3-Vinylazetidin-2-Ones: Synthesis, Antiproliferative and Tubulin Destabilizing Activity in MCF-7 and MDA-MB-231 Breast Cancer Cells

Microtubule-targeted drugs are essential chemotherapeutic agents for various types of cancer. A series of 3-vinyl-β-lactams (2-azetidinones) were designed, synthesized and evaluated as potential tubulin polymerization inhibitors, and for their antiproliferative effects in breast cancer cells. These compounds showed potent activity in MCF-7 breast cancer cells with an IC50 value of 8 nM for compound 7s 4-[3-Hydroxy-4-methoxyphenyl]-1-(3,4,5-trimethoxyphenyl)-3-vinylazetidin-2-one) which was comparable to the activity of Combretastatin A-4. Compound 7s had minimal cytotoxicity against both non-tumorigenic HEK-293T cells and murine mammary epithelial cells. The compounds inhibited the polymerisation of tubulin in vitro with an 8.7-fold reduction in tubulin polymerization at 10 μM for compound 7s and were shown to interact at the colchicine-binding site on tubulin, resulting in significant G2/M phase cell cycle arrest. Immunofluorescence staining of MCF-7 cells confirmed that β-lactam 7s is targeting tubulin and resulted in mitotic catastrophe. A docking simulation indicated potential binding conformations for the 3-vinyl-β-lactam 7s in the colchicine domain of tubulin. These compounds are promising candidates for development as antiproiferative microtubule-disrupting agents.


Introduction
Antimitotic agents such as taxol and the vinca alkaloids vinblastine and vincristine are a major class of drugs used clinically in the treatment of many cancers [1][2][3]. Microtubule-destabilizing agents (e.g., vinblastine) typically bind with tubulin at the vinca alkaloid site [4], while colchicine 1 exerts its biological effects at the intrasubunit interface within a tubulin dimer [5]. Stilbene-based compounds have attracted the attention of chemists and pharmacologists due to their many biological properties such as structure-activity relationships for the series. On this basis, we now aimed to investigate a new series of novel 3-vinylazetidinones compounds with an improved biochemical profile particularly in triple negative breast cancer for potential development in preclinical study of breast cancer as tubulin destabilising agents. Therefore, we focused our efforts on the preparation of a library of 1,4-diarylazetidin-2-ones which contain a vinyl substituent at C-3. The synthesis of phosphate esters and amino acid amide type prodrugs of the most potent 1,4-diarylazetidin-2-ones were examined, together with the antiproliferative and tubulin targeting effects.

Chemistry: Synthesis of β-lactams
There are many synthetic routes available for the construction of the β-lactam ring [43,48]. The choice of route depends on the structural features required in the final product. In the present work, the Staudinger reaction between an imine and a ketene was chosen for the formation of the βlactam ring because of its ease of use, adaptability for use with structurally diverse imines and acid chlorides, and readily available starting materials. A series of analogues with a variety of substituents at C4 of the β-lactam ring B was synthesized from the appropriate imines. The preparation of the Schiff bases 5a-5r was achieved by the condensation of the appropriately substituted benzaldehyde with the 3,4,5-trimethoxyaniline in ethanol in the presence of a catalytic amount of sulphuric acid, (Scheme 1). The 3,4,5-trimethoxy substituted A-Ring of CA-4 plays an important role in inhibiting tubulin polymerisation, and is confirmed in the docking of CA-4 in tubulin [49]. The substituents located at the para-position of C-4 aryl Ring B included halogens (compounds 5a-5c), nitro (5d), dimethylamino (5e), methyl (5g), alkoxyl (5h-5j), phenoxy (5k), benzyloxy (5l), nitrile (5q) and thiomethyl (5r) together with naphthyl (compounds 5m and 5n). 5s was similarly obtained by reaction of 4-methoxybenzaldehyde with 3,5-dimethoxyaniline. For the synthesis of β-lactam derivatives with a phenolic hydroxy group to mimic Ring B of CA-4, it was necessary to use the benzyl ether 5l and tert-butyldimethylsilyl ether 5o. A further series of Schiff bases (6a-6k) was obtained from 3,4,5-trimethoxybenzaldehyde with appropriate anilines using the same procedure as above, (Scheme 1). An example of the crystal structure of the imine 6k is displayed in Figure 2, showing the E configuration of the imine N1-C2 bond (bond length 1.278(2) Å) (Table 1).
We have previously investigated the antiproliferative and SERM (selective estrogen receptor modulator) activity of the azetidin-2-one(β-lactam) scaffold [44] and also demonstrated the effectiveness of 1,4-diarylazetidin-2-ones in breast cancer cell lines as tubulin targeting agents. [45,46]. These compounds also demonstrated both anti-angiogenic effects in MDA-MB-231 breast adenocarcinoma cells. In addition, we established that these compounds inhibited the migration of MDA-MB-231 cells indicating a potential anti-metastatic function for these compounds [47]. To further our understanding of the antiproliferative activity of these compounds, we wished to investigate the design, synthesis and evaluation of a series of azetidin-2-ones containing a vinyl substituent at C3 of the azetidin-2-one ring, and to explore the effect of this hydrophobic substituent on the biological activity of these compounds in which the cis configuration (Rings A and B) is locked into the azetidin-2-one ring structure. The introduction of this vinyl substituent at C-3 also allowed us to examine further chemical transformations of the alkene, and to determine structure-activity relationships for the series. On this basis, we now aimed to investigate a new series of novel 3-vinylazetidinones compounds with an improved biochemical profile particularly in triple negative breast cancer for potential development in preclinical study of breast cancer as tubulin destabilising agents. Therefore, we focused our efforts on the preparation of a library of 1,4-diarylazetidin-2-ones which contain a vinyl substituent at C-3. The synthesis of phosphate esters and amino acid amide type prodrugs of the most potent 1,4-diarylazetidin-2-ones were examined, together with the antiproliferative and tubulin targeting effects.
Many potentially useful CA-4 derivatives contain the amino substituent replacing the phenol on ring B and have shown interesting biochemical activity [52]. We were interested in the preparation of β-lactam CA-4 type compounds containing an amino substituent in Ring B, and the subsequent conversion to a water-soluble prodrug by conjugation with an amino acid. The nitro containing C-3-vinyl-β-lactam 7p was successfully reduced to the amino product 7t using zinc dust in the presence of acetic acid (Scheme 2). To investigate the effect of replacement of the 3,4,5-trimethoxy ring A of CA-4 with 3,5-dimethoxy substituted ring A, the β-lactam 7u was prepared in a similar route from the imine 5s. Tripodi et al. reported that 3,5-dimethoxy substituted ring A compounds demonstrated comparable activity to the β-lactam compounds containing the 3,4,5-trimethoxy ring A of CA-4 [53]. A further series of β-lactam compounds (8a-8k), was also prepared containing the 3,4,5-trimethoxyphenyl substituent (Ring A of the Combretastatin A-4) at C-4 position, (Scheme 2).
The products of the Staudinger reaction with imines and crotonyl chloride show IR absorptions at approximately ν 1750 cm −1 characteristic of the carbonyl group of the β-lactam ring. All of the β-lactams were obtained with exclusively trans stereochemistry, with coupling constants of 1-3 Hz for the β-lactam ring protons (e.g., for compound 7s, H-4 is identified as a doublet δ 4.69, J 3,4 = 2.52 Hz). Coupling constants of 5-6 Hz are usually observed for β-lactams with cis stereochemistry [46]. Subsequent to our initial biochemical evaluation of the 3-vinyl-β-lactam CA-4 analogues, a further series of 3-substituted β-lactams was prepared from 3-unsubstituted β-lactams by aldol type reaction with a suitable electrophile [54,55]. We were particularly interested in the introduction of modified alkene substituents at C-3, due to the exceptional biochemical activity displayed by the 3-vinyl β-lactam 7s. Lithium enolates of 3-unsubstituted β-lactams 9a and 9b were reacted with selected aldehydes and ketones to provide alcohol products 10a-10i, (Scheme 3). The β-lactams 9a and 9b were obtained via the Reformatsky reaction of ethyl bromoacetate with imines 5h and 5o using microwave conditions. Treatment of 9b with tBAF afforded the phenol 9c. Similarly, for the preparation of compounds 10a, 10c, 10d, 10f, 10g, 10h the initially obtained tBDMS ether intermediate was subsequently deprotected in situ using tBAF to yield the desired phenolic product. The enolate chemistry is stereoselective, favouring trans stereochemistry for the products. The presence of a diastereomeric mixture for products is confirmed from the 1 H NMR spectra (e.g., for 10h where H-3 and H-4 appear as two sets of doublets, δ 3.20 and δ 4.83 respectively, with J = 2.4 Hz, ratio H 3 /H 4 1.14:1.00). To investigate the role of the alcohol group at C-5 in the biochemical activity of the products 10a-10h, the alcohol 10i was oxidised to the corresponding ketone 11 using pyridinium chlorochromate. An alternative route to 11 was identified where treatment of the 3-unsubstituted β-lactam 9b with LDA followed by addition of acetyl chloride to gave the desired product 11 but only in low yields (11%) with the alcohol 10a also isolated (22%), (Scheme 3).
To further investigate the role of the 3-vinyl substitution pattern in the biochemical activity of. β-lactams, a 3-ethylidene product 12 was investigated. The initial route attempted involved the chlorination of the alcohol 10i using thionyl chloride followed by dehydrohalogentation with a suitable base such as DBU. However, a more successful method to give the 3-ethylidene β-lactams was the dehydration of the alcohol 10i under Mitsunobu conditions and subsequent deprotection by treatment with tBAF to yield 12 in 63% yield overall, (Scheme 3). The Peterson olefination of 3-unsubstituted β-lactams has also been reported by Kano et al. as an alternative route to 3-ethylidene β-lactams [56], while the Mitsunobu reaction for the dehydration of alcohols has been described by Plantan et al. in the synthesis of a trinem β-lactamase inhibitor [57]. The product 12 was obtained as a mixture of Z/E isomers in a 1:1 ratio. The configuration of the separated isomers was determined by examining the chemical shifts associated with the C-6 methyl protons. The further downfield doublet signal (δ 2.05, J = 4.16 Hz) is more deshielded, and so is assigned to the Z isomer while the signal at δ 1.62, (J = 4.40 Hz) is assigned to the E isomer [51].
The introduction of a diol functionality at C-3 was now explored. The diol 13 was synthesised in 39% yield by the oxidation of the alkene 7s with osmium tetroxide (Scheme 3). The 1 H NMR spectrum for 13 clearly illustrates the formation of a diastereomeric product. H-3 appears as a pair of double doublets at δ 3.16 (0.7H) and δ 3.19 (0.3H) with coupling constants of 2.42 Hz and 5.55 Hz, while H-4 appears as two separate doublets at δ 4.90 (0.3H) and δ 5.00 (0.7H), J = 2.37 Hz. To further investigate the role of the 3-vinyl substitution pattern in the biochemical activity of. β-lactams, a 3-ethylidene product 12 was investigated. The initial route attempted involved the chlorination of the alcohol 10i using thionyl chloride followed by dehydrohalogentation with a suitable base such as DBU. However, a more successful method to give the 3-ethylidene β-lactams was the dehydration of the alcohol 10i under Mitsunobu conditions and subsequent deprotection by treatment with tBAF to yield 12 in 63% yield overall, (Scheme 3). The Peterson olefination of 3unsubstituted β-lactams has also been reported by Kano et al. as an alternative route to 3ethylidene β-lactams [56], while the Mitsunobu reaction for the dehydration of alcohols has been described by Plantan et al. in the synthesis of a trinem β-lactamase inhibitor [57]. The product 12 was obtained as a mixture of Z/E isomers in a 1:1 ratio. The configuration of the separated isomers was determined by examining the chemical shifts associated with the C-6 methyl protons. The further downfield doublet signal (δ 2.05, J = 4.16 Hz) is more deshielded, and so is assigned to the Z isomer while the signal at δ 1.62, (J = 4.40 Hz) is assigned to the E isomer [51].
The introduction of a diol functionality at C-3 was now explored. The diol 13 was synthesised in 39% yield by the oxidation of the alkene 7s with osmium tetroxide (Scheme 3). The 1 H NMR spectrum for 13 clearly illustrates the formation of a diastereomeric product. H-3 appears as a pair The amino acid alanine was chosen for prodrug formation of the β-lactam 7t [58]. The protected amino acid prodrug 14 was obtained from 7t using the coupling agent DCC with HOBt in dry DMF (Scheme 4). The FMOC protecting group was easily removed from 14 by treatment with 2N sodium hydroxide over 24 h to afford the amino acid prodrug conjugate 15 (57%). Controlled esterification of the phenolic β-lactams 7s and 9c with dibenzyl phosphite using diisopropylethylamine and dimethylaminopyridine afforded dibenzyl phosphate β-lactams 16a and 16b respectively, (Scheme 4). The dimethyl and diethylethyl phosphate esters of compound 9c, 16c and 16d respectively, were also prepared (Scheme 4). The phosphate 17a was obtained by treatment of dibenzylphosphate ester 16a with bromotrimethylsilane. Hydrogenation of the dibenzylphosphate ester 16b with palladium/carbon catalyst removed the dibenzyl protecting groups and also reduced the double bond at C-3 position of the β-lactam ring to afford phosphate 17c. For the preparation of compound 17b, where removal of the benzyl protecting groups and retention of the double bond was required, reaction of the dibenzyl phosphate ester 16b with bromotrimethylsilane was effective.
Preliminary stability studies of the representative β-lactam 7s were carried out at acidic, neutral and basic conditions (pH 4, 7.4 and 9) and in plasma using HPLC. The half-life (t 1 2 ) was determined to be greater than 24 h at pH 4, 7.4 and 9 and in plasma for compound 7s. The phosphate esters 17b and 17c were also found to be stable over the range of pH and in plasma, with half-life (t 1 2 ) determined to be greater than 24 h. The cleavage of phosphate prodrugs 17b and 17c was also investigated in whole blood. They were cleaved much more rapidly in whole blood (62% and 34% remaining after 6 h respectively) than in human plasma (94% and 92% remaining after 6 h respectively). Based on this stability study the β-lactam 7s would be suitable for further development.

