CA-4. that these two isomers caused an obvious G2/M arrest with a similar mechanism to that of CA-4.

Figure 4. Analysis of cell cycle distribution in HepG-2 cells by flow cytometry. The cultured cells wereFigure treated 4. Analysis with 6b-( ofZ cell), 6b-( cycleE) and distribution CA-4 at thein HepG indicated-2 cells concentrations by flow cytometry for 24 h,. The then cultured harvested, cells andFigurewere analyzed treated 4. Analysis using with a6bof FACS -cell(Z) , cycle Calibur6b-(E distribution) and (BD CA Biosciences).-4 atin theHepG indicated-2 cells concentrationsby flow cytometry for 24. The h, then cultured harvested, cells wereand analyzedtreated with using 6b a-( FACSZ), 6b -Calibur(E) and (BDCA -Biosciences).4 at the indicated concentrations for 24 h, then harvested, and analyzed using a FACS Calibur (BD Biosciences).

Int. J. Mol. Sci. 2020, 21, 1817 8 of 17 Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 8 of 17 Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 8 of 17 To visualize the tubulin polymerization inhibition of these compounds in cells, To visualize the tubulin polymerization inhibition of these compounds in cells, immTounofluorescence visualize the analysis tubulin was polymerization investigated on microtubuleinhibition ofnetwork these in compoundsHepG2 cells . inAs showncells, immunofluorescence analysis was investigated on microtubule network in HepG2 cells. As shown in immunofluorescencein Figure 5, control cellsanalysis displayed was investigated a normal distribution on microtubule of microtubular network in networkHepG2 cells. with Asa slim shown and Figure5, control cells displayed a normal distribution of microtubular network with a slim and fibrous infibrous Figure shape.5, control However, cells displayed cells treated a normal with distribution CA-4 and of6b microtubular-(E) showed network a disrupted with microtubulea slim and shape. However, cells treated with CA-4 and 6b-(E) showed a disrupted microtubule network (green) fibrousnetwork shape. (green However,) becoming cells short treated and wrapped with CA-4 around and cell 6b-( nucleus,E) showed demonstrating a disrupted that microtubule 6b-(E) could becoming short and wrapped around cell nucleus, demonstrating that 6b-(E) could inhibit the cellular networkinhibit the (green) cellular becoming tubulin short polymerization. and wrapped around cell nucleus, demonstrating that 6b-(E) could tubulin polymerization. inhibit the cellular tubulin polymerization.

.

FigureFigure 5. 5Immunofluorescence. Immunofluorescence eff effectsects of of6b-( 6bE-()E(0.3) (0.3µ M)μM) and and CA-4 CA- (0.024 (0.02µ M)μM) on on inhibition inhibition of of tubulin tubulin Figure 5. Immunofluorescence effects of 6b-(E) (0.3 μM) and CA-4 (0.02 μM) on inhibition of tubulin polymerizationpolymerization in in HepG-2 HepG- cells.2 cells. After After treatment treatment for for 48 48 h, h, the the cells cells were were stained stained with with DAPI DAPI (blue, (blue, left left polymerization in HepG-2 cells. After treatment for 48 h, the cells were stained with DAPI (blue, left panel)panel) for for nucleus nucleus and and an an anti- antiα--tubulinα-tubulin antibody antibody (TRITC, (TRITC, green, green, middle middle panel) panel) for for . microtubules. panel) for nucleus and an anti-α-tubulin antibody (TRITC, green, middle panel) for microtubules. TheThe right right panel panel represents represents a merge a merge of the ofcorresponding the corresponding left and leftmiddle and panels. middle Images panels. were Images acquired were The right panel represents a merge of the corresponding left and middle panels. Images were withacquired an Olympus with an fluorescence Olympus fluorescence microscope microscope using a 20 usingobjective. a 20× objective. acquired with an Olympus fluorescence microscope ×using a 20× objective. Furthermore,Furthermore, to to investigate investigate whether whether there there is is such such an an interaction interaction of of the the active active analogues analogues with with Furthermore, to investigate whether there is such an interaction of the active analogues with tubulintubulin responsible responsible for for the the activity, activity, tubulin tubulin polymerization polymerizationin vitro in vitrowas was measured measured for 50for min 50 min at 37 at◦C 37 tubulin responsible for the activity, tubulin polymerization in vitro was measured for 50 min at 37 with°C paclitaxelwith paclitaxel as a promoting-assembly as a promoting-assembly control control and CA-4 and as CA a depolymerization-4 as a depolymeriz control.ation Ascontrol. shown As °C with paclitaxel as a promoting-assembly control and CA-4 as a depolymerization control. As inshown Figure 6 in, paclitaxelFigure 6, promotedpaclitaxel thepromot assemblyed the of assemblytubulin, while of tubulin, CA-4 inhibited while CA it- well.4 inhibited Compound it well. shown in Figure 6, paclitaxel promoted the assembly of tubulin, while CA-4 inhibited it well. 6b-(CompoundE) displayed 6b- an(E) inhibitiondisplayed e ffanect inhibition for tubulin effect polymerization for tubulin inpolymerization a dose-dependent in a dosemanner.-dependent From Compound 6b-(E) displayed an inhibition effect for tubulin polymerization in a dose-dependent thesemanner results,. From it could these be results concluded, it could that be6b-( concludeE) couldd that inhibit 6b- the(E) assemblycould inhibit of tubulin, the assembly and the of cellular tubulin, manner. From these results, it could be concluded that 6b-(E) could inhibit the assembly of tubulin, activitiesand the ofcellular6b-(E )activitmighties be of generated 6b-(E) might from be this generated interaction from with this tubulin. interaction with tubulin. and the cellular activities of 6b-(E) might be generated from this interaction with tubulin.

Figure 6. Effects of 6b-(E) on tubulin polymerization in vitro. Tubulin (10 mg/mL) was incubated Figure 6. Effects of 6b-(E) on tubulin polymerization in vitro. Tubulin (10 mg/mL) was incubated withFigure6g-(E) 6. Effects(5, 20 andof 6b-(E) 40 µM) on fortubulin 50 min polymerization at 37 C. Paclitaxel in vitro. (3 µTubulinM) and (10 CA-4 mg/mL) (3 µM) was were incubated used as with 6g-(E) (5, 20 and 40 μM) for 50 min at 37◦ °C. Paclitaxel (3 μM) and CA-4 (3 μM) were used as positivewith 6g-(E) controls. (5, 20 The and fluorescence 40 μM) for intensity50 min at was 37 monitored°C. Paclitaxel every (3 minuteμM) and for CA-4 characterization (3 μM) were of tubulinused as positive controls. The fluorescence intensity was monitored every minute for characterization of polymerizationpositive controls. at aThe 360 nmfluorescence excitation intensity wavelength was and monitored 460 nm emissionevery minute wavelength. for characterization of tubulin polymerization at a 360 nm excitation wavelength and 460 nm emission wavelength. tubulin polymerization at a 360 nm excitation wavelength and 460 nm emission wavelength.