X-Ray Structural Study
The X-ray crystal structures of compounds 7h, 8i, 8k and 8h are displayed in Figure 3 and confirm the structural assignment. The crystal data for the compounds are shown in Tables 1 and  2. For each compound the two aryl rings at N-1 and C-4 position are in a pseudo cis arrangement while the phenyl ring at C4 and the alkene group are on opposite sides of the β-lactam (trans configuration). The structure of the compounds 7h, 8h, 8i and 8k clearly demonstrated a noncoplanar configuration for rings A and B of the β-lactams, with the β-lactam ring providing a rigid scaffold. For compound 7h even though both enantiomers are present, the compound crystallizes

X-Ray Structural Study
The X-ray crystal structures of compounds 7h, 8i, 8k and 8h are displayed in Figure 3 and confirm the structural assignment. The crystal data for the compounds are shown in Tables 1 and 2. For each compound the two aryl rings at N-1 and C-4 position are in a pseudo cis arrangement while the phenyl ring at C4 and the alkene group are on opposite sides of the β-lactam (trans configuration). The structure of the compounds 7h, 8h, 8i and 8k clearly demonstrated a non-coplanar configuration for rings A and B of the β-lactams, with the β-lactam ring providing a rigid scaffold. For compound 7h even though both enantiomers are present, the compound crystallizes out in a chiral space group. The trans configuration of the aryl rings A and B at C-3 and C-4 is also evident. The dihedral angle H3/H4 is observed for compounds 7h, 8h, 8i and 8k respectively, which is consistent with the small trans coupling constant observed in the 1 H NMR spectrum of 2.00 Hz, 2.52 Hz, 2.48 Hz and 2.44 Hz respectively for these compounds.  The ring A/B torsional angles for compounds 7h, 8h, 8i and 8k were observed as −59.5 • , 59.7 • , −73.5 • and −77.0 • respectively; these values are significantly greater than those observed for the corresponding rings A/B in the DAMA-colchicine 1b [5], Combretastatin A-4 2a [61] and related 4-arylcoumarin [62] as 53 • , 55 • and 48.3 • respectively ( Table 2). The azetidinone N1-C4 bond length was observed at 1.372(3) Å, 1.376(3) Å, 1.367(2) Å and 1.3767(18) Å for compounds 7h, 8h, 8i and 8k respectively, which compares with 1.334(4)Å reported for the alkene C=C of combretastatin A-4 [61]. The C26-C27 alkene bond length for 7h, 8h, 8i and 8k were observed at 1.303(3) Å, 1.3174 Å, 1.308(3) Å and 1.316(2) Å respectively, while the alkene C=C bond length for iso-combretastatin CA-4 has been reported as 1.329(3) Å [63]. The C-N bonds lengths in the β-lactam ring are unequal with N1-C4 bond lengths of 1.487 (3)

In vitro Antiproliferative Activities
The synthesized compounds were first evaluated for their antiproliferative activity against the human breast cancer cell line MCF-7 and compared with CA-4 as a reference compound (IC 50 = 3.9 nM) [64,65]. The results are shown in Table 3 (7a-7n, 7p-7t, 8a-8k), and Table 4 (10a-h, 11-13, 15 and 17a-c). All β-lactams were evaluated as the trans isomer. The most potent compounds were identified as 7s and 7t, with IC 50 values of 8 νM and 17 nM respectively. Compound 7s is a direct analogue of CA-4, while 7t is the corresponding amino compound and this type of substitution has been demonstrated to confer potency in many CA-4 analogues [52]. Compounds having the methoxy, ethoxy and thiomethyl substituents at C-4 of Ring B displayed potent antiproliferative effects, with IC 50 values of 20 nM, 37 nM and 51 nM respectively for compounds 7h, 7i and 7r respectively. The halo substituted compounds, 7b and 7c and 4-methyl compound 7g were less effective with IC 50  The IC 50 of compound 7u containing the 3,5-dimethoxyphenyl Ring A was determined as 170 nM in MCF-7 cells, demonstrating retention of antiproliferative potency with slightly reduced activity compared to the 3,4,5-trimethoxy ring A substituted compound 7h. This observation could infer that the para-methoxy aryl group is less important for activity and the 3,5-dimethoxyaryl substituted Ring A is favourable for interaction of the molecule with the colchicine binding site of tubulin [53]. Compounds 8a-8k containing the 3,4,5-trimethoxyphenyl substituent (Ring A of CA-4) at the C-4 position were generally observed to have poorer antiproliferative activity than the corresponding compounds 7a-7t, containing the 3,4,5-trimethoxyphenyl substituent (Ring A of the Combretastatin A-4) at the N-1 position, ( Table 3). The exceptions were compounds 8a (4-fluoro) and 8j (4-NHCOCH 3 ) with IC 50 values of 1.066 µM and 4.024 µM respectively. The relative positions of the 3,4,5-trimethoxyphenyl Ring A and Ring B on the β-lactam ring at positions N-1 and C-4 have a significant effect on the antiproliferative activity of the compounds as we previously reported [45].
The effects of various structural modifications on the activity of the more potent 3-vinylazetidinones were next explored, ( Table 4). The most potent compound in this series is the 3-styryl containing compound 10f, with IC 50 = 46 nM. The alcohol 10a showed interesting activity (65 nM) while the introduction of an additional methyl group at C-5 to afford the alcohol 10d resulted in reduced efficacy with IC 50 = 544 nM. The diol 13 also proved noteworthy with IC 50 = 69 nM. The 3-acetyl compound 11 and 3-ethylidene compound 12 resulted in similar antiproliferative effects (IC 50 = 414 nM and 502 nM respectively). Additional compounds containing the hydroxyalkene substituent at C-3 (e.g., compounds 10b, 10c, 10e, 10g, 10h) were found to be moderately active (IC 50 values 288-570 nM).
The amino acid prodrug amide 15 was evaluated in MCF-7 breast cancer cells to determine if it retained any antiproliferative activity when compared with the parent compound 7t which was extremely potent with IC 50 = 17 nM. The IC 50 for 15 (3.251 µM) was lower than expected; however metabolic activation in vivo may be required for the hydrolysis of the amide [67]. The phosphate esters 17a-17c displayed impressive antiproliferative activity, with IC 50 values of 22 nM, 27 nM and 21 nM respectively ( Table 4). The IC 50 values for the corresponding phenols 7s and 9c in MCF-7 cells are 8 nM and 17 nM respectively. Comparison of the 3-vinyl 17b (IC 50 = 27 nM) with the 3-ethyl 17c (IC 50 = 21 nM) indicated that introduction of the 3-vinyl or 3-ethyl substituent, together with the 3-unsubstituted 17a (22 nM) retains potency and optimum activity. The potent activity displayed for the phosphate esters 17a-17c, together with the predicted improvement in water solubility, indicate that these compounds are useful prodrugs for future development. Rapid in vivo dephosphorylation would be expected to occur for the β-lactam phosphates 17a-c as observed for CA-4P [11]. Table 3. Antiproliferative activities of β-lactams 7a-7n, 7p-7u, 8a-8k in human MCF-7 breast cancer cells.

Compound Number
Antiproliferative Activity IC 50   Triple-negative breast cancers (TNBC) are characterised by the absence of estrogen receptors (ER-), progesterone receptors (PR-) and human epidermal growth factor receptor 2 (HER2-). TNBC does not respond to hormonal therapy (such as tamoxifen or aromatase inhibitors) or therapies that target HER2 receptors, such as Herceptin. Treatment options are limited leading to poor prognosis, as indicated by low 5-year survival rates. A number of the more potent compounds were evaluated in the triple negative MDA-MB-231 cell line (Table 5). Compound 7s was the most effective of the series with an IC 50 value of 10 nM. Compounds 7h, 7t, 17a, 17b and 17c were also seen to be effective with IC 50 values of 31 nM, 30 nM, 30 nM, 49 nM and 44 nM respectively, and compared favourably with the positive CA-4 (control IC 50 = 43 nM) [34,63,68]. Table 5. Antiproliferative activities of selected β-lactams in human MDA-MB-231 breast cancer cells a .

Compound Number
Antiproliferative Activity IC 50  Compound 7h was also evaluated in the triple-negative Hs578T breast cancer cell line and its isogenic subclone Hs578Ts(i)8 cells to examine the activity of β-lactams as CA-4 analogues and as anti-tubulin agents for metastasis. Hs578Ts(i)8 cells are 3-fold more invasive and 2.5-fold more migratory than the parental cell line (Hs578T). In addition, Hs578Ts(i)8 cells had 30% more CD44+/CD24-/low cells that could enhance the invasive properties but with a significantly increased capacity to proliferate, migrate and produce tumours in vivo in nude mice [69]. Compound 7h exhibited an excellent anti-proliferative activity in Hs578T cells (IC 50 31 nM) and interestingly retained potency in invasive Hs578Ts(i)8 cells (IC 50 76 nM). The values for CA4 in these cells were 8 nM and 20 nM respectively. These results could indicate the ability of β-lactams as CA-4 analogues to inhibit tumour invasion and angiogenesis which are characteristic of tumour growth and metastasis. These β-lactam compounds may provide potential development leads for this subset of aggressive breast cancers. Compound 7s was also evaluated in the leukemia cell lines HL-60 and K562 and was found to be extremely potent with IC 50  The novel compounds 7h, 7s, 7t, 17b and 17c were selected for further investigation based on analysis of their drug-like properties (Lipinski) from a Tier-1 profiling screen, together with predictions of blood brain barrier partition, permeability, plasma protein binding, metabolic stability and human intestinal absorption properties which confirmed that these compounds are moderately lipophilic-hydrophilic drugs and are suitable candidates for further investigation (Tables S1 and S2, Supporting information).