Int. J. Mol. Sci. 2020, 21, 1817 9 of 17 Int. J. Mol. Sci. 2020, 21, x FOR PEER REVIEW 9 of 17

2.3. Molecular Docking 2.3. Molecular Docking To elucidate the binding model of the analogues 6b-(Z) and 6b-(E) with the tubulin colchicine To elucidate the binding model of the analogues 6b-(Z) and 6b-(E) with the tubulin colchicine binding site, the X-ray crystal structure (PDB:1SA0) was employed for performing the docking studies binding site, the X-ray crystal structure (PDB:1SA0) was employed for performing the docking using the Surflex-Dock program in the Sybyl-X 2.0 software. Figure7a,b exhibits the docking studies of studies using the Surflex-Dock program in the Sybyl-X 2.0 software. Figure 7a,b exhibits the docking 6b-(Z) 6b-(E) studiesand of 6b-(Z) withand 6b the-( tubulinE) with colchicinethe tubulin binding colchicine site ofbinding tubulin, site respectively. of tubulin, Figurerespectively.7a shows Fig thature the meta-methoxy groups on alkene of 6b-(E) formed a hydrogen bond with the key amino acid Ser178 7a shows that the meta-methoxy groups on alkene of 6b-(E) formed a hydrogen bond with the key about 1.96 Å bond distance, and the oxygen atoms of the hydroxyl groups on ring-B of 6b-(Z) formed a amino acid Ser178 about 1.96 Å bond distance, and the oxygen atoms of the hydroxyl groups on meta hydrogenring-B of bond6b-(Z with) formed the key a hydrogen amino acid bond Cys241 with about the key 2.35 amino Å. Figure acid7b Cys241 exhibits about that the2.35 Å. -methoxyFigure 7b 6b-(E) groupsexhibits on that ring-A the meta of -methoxyformed groups a hydrogen on ring bond-A of with6b-(E the) formed key amino a hydrogen acid Cys241 bond aboutwith the 1.93 key Å, while the oxygen atoms of the hydroxyl groups on ring-B formed a hydrogen bond with the key amino amino acid Cys241 about 1.93 Å, while the oxygen atoms of the hydroxyl groups on ring-B formed a acid Thr193 about 2.06 Å. From the simulated bond distance results, it can be concluded that 6b-(E) hydrogen bond with the key amino acid Thr193 about 2.06 Å. From the simulated bond distance 6b-(Z) shouldresults,have it can a be better concluded biological that activity6b-(E) should than have ,a consistingbetter biological with cytotoxicity activity than test 6b- results.(Z), consisting These 6b-(Z) 6b-(E) dockingwith cytotoxicity studies suggested test results. that These the binding docking site studies of suggestedand thatmight the binding be the tubulin site of 6b colchicine-(Z) and binding6b-(E) might site. be the tubulin colchicine binding site.

(a) (b)

FigureFigure 7. 7.Molecular Molecular docking docking simulations simulations obtained obtained at lowest at lowestenergy conformation, energy conformation, highlighting highlighting potential hydrogenpotential contactshydrogen of contacts (a) 6b-(Z o)f and(a) 6b (b)-(6b-(Z) andE) with (b) 6b the-(E colchicine’s) with the colchicine’s binding site binding of tubulin site (PDB of tubulin code: 1SA0).(PDB code: Surrounding 1SA0). Surrounding key amino acidkey residuesamino acid are residues labeled, are and labeled, hydrogen and bonds hydrogen are shown bonds byare orange shown dashedby orange lines. dashed lines. 3. Materials and Methods 3. Materials and Methods 3.1. Chemistry 3.1. Chemistry 3.1.1. General 3.1.1. General All the compounds were recorded by a 500 MHz Bruker NMR spectrometer (Bruker Daltonics 1 13 Inc., Bremen,All the compounds Germany) withwere recordedH, C and by DEPT135a 500 MHz NMR Bruker in CDClNMR3 spectrometersolution with (Bruker TMS as Daltonics internal standardInc., Bremen, for protons. Germany The) with 2D 1 NOSEYH, 13C and experiment DEPT135 was NMR carried in CDCl out3 solution in the 500 with MHz TMS Bruker as internal NMR spectrometerstandard for with protons 32. ofThe the 2D scans NOSEY number. experiment Chemical was shift carried values out are in mentionedthe 500 MHz in δ Bruker(ppm) NMR and couplingspectrometer constants with (J32) in of Hz. the MS scans (ESI) number was measured. Chemical on anshift ESI-Thermo values are Fisher mentioned LTQ Fleet in δ instrument(ppm) and spectrometercoupling constants (Thermo (J) in Fisher Hz. MS Scientific (ESI) was Inc., measured Waltham, on MA)The an ESI- reactionThermo progressFisher LTQ was Fleet supervised instrument by thin-layerspectrometer chromatography (Thermo Fisher (TLC), Scientific spots Inc. were, Waltham, observed MA under)The UVreaction light progress (254 nm was and supervised 365 nm) and by colorizedthin-layer with chromatography 5% phosphomolybdic (TLC), spots acid. were Silica observed gels (300-400 under mesh, UV Qingdaolight (254 Marine nm and Chemical 365 nm) Ltd., and colorized with 5% phosphomolybdic acid. Silica gels (300-400 mesh, Qingdao Marine Chemical Ltd., Qingdao, China) were used for column chromatography. Waters 1525 series equipped with a PDA detector (Waters Corporation, Milford, MA) was used for analytical HPLC, and YMS-pack ODS-A

Int. J. Mol. Sci. 2020, 21, 1817 10 of 17

Qingdao, China) were used for column chromatography. Waters 1525 series equipped with a PDA detector (Waters Corporation, Milford, MA) was used for analytical HPLC, and YMS-pack ODS-A column (250 mm 10 mm, 5 µm) was used for semipreparation. All reagents were purchased from × commercial sources, and purifications and desiccations of available solvents were accomplished by standard techniques prior to use. Positive control drug CA-4 was purchased from Sigma, and isoCA-4 was obtained refer to the previous report [37]. The purity of all compounds was confirmed to be greater than 95% through analytical HPLC.