Evaluation of β-Lactams in the NCI60 Cell Line Screen
A series of the more potent compounds 7h, 7s, 7t, 17b and 17c were evaluated in the National Cancer Institute (NCI)/Division of Cancer Treatment and Diagnosis (DCTD)/Developmental Therapeutics Program (DTP) [70], in which the activity of each compound was determined using approximately 60 different cancer cell lines of diverse tumor origins. The results are summarized in Tables S3-S5,  Supplementary Information. The compounds were tested for inhibition of growth (GI 50 ) and cytotoxicity (LC 50 ) in the NCI panel of cancer cell lines and showed excellent broad-spectrum antiproliferative activity against tumor cell lines derived from leukemia, non-small-cell lung cancer, colon cancer, CNS cancer, melanoma, ovarian cancer, renal cancer, breast cancer and prostate cancer [71] using the sulforhodamine B (SRB) protein assay [72], (Tables S3-S5 Supplementary Information The NCI COMPARE algorithm allows a comparison of the activities of β-lactams 7h and 7s with compounds of a known mechanism of antiproliferative action in the NCI Standard Agents Database. Compounds 7h and 7s showed high correlation to tubulin targeting agents such as maytansine, rhizoxin and the clinically important anticancer drugs vincristine and vinblastine, (see Supplementary  Information, Tables S6 and S7).
The cytotoxicity of the most potent compound 7s on non-tumourigenic cell line HEK-293 (normal human embryonic kidney) was also investigated. We demonstrated an IC 50 value greater than 5 µM in HEK-293T cells for 7s. Cell viability of HEK-293T cells was significantly higher than MCF-7 cells at 10, 1 and 0.5 µM concentrations of compound 7s ( Figure 4A), demonstrating the lack of cellular toxicity of the compounds in these non-cancerous cells.
Further toxicity studies were carried out on the most potent compound β-lactam 7s in primary cells (mouse mammary healthy epithelial cells) at two different cell concentrations (25,000 and 50,000 cells/mL), with CA4 as a positive control. The cells were harvested from mid-to late-pregnant CD-1 mice and were cultured as previously reported [74,75]. Both CA-4 [76] and 7s were not cytotoxic at concentrations up to 10 µM in the NCI cell line panel (See Tables S8 and S9 Supplementary Information). The IC 50 values for both compounds 7s and CA-4 evaluated in normal murine mammary epithelial cells was greater than 10 µM which indicated a minimal toxicity for these compounds ( Figure 4B). At both 25,000 cells/mL and 50,000 cells/mL and a concentration of 10 µM, CA-4 was lethal to the highest percentage of cells. The percentage of viable murine mammary epithelial cells at the IC 50 value of each compound in MCF-7 cells (see Table 3) was calculated in order to give an estimation of the toxicity at this value. At 50,000 cells/mL, over 90% of cells were viable after 72 h for compound 7s, ( Figure 4B). At 25,000 cells/mL, the percentage of cells remaining viable after treatment with compound 7s for 72 h was 93%, compared to 74% for CA-4. (Supplementary Information Tables S8 and S9).
These results indicate a favourable toxicity profile for 7s in comparison to CA4. This provides further evidence, in addition to the NCI60 LC 50 values for 7s, that the β-lactam compound developed in this study is minimally toxic to cells that are not proliferating. normal murine mammary epithelial cells was greater than 10 M which indicated a minimal toxicity for these compounds ( Figure 4B). At both 25,000 cells/mL and 50,000 cells/mL and a concentration of 10 µM, CA-4 was lethal to the highest percentage of cells. The percentage of viable murine mammary epithelial cells at the IC50 value of each compound in MCF-7 cells (see Table 3) was calculated in order to give an estimation of the toxicity at this value. At 50,000 cells/mL, over 90% of cells were viable after 72 h for compound 7s, ( Figure 4B). At 25,000 cells/mL, the percentage of cells remaining viable after treatment with compound 7s for 72 h was 93%, compared to 74% for CA-4. (Supplementary Information Tables S8, S9). These results indicate a favourable toxicity profile for 7s in comparison to CA4. This provides further evidence, in addition to the NCI60 LC50 values for 7s, that the β-lactam compound developed in this study is minimally toxic to cells that are not proliferating.  Statistical analysis was performed using a non-paired two-tailed t-test (ns, not significant; **, p < 0.01). (B) Cell viability for compound 7s and CA-4 in murine mammary epithelial cells. Mouse mammary epithelial cells were harvested from mid-to late-pregnant CD-1 mice and cultured. The isolated mammary epithelial cells were seeded at 50,000 cells/mL. After 24 h, they were treated with 2 L volumes of test compound which had been pre-prepared as stock solutions in ethanol to furnish the concentration range of study, 1 nM-100 M, and re-incubated for a further 72 h. Control wells contained the equivalent volume of the vehicle ethanol (1%, v/v). The cytotoxicity was assessed using alamar blue dye.

Effect of β-Lactam 7s on Cell Cycle and Apoptosis
It is well recognised that tubulin destabilizing agents arrest the cell cycle in the G2/M phase due to cytoskeleton disruption and microtubule depolymeriztion. The effects of β-lactam 7s on cell cycle events and induction of apoptosis in MCF-7 cells were next explored. Initial analysis by flow cytometry of propidium iodide stained MCF-7 cells showed G2M arrest at 24 h by compound 7s [64% (10 nM) and 82% (100 nM  Values represent the mean for two independent experiments. Statistical analysis was performed using a non-paired two-tailed t-test (ns, not significant; **, p < 0.01). (B) Cell viability for compound 7s and CA-4 in murine mammary epithelial cells. Mouse mammary epithelial cells were harvested from mid-to late-pregnant CD-1 mice and cultured. The isolated mammary epithelial cells were seeded at 50,000 cells/mL. After 24 h, they were treated with 2 µL volumes of test compound which had been pre-prepared as stock solutions in ethanol to furnish the concentration range of study, 1 nM-100 µM, and re-incubated for a further 72 h. Control wells contained the equivalent volume of the vehicle ethanol (1%, v/v). The cytotoxicity was assessed using alamar blue dye.

Effect of β-Lactam 7s on Cell Cycle and Apoptosis
It is well recognised that tubulin destabilizing agents arrest the cell cycle in the G 2 /M phase due to cytoskeleton disruption and microtubule depolymeriztion. The effects of β-lactam 7s on cell cycle events and induction of apoptosis in MCF-7 cells were next explored. Initial analysis by flow cytometry of propidium iodide stained MCF-7 cells showed G 2 M arrest at 24 h by compound 7s [64% (10 nM) and 82% (100 nM)] ( Figure 5C). A time dependent increase in the percentage of apoptotic cells (sub-G 0 G 1 ) after 72 h (14% and 26% respectively for 10 nM and 100 nM concentration) was also evident compared to the vehicle control (6% at 72 h), ( Figure 5A), with a corresponding decrease of cells in the G 0 -G 1 phase of the cell cycle, ( Figure 5C). The positive control CA-4 (100 nM) showed 52% of cells in G 2 M arrest at 48 h, and 9.4% in the sub-G 0 G 1 population.
To characterize the mode of cell death induced by 7s in MCF-7 cells, analysis of apoptosis was performed using propidium iodide (PI), which stains DNA and enters only dead cells, and annexin-V, which binds selectively to phosphatidyl serine ( Figure 6). Dual staining for annexin-V and PI facilitates discrimination between live cells (annexin-V-/PI-), early apoptotic cells (annexin-V+/PI-), late apoptotic cells (annexin-V+/PI+) and necrotic cells (annexin-V-/PI+). Each concentration induced an accumulation of annexin-V positive cells when compared to the vehicle control (5%), Figure 6. About 13.6% of cells were found to be apoptotic (annexin-V positive) when treated with compound 7s at 10 nM for 72 h. With an increase in concentration of 7s, 31.9% of cells were found to be apoptotic at 100 nM. The positive control CA-4 (50 nM) resulted in 34.6% apoptotic cells. The observed effect of compound 7s on cell cycle resulting in G 2 M arrest followed by apoptosis is typical of tubulin targeting compounds. However, we have previously reported that prolonged exposure of colon cancer cells CT-26, CaCo-2 and HT-29 to our structurally related 3-aryl-β-lactams induced autophagy [77]; it is possible that autophagy may be the cell death mechanism in the present case, because of the level of apoptosis observed. 0.01). (B) Cell viability for compound 7s and CA-4 in murine mammary epithelial cells. Mouse mammary epithelial cells were harvested from mid-to late-pregnant CD-1 mice and cultured. The isolated mammary epithelial cells were seeded at 50,000 cells/mL. After 24 h, they were treated with 2 L volumes of test compound which had been pre-prepared as stock solutions in ethanol to furnish the concentration range of study, 1 nM-100 M, and re-incubated for a further 72 h. Control wells contained the equivalent volume of the vehicle ethanol (1%, v/v). The cytotoxicity was assessed using alamar blue dye.

Effect of β-Lactam 7s on Cell Cycle and Apoptosis
It is well recognised that tubulin destabilizing agents arrest the cell cycle in the G2/M phase due to cytoskeleton disruption and microtubule depolymeriztion. The effects of β-lactam 7s on cell cycle events and induction of apoptosis in MCF-7 cells were next explored. Initial analysis by flow cytometry of propidium iodide stained MCF-7 cells showed G2M arrest at 24 h by compound 7s [64% (10 nM) and 82% (100 nM To characterize the mode of cell death induced by 7s in MCF-7 cells, analysis of apoptosis was performed using propidium iodide (PI), which stains DNA and enters only dead cells, and annexin-V, which binds selectively to phosphatidyl serine ( Figure 6). Dual staining for annexin-V and PI facilitates discrimination between live cells (annexin-V-/PI-), early apoptotic cells (annexin-V+/PI-), late apoptotic cells (annexin-V+/PI+) and necrotic cells (annexin-V-/PI+). Each concentration induced an accumulation of annexin-V positive cells when compared to the vehicle control (5%), Figure 6. About 13.6% of cells were found to be apoptotic (annexin-V positive) when treated with compound 7s at 10 nM for 72 h. With an increase in concentration of 7s, 31.9% of cells were found to be apoptotic at 100 nM. The positive control CA-4 (50 nM) resulted in 34.6% Values represent the mean ± SEM for three independent experiments. Statistical analysis was performed using two-way ANOVA (**, p < 0.01; ***, p < 0.001).