3.1.2. Synthesis of Diphenylethanone Analogue 5a, 5b and by-Products 6a, 6b Preparation of compound 4 refers to our previous report method [37]. To a solution of 4 (11.2 g, 26 mmol, 1 eq) in MeOH (85 mL) was added PhI(OAc)2 (8.8 g, 27.3 mmol, 1.05 eq) at -20 ◦C. Then, 50% H2SO4 in MeOH (85 mL) was added by drops. The mixture was stirred for 30 min. The reaction was then quenched with saturated aqueous NaHCO3, and diluted by H2O (100 mL). The resulting mixture was extracted with EtOAc (3 100 mL). The combined organic layers were dried over anhydrous × Mg2SO4, filtered and concentrated under reduced pressure. The residue was purified by flash column chromatography on silica gel with petroleum ether/EtOAc (15:1) to give 5a and 6a. 5a or 6a with the TBS protective group was dissolved in THF (2 mL), and treated with 1.0 M TBAF (1.1 eq) in THF. The mixture was stirred for 0.5 h and diluted by H2O (10 mL), and extracted with EtOAc (3 10 mL). The combined organic layers were dried over anhydrous Na SO , filtered and × 2 4 concentrated under reduced pressure. The residue was purified by flash column chromatography on silica gel with petroleum ether/EtOAc (4:1) to give compounds 5b or 6b, respectively. The mixture 6a was separated by preparative HPLC to afford 6b-(Z) and 6b-(E) with ratio of 1:1. 6b-(Z): tR = 18.6 min, purity = 90.4%; 6b-(E): tR = 20.6 min, purity = 91.8% at 220 nm, 30 min gradient, 40% MeOH: 60% H2O to 100% MeOH. 2-(3-((tert-butyldimethylsilyl)oxy)-4-methoxyphenyl)-1-(3,4-dimethoxyphenyl)ethan-1-one (5a): 1 yellow viscosity liquid, 61% yield. H NMR (500 MHz, CDCl3): δ 7.28 (s, 2H), 6.83 (s, 2H), 6.79 (d, J = 1.0 Hz, 1H), 4.16 (s, 2H), 3.93 (s, 3H), 3.91 (s, 6H), 3.81 (s, 3H), 1.00 (s, 9H), 0.15 (s, 6H); 13C NMR (125 MHz, CDCl3): δ 196.77, 153.00 (2C), 149.96, 145.15, 142.47, 131.79, 127.35, 122.30, 122.01, 112.33, 106.31 (2C), 60.93, 56.25 (2C), 55.52, 45.04, 25.72 (3C), 18.45, -4.62 (2C). 2-(3-hydroxy-4-methoxyphenyl)-1-(3,4,5-trimethoxyphenyl)ethan-1-one (5b): yellow solid, 95% 1 yield. H NMR (500 MHz, CDCl3): δ 7.24 (s, 2H), 6.83 (d, J = 2.1 Hz, 1H), 6.76 (d, J = 8.3 Hz, 1H), 6.71 (dd, J = 8.3, 2.1 Hz, 1H), 5.67 (s, 1H), 4.50–4.46 (m, 1H), 3.85 (s, 3H), 3.84 (s, 6H), 3.82 (s, 3H), 13 2.40–2.31 (m, 5H), 2.08 – 2.02 (m, 1H), 1.01 (t, J = 7.3 Hz, 3H); C NMR (126 MHz, CDCl3): δ 211.39, 198.28, 152.90 (2C), 146.08, 145.82, 142.35, 132.55, 131.74, 119.70, 114.42, 111.10, 106.43 (2C), 60.86, 56.22 + + (2C), 55.93, 51.66, 39.53, 35.96, 27.64, 7.86; MS (ESI): m/z calcd for C18H21O6 [M+H] 333.13, found 333.34. (Z)-2-methoxy-5-(2-methoxy-1-(3,4,5-trimethoxyphenyl)vinyl)phenol (6b-(Z)): a yellow solid. 1H NMR (500 MHz, CDCl3): δ 6.76 (d, J = 2.1 Hz, 1H), 6.71 (d, J = 8.3 Hz, 1H), 6.64 (d, J = 2.1 Hz, 1H), 6.54 13 (s, 2H), 6.32 (s, 1H), 3.82 (s, 3H), 3.79 (s, 3H), 3.72 (s, 6H), 3.68 (s, 3H); C NMR (126 MHz, CDCl3): δ 152.74 (2C), 145.49, 145.35, 140.12, 136.79, 133.73, 133.16, 120.14, 119.89, 114.39, 110.48, 107.28 (2C), + + 60.85, 56.10 (2C), 56.02, 50.89; MS (ESI): m/z calcd for C19H23O6 [M+H] 347.15, found 347.05. (E)-2-methoxy-5-(2-methoxy-1-(3,4,5-trimethoxyphenyl)vinyl)phenol (6b-(E)): a yellow solid. 1H NMR (500 MHz, CDCl3): δ 7.06 (d, J = 2.1 Hz, 1H), 6.90 6.86 (m, 1H), 6.81 (d, J = 8.4 Hz, 1H), 6.43 (s, − + 2H), 6.35 (s, 1H), 3.89 (s, 3H), 3.85 (s, 3H), 3.81 (s, 6H), 3.76 (s, 3H); MS (ESI): m/z calcd for C19H23O6 [M+H] + 347.15, found 347.06. Int. J. Mol. Sci. 2020, 21, 1817 11 of 17

3.1.3. Synthesis of Cis-Locked Dihydrofuran Analogues 7a, 7b and α-Substituted Diphenylethanones 8a, 8b

To a solution of compound 5a (1.0 eq) and Cu(OAc)2 (3.0 eq) in DCE was added olefin (3.0 eq), then the mixture was heated to 110 ◦C, and stirred for 4 h. The reaction was filtered by diatomite, and diluted by H2O. The resulting mixture was extracted with EtOAc. The combined organic layers were dried over anhydrous Na2SO4, filtered and concentrated under reduced pressure. The residue was purified by flash column chromatography on silica gel with petroleum ether/EtOAc to give target compounds with the TBS protective group. Next, compound with TBS group was dissolved in THF (2 mL), and treated with 1.0 M TBAF (1.1 eq) in THF. The mixture was stirred for 0.5 h. The reaction was diluted by H O (10 mL), and extracted with EtOAc (3 10 mL). The combined organic layers were 2 × dried over anhydrous Na2SO4, filtered, and concentrated under reduced pressure. The residue was purified by flash column chromatography on silica gel with petroleum ether/EtOAc (4:1) to give 7a and 7b. 2-methoxy-5-(5-phenyl-2-(3,4,5-trimethoxyphenyl)-4,5-dihydrofuran-3-yl)phenol (7a): yellow solid. 1 Two steps yield: 65%. H NMR (500 MHz, CDCl3): δ 7.48 (d, J = 7.7 Hz, 2H), 7.39 (t, J = 7.5 Hz, 2H), 7.32 (t, J = 7.0 Hz, 1H), 6.90 (s, 1H), 6.83 (s, 2H), 6.76 (q, J = 8.4 Hz, 2H), 5.67 (t, J = 9.5 Hz, 1H), 5.54 (s, 1H), 3.86 (s, 3H), 3.86 (s, 3H), 3.72 (s, 6H), 3.54 (dd, J = 15.0, 10.4 Hz, 1H), 3.17 (dd, J = 15.0, 8.6 Hz, 1H); 13 C NMR (125 MHz, CDCl3): δ 152.89 (2C), 148.74, 145.44, 145.13, 143.02, 128.62 (2C), 127.78, 126.91, 125.79 (2C), 119.65, 114.01, 110.58, 108.88, 105.40 (2C), 80.42, 60.91, 56.04 (3C), 44.85; MS (ESI): m/z calcd + for NaC26H26O6 [M+Na] 457.16, found 457.49. 2-methoxy-5-(5-methyl-2-(3,4,5-trimethoxyphenyl)-5-vinyl-4,5-dihydrofuran-3-yl)phenol (7b): 1 yellow solid. Two steps yield: 45%. H NMR (500 MHz, CDCl3): δ 6.85 (s, 1H), 6.76 (s, 2H), 6.73 (s, 2H), 6.11 (dd, J = 17.3, 10.7 Hz, 1H), 5.52 (s, 1H), 5.34 (dd, J = 17.3, 1.1 Hz, 1H), 5.12 (dd, J = 10.7, 1.1 Hz, 1H), 3.86 (s, 3H), 3.84 (s, 3H), 3.72 (s, 6H), 3.09 (d, J = 14.9 Hz, 1H), 2.98 (d, J = 14.9 Hz, 1H), 1.58 (s, 3H); 13 C NMR (125 MHz, CDCl3): δ 152.87 (2C), 147.82, 145.36, 144.94, 142.49, 138.31, 129.50, 127.40, 119.46, 113.83, 112.23, 110.53, 108.37, 105.36 (2C), 83.41, 60.90, 56.04, 56.03 (2C), 48.00, 26.27; MS (ESI): m/z calcd + for NaC23H26O6 [M+Na] 421.16, found 421.67. 2-(3-((tert-butyldimethylsilyl)oxy)-4-methoxyphenyl)-1-(3,4,5-trimethoxyphenyl)heptane-1,5-dione 1 (8a): yellow viscosity liquid, 36% yield. H NMR (500 MHz, CDCl3): δ 7.22 (s, 2H), 6.77 (s, 2H), 6.71 (s, 1H), 4.46 (s, 1H), 3.85 (s, 3H), 3.84 (s, 6H), 3.75 (s, 3H), 2.40–2.32 (m, 5H), 2.04 (d, J = 7.2 Hz, 1H), 1.02 (s, 13 3H), 0.95 (s, 9H), 0.09 (s, 6H); C NMR (125 MHz, CDCl3 ): δ 211.40, 198.48, 152.89 (2C), 150.14, 145.39, 142.28, 131.90, 131.87, 121.33, 120.91, 112.47, 106.41 (2C), 60.87, 56.21 (2C), 55.46, 51.56, 39.50, 35.96, + 27.61, 25.69 (3C), 18.44, 7.87, -4.460, -4.61; MS (ESI): m/z calcd for SiC29H43O7 [M+H] 531.28, found 531.62. 2-(3-hydroxy-4-methoxyphenyl)-1-(3,4,5-trimethoxyphenyl)heptane-1,5-dione (8b): white solid, 95% 1 yield. H NMR (500 MHz, CDCl3): δ 7.24 (s, 2H), 6.83 (d, J = 2.1 Hz, 1H), 6.76 (d, J = 8.3 Hz, 1H), 6.71 (dd, J = 8.3, 2.1 Hz, 1H), 5.67 (s, 1H), 4.50–4.46 (m, 1H), 3.85 (s, 3H), 3.84 (s, 6H), 3.82 (s, 3H), 2.40–2.31 13 (m, 5H), 2.08–2.02 (m, 1H), 1.01 (t, J = 7.3 Hz, 3H); C NMR (125 MHz, CDCl3): δ 211.39, 198.28, 152.90 (2C), 146.08, 145.82, 142.35, 132.55, 131.74, 119.70, 114.42, 111.10, 106.43 (2C), 60.86, 56.22 (2C), 55.93, + 51.66, 39.53, 35.96, 27.64, 7.86; MS (ESI): m/z calcd for NaC23H28O7 [M+Na] 439.17, found 439.46.