Tubulin Polymerization Studies
The effect of selected β-lactam CA-4 compounds (7h, 7i, 7s, 7t) which demonstrated the most potent antiproliferative effects in vitro was assessed on the assembly of purified bovine tubulin. CA-4 which effectively inhibits the assembly of tubulin was used as a positive control, while paclitaxel was used to demonstrate effective tubulin polymerization. Tubulin polymerization was determined for compounds 7h, 7i and 7t at 10 µM for 30 min and compound 7s at 1, 5 and 10 µM for 60 min by measuring the increase in absorbance at 340 nm, (Figure 7A,B) [78]. The degree of light scattering by microtubules is proportional to their degree of polymerization. For the paclitaxel control the v max was found to be 89.4 mOD/min. The v max value provides a sensitive indication of the tubulin/ligand interactions for the tubulin polymerization. The most potent antiproliferative compound 7s (10 µM) demonstrated a significant 8.7-fold reduction in v max value while exposure to CA-4 (10 µM) brings about a 5.28-fold reduction in the v max value. Compound 7s compares very favourably to CA-4 in this respect. These effects are in good agreement with the antiproliferative data recorded for both CA-4 (IC 50 = 4.2 nM) and 7s (IC 50 = 8 nM) in the MCF-7 cell line. The v max value for compounds 7h, 7i and 7t was determined as 3.43, 3.84 and 0.92 mOD/min respectively, together with the fold-reduction in the v max values of 2.45, 2.19 and 9.15 respectively for the tubulin polymerization with reference to ethanol control. These results confirm that the molecular target of these antiproliferative 3-vinyl-β-lactams is tubulin and that they are microtubule-destabilising agents. cells [81,82], human endothelial cells (HUVEC) [83], human lung carcinoma cells (H460) [83] and human breast cancer cells (MCF-7) [84]. Taken together with the effects demonstrated above in section 2.2.5 on the inhibition of polymerisation of isolated tubulin, the confocal imaging results confirm that β-lactam 7s is targeting tubulin. The dose-dependent effect of 7s on tubulin polymerization is illustrated in Figure 7. Exposure of the tubulin to 10 µM, 5 µM and 1 µM of 7s resulted in a dose-dependent fold reduction of v max of 8.70, 7.31 and 2.61 respectively while the IC 50 value for 7s for the inhibition of polymerization was calculated to be 1.37 µM, Figure 7. Taken together, these results demonstrate that for these novel β-lactam containing CA4 analogues, antiproliferative activity against the MCF-7 cell line and the inhibition of tubulin polymerization are closely related. It has also been shown that the most potent antiproliferative compound synthesised (7s) inhibits tubulin polymerization to a greater extent than CA-4.

Immunofluorescence Microscopy
Alterations in the microtubule network induced by β-lactam 7s in MCF-7 cells were investigated using immunofluorescence and confocal microscopy ( Figure 8). A well organised microtubular network was observed in MCF-7 control cells when stained with α-tubulin mAb ( Figure 8) and in untreated cells (data not shown). Formation of microtubule bundles and pseudo asters was demonstrated for cells when exposed to paclitaxel (a microtubule-stabilising agent), Figure 8 [79]. A complete loss of microtubule formation was induced in cells exposed to CA-4 or β-lactam 7s for 16 h. This effect is consistent with depolymerised microtubules. Following treatment with CA-4 or β-lactam 7s, MCF-7 cells were observed to contain multiple micronuclei. Mitotic catastrophe resulting from premature or inappropriate entry of cells into mitosis is a type of programmed cell death in response to DNA damage, and is characterised by multinucleated cells [80]. CA-4 induced mitotic catastrophe has also been reported in non-small cell lung cancer cells [81,82], human endothelial cells (HUVEC) [83], human lung carcinoma cells (H460) [83] and human breast cancer cells (MCF-7) [84]. Taken together with the effects demonstrated above in Section 2.3.5 on the inhibition of polymerisation of isolated tubulin, the confocal imaging results confirm that β-lactam 7s is targeting tubulin. The binding of the lead compound 7s to the colchicine binding site of tubulin was confirmed in a whole cell-based assay. N,N'-ethylene-bis(iodoacetamide) (EBI) is an alkylating agent that cross-links cysteine residues 239 and 354 in the colchicine-binding site of tubulin to form theβtubulin-EBI adduct that migrates faster than β-tubulin [85,86], and is detected by Western blotting. However, when the MCF-7 cells are pre-treated with colchicine or a colchicine-site ligand such as CA-4, the formation of the β-tubulin-EBI adduct is prevented. The MCF-7 cells were initially treated with selected β-lactam 7s (10 M) or CA-4 for 2 h, then followed by addition of EBI for a further 1.5 h (Figure 9). The presence of the β-tubulin-EBI adduct was demonstrated for the control samples (no drug) at a lower position on the gel, indicating that EBI has cross-linked Cys239 and The binding of the lead compound 7s to the colchicine binding site of tubulin was confirmed in a whole cell-based assay. N,N -ethylene-bis(iodoacetamide) (EBI) is an alkylating agent that cross-links cysteine residues 239 and 354 in the colchicine-binding site of tubulin to form the β-tubulin-EBI adduct that migrates faster than β-tubulin [85,86], and is detected by Western blotting. However, when the MCF-7 cells are pre-treated with colchicine or a colchicine-site ligand such as CA-4, the formation of the β-tubulin-EBI adduct is prevented. The MCF-7 cells were initially treated with selected β-lactam 7s (10 µM) or CA-4 for 2 h, then followed by addition of EBI for a further 1.5 h (Figure 9). The presence of the β-tubulin-EBI adduct was demonstrated for the control samples (no drug) at a lower position on the gel, indicating that EBI has cross-linked Cys239 and Cys354 on β-tubulin. When the cells are treated with β-lactam 7s and CA-4, the EBI adduct formation is inhibited, indicating that 7s is interacting with tubulin at the colchicine site of tubulin. The binding of the lead compound 7s to the colchicine binding site of tubulin was confirmed in a whole cell-based assay. N,N'-ethylene-bis(iodoacetamide) (EBI) is an alkylating agent that cross-links cysteine residues 239 and 354 in the colchicine-binding site of tubulin to form theβtubulin-EBI adduct that migrates faster than β-tubulin [85,86], and is detected by Western blotting. However, when the MCF-7 cells are pre-treated with colchicine or a colchicine-site ligand such as CA-4, the formation of the β-tubulin-EBI adduct is prevented. The MCF-7 cells were initially treated with selected β-lactam 7s (10 M) or CA-4 for 2 h, then followed by addition of EBI for a further 1.5 h (Figure 9). The presence of the β-tubulin-EBI adduct was demonstrated for the control samples (no drug) at a lower position on the gel, indicating that EBI has cross-linked Cys239 and Cys354 on β-tubulin. When the cells are treated with β-lactam 7s and CA-4, the EBI adduct formation is inhibited, indicating that 7s is interacting with tubulin at the colchicine site of tubulin.

Molecular Modelling Studies
The 3-vinyl-β-lactam compound 7s represents the most potent compound synthesised in the study with IC 50 value of 8 nM in MCF-7 breast cancer cells. The tubulin binding and immunofluorescence studies of 3-vinyl-β-lactam 7s have demonstrated that the colchicine binding site of tubulin is the target for the compound. Flexible alignment of compound 7s with CA-4 resulted in a good degree of overlap between the trimethoxyphenyl rings (Ring A) and the phenolic hydroxyl group of ring B ( Figure 10A). The energy minimised structure of compound 7s demonstrates the inter-atomic distances of the oxygens of the methoxy groups of ring A and ring B as 9.17 Å, which is similar to that calculated for CA-4 (9.27 Å).
The X-ray structure of CA-4 co-crystallised with tubulin has been determined suggesting that cis-CA-4 inhibits tubulin polymerization by preventing the transition from curved to straight tubulin [49]. The X-ray structure of cis and trans stereoisomers of a 3-methyl-1,4-diarylazetidinone [87] co-crystallised with tubulin was reported by Zhou et al. [37,38]. In the present study the potential interaction of our novel synthesised 3-vinyl-β-lactams with the colchicine binding site of tubulin, a series of docking calculations using MOE 2018.0101 [88] was undertaken on both the 3S/4R and 3R/4S enantiomers of the β-lactams 7s and 7t using the tubulin co-crystallised with DAMA-colchicine X-ray crystal structure (PDB entry 1SA0) [5]. Only results for the 3S/4R studies will be discussed as these stereoisomers were more highly ranked than the 3R/4S enantiomer and this is also supported by the crystallographic evidence [37,38]. The 3S/4R enantiomers of the hydroxyl 7s and amino 7t substituted analogues overlay their B-rings on the C-ring of DAMA-colchicine, collocate the trimethoxyphenyl substituents, overlap the 3-hydroxyl/amino groups on the DAMA-colchicine carbonyl oxygen atom and form HBA interactions with Lys β352 as shown in Figure 10B and 10C. The 3,4,5-trimethoxyphenyl groups of all analogues are able to make favourable van der Waals contacts within the lower subpocket delineated by Val β318 and Cys β241. The β-lactam carbonyl oxygen atom can make an HBA interaction with the backbone amine of Asp β251 for both analogues. For both compounds, the trans geometry at C3/C4 facilitates a more favourable interaction of rings A and B with the residues of the β-tubulin colchicine binding site. Protein-ligand interactions for 7s are illustrated in Figure 11. The enantioselective synthesis of 7s and 7t are in progress which will provide the optimum configuration of these compounds to be determined for biological activity.