3.1.4. Synthesis of Cyclobutane Analogue 9 To a solution of 8a (0.5 mmol, 1.0 eq) in dry THF (10 mL) was added 1 M LiHMDS solution in THF (1.0 mL, 1.0 mmol, 2.0 eq) at 78 C under argon atmosphere. After stirred for 1 h, the − ◦ reaction was added p-toluensulfonyl chloride (TsCl, 191 mg, 1.88 mmol, 2.0 eq) solution in THF (2 mL). The reaction stirred overnight, and quenched with saturated aqueous NH4Cl, diluted by H2O (10 mL). The resulting mixture was extracted with EtOAc (3 10 mL). The combined organic layers were × Int. J. Mol. Sci. 2020, 21, 1817 12 of 17

dried over anhydrous Mg2SO4, filtered, and concentrated under reduced pressure. The flash column chromatography on silica gel with petroleum ether/EtOAc (8:1) gave 9. 1-(3-(3-((tert-butyldimethylsilyl)oxy)-4-methoxyphenyl)-2-hydroxy-2-(3,4,5-trimethoxyphenyl) 1 cyclobutyl)propan-1-one (9): yellow solid, 40% yield. H NMR (500 MHz, CDCl3): δ 6.70 (d, J = 0.8 Hz, 2H),6.63 (s, 1H), 6.45 (s, 2H), 3.80 (s, 3H), 3.78 (s, 6H), 3.73 (s, 3H), 3.69 (d, J = 3.4 Hz, 1H), 3.50 (dd, J = 12.1, 3.3 Hz, 1H), 2.72–2.52 (m, 4H), 2.10 (s, 1H), 0.92 (s, 12H), 0.03 (s, 6H); 13C NMR (126 MHz, CDCl3): δ 213.47, 152.69 (2C), 150.01, 144.79, 139.75, 136.98, 132.28, 122.09, 122.02, 111.97, 103.82 (2C), 80.85, 0.81, 56.89, 56.17 (2C), 55.42, 47.27, 38.30, 27.28, 25.66 (3C), 18.36, 12.99, 4.66, 4.72; + − − MS (ESI): m/z calcd for SiC29H43O7 [M+H] 531.28, found 531.74.

3.1.5. Synthesis of Cyclohexane Analogue 10a, 10b and 11, 12 To a solution of 8a (1.87 g, 35.2 mmol, 1.0 eq) in MeOH (10 mL) was added LiOH (844 mg, 35.2 mmol, 1.0 eq) at 0 ◦C. After stirring for 0.5 h, the solution was quenched with saturated aqueous NH Cl, diluted by H O (20 mL). The resulting mixture was extracted with EtOAc (3 20 mL). 4 2 × The combined organic layers were dried over anhydrous Na2SO4, filtered, and concentrated under reduced pressure. The residue was purified by flash column chromatography on silica gel with petroleum ether/EtOAc (10: 1) to give 560 mg compounds 10a. To a solution of 10a (20 mg, 0.38 mmol, 1.0 eq) in THF (2 mL) was added 1.0 M TBAF (42 µL, 0.42 mmol, 1.1 eq) in THF. The mixture was stirred for 0.5 h. The reaction was diluted by H2O (10 mL), and extracted with EtOAc (3 10 mL). The combined organic layers were dried over anhydrous × Na2SO4, filtered and concentrated under reduced pressure. The residue was purified by flash column chromatography on silica gel with petroleum ether/EtOAc (4:1) to give 15 mg compounds 10b. To a solution of 10a (30 mg, 56 µmol, 1.0 eq) in dry DCM (5 mL) was added NaBH4 (7.6 mg, 113 µmol, 2.0 eq) at 0 ◦C. After stirred for 2 h, the reaction was then quenched with saturated aqueous NH Cl, and diluted by H O (5 mL). The resulting mixture was extracted with EtOAc (5 10 mL). 4 2 × The combined organic layers were dried over anhydrous Mg2SO4, filtered and concentrated under reduced pressure. The residue was redissolved in THF (3 mL), and treated with 1.0 M TBAF (62 µL, 62 µmol, 1.1 eq) in THF. The mixture was stirred for 0.5 h. The reaction was diluted by H2O (5 mL), and extracted with EtOAc (3 5 mL). The combined organic layers were dried over anhydrous × Na2SO4, filtered and concentrated under reduced pressure. The residue was purified by flash column chromatography on silica gel with petroleum ether/EtOAc (2: 1) to give 10 mg compounds 11. To a solution of 10a (500 mg, 0.94 mmol, 1.0 eq) in dry THF (10 mL) was added 1 M LiHMDS solution in THF (1.88 mL, 1.88 mmol, 2.0 eq) at -78 ◦C under argon atmosphere. After being stirred for 1 h, the reaction was added to TsCl (360 mg, 1.88 mmol, 2.0 eq) solution in THF (3 mL). The reaction stirred overnight, and quenched with saturated aqueous NH4Cl, diluted by H2O (10 mL). The resulting mixture was extracted with EtOAc (3 10 mL). The combined organic layers were dried over anhydrous × Mg2SO4, filtered and concentrated under reduced pressure. The residue was redissolved in DMF (3 mL), and added DMAP (12 mg, 94 µmol, 0.1 eq) and Et3N (1.3 mL, 9.4 mmol, 10 eq). The reaction was heated to 80 ◦C, and stirred overnight, then quenched with saturated aqueous NH4Cl, diluted by H2O (30 mL). The resulting mixture was extracted with EtOAc (3 20 mL). The combined organic layers × were dried over anhydrous Na2SO4, filtered and concentrated under reduced pressure. The residue was redissolved in THF (5 mL), and treated with 1.0 M TBAF (940 µL, 0.94 mmol, 1.0 eq) in THF. The mixture was stirred for 0.5 h. The reaction was diluted by H2O (10 mL), and extracted with EtOAc (3 10 mL). The combined organic layers were dried over anhydrous Na SO , filtered, and × 2 4 concentrated under reduced pressure. The residue was purified by flash column chromatography on silica gel with petroleum ether/EtOAc (4:1) to give 93 mg compounds 12 as a white solid. 4-(3-((tert-butyldimethylsilyl)oxy)-4-methoxyphenyl)-3-hydroxy-2-methyl-3-(3,4,5-trimethoxyphenyl) 1 cyclohexan-1-one (10a): white solid, 65% yield. H NMR (500 MHz, CDCl3 ): δ 6.56 (d, J = 7.9 Hz, 1H), 6.48 (d, J = 8.3 Hz, 2H), 3.77 (s, 3H), 3.68 (s, 3H), 3.35 (d, J = 11.2 Hz, 1H), 3.08 (d, J = 6.5 Hz, 1H), Int. J. Mol. Sci. 2020, 21, 1817 13 of 17

2.72 – 2.52 (m, 3H), 2.09 (d, J = 11.8 Hz, 1H), 1.97 (s, 1H), 0.92 (s, 9H), 0.80 (d, J = 6.3 Hz, 3H), 0.02 (s, 13 3H), -0.02 (s, 3H); C NMR (125 MHz, CDCl3 ): δ 210.40, 149.77, 144.33, 140.22, 136.61, 132.41, 121.77, 121.76, 111.50, 82.95, 60.84, 56.18, 56.16, 55.48, 53.62, 52.88, 41.56, 28.97, 25.67 (3C), 18.36, 8.15, -4.75, + -4.84; MS (ESI): m/z calcd for SiC29H43O7 [M+H] 531.28, found 531.63. 3-hydroxy-4-(3-hydroxy-4-methoxyphenyl)-2-methyl-3-(3,4,5-trimethoxyphenyl)cyclohexan-1-one (10b): white solid, 95% yield. 1H NMR (500 MHz, DMSO-D6): δ 8.49 (s, 1H), 6.57 (d, J = 1.5 Hz, 1H), 6.55 (d, J = 8.3 Hz, 1H), 6.35 (d, J = 8.2 Hz, 1H), 4.94 (s, 1H), 3.62 (s, 3H), 3.56 (s, 3H), 3.39 (dd, J = 12.6, 3.3 Hz, 1H), 3.34 - 3.31 (m, 1H), 2.71 (td, J = 13.3, 6.6 Hz, 1H), 2.46 2.31 (m, 2H), 1.90–1.83 (m, 1H), − 0.66 ppm (d, J = 6.7 Hz, 3H); 13C NMR (125 MHz, DMSO-D6 ): δ 211.12, 146.07, 145.68, 141.49, 135.85, 135.35, 120.18, 116.97, 111.59, 82.75, 60.42, 55.96, 53.75, 52.64, 41.40, 29.85, 8.79; MS (ESI): m/z calcd for + NaC23H28O7 [M+Na] 439.17, found 439.59. 6-(3-hydroxy-4-methoxyphenyl)-2-methyl-1-(3,4,5-trimethoxyphenyl)cyclohexane-1,3-diol (11): 1 white solid, two steps yield: 66%. H NMR (500 MHz, CDCl3): δ 7.66 (dd, J = 259.2, 6.5 Hz, 1H), 6.52 (s, 1H), 6.50 (s, 2H), 6.20 (d, J = 6.9 Hz, 1H), 6.11 (s, 1H), 5.48 (s, 1H), 3.90 (s, 3H), 3.84 (d, J = 10.0 Hz, 1H), 3.79 (s, 3H), 3.76 (s, 3H), 3.54 (s, 3H), 2.86 (d, J = 12.1 Hz, 1H), 2.48 (s, 1H), 2.24 (dd, J = 27.5, 12.3 13 Hz, 2H), 1.98 (s, 1H), 1.63 (dd, J = 23.6, 11.6 Hz, 2H), 0.86 (s, 3H); C NMR (125 MHz, CDCl3): δ 145.10, 144.96, 141.62, 136.48, 134.03, 130.25, 127.08, 120.49, 114.73, 110.00, 103.80, 101.65, 79.72, 72.37, 60.92, + 56.38, 56.09, 55.86, 53.40, 48.04, 35.44, 26.56, 11.48; MS (ESI): m/z calcd for NaC23H30O7 [M+Na] 441.19, found 441.68. 5-(3-hydroxy-4-methoxyphenyl)-1-methyl-6-(3,4,5-trimethoxyphenyl)-7-oxabicyclo[4.1.0]heptan-2- 1 one (12): white solid, three steps yield: 28%. H NMR (500 MHz, CDCl3): δ 6.80 (s, 1H), 6.60 (t, J = 5.6 Hz, 2H), 6.32 (s, 2H), 5.48 (s, 1H), 3.79 (s, 3H), 3.76 (s, 3H), 3.43 (d, J = 10.9 Hz, 1H), 2.73 (d, 13 J = 15.3 Hz, 1H), 2.44 – 2.29 (m, 2H), 1.91 – 1.84 (m, 1H), 1.13 (s, 3H); C NMR (125 MHz, CDCl3): δ 205.58, 152.78 (2C), 145.34, 145.26, 137.09,134.93, 132.67, 119.84, 114.35, 110.31, 72.21, 65.16, 60.82, 56.19, + 55.84 (2C), 45.67, 37.13, 26.33, 13.05; MS (ESI): m/z calcd for NaC23H26O7 [M+Na] 437.16, found 437.34.

3.2. Biological Activity Evaluation

3.2.1. Cell Culture The HepG2 cell line was originally acquired from Shanghai Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences. This cell line was grown in RPMI-1640 (Gibco) including 10% (v/v) thermally inactivated fetal bovine serum (FBS), penicillin (100 KU/L) and streptomycin (100 KU/L) at 37 ◦C in a 5% CO2 humidified incubator.

3.2.2. Cytotoxic Activity (SRB Assay) In vitro cytotoxic activity was evaluated by using the SRB colorimetric assay as previous methods [38,41]. HepG2 cells were treated with analogues for 48 h at various concentrations.

3.2.3. Apoptosis-Inducing Activity (PI/Annexin V-FITC Assay) The apoptosis in HepG2 cells was detected with a propidium iodide (PI)/Annexin V-FITC apoptosis detection kit (KeyGEN Biotech, Nanjing, China) as described previously [42]. After the treatment of 6b-(Z), 6b-(E) and CA-4 at the indicated concentrations for 48 h, HepG-2 cells were stained with PI and Annexin V-FITC and then were resuspended in 400 µL of binding buffer. Flow cytometric analysis (FACS Calibur; Becton-Dickinson, San Jose, CA, US) was performed. A total of 10000 cells were acquired per sample and data were analyzed using Cellquest software (Becton-Dickinson, San Jose, CA, US). Int. J. Mol. Sci. 2020, 21, 1817 14 of 17

3.2.4. Cell Cycle Distribution (PI Staining Assay) Cell cycle distribution on HepG2 cells was assessed by flow cytometry. PI staining assay (Sigma) was used for the cell cycle arrest of cells as described previously [40]. After treatment with test compound 1 at the indicated concentrations for 24 h, cells were centrifuged and fixed in 70% ethanol at 4 ◦C overnight and subsequently resuspended in PBS containing 100 µL RNase A and 400 µL PI. Cellular DNA content, for cell cycle distribution analysis, was measured using a FACS Calibur flow cytometer (Becton-Dickinson, San Jose, CA, US) and analyzed using a CellQuest software package (Becton-Dickinson, San Jose, CA, US). Twenty thousand events were collected per sample. Mean values from 3 independent experiments are presented. The results were analyzed using Modfit LT 3.0 software.

3.2.5. Immunofluorescence Analysis (DAPI /α-Tubulin-TRITC Assay) Immunofluorescence effects of 6b-(E) and CA-4 on the inhibition of tubulin polymerization in HepG-2 cells were measured as described previously [37]. HepG-2 cells were incubated in 8-well cell slide (Eppendorf AG, Hauppauge, NY). After treatment for 48 h at the indicated concentrations of 6b-(E) (0.3 µM) and CA-4 (0.02 µM), the cells was fixed with 4% paraformaldehyde, blocked with PBS containing 5% BSA and incubated with anti-α-tubulin-TRITC antibodies (1:500, Abcam Trading Company Ltd., Shanghai, China) and the secondary antibody (goat anti-rabbit IgG, 1:500, ZSGB-BIO Inc., Beijing, China). At the last step, the cells were stained with DAPI. Images of cells were acquired using an Olympus fluorescence microscope with a 20 objective. × 3.2.6. Tubulin Polymerization Assay Tubulin polymerization assay were performed in the light of the fluorescence based tubulin polymerization assay kit’s protocol (Cat. #BK011P, Inc., Denver, CO). Simplify, tubulin (10 mg/mL) containing buffer 1, tubulin glycerol buffer and GTP stock (100 mM) were incubated with 6b-(E) (5, 20 and 40 µM), paclitaxel (3 µM), CA-4 (3 µM) and DMSO for 50 min at 37 ◦C. The fluorescence increase was monitored every minute for the characterization of tubulin polymerization at 360 nm excitation wavelength and 460 nm emission wavelengths by microplate spectrophotometer (Synergy HTX, BioTek Instruments, Inc., Winooski, VT).

3.2.7. Statistics Analysis All the data were shown as mean standard deviation (SD) from at least three separate ± determinations with three duplicates. Statistical analysis was performed using GraphPad Prism software with the 2-tailed Student t-test.

3.3. Molecular Docking All docking analysis was performed by using the Surflex-Dock program in the Sybyl-X 2.0 software (Tripos Inc., St. Louis, MO). The X-ray crystal structure of tubulin protein (PDB ID: 1SA0) was acquired from the RSCB Protein Data Bank (http://www.rcsb.org/pdb). All bound waters and ligands were eliminated from the protein and the polar hydrogen was added to the proteins. Compounds 6b-(Z) and 6b-(E) were drawn and optimized automatically in the Sybyl-X 2.0 software, and saved as a mol2 file. Default parameters were used for Docking. The analysis of results was carried out by USCF Chimera software, which was downloaded from http://www.cgl.ucsf.edu/chimera.

4. Conclusions In this work, we constructed a series of novel CA-4 analogues on the cis-double bond containing diphenylethanone, cis-locked dihydrofuran, α-substituted diphenylethanone, cyclobutane and cyclohexane. Yet, most compounds exhibited weak cytotoxic activity against HepG2 cells. Unexpectedly, a by-product 6b (cis–trans isomers) in the preparation of diphenylethanone analogues Int. J. Mol. Sci. 2020, 21, 1817 15 of 17

was found to be the most potent cytotoxic agents against HepG2 cells with IC50 values of less than 0.5 µM. The two isomers of 6b both induced cellular apoptosis, significantly arrested cells in the G2/M phase, severely disrupted microtubule network in HepG2 cells. An in vitro tubulin polymerization assay revealed that 6b-(E) was a dose-dependent and effective inhibitor of the tubulin assembly. In addition, molecular docking showed that two stereoisomers of 6b bound similarly and efficiently at the colchicine binding site on tubulin well with CA-4 in similar distorted conformations. Although the anticancer activity of all novel compounds did not behave as excellently as CA-4, these results provided fundamental data for new developments of tubulin inhibitors. Additional insight on 6b will expand our research of olefin modification of CA-4 or isoCA-4 as tubulin polymerization agents.

Supplementary Materials: Supplementary materials can be found at http://www.mdpi.com/1422-0067/21/5/1817/ s1. Author Contributions: Conceptualization, M.-Y.S. and J.-J.T.; methodology, M.-Y.S., Y.-L.W., H.-R.W. and Q.-R.H.; software, M.-Y.S.; validation, J.-M.G.; formal analysis, Q.-R.H. and Y.-L.W.; investigation, H.-R.W. and T.-C.J.; resources, J.-J.T.; data curation, M.-Y.S. and T.-C.J.; writing—original draft preparation, M.-Y.S.; writing—review and editing, J.-J.T.; supervision, J.-J.T. and J.-M.G.; project administration, Q.-R.H.; funding acquisition, J.-M.G. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by the National Natural Science Foundation of China (21542016) and Yangling Demonstration Zone Science and Technology Plan Project (2017NY-07), China. Acknowledgments: This work was given some technical support by Instrument Shared Platform of Northwest A&F University. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Pettit, G.R.; Singh, S.B.; Hamel, E.; Lin, C.M.; Alberts, D.S.; Garciakendall, D. Isolation and structure of the strong cell growth and tubulin inhibitor combretastatin A-4. Experientia 1989, 45, 209–211. [CrossRef] [PubMed] 2. Hamel, E. Antimitotic natural products and their interactions with tubulin. Med. Res. Rev. 1996, 16, 207–231. [CrossRef] 3. Woods, J.A.; Hadfield, J.A.; Pettit, G.R.; Fox, B.W.; Mcgown, A.T. The Interaction with Tubulin of a Series of Stilbenes Based on Combretastatin a-4. Brit. J. Cancer 1995, 71, 705–711. [CrossRef] 4. Young, S.L.; Chaplin, D.J. Combretastatin A4 phosphate: Background and current clinical status. Expert Opin. Investig. Drugs 2004, 13, 1171–1182. [CrossRef][PubMed] 5. Hori, K.; Saito, S. Microvascular mechanisms by which the combretastatin A-4 derivative AC7700 (AVE8062) induces tumour blood flow stasis. Br. J. Cancer 2003, 89, 1334–1344. [CrossRef][PubMed] 6. Available online: www.mateon.com (accessed on 15 February 2020). 7. Lipeeva, A.V.; Shults, E.E.; Shakirov, M.M.; Pokrovsky, M.A.; Pokrovsky, A.G. Synthesis and Cytotoxic Activity of a New Group of Heterocyclic Analogues of the Combretastatins. Molecules 2014, 19, 7881–7900. [CrossRef] 8. Jin, Y.H.; Qi, P.; Wang, Z.W.; Shen, Q.R.; Wang, J.; Zhang, W.G.; Song, H.R. 3D-QSAR Study of Combretastatin A-4 Analogs Based on Molecular Docking. Molecules 2011, 16, 6684–6700. [CrossRef] 9. Piekus-Slomka, N.; Mikstacka, R.; Ronowicz, J.; Sobiak, S. Hybrid cis-stilbene Molecules: Novel Anticancer Agents. Int. J. Mol. Sci. 2019, 20, 1300. [CrossRef] 10. Schmitt, F.; Gosch, L.C.; Dittmer, A.; Rothemund, M.; Mueller, T.; Schobert, R.; Biersack, B.; Volkamer, A.; Hopfner, M. Oxazole-Bridged Combretastatin A-4 Derivatives with Tethered Hydroxamic Acids: Structure-Activity Relations of New Inhibitors of HDAC and/or Tubulin Function. Int. J. Mol. Sci. 2019, 20, 383. [CrossRef] 11. Tron, G.C.; Pirali, T.; Sorba, G.; Pagliai, F.; Busacca, S.; Genazzani, A.A. Medicinal Chemistry of Combretastatin A4: Present and Future Directions. J. Med. Chem. 2006, 49, 3033–3044. [CrossRef] 12. Nam, N.H. Combretastatin A-4 analogues as antimitotic antitumor agents. Curr. Med. Chem. 2003, 10, 1697–1722. [CrossRef][PubMed] Int. J. Mol. Sci. 2020, 21, 1817 16 of 17

13. Aprile, S.; Del Grosso, E.; Tron, G.C.; Grosa, G. In vitro metabolism study of combretastatin A-4 in rat and human liver microsomes. Drug Metab. Dispos. 2007, 35, 2252–2261. [CrossRef][PubMed] 14. Nakamura, M.; Kajita, D.; Matsumoto, Y.; Hashimoto, Y. Design and synthesis of silicon-containing tubulin polymerization inhibitors: Replacement of the moiety of combretastatin A-4 with a silicon linker. Bioorg. Med. Chem. 2013, 21, 7381–7391. [CrossRef] 15. Pang, Y.Q.; Yan, J.; An, B.J.; Huang, L.; Li, X.S. The synthesis and evaluation of new butadiene derivatives as tubulin polymerization inhibitors. Bioorg. Med. Chem. 2017, 25, 3059–3067. [CrossRef][PubMed] 16. Chen, H.; Li, Y.; Sheng, C.; Lv, Z.; Dong, G.; Wang, T.; Liu, J.; Zhang, M.; Li, L.; Zhang, T.; et al. Design and synthesis of cyclopropylamide analogues of combretastatin-A4 as novel microtubule-stabilizing agents. J. Med. Chem. 2013, 56, 685–699. [CrossRef][PubMed] 17. Ty, N.; Pontikis, R.; Chabot, G.G.; Devillers, E.; Quentin, L.; Bourg, S.; Florent, J.C. Synthesis and biological evaluation of enantiomerically pure cyclopropyl analogues of combretastatin A4. Bioorg. Med. Chem. 2013, 21, 1357–1366. [CrossRef] 18. Zhou, P.; Liu, Y.; Zhou, L.; Zhu, K.; Feng, K.; Zhang, H.; Liang, Y.; Jiang, H.; Luo, C.; Liu, M.; et al. Potent Antitumor Activities and Structure Basis of the Chiral beta-Lactam Bridged Analogue of Combretastatin A-4 Binding to Tubulin. J. Med. Chem. 2016, 59, 10329–10334. [CrossRef] 19. Theeramunkong, S.; Caldarelli, A.; Massarotti, A.; Aprile, S.; Caprioglio, D.; Zaninetti, R.; Teruggi, A.; Pirali, T.; Grosa, G.; Tron, G.C.; et al. Regioselective Suzuki coupling of dihaloheteroaromatic compounds as a rapid strategy to synthesize potent rigid combretastatin analogues. J. Med. Chem. 2011, 54, 4977–4986. [CrossRef] 20. Pirali, T.; Busacca, S.; Beltrami, L.; Imovilli, D.; Pagliai, F.; Miglio, G.; Massarotti, A.; Verotta, L.; Tron, G.C.; Sorba, G.; et al. Synthesis and cytotoxic evaluation of combretafurans, potential scaffolds for dual-action antitumoral agents. J. Med. Chem. 2006, 49, 5372–5376. [CrossRef] 21. Assadieskandar, A.; Amini, M.; Ostad, S.N.; Riazi, G.H.; Cheraghi-Shavi, T.; Shafiei, B.; Shafiee, A. Design, synthesis, cytotoxic evaluation and tubulin inhibitory activity of 4-aryl-5-(3,4,5- trimethoxyphenyl)-2-alkylthio-1H-imidazole derivatives. Bioorg. Med. Chem. 2013, 21, 2703–2709. [CrossRef] 22. Wang, Z.; Yang, Q.; Bai, Z.; Sun, J.; Jiang, X.; Song, H.; Wu, Y.; Zhang, W. Synthesis and biological evaluation of 2,3-diarylthiophene analogues of combretastatin A-4. Med. Chem. Commun. 2015, 6, 971–976. [CrossRef] 23. Simoni, D.; Grisolia, G.; Giannini, G.; Roberti, M.; Rondanin, R.; Piccagli, L.; Baruchello, R.; Rossi, M.; Romagnoli, R.; Invidiata, F.P.; et al. Heterocyclic and phenyl double-bond-locked combretastatin analogues possessing potent apoptosis-inducing activity in HL60 and in MDR cell lines. J. Med. Chem. 2005, 48, 723–736. [CrossRef][PubMed] 24. Li, Y.H.; Zhang, B.; Yang, H.K.; Li, Q.; Diao, P.C.; You, W.W.; Zhao, P.L. Design, synthesis, and biological evaluation of novel alkylsulfanyl-1,2,4-triazoles as cis-restricted combretastatin A-4 analogues. Eur. J. Med. Chem. 2017, 125, 1098–1106. [CrossRef][PubMed] 25. Subba Rao, A.V.; Swapna, K.; Shaik, S.P.; Lakshma Nayak, V.; Srinivasa Reddy, T.; Sunkari, S.; Shaik, T.B.; Bagul, C.; Kamal, A. Synthesis and biological evaluation of cis-restricted triazole/tetrazole mimics of combretastatin-benzothiazole hybrids as tubulin polymerization inhibitors and apoptosis inducers. Bioorg. Med. Chem. 2017, 25, 977–999. [CrossRef] 26. Salehi, M.; Amini, M.; Ostad, S.N.; Riazi, G.H.; Assadieskandar, A.; Shafiei, B.; Shafiee, A. Synthesis, cytotoxic evaluation and molecular docking study of 2-alkylthio-4-(2,3,4-trimethoxyphenyl)-5-aryl-thiazoles as tubulin polymerization inhibitors. Bioorg. Med. Chem. 2013, 21, 7648–7654. [CrossRef][PubMed] 27. Wang, F.; Yang, Z.; Liu, Y.; Ma, L.; Wu, Y.; He, L.; Shao, M.; Yu, K.; Wu, W.; Pu, Y.; et al. Synthesis and biological evaluation of diarylthiazole derivatives as antimitotic and antivascular agents with potent antitumor activity. Bioorg. Med. Chem. 2015, 23, 3337–3350. [CrossRef] 28. Xu, Q.; Wang, Y.; Xu, J.; Sun, M.; Tian, H.; Zuo, D.; Guan, Q.; Bao, K.; Wu, Y.; Zhang, W. Synthesis and Bioevaluation of 3,6-Diaryl-[1,2,4]triazolo[4,3-b] Pyridazines as Antitubulin Agents. ACS Med. Chem. Lett. 2016, 7, 1202–1206. [CrossRef] 29. Messaoudi, S.; Treguier, B.; Hamze, A.; Provot, O.; Peyrat, J.F.; De Losada, J.R.; Liu, J.M.; Bignon, J.; Wdzieczak-Bakala, J.; Thoret, S.; et al. Isocombretastatins a versus combretastatins a: The forgotten isoCA-4 isomer as a highly promising cytotoxic and antitubulin agent. J. Med. Chem. 2009, 52, 4538–4542. [CrossRef] Int. J. Mol. Sci. 2020, 21, 1817 17 of 17

30. Tréguier, B.; Hamze, A.; Provot, O.; Brion, J.-D.; Alami, M. Expeditious synthesis of 1,1-diarylethylenes related to isocombretastatin A-4 (isoCA-4) via palladium-catalyzed arylation of N-tosylhydrazones with aryl triflates. Tetrahedron Lett. 2009, 50, 6549–6552. [CrossRef] 31. Stocker, V.; Ghinet, A.; Leman, M.; Rigo, B.; Millet, R.; Farce, A.; Desravines, D.; Dubois, J.; Waterlot, C.; Gautret, P. On the synthesis and biological properties of isocombretastatins: A case of ketone homologation during Wittig reaction attempts. Rsc Adv. 2013, 3, 3683–3696. [CrossRef] 32. Aziz, J.; Brachet, E.; Hamze, A.; Peyrat, J.F.; Bernadat, G.; Morvan, E.; Bignon, J.; Wdzieczak-Bakala, J.; Desravines, D.; Dubois, J.; et al. Synthesis, biological evaluation, and structure-activity relationships of tri- and tetrasubstituted olefins related to isocombretastatin A-4 as new tubulin inhibitors. Org. Biomol. Chem. 2013, 11, 430–442. [CrossRef][PubMed] 33. Rasolofonjatovo, E.; Provot, O.; Hamze, A.; Rodrigo, J.; Bignon, J.; Wdzieczak-Bakala, J.; Desravines, D.; Dubois, J.; Brion, J.D.; Alami, M. Conformationnally restricted naphthalene derivatives type isocombretastatin A-4 and isoerianin analogues: Synthesis, cytotoxicity and antitubulin activity. Eur. J. Med. Chem. 2012, 52, 22–32. [CrossRef][PubMed] 34. Renko, D.; Provot, O.; Rasolofonjatovo, E.; Bignon, J.; Rodrigo, J.; Dubois, J.; Brion, J.D.; Hamze, A.; Alami, M. Rapid synthesis of 4-arylchromenes from ortho-substituted alkynols: A versatile access to restricted isocombretastatin A-4 analogues as antitumor agents. Eur. J. Med. Chem. 2015, 90, 834–844. [CrossRef][PubMed] 35. Soussi, M.A.; Provot, O.; Bernadat, G.; Bignon, J.; Wdzieczak-Bakala, J.; Desravines, D.; Dubois, J.; Brion, J.D.; Messaoudi, S.; Alami, M. Discovery of azaisoerianin derivatives as potential antitumors agents. Eur. J. Med. Chem. 2014, 78, 178–189. [CrossRef] 36. Rasolofonjatovo, E.; Provot, O.; Hamze, A.; Rodrigo, J.; Bignon, J.; Wdzieczak-Bakala, J.; Lenoir, C.; Desravines, D.; Dubois, J.; Brion, J.-D.; et al. Design, synthesis and anticancer properties of 5-arylbenzoxepins as conformationally restricted isocombretastatin A-4 analogs. Eur. J. Med. Chem. 2013, 62, 28–39. [CrossRef] 37. Song, M.Y.; Cao, C.Y.; He, Q.R.; Dong, Q.M.; Li, D.; Tang, J.J.; Gao, J.M. Constructing novel dihydrofuran and dihydroisoxazole analogues of isocombretastatin-4 as tubulin polymerization inhibitors through [3+2] reactions. Bioorgan. Med. Chem. 2017, 25, 5290–5302. [CrossRef] 38. Skehan, P.; Storeng, R.; Scudiero, D.; Monks, A.; McMahon, J.; Vistica, D.; Warren, J.T.; Bokesch, H.; Kenney, S.; Boyd, M.R. New colorimetric cytotoxicity assay for anticancer-drug screening. J. Natl. Cancer Inst. 1990, 82, 1107–1112. [CrossRef] 39. Simoni, D.; Romagnoli, R.; Baruchello, R.; Rondanin, R.; Rizzi, M.; Pavani, M.G.; Alloatti, D.; Giannini, G.; Marcellini, M.; Riccioni, T.; et al. Novel combretastatin analogues endowed with antitumor activity. J. Med. Chem. 2006, 49, 3143–3152. [CrossRef] 40. Tang, J.J.; Fan, G.J.; Dai, F.; Ding, D.J.; Wang, Q.; Lu, D.L.; Li, R.R.; Li, X.Z.; Hu, L.M.; Jin, X.L.; et al. Finding more active antioxidants and cancer chemoprevention agents by elongating the conjugated links of resveratrol. Free Radic. Biol. Med. 2011, 50, 1447–1457. [CrossRef] 41. Dong, S.; Tang, J.-J.; Zhang, C.-C.; Tian, J.-M.; Guo, J.-T.; Zhang, Q.; Li, H.; Gao, J.-M. Semisynthesis and in vitro cytotoxic evaluation of new analogues of 1-O-acetylbritannilactone, a sesquiterpene from Inula britannica. Eur. J. Med. Chem. 2014, 80, 71–82. [CrossRef] 42. Tang, J.J.; Dong, S.; Han, Y.Y.; Lei, M.; Gao, J.M. Synthesis of 1-O-acetylbritannilactone analogues from Inula britannica and in vitro evaluation of their anticancer potential. Med. Chem. Comm. 2014, 5, 1584–1589. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).