Conclusion
We have developed an interesting series of 3-vinylazetidinones which selectively modulate the activity of the tubulin protein, resulting in significant cytotoxicity to cancer cells and minimum cytotoxic effects to normal cells. Molecular modelling studies indicated that these compounds could interact with the colchicine binding site of tubulin, and consequently disrupt tubulin polymerization. X-Ray crystallographic studies confirmed that the torsional angle between Ring A and Ring B of the β-lactam was similar to CA-4 and was important in maintaining antiproliferative and tubulin disrupting activity. Biochemical evaluation of these compounds coupled with a molecular modeling study contributes to our understanding of the attributes of the 3vinylazetidinones such as 7s and 7t that result in exceptional antiproliferative activity and dosedependent microtubule assembly inhibition. Analysis of DNA content by flow cytometry demonstrated that the cells were arrested in the G2/M phase; induction of apoptosis was confirmed by an increase in the sub-G0G1 population, which was confirmed by Annexin-V staining. Immunofluorescence staining with -tubulin antibodies in MCF-7 cells demonstrated disorder and

Conclusions
We have developed an interesting series of 3-vinylazetidinones which selectively modulate the activity of the tubulin protein, resulting in significant cytotoxicity to cancer cells and minimum cytotoxic effects to normal cells. Molecular modelling studies indicated that these compounds could interact with the colchicine binding site of tubulin, and consequently disrupt tubulin polymerization. X-Ray crystallographic studies confirmed that the torsional angle between Ring A and Ring B of the β-lactam was similar to CA-4 and was important in maintaining antiproliferative and tubulin disrupting activity. Biochemical evaluation of these compounds coupled with a molecular modeling study contributes to our understanding of the attributes of the 3-vinylazetidinones such as 7s and 7t that result in exceptional antiproliferative activity and dose-dependent microtubule assembly inhibition. Analysis of DNA content by flow cytometry demonstrated that the cells were arrested in the G 2 /M phase; induction of apoptosis was confirmed by an increase in the sub-G 0 G 1 population, which was confirmed by Annexin-V staining. Immunofluorescence staining with α-tubulin antibodies in MCF-7 cells demonstrated disorder and fragmentation of the microtubule network and disruption of mitotic spindle formation. The phosphate prodrugs 17a-c were found to retain antitumour potency. The potent antiproliferative activity of the 3-vinylazetidinones 7s and 7t in breast cancer cells MCF-7 and notably in the triple negative MDA-MB-231 cell line reported in the present study compare very favourably with examples from the related series of 3-arylazetidinone compounds previously reported by our research group [45]. Vinyl substitution at C-3 of these azetidinones also results potent tubulin destabilizing effects in these derivatives of combretastatin A-4.
In summary, these novel 3-vinyl-β-lactam analogues of CA-4 which we now report show potent antiproliferative effects in preliminary in vitro investigations on MCF-7 and MDA-MB231 breast cancer cells. Further studies to establish the long-term effect of these compounds on cancer cell growth, migration and the potential vascular disrupting effects of these molecules are ongoing.

3-(tert-Butyldimethylsilyloxy)-4-methoxybenzaldehyde
To a solution of 3-hydroxy-4-methoxybenzaldehyde (20 mmol) and tert-butyl-dimethylsilylchloride (24 mmol) in dry CH 2 Cl 2 (60 mL) under a nitrogen atmosphere, 1,8-diazabicyclo [5.4.0]undec-7-ene (DBU) (32 mmol) was added dropwise via syringe. The resulting mixture was stirred at room temperature under a nitrogen atmosphere until reaction was complete on thin layer chromatography. The solution was then diluted with CH 2 Cl 2 (80 mL) and washed successively with water (60 mL), 0.1M HCl (60 mL) and saturated aqueous NaHCO 3 (60 mL), retaining the organic layer each time,  5a-5s, 6a-6k The appropriately substituted benzaldehyde (10 mmol) and corresponding substituted aniline (10 mmol) were heated reflux in ethanol (40 mL) for 4 h with a catalytic amount of concentrated sulphuric acid. The volume of reaction was then reduced to approximately 10 mL in vacuo. The Schiff base precipitated from solution upon standing at room temperature overnight. The solid product obtained was filtered and purified by recrystallisation from ethanol.

General method II: Preparation of 2-azetidinones 7a-7u, 8a-8k
To a stirring, refluxing solution of the imine (5 mmol) and triethylamine (6 mmol) in anhydrous dichloromethane (40 mL), a solution of crotonyl chloride (6 mmol) in anhydrous dichloromethane (10 mL) was injected dropwise through a rubber septum over 45 min under nitrogen. The reaction was heated at reflux for 5 h and stirred at room temperature overnight, continuously under nitrogen. The reaction mixture coled and washed with water (2 × 100 mL), with the organic layer being retained each time. The reaction was dried over anhydrous sodium sulfate and the solvent was then removed under reduced pressure. The crude product was purified by flash chromatography over silica gel (eluent: n-hexane: ethyl acetate, 4:1).  13 13 13 13 13

4-[3-Hydroxy-4-methoxyphenyl]-1-(3,4,5-trimethoxyphenyl)-3-vinylazetidin-2-one (7s)
To a stirring, refluxing solution of the TBDMS protected imine 5o (5 mmol) and triethylamine (6 mmol) in anhydrous dichloromethane (40 mL), a solution of crotonyl chloride (6 mmol) in anhydrous dichloromethane (10 mL) was added over 45 min under nitrogen. The reaction was kept at reflux for 5 h and then at room temperature overnight (16 h), until the starting material had disappeared as monitored by TLC in (1:1 n-hexane: ethyl acetate). The reaction mixture was washed with water (2 × 100 mL). The combined organic extract was dried over anhydrous Na 2 SO 4 before the solvent was removed under reduced pressure. The crude product was purified by flash chromatography over silica gel (eluant: n-hexane: ethyl acetate, 4:1) to afford the β-lactam 7o as an oil. To a stirring solution of the protected β-lactam 7o (5 mmol) under N 2 and 0 • C in dry THF was added dropwise 1.5 equivalents of 1.0 M tert-butylammonium fluoride (t-BAF) solution in hexanes (5 mmol). The resulting solution was left to stir at 0 • C until reaction was complete as monitored by TLC. The reaction mixture was diluted with ethyl acetate (75 mL) and washed with 0.1M HCl (100 mL). The aqueous layer was further extracted with ethyl acetate (2 × 25 mL). All organic layers were combined and washed with water (100 mL) and saturated brine (100 mL) before being dried over anhydrous sodium sulphate. The solvent was removed under reduced pressure to yield the phenol which was further purified by flash chromatography over silica gel (eluent: n-hexane: ethyl acetate, 4:1) to afford the product as a yellow oil, yield 20%. IR (NaCl, film) Zinc powder (9 mmol) was activated using trimethylchlorosilane (0.5 mmol) in anhydrous benzene (1 mL) by heating for 15 min at 40 • C and followed by 5 min at 100 • C in a microwave. After cooling, the imine 5h (2 mmol) and ethyl bromoacetate (2.4 mmol) were added to the reaction vessel and the mixture was placed in the microwave for 30 min at 100 • C. The reaction mixture was filtered through Celite to remove the zinc catalyst and then diluted with dichloromethane. This solution was washed with saturated ammonium chloride solution (20 mL) and 25% ammonium hydroxide (20 mL) and then with dilute HCl (40 mL), followed by water (40 mL). The organic phase was dried over anhydrous sodium sulphate and the solvent was removed under reduced pressure. The crude product was purified by flash column chromatography over silica gel (eluent: n-hexane: ethyl acetate, 2:1) to afford the product as a yellow solid, 39%, 267 mg, Mp   (i) Zinc powder (458 mg, 7 mmol (method A) or 21 mmol (method B)) and chlorotrimethylsilane (0.32 mL, 2.5 mmol) were refluxed for 3 min in anhydrous benzene (10 mL) under N 2 and then allowed to cool. To the cooled stirring solution, the appropriately substituted imine (5o) (5 mmol) and ethylbromoacetate (0.66 mL, 6 mmol) were added and refluxed for 7 h. The reaction was cooled to 0 • C and poured onto NH 4 Cl (sat), (10 mL) and 30% NH 4 OH (10 mL). The resulting solution was extracted with DCM (2 × 20 mL) and the organic layer further washed with 0.1N HCl (20 mL) and water (20 mL) before being dried over Na 2 SO 4 , filtered and the solvent removed under reduced pressure to afford the protected product (9b), yield 37%, 876 mg (method A), 77%, 1.823 g (method B) as a pale brown resin which was used immediately in the following reaction. IR (NaCl  (30 mL) was added a solution of 1.0M tBAF in hexanes (4 mL, 4 mmol) under N 2 at 0 • C. The reaction mixture was stirred for a further 90 min. Reaction was diluted with ethyl acetate (150 mL) and washed with 0.1M HCl (200 mL). The aqueous layer was further extracted with ethyl acetate (2 × 50 mL). All the organic layers were collected and washed with water (200 mL) and saturated brine (200 mL) before being dried over Na 2 SO 4 . Solvent was removed under reduced pressure and the phenol was isolated by flash chromatography over silica gel (eluent: n-hexane: ethyl actetate, 1:1) to afford the desired product, yield 73%, 1.05 g, as a yellow solid, Mp

3-(1-Hydroxyethyl)-4-(3-hydroxy-4-methoxyphenyl)-1-(3,4,5-trimethoxyphenyl) azetidin-2-one (10a)
To a solution of the TBDMS protected 3-unsubstituted β-lactam (9b) (125 mg, 0.264 mmol) in dry THF (3 mL) under N 2 at −78 • C (dry ice and acetone) was added 2.0 M LDA solution (0.264 mL, 0.528 mmol). The resulting solution was left to stir for 5 min before a solution of acetaldehyde (49 mg, 0.396 mmol) in dry THF (1.5 mL) was added. The reaction was left to stir for 30 min at −78 • C, then poured onto saturated NaCl solution (25 mL). The resulting solution was extracted with ethyl acetate (50 mL) and the solvent was dried over Na 2 SO 4 before being removed under reduced pressure. Preliminary purification was achieved by passage through a short pad (5 cm) of silica (eluent: DCM) to yield the OTBDMS protected ether 10i as an oil. To a stirring solution of the OTBDMS protected ether 10i (2 mmol) in dry THF (10 mL) was added a solution of 1.0 M TBAF in hexanes (2 mL, 2 mmol) under N 2 at 0 • C. The reaction mixture was stirred for a further 90 min then diluted with ethyl acetate (75 mL) and washed with 0.1 M HCl (100 mL). The aqueous layer was further extracted with ethyl acetate (2 × 25 mL). All the organic extracts were collected and washed with H 2 O (100 mL), and saturated brine (100 mL) before being dried over Na 2 SO 4 and solvent was removed under reduced pressure. Purification was carried out by chromatography using a Biotage SP1 chromatography system using a +12M column and detection set at 280 nM and a fraction volume of 12 mL. A gradient elution of 2% ethyl acetate in n-hexane to 100% ethyl acetate over 15 column volumes was used. The desired product was obtained as a brown oil, 36 mg, yield 17%    25, 55.53, 55.84, 64.38, 60.47, 64.65, 68.42, 70.60, 94.26, 110.50, 111.75, 111.82, 117.42, 117.48, 127.62, 129.22, 129.78, 130.47, 133.25, 133.27, 146.17, 146.32, 152.93, 165.21 36, 27.62, 55.50, 55.53, 56.73, 60.47, 69.50, 94.27, 110.57, 111.78, 117.48, 130.51, 133.18, 133.79, 145.80, 146.25, 152.93, 165.29 13 53, 55.55, 55.62, 60.48, 56.85, 63.99, 68.28, 70.76, 94.34, 110.47, 110.52, 111.74 111.81, 115.56, 117.47, 117.51, 129.93, 130.20, 133.20, 136.59, 136.80, 145.79, 146.14, 146.33, 152.96, 164.89 26.19, 53.00, 56.01, 60.46, 55.51, 55.54, 67.47, 68.46, 71.70, 71.76, 94.28, 110.46, 110.54, 111.84, 111.92, 117.51, 117.60, 130 Method A: To a stirring solution of pyridinium chlorochromate (132 mg, 0.57 mmol) in dry DCM (2 mL) under N 2 at room temperature was added quickly a solution of the silyl protected β-lactam (10i) (195 mg, 0.38 mmol). The reaction was stirred at room temperature for 18 h and then diluted with diethyl ether (25 mL) and the resulting suspension allowed to settle and the diethyl ether layer decanted off. The remaining solid was washed and decanted twice with two further 25 mL portions of diethylether. The organic extracts were combined and dried over MgSO 4 , filtered, and the solvent removed under reduced pressure. The tBDMS ether was removed by treatment with tBAF as previously described, to afford the title compound as an oil, 7%, 11mg. Method B: The 3-unsubstituted β-lactam (9b) (378 mg, 0.80 mmol) was dissolved in THF (7 mL) in a dry flask flushed with N 2 and cooled to −78 • C. To this stirring solution LDA (1.0 M solution, 0.8 mL, 0.8 mmol) was added all at once and the reaction left to stir for 5 min prior to the dropwise addition of acetyl chloride (0.085 mL, 1.2 mmol), in THF (2 mL). The reaction mixture was allowed to stir at −78 • C for 30 min then stirred at room temperature for 5 min before being poured into saturated brine (50 mL). The brine solution was extracted with ethyl acetate (2 × 50 mL), the organic layers combined, dried over MgSO 4 , filtered, and the solvent removed under reduced pressure. Purification by flash column chromatography over silica gel (eluent: n-hexane: ethyl acetate, 1:1) followed by removal of the TBDMS ether by treatment with tBAF as previously described afforded the title compound as an oil, 35   To a solution of the silyl ether protected β-lactam 10i (1 mmol) in DCM (10 mL), stirring at 0 • C under N 2 , was added PPh 3 (1 mmol) and DEAD (1.2 mmol). Stirring at 0 • C was continued for a further 3 min before the reaction was allowed to warm to room temperature. Diethyl ether (30 mL) was added to the reaction mixture to precipitate the triphenylphosphine oxide side product which was removed by filtration. The filtrate was collected and evaporated to dryness under reduced pressure to afford the product. Separation of the E/Z isomers was carried out on a Biotage SP1 system using a gradient elution from 2% ethyl acetate in hexanes to 100% ethyl acetate over 20 column volumes, and detection at 280 nm. The product was obtained as a colourless resin, [99], [ (13) To a solution of the silyl ether protected azetidin-2-one (7o) (156 mg, 0.312 mmol) in pyridine (0.5 mL) stirring under N 2 at room temperature was added osmium tetroxide, OsO 4 (80 mg, 0.312 mmol). The reaction darkened in colour and became hot to the touch upon completion of the addition. The flask was immersed in ice-water for 60 s, then left to stir at room temperature under N 2 for 22 h. A solution of Na 2 (SO 3 ) 2 (1.343 g, 6.8 mmol) in a 1:4 mixture of pyridine/water (20 mL) was added and the reaction was stirred at room temperature for a further 7 h. The reaction mixture was extracted with warm ethyl acetate (100 mL). The organic layer was collected and washed with 0.1M HCl (100 mL), saturated NaHCO 3 (100 mL), and water (100 mL). The organic layer was collected and dried over MgSO 4 , filtered and the solvent removed under reduced pressure. The product was purified by passage through a short silica column (5 cm) and eluted with DCM. The tBDMS group was cleaved by treatment with tBAF as described above, to afford the product as a colourless resin, yield 39%, 51  (1-((2-Methoxy-5-(4-oxo-1-(3,4,5-trimethoxyphenyl)-3-vinylazetidin-2-yl)phenyl)-amino)-1 -oxopropan-2-yl) carbamaic acid 9H-fluoren-9-ylmethyl ester (14) To a stirred solution of β-lactam 7t (4.76 mmol) in anhydrous DMF (30 mL) were added DCC (5.7 mmol), Fmoc-protected alanine (5.6 mmol) and HOBt.H 2 O (7.3 mmol) at room temperature. The mixture was stirred for 24 h, then ethyl acetate (50 mL) was added and the reaction mixture was filtered. The DMF was removed by washing with water (5 × 50 mL). The organic solvent was removed under reduced pressure, and the product was isolated by flash column chromatography over silica gel (eluent: dichloromethane: methanol gradient) as a brown oil, yield, 58%, 173 mg. IR (NaCl, film)   (15) To amino acid amide 14 (1.56 mmol) in methanol (10 mL)/CH 2 Cl 2 (10 mL) was added 2N NaOH (3.4 mmol) at room temperature and the mixture was stirred for 24 h. Saturated aq. NaHCO 3 was added and the mixture was extracted with CH 2 Cl 2 three times. The organic solution was dried and evaporated. The product was dissolved diethyl ether and extracted with 2N HCl (5 × 50 mL). 2N NaOH was added to the HCl mixture solution and the mixture was washed with diethyl ether (5 × 50 mL). The organic solution was dried and the solvent was removed under reduced pressure to afford the product as an off-yellow oil, yield 57%. IR (NaCl) ν max : 3307 cm −1 (NH 2  To a solution of phenol 7s, 9a (17 mmol) in acetonitrile (100 mL cooled to 0 • C) was added carbon tetrachloride (85 mmol). The resulting solution was stirred for 10 min prior to adding diisopropylethylamine (35 mmol) and dimethylaminopyridine (1.7 mmol). The dibenzyl phosphite (24.5 mmol) was then added dropwise to the mixture. When the reaction was complete, 0.5M KH 2 PO 4 (aq) was added and the reaction mixture was allowed to warm to room temperature. An ethyl acetate extract (3 × 50 mL) was washed with saturated sodium chloride (aqueous, 100 mL) followed by water (100 mL) and dried using anhydrous sodium sulfate. The organic solvent was removed under reduced pressure and the product was isolated by flash column chromatography over silica gel (n-hexane: ethyl acetate gradient).   13  A solution of β-lactam phenol 7s (280 mg, 0.64 mmol), acetonitrile (5 mL) and carbon tetrachloride (0.62 mL, 0.64 mmol) was cooled to −10 • C and stirred under a nitrogen atmosphere for ten minutes. Diisopropyl ethylamine (1.28 mmol) and dimethylaminopyridine (0.06 mmol) were added. After one minute, dimethyl phosphite (0.96 mmol) was added over three minutes. The mixture was stirred for a further 3 h allowing the reaction to come to ambient temperature slowly. The reaction was terminated via the addition of 0.5 M potassium dihydrogen phosphate. The mixture was extracted with ethyl acetate. The organic phases were combined and evaporated to dryness under reduced pressure. The residue was purified by flash chromatography on silica gel to afford the product (155 mg, 52%

Biochemical Evaluation
All biochemical assays were performed in triplicate on at least three independent occasions for the determination of mean values reported.

Cell Culture
The human breast carcinoma cell line MCF-7, was purchased from the European Collection of Animal Cell Cultures (ECACC) and was cultured in Eagle's minimum essential medium with 10% fetal bovine serum, 2 mM L-glutamine and 100 µg/mL penicillin/streptomycin. The medium was supplemented with 1% non-essential amino acids. The human breast carcinoma cell line MDA-MB-231 was purchased from the European Collection of Animal Cell Cultures (ECACC). MDA-MB-231 cells were maintained in Dulbecco's modified Eagle's medium (DMEM) supplemeted with 10% (v/v) fetal bovine serum, 2 mM L-glutamine and 100 µg/mL penicillin/streptomycin (complete medium). All media contained 100 U/mL penicillin and 100 µg/mL streptomycin. Triple negative breast cancer Hs578T cells and its invasive variant Hs578Ts(i) 8 were obtained as a kind gift from Dr. Susan McDonnell, School of Chemical and Bioprocess Engineering, University College Dublin and were cultured in Dulbecco's Modified Eagle's Media (DMEM) with GlutaMAX TM -I, with the same supplement as for MDA-MB-231 cells in the absence of non-essential amino acids. HEK-293T normal epithelial embryonic kidney cells were cultured in Dulbecco's Modified Eagle's Medium (DMEM) with GlutaMAX TM -I in the absence of non-essential amino acids. K562 and HL-60 cells were originally obtained from the European Collection of Cell Cultures (Salisbury, UK).The K562 cells were derived from a patient in the blast crisis stage of CML HL-60 cells were derive from a patient with acute myeloid leukaemia. Cells were cultured in RPMI-1640 Glutamax medium supplemented with 10% FCS media, and 100 µg/mL penicillin/streptomycin. Cells were maintained at 37 • C in 5% CO 2 in a humidified incubator. All cells were sub-cultured three times/week (adherent cells by trypsinisation).

Cell Viability Assay
Cells were seeded at a density of 5 × 10 3 cells/well (MCF-7), in triplicate in 96-well plates. After 24 h, cells were then treated with either medium alone, vehicle [1% ethanol (v/v)] or with serial dilutions of CA-4 or β-lactam analogue. Cell viability for MCF-7 and MDA-MB-231 was analysed using the Alamar Blue assay (Invitrogen Corp, Thermo Fisher Scientific, 168 Third Avenue, Waltham, MA, USA) according to the manufacturer's instructions. After 72 h, Alamar Blue [10% (v/v)] was added to each well and plates were incubated for 3-5 h at 37 • C in the dark. Fluorescence was read using a 96-well fluorimeter with excitation at 530 nm and emission at 590 nm. Results were expressed as percentage viability relative to vehicle control (100%). Dose response curves were plotted and IC 50 values (concentration of drug resulting in 50% reduction in cell survival) were obtained using the commercial software package Prism (GraphPad Software, Inc., 2365 Northside, Suite 560, San Diego, CA, USA). Experiments were performed in triplicate on at least three separate occasions.

Lactate Dehydrogenase Assay for Cytotoxicity
Cytotoxicity was determined using the CytoTox 96 non-radioactive cytotoxicity assay (Promega Corporation; 2800 Woods Hollow Road, Madison, WI, USA) [101] following the manufacturer's protocol. Briefly, MCF-7 cells were seeded in 96-well plates, incubated for 24 hr and then treated with test compounds (7d, 7h, 7i, 7q, 7r, 7u, 7t, 10d, 10f, 12) as described in the cell viability assay above. After 72 h, 20 µL of 'lysis solution (10X)' was added to control wells and the plate was incubated for a further 1 hr to ensure 100% death. 50 µL of supernatant was carefully removed from each well and transferred to a new 96-well plate. 50 µL of reconstituted 'substrate mix' was added and the plate was placed in the dark at room temperature for 30 min. After this period, 50 µL of 'stop solution' was added to each well and the absorbance was read at a wavelength of 490 nm using a Dynatech MR5000 plate reader. The percentage cell death at 10 µM was calculated.

Cytotoxicity Assay
As previously reported [45,74,75] mammary glands from 14-18 day pregnant CD-1 mice were used as source and primary mammary epithelial cell cultures were prepared from these. The isolated mammary epithelial cells were seeded at two concentrations (25,000 cells/mL and 50,000 cells/mL). Initially a third concentration of 100,000 cells/mL was also used, but this proved to be too high to give meaningful results. After 24 h, the cells were treated with 2 µL volumes of test compound 7s which had been pre-prepared as stock solutions in ethanol to furnish the concentration range of study, 1 nM-100 µM, and re-incubated for a further 72 h. Control wells contained the equivalent volume of the vehicle ethanol (1% v/v). The cytotoxicity was assessed using alamar blue dye. The assembly of purified bovine tubulin was monitored using a kit, BK006, purchased from Cytoskeleton Inc., 1830 S Acoma St, Denver, CO, 80223, USA. [78]. The assay was carried out in accordance with the manufacturer's instructions in the tubulin polymerisation assay kit manual using the standard assay conditions. The values reported represent the average values from two independent assays. Purified (>99%) bovine brain tubulin (3 mg/mL) in a buffer consisting of 80 mM PIPES (pH 6.9), 0.5 mM EGTA, 2 mM MgCl 2 , 1 mM GTP and 10% glycerol was incubated at 37 • C in the presence of either vehicle (2% (v/v) ddH 2 O) or compounds 7h, 7i, 7s, 7t (initially 10 µM in EtOH); CA-4 and Paclitaxel were used as controls. Light is scattered proportionally to the concentration of polymerised microtubules in the assay. Therefore, tubulin assembly was monitored turbidimetrically at 340 nm at 37 • C in a Spectramax 340 PC spectrophotometer (Molecular Devices, 3860 N 1st St, San Jose, CA, USA). The absorbance was measured at 30 s intervals for 60 min.
Supplementary Materials: The following are available online at http://www.mdpi.com/1424-8247/12/2/56/s1, Experimental procedures and spectroscopic data for intermediate compounds 5a-i, 5m-s, 6a-c, 6f-k, additional cytotoxicity data in normal murine epithelial cells for compound 7s, Tier-1 Profiling Screen of Selected β-Lactams, results of comparative antitumor evaluations of compounds 7h, 7s, 7t, 17b, 17c in the NCI60 cell line in vitro primary screen, NCI 60 cell line mean screening results for selected compounds and results of standard COMPARE analysis of 7h and 7s.
Author Contributions: S.W. synthesised and characterised compounds in the studies according to Schemes 1, 2 and 4 performed cell studies and generated data in Tables 3-5, performed the HPLC analytical study and the stability study. and performed data analysis and interpreted data. A.M.M. synthesised and characterised some molecules in Schemes 1 and 2, performed cell studies and generated the the data for Figures 4-6 and 9. T.F.G. synthesised compounds in Schemes 1-4, characterised these compounds, performed the cell studies and generated data in Tables 3-5

Conflicts of Interest:
The authors declare no conflict of interest.

Abbreviations
The following abbreviations are used in this manuscript: