Borylated 2,3,4,5-Tetrachlorophthalimide and Their 2,3,4,5-Tetrachlorobenzamide Analogues: Synthesis, Their Glycosidase Inhibition and Anticancer Properties in View to Boron Neutron Capture Therapy

Tetrachlorinated phthalimide analogues bearing a boron-pinacolate ester group were synthesised via two synthetic routes and evaluated in their glycosidase modulating and anticancer properties, with a view to use them in boron neutron capture therapy (BNCT), a promising radiation type for cancer, as this therapy does little damage to biological tissue. An unexpected decarbonylation/decarboxylation to five 2,3,4,5-tetrachlorobenzamides was observed and confirmed by X-ray crystallography studies, thus, giving access to a family of borylated 2,3,4,5-tetrachlorobenzamides. Biological evaluation showed the benzamide drugs to possess good to weak potencies (74.7–870 μM) in the inhibition of glycosidases, and to have good to moderate selectivity in the inhibition of a panel of 18 glycosidases. Furthermore, in the inhibition of selected glycosidases, there is a core subset of three animal glycosidases, which is always inhibited (rat intestinal maltase α-glucosidase, bovine liver β-glucosidase and β-galactosidase). This could indicate the involvement of the boron atom in the binding. These glycosidases are targeted for the management of diabetes, viral infections (via a broad-spectrum approach) and lysosomal storage disorders. Assays against cancer cell lines revealed potency in growth inhibition for three molecules, and selectivity for one of these molecules, with the growth of the normal cell line MCF10A not being affected by this compound. One of these molecules showed both potency and selectivity; thus, it is a candidate for further study in this area. This paper provides numerous novel aspects, including expedited access to borylated 2,3,4,5-tetrachlorophthalimides and to 2,3,4,5-tetrachlorobenzamides. The latter constitutes a novel family of glycosidase modulating drugs. Furthermore, a greener synthetic access to such structures is described.

In our group, we are interested in the use of organic boron as a pharmacophoric group in its boronic acid (R-B(OH)2) and boronate ester (R-B(OR′)2) [7,8] functional groups and in the development of synthetic methodologies for the installation of this pharmacophore on biologically active molecules to study and expand the palette of enzyme-drug interactions [9][10][11][12][13].
Glycosidase enzymes are involved in a number of disease states, ranging from diabetes and lysosomal storage disorders to viral infections, with their modulation being paramount in the management of these diseases [14][15][16][17][18][19][20][21][22]. The introduction of boron atoms to drug molecules also provides access to potential boron neutron capture therapy (BNCT) agents. BNCT provides the opportunity to utilise a type of radiotherapy that causes minimal damage to healthy tissue [23][24][25].
We report the synthesis of a family of novel drugs consisting of borylated 2,3,4,5tetrachlorophthalimides and 2,3,4,5-tetrachlorobenzamides. The latter group arose from a decarbonylation side reaction, giving expedited access to them. Drugs have been characterized, including by X-ray crystallographic analysis in two instances. These confirmed the structural integrity, the outcome of the side reaction and the conformation of the boronate ester groups in the solid state. Furthermore, biological assays against glycosidase enzymes and cancer cell lines highlighted a good inhibitor for bovine liver β-galactosidase and three potent growth inhibitors and, of these, one selective growth inhibitor for cancer versus healthy cell lines in the cancer assay. These drugs represent an optimal set for further derivatisations.

Summary of Synthetic Work
Synthesis of the N-borylated 2,3,4,5-tetrachlorophthalimides was attempted via two synthetic strategies: the double acyl substitution route (reaction of 1 with meta 2, para 2, Phthalimide analogues have also been shown to display glycosidase inhibition [3][4][5], with the 2,3,4,5-tetrachlorophthalimide scaffold deemed necessary for potent activity and the corresponding unsubstituted phthalimide derivatives showing reduced activity [6]. In our group, we are interested in the use of organic boron as a pharmacophoric group in its boronic acid (R-B(OH) 2 ) and boronate ester (R-B(OR ) 2 ) [7,8] functional groups and in the development of synthetic methodologies for the installation of this pharmacophore on biologically active molecules to study and expand the palette of enzyme-drug interactions [9][10][11][12][13].
Glycosidase enzymes are involved in a number of disease states, ranging from diabetes and lysosomal storage disorders to viral infections, with their modulation being paramount in the management of these diseases [14][15][16][17][18][19][20][21][22]. The introduction of boron atoms to drug molecules also provides access to potential boron neutron capture therapy (BNCT) agents. BNCT provides the opportunity to utilise a type of radiotherapy that causes minimal damage to healthy tissue [23][24][25].
We report the synthesis of a family of novel drugs consisting of borylated 2,3,4,5tetrachlorophthalimides and 2,3,4,5-tetrachlorobenzamides. The latter group arose from a decarbonylation side reaction, giving expedited access to them. Drugs have been characterized, including by X-ray crystallographic analysis in two instances. These confirmed the structural integrity, the outcome of the side reaction and the conformation of the boronate ester groups in the solid state. Furthermore, biological assays against glycosidase enzymes and cancer cell lines highlighted a good inhibitor for bovine liver β-galactosidase and three potent growth inhibitors and, of these, one selective growth inhibitor for cancer versus healthy cell lines in the cancer assay. These drugs represent an optimal set for further derivatisations.
An unexpected decarbonylation reaction gave products ortho 5, meta 5 and para 5, and decarboxylation reaction gave ortho 8 and para 8.

Decarbonylation Reaction
To our knowledge, there are limited literature reports to the synthesis of 2,3,4,5tetrachlorobenzyl scaffolds. Two main synthetic strategies were employed.
A second method, described by Harvey et al., achieves the synthesis of the same set of target molecules via a two-/three-step sequence from aromatic starting materials (e.g., toluene, ethylbenzene), firstly via perchlorination with chlorine bubbling through the aromatic starting material, iron powder and anhydrous ferric chloride boiling in carbon tetrachloride for 8 h, followed by reaction of the perchlorinated aromatics with sulfuric acid at 180-200 • C for heptachlorotoluene or 260-280 • C for nonachloroethylbenzene starting material used. The reaction with the nonachloroethylbenzene required a further step, namely the reaction with potassium permanganate in 2N sodium hydroxide at 80 • C. The yields for 2,3,4,5-tetrachlorobenzoic acid were 29% and 14%, respectively [27].
Earlier synthetic strategies include the reaction of tetrachlorophthalic acid in a sealed vessel at 300 • C in subcritical acetic acid [28], of tetrachloro(trichloromethyl)benzene (unknown isomer/s) in a sealed vessel at 280 • C in subcritical water [29]. A later publication, reporting on the phytotoxicity activity of benzoic acid derivatives, includes production of a small library of 2,3,4,5-tetrachlorobenzamides; however, the synthetic details to this sub-family are scarce [30].
Here, decarbonylation occurred during the reaction of boron-bearing amines with phthalic anhydride to produce the corresponding secondary amides (ortho 5, meta 5 and para 5). The reaction mechanism is hypothesised to proceed through an acyl substitution reaction occurring at an anhydride carbonyl by the nitrogen atom of the amine reagent. This is followed by decarbonylation of the adjacent carboxylate and formation of an anionic intermediate, with the resulting electron pair held in the aromatic C-sp 2 orbital bearing the carboxylate group. This lone pair is thought to be stabilised by inductive and resonance effects via the nearby four electronegative chlorine atoms. This lone pair then strips off a proton intramolecularly from the quaternary nearby nitrogen atom, thus, providing the 2,3,4,5-tetrachlorobenzamide products.

Decarboxylation Reaction to Benzamides
Our methodology achieves the transformation of 2,3,4,5-tetrachlorophthalic anhydride and 2,3,4,5-tetrachlorophthalimide to the corresponding decarbonylated secondary amides in one synthetic step, avoiding the use of corrosive reagents and harsh conditions, and at lower temperature by heating the reaction mixture to 100 • C in DMF for 48 h to give the products ortho 8 and para 8 in moderate yields.

X-ray Crystallography Commentary
The structure of compound meta 5 is shown in Figure 2. The asymmetric unit comprises two molecules, which exhibit essentially the same conformations and only one of these is shown. As expected, the trans-amide group is essentially planar (O1a-C-N1a-H 172.4 • ), while the two phenyl substituents are twisted (C Ar -C Ar -C-O1a 75.5 • ; C Ar -C Ar -N1a-C 33.1 • ) to minimise repulsion with the amide functional group.
In the structure of ortho 8 (Figure 3), the substituents on the B-substituted ring are in ortho positions. In comparison with meta 5, the insertion of a methylene group between the amide and B-substituted phenyl ring relieves torsional strain (C Ar -C Ar -C(H 2 )-N1 4.6 • ). The other structural features resemble those found in meta 5, defined by the dihedral angles O1-C-N1-H 173.8 • and C Ar -C Ar -C-O1 76.6 • , C Ar -C-N1a-C 61.1 • .
The boronic ester groups in both structures are close to coplanar with the adjacent phenyl ring (out of plane twist <10 • ) and the C-B bonds (ortho 8 1.561(6) Å; meta 5 1.558(6) and 1.557(6) Å) are reinforced due to π-bonding with the sp 2 -hybridised B-atom. In twisted (purely σ-bonded) aromatic boronate esters (C Ar -C Ar -B-O~90 • ), the C Ar -B bond is typically in a range 1.57-1.59 Å [31,32]. In the structure of ortho 8 (Figure 3), the substituents on the B-substituted ring are in ortho positions. In comparison with meta 5, the insertion of a methylene group between the amide and B-substituted phenyl ring relieves torsional strain (CAr-CAr-C(H2)-N1 4.6°). The other structural features resemble those found in meta 5, defined by the dihedral angles O1-C-N1-H 173.8° and CAr-CAr-C-O1 76.6°, CAr-C-N1a-C 61.1°. The boronic ester groups in both structures are close to coplanar with the adjacent phenyl ring (out of plane twist <10°) and the C-B bonds (ortho 8 1.561(6) Å; meta 5 1.558(6) and 1.557(6) Å) are reinforced due to π-bonding with the sp 2 -hybridised B-atom. In In the structure of ortho 8 (Figure 3), the substituents on the B-substituted ring are in ortho positions. In comparison with meta 5, the insertion of a methylene group between the amide and B-substituted phenyl ring relieves torsional strain (CAr-CAr-C(H2)-N1 4.6°). The other structural features resemble those found in meta 5, defined by the dihedral angles O1-C-N1-H 173.8° and CAr-CAr-C-O1 76.6°, CAr-C-N1a-C 61.1°. The boronic ester groups in both structures are close to coplanar with the adjacent phenyl ring (out of plane twist <10°) and the C-B bonds (ortho 8 1.561(6) Å; meta 5 1.558(6) and 1.557(6) Å) are reinforced due to π-bonding with the sp 2 -hybridised B-atom. In Intermolecular H-bonding in both structures comprises one-dimensional N-H . . . O chains of the amide functional group (in its trans H-N-C=O conformation). There are differences in the symmetry of the chains in the two structures. When the boronate ester is ortho on the benzamide ring, adjacent molecules are related by the c glide place ( Figure 4A), orthogonal to the place of the page and propagating right to left, leading to an alternating (zig-zag) array of H-bonds along the chain.  31,32]. Intermolecular H-bonding in both structures comprises one-dimensional N-H…O chains of the amide functional group (in its trans H-N-C=O conformation). There are differences in the symmetry of the chains in the two structures. When the boronate ester is ortho on the benzamide ring, adjacent molecules are related by the c glide place ( Figure  4A), orthogonal to the place of the page and propagating right to left, leading to an alternating (zig-zag) array of H-bonds along the chain.
In the structure of the meta 5 isomer ( Figure 4B), the presence of two molecules in the asymmetric unit (molecules A and B) breaks any symmetry relationship between adjacent molecules within the H-bonded chain, and the orientations of the N-H…O bonds are approximately the same along the chain, with rotation of the nearby aromatic rings facilitating closer packing.

Glycosidase Assay
In our laboratory we are interested in glycosidase modulation [13,18,19,21]. Screening for selectivity, as well as potency, is of paramount importance in carbohydrate-active enzyme research.
Our drugs and controls are, therefore, screened against two panels of glycosidases, respectively, in methanol (Table 1) and water (Table 2). This allows identification of the glycosidase-related disease area(s), selectivity profile and potency of biological action for each drug.
Following the glycosidase inhibition range recommendations [21], an IC50 value >250 μm denotes weak inhibition, 100-249 μm denotes moderate inhibition, 10-99 μm good inhibition, 0.1-9 μm potent inhibition and <0.1 μm very potent inhibition. Our 2,3,4,5tetrachlorobenzamides enter as a novel family of glycosidase inhibitors. In the structure of the meta 5 isomer (Figure 4B), the presence of two molecules in the asymmetric unit (molecules A and B) breaks any symmetry relationship between adjacent molecules within the H-bonded chain, and the orientations of the N-H . . . O bonds are approximately the same along the chain, with rotation of the nearby aromatic rings facilitating closer packing.

Glycosidase Assay
In our laboratory we are interested in glycosidase modulation [13,18,19,21]. Screening for selectivity, as well as potency, is of paramount importance in carbohydrate-active enzyme research.
Our drugs and controls are, therefore, screened against two panels of glycosidases, respectively, in methanol (Table 1) and water (Table 2). This allows identification of the glycosidase-related disease area(s), selectivity profile and potency of biological action for each drug.

Biological Activities for Phthalimides and Benzamides in the Literature
An overview of the literature in the field provides the following studies for phthalamides, 2,3,4,5-chlorophthalimides, and benzamides.
SAR studies on phthalimide analogues with Saccharomyces cerevisiae (yeast) α-glucosidase highlight good inhibitions for N-phenylphthalimides derivatised at the ortho-position with non-polar groups [59]. This was not found by us with our phthalimide drugs, but with our benzamide drugs.
On the other side of the molecule, substitutions of the phthalimide scaffold H atoms with other groups, such as amine or hydroxyl, tend to largely abrogate potency, but the introduction of nitro or alkyl groups tend to produce good inhibitors.
Molecules 2021, 26, x FOR PEER REVIEW 8 of 29 Molecules 2021, 26, x FOR PEER REVIEW 8 of 29 For BSH and BPA.
For BSH and BPA.   Another study on a library of N-phenylphthalimide derivatives showed the strongest potency against the three α-glucosidases screened was displayed by N-(2,4-dinitrophenyl)phthalimide, having two nitro groups in ortho and para positions. This is a moderate inhibitor of yeast α-glucosidase (IC 50 158 µM) and a good inhibitor of maltase (IC 50 51 µM), displaying no inhibition of sucrase [60].
N-Phenylphthalimide derivatives substituted with non-polar groups departing from the ortho position of the phenyl group keep providing inhibition of Saccharomyces cerevisiae α-glucosidase as low as 16 µM [61].
Other examples include the Hashimoto papers [62][63][64]. Several members of a phthalimide moiety connected by an alkyl chain to variously substituted phenoxy rings were screened against α-glucosidase. The inhibition potency appeared to be governed by the chain length of the substrate. Substrates possessing 10 carbons afforded the highest levels of activity, which were one to two orders of magnitude more potent than the known inhibitor 1-DNJ [4,65].
Bian and coworkers screened, against α-glucosidase, a series of N-substituted-(ptoluenesulfonylamino)phthalimides. Many analogues provide good inhibitions of the enzymes, with aromatic pendants and tethers containing 1-3 atoms generally producing the most potent inhibitions [3].
Another study of N-phenoxy-substituted phthalimides showed that the presence of a thiazolidine-2,4-dione or a rhodanine group, located at the 4-position of the phenyl ring, resulted in the best activity, with IC 50 values as low as 5 µM against Saccharomyces cerevisiae α-glucosidase [66].
2,3,4,5-Tetrachlorophthalimides N-derivatised with a phenyl group attached directly to the N or through a linear alkyl tether (1-6 CH 2 units) all displayed potent IC 50 (3-11 µM) towards one α-glucosidase screened and more potent than the \1-DNJ control. The replacement of the four chlorine atoms with hydrogen atoms and the replacement of hydrogen atoms with other groups (e.g., nitro, amine) partially or completely abrogated the inhibitory activity of the drugs [6].
The same group also investigated other groups bonded to the phthalimide N, namely branched and cyclical alkyl groups and a dodecaborane group. All drugs displayed comparable or more potent activity (1-49 µM) than 1-DNJ. Cyclical alkyl groups and the borane group produced the most potent inhibitions [5].
The pendant groups attached to the phthalimide unit can clearly interact effectively with a number of sites in the vicinity of the active site, which is presumably occupied by the 2,3,4,5-tetrachlorophthalmide scaffold. This is highlighted by the variety and length of pendant groups. Hydrophobicity seems to be the common motif.
N-Phenyl-2,3,4,5-tetrachlorophthalimide derivatives substituted with non-polar groups departing from the ortho, meta and para positions of the phenyl group keep providing inhibition of α-glucosidase as low as 13 µM against Saccharomyces cerevisiae [61].

Benzamides
A series of N-substituted 1-aminomethyl-β-D-glucopyranoside derivatives was screened against Saccharomyces cerevisiae α-glucosidase, rat intestinal maltase α-glucosidase and sucrase. The most potent inhibitions were produced when the benzamide aromatic ring displayed groups in the para position to the amide. The three most potent compounds comprised O-acetyl groups in the 3-, 4-and 5-positions (IC 50 7.7 and 15.6 µM against rat intestinal maltase α-glucosidase and sucrase), a nitro group in the 4-position (IC 50 36.2 µM against Saccharomyces cerevisiae α-glucosidase) and an O-acetyl group in the 4-position (IC 50 96.5 µM against Saccharomyces cerevisiae α-glucosidase) [68].

Biological Activities for Our Drug Library Screened in Methanol
Biological activities for controls, and our 2,3,4,5-tetrachloro phthalimides and benzamides, follow in Tables 1 and 2. For Table 1:

Controls
Borocaptate sodium (BSH) and 4-borono-L-phenylalanine (BPA), and their 10 B-enriched congeners 10 B-BSHand 10 B-BPA, were the controls. To our knowledge, these drugs have never been reported in a glycosidase assay. BSH and BPA are the drugs currently clinically used in BNCT. It is possible to see that none of them significantly inhibit any of the glycosidases in the panel at 100 or 1000 µM. In the panel, percent inhibitions range from a minimum value of 0 to a maximum value of 19.6.

2,3,4,5-Tetrachlorophthalimides
The 2,3,4,5-tetrachlorophthalimide drugs presented in this work do not provide any appreciable degree of inhibition, most likely because the spatial geometry and length of tether extending from the phthalimide scaffold are probably not able to reach the sites of interaction. para 3 and meta 3 possess the CH 2 spacer between the phthalimide and the aromatic boron group, with para 3 displaying the boronate ester group in para position and meta 3 in meta position to the phthalimide.
Of the benzamides, para 5, with no CH 2 spacer between the aromatic boron group and the benzamide, shows moderate inhibition towards maltase α-glucosidase, with an IC 50 188 µM. Weak inhibitions are displayed towards bovine liver β-glucosidase (IC 50 543 µM) and bovine liver β-galactosidase (IC 50 333 µM). meta 5, with no CH 2 spacer between the aromatic boron group and the benzamide, again, shows moderate inhibition towards bovine liver β-glucosidase (IC 50 175 µM) and bovine liver β-galactosidase (IC 50 213 µM). Weak inhibition is observed towards maltase α-glucosidase, with an IC 50 of 274 µM, and E. coli β-glucuronidase (IC 50 932 µM). para 5 and meta 5 display the boronate ester group, respectively, in para and meta position from the benzamide. Both display partially selective inhibition profiles, inhibiting only 3-4 within the panel of 16 enzymes. Interestingly, their inhibitory profiles show a swap in potency, with para 5 preferentially selecting one α-glucosidase and meta 5 selecting β-glycosidases. ortho 5, the ortho congener to para 5 and meta 5, displays no significant inhibition of any of the enzymes. Hence, the location of the boronate ester group has a negative effect on drug-enzyme interactions, presumably either by preventing the drug from sitting in the active site as para 5 and meta 5 and/or abrogating any further interactions the aromatic boronate group may have with the enzyme.

2,3,4,5-Tetrachlorobenzamides
Benzamides para 8 and ortho 8, possessing the CH 2 spacer between the aromatic boron group and the benzamide, show partially selective inhibition of the same enzymes. para 8 has the boronate ester group para to the benzamide, whereas ortho 8 displays the boronate ester group ortho to the benzamide. Presumably, the lack of one of the carbonyl groups allows the pendant group to reach sites of favourable interactions with the enzymes. Since the enzymes inhibited are the same, it is surmised that the drug SAR profiles are similar. para 8 elicits moderate inhibition of maltase α-glucosidase, with an IC 50 of 207 µM. ortho 8 inhibits the same enzyme weakly, with a similar value of 305 µM. Weak inhibitions are displayed towards bovine liver β-glucosidase (IC 50 = 702 and 922 µM, respectively).
The two main differences are seen: (a) In the inhibition of bovine liver β-galactosidase, which is inhibited weakly by ortho 8 (IC 50 278 µM), but para 8 inhibits the same enzyme with a good IC 50  The drugs that show appreciable inhibition all inhibit the same glycosidase enzymes (maltase α-glucosidase, bovine liver β-glucosidase and bovine liver β-galactosidase). Furthermore, all drugs selectively inhibit glycosidases of animal origin vs. glycosidases of plant or bacterial origin within the same glycosidase class. This is a positive result for applications in a human disease medicinal chemistry context.

Biological Activities for Our Drug Library Screened in Water
For Table 2: Some differences were noted in the results obtained from this second laboratory that used just water to suspend or dissolve the compounds. The compounds did not fully go into solution but nonetheless, activities were observed and so they are included here for comparison.
No appreciable inhibition was detected for any of the controls and compounds against Bacillus α-glucosidase, Jack bean α-glucosidase, bovine kidney A-acetyl-β-glucosaminidase and bovine liver β-glucuronidase.

Cancer Assay and Structure Activity Relationships
In our laboratory, we are interested in BNCT as a potentially broad-spectrum approach to cancer management. It would be advantageous upon irradiation, if the boron-containing drugs accumulate more selectively in cancer cells vs. healthy cells [72,73]. In case the drugs do not accumulate selectively in cancer vs. healthy cells, the delivery of radiation is required with greater precision.
BNCT is essentially a non-invasive radiation technique and the least destructive currently available [23][24][25]. Use of a borylated drug in BNCT would ideally require that it is non-toxic in the absence of radiation. Following a first study of synthesis, purification and toxicity of organic-boron-containing drugs for BNCT applications [13], we report here, two further families of potential BNCT agents.
BNCT agents that contain organic boron groups are preferable to ones containing inorganic boron. The currently utilised sodium borocaptate, BSH, with its inorganic boron atoms, raises several toxicity concerns [74,75]. Boronophenylalanine, BPA, which contains the organic boronic acid moiety, has long been known to show no discernible toxicity [76]. Similarly, in this area, candidate BNCT agents containing an organic boron group should be more likely to reach the clinic.
It has long been known that organic boron is an essential element for plants [77,78] and is likely to be essential for human and animal health [79].
When comparing toxicological data for organic boron-containing molecules with their non-borylated congeners, the trend is that the presence of organic boron lowers toxicity profiles. For example, benzene has an LD 50 (lethal dose) of 125 mg/kg (human, oral) [80] and an LCLO (lethal concentration) of 20,000 ppm (human, 5 min); it is carcinogenic, and also possibly mutagenic. The NIOSH Permissible Exposure Limit for benzene is 1 ppm, the Recommended Exposure Limit is 0.1 ppm, and the Immediately Dangerous to Life and Health concentration is at 500 ppm [81].
If a BNCT agent also has growth inhibition capability against cancer cells, then it is important to screen them in more complex biological systems, such as spheroids, as we did recently [72,73]. Table 3 shows, on a blue background, the percentage cell growth inhibition in response to 25 µM of the drug. In this case, a higher value correlates with a greater growth inhibition. Inhibition value ranges have been colour coded, according to potency, with 80-100% inhibition in red, 60-79% in orange, 35-59% in green, 10-34% in blue and 0-9% in black.
On the green background, the GI 50 values are provided for the most potent drugs. The GI 50 value provides the concentration in µM that induces a 50% cell growth inhibition. In this case, a lower value correlates with greater growth inhibition. It is evident that more efficacious BNCT agents are required. When analysing the data for the borylated drugs, a number of Structure-Activity Relationship considerations can be evinced.

Analysis of the Percent Cell Growth Inhibition Data
A general overarching consideration is that the vast majority of percent cell growth inhibitions for the borylated drugs are significantly greater than the percent cell growth inhibitions for the BSH and BPA controls.
There are three drugs that possess potent inhibitions, namely tetrachlorophthalimides para 3 and meta 3, and tetrachlorobenzamide ortho 5. para 3 is the only tetrachlorophthalimide that displays potent inhibition. It displays the boronate ester group in para position to the phthalimide and it has the CH 2 spacer between the phthalimide and the aromatic boron group. Inhibitions range from 98% to >100% for 9 out of 10 cancer cell lines and for the normal cell line. Only the A431 cancer cell line displays a lower inhibition (85%).
Tetrachlorophthalimide meta 3 possesses a similar structure to para 3, with the boronate ester group in meta position to the phthalimide. The installation of the boronate ester group in the meta position reduces potency significantly for all cancer lines and the normal cell line, though to varying extents. The smallest reduction in inhibition is seen in cell lines U87 (70%), MCF-7 (67%), A2780 (70%) and MCF10A (69%). A further loss in growth inhibition is seen in HT29 (53%), H460 (42%), A431 (48%), Du145 (43%) and BE2-C (45%). The greatest loss of growth inhibition is displayed in cell lines SJ-G2 (35%) and MIA-Pa-Ca2 (29%). Hence, the location of the boron ester group in para position greatly favours cell growth inhibition. These two drugs likely interact in similar ways with the cells, with the boronate ester group likely trying to interact with the same site/s (designated Site A for discussion purposes), but not managing quite as effectively when it is in the meta position. Table 3. Cancer Screening. On blue background DOSE SCREEN: Percentage (%) Cell Growth Inhibition in response to 25 µM of Drug (the higher the value the greater the growth inhibition) and inhibition value ranges: 80-100% (red), 60-79% (orange), 35-59% (green), 10-34% (blue) and 0-9% (black). On green background DOSE RESPONSE: GI 50 = Concentration (µM) that inhibits cell growth by 50% (the lower the value the greater the growth inhibition). In bold are highlighted the values of the three most potent drugs.      Table 3. Cancer Screening. On blue background DOSE SCREEN: Percentage (%) Cell Growth Inhibition in response to 25 μM of Drug (the higher the value the greater the growth inhibition) and inhibition value ranges: 80-100% (red), 60-79% (orange), 35-59% (green), 10-34% (blue) and 0-9% (black). On green background DOSE RESPONSE: GI50 = Concentration (μM) that inhibits cell growth by 50% (the lower the value the greater the growth inhibition). In bold are highlighted the values of the three most potent drugs.  Percentage (%) cell growth inhibition in response to 25 µM of drug after 72 h exposure using the MTT cell growth assay. The experiment was carried out in duplicate and replicated on 3 separate occasions.

Compound
Tetrachlorobenzamide ortho 5, displaying the boronate ester group in ortho position and having no CH 2 spacer between the benzamide and the aromatic boron group, also displays potent cell growth inhibition and a significant level of cell selectivity. This capability makes this drug the most interesting from a medicinal chemistry perspective. Cell selectivity (in particular, cancer versus healthy cell selectivity) is an area of active research in our group [72]. The removal of one of the carbonyl groups allows for greater conformational flexibility to this molecule, which may allow the boronate ester to interact with a different site than para 3 and meta 3 (designated Site B for discussion purposes). The ortho-phthalimide congener has not been synthetically achievable so far, and so it was not tested. Percent cell growth inhibitions range from 97 to >100 for most cancer cell lines (HT29, MCF-7, A2780, H460, A431, Du145, BE2-C and MIA-Pa-Ca2). It is 92% for SJ-G2 and drops dramatically for U87 (58%) and for the normal cell line MCF10A (58%). This selectivity between cancer versus healthy cells is a highly desirable drug capability.
The comparison in percent inhibition between benzamide ortho 5 and its congener ortho 8 is particularly interesting. The only structural difference is the CH 2 spacer between the benzamide and the aromatic boron group. However, there is a complete abrogation of cell growth inhibition produced by meta 3. It can be evinced that ortho 8 likely interacts with the same sites para 3 and meta 3 interact with, whereas ortho 5 has a different mode of action, interacting with another site that seems to be overexpressed in all cancer cell lines, apart from U87 and the normal cancer cell line MCF10A.
Benzamide para 5, not possessing the CH 2 spacer between the benzamide and the para-aromatic boron group, is thought to interact with Site A, due to structural similarities with its phthalimide congeners; however, it does not efficiently interact with Site A. This may be due to the boronate ester not reaching Site A due to the lack of the CH 2 spacer and the greater degree of conformational flexibility deriving from the removal of the carbonyl group from the phthalimide scaffold. This results in an overall reduction in cell growth inhibition. In this case as well, cell selectivity is displayed in inhibition. Normal cells MCF10A (2%) and Du145 (1%) were not inhibited, whereas U87 (21%) and SJ-G2 (28%) were somewhat more inhibited, MCF-7 (42%), A2780 (50%), H460 (54%), BE2-C (54%) and MIA-Pa-Ca2 (52%) were significantly inhibited, and finally, HT29 (69%) and A431 (64%) were inhibited the most.
Benzamide meta 5, not possessing the CH 2 spacer between the benzamide and the aromatic boron group, shows selective and significant inhibition for A2780 (49%). The benzamide structure provides a greater degree of conformational flexibility, likely placing the meta-positioned boronate ester somewhere in between Site A and Site B, and preventing it from interacting efficiently with either.
Benzamide ortho 8, possessing the CH 2 spacer between the benzamide and the aromatic boron group, is thought to interact with Site A, due to structural similarities with its phthalimide congeners; however, it does not efficiently interact with Site A, thus, showing almost complete abrogation of cell growth inhibition.
Based on Structure-Activity Relationship data obtained, two modes of cell growth inhibition are put forward, Mode of Action A, which arises from drugs interacting at Site A, and Mode of Action B, arising from drug ortho 5 interacting at Site B. Both Modes of Action can be elicited in selective ways by drugs para 8 (for Mode of Action A) and ortho 5 (for Mode of Action B) and displaying minimal or zero inhibition on the normal cell line.
Brush border membranes were prepared from the rat small intestine according to the method of Kessler et al. [83] and were assayed at pH 6.8 for rat intestinal maltase using maltose. For rat intestinal maltase, porcine kidney trehalase, and A. niger amyloglucosidase activities, the reaction mixture contained 25 mM maltose and the appropriate amount of enzyme, and the incubations were performed for 10-30 min at 37 • C. The reaction was stopped by heating at 100 • C for 3 min. After centrifugation (600× g; 10 min), the resulting reaction mixture was added to the Glucose CII-test Wako (Wako Pure Chemical Ind., Osaka, Japan). The absorbance at 505 nm was measured to determine the amount of the released D-glucose. Other glycosidase activities were determined using an appropriate para-nitrophenyl glycoside as substrate at the optimum pH of each enzyme. The reaction mixture contained 2 mM of the substrate and the appropriate amount of enzyme. The reaction was stopped by addition of 400 mM Na 2 CO 3 . The released para-nitrophenol was measured spectrometrically at 400 nm. All reactions run in methanol.

Glycosidase Inhibition Experimental from Laboratory 2
In Table 2 All enzymes and para-nitrophenyl substrates were purchased from Sigma. Enzymes were assayed at 27 • C in 0.1 M citric acid/0.2 M disodium hydrogen phosphate buffers at the optimum pH for the enzyme. The incubation mixture consisted of 10 µL enzyme solution, 10 µL of 1 mg/mL aqueous solution of extract and 50 µL of the appropriate 5 mM para-nitrophenyl substrate made up in buffer at the optimum pH for the enzyme. The reactions were stopped by addition of 70 µL 0.4 M glycine (pH 10.4) during the exponential phase of the reaction, which had been determined at the beginning using uninhibited assays in which water replaced inhibitor. Final absorbances were read at 405 nm using a Versamax microplate reader (Molecular Devices). Assays were carried out in triplicate, and the values given are means of the three replicates per assay. All reactions run in water.
Growth inhibition was determined by plating cells in duplicate in medium (100 µL) at a density of 2500-4000 cells per well in 96-well plates. On day 0 (24 h after plating), when the cells were in logarithmic growth, medium (100 µL) with or without the test agent was added to each well. After 72 h drug exposure, growth inhibitory effects were evaluated using the MTT (3-(4,5-dimethyltiazol-2-yl)-2,5-diphenyltetrazolium bromide) assay and absorbance read at 540 nm. The percentage growth inhibition was calculated at a fixed concentration of 25 µM, based on the difference between the optical density values on day 0 and those at the end of drug exposure. Each data point is the mean ± the standard error of the mean (SEM) calculated from three replicates which were performed on separate occasions and separate cell line passages.

General Experimental
Reaction solvents were purchased from the Aldrich Chemical Company (St Louis, MO, USA) in sure-seal TM reagent bottles. All other solvents (analytical or HPLC grade) were used as supplied without further purification. Deuterated chloroform (CDCl 3 ) and water (D 2 O) were used as NMR solvent. Triethylamine, sodium hydride (60% dispersion in mineral oil), tetrachlorophthalimide and tetrachlorophthalic anhydride were purchased from Sigma Aldrich. All boron-containing reagents were purchased from Boron Molecular, apart from BSH (>97%), 10 B-BSH (>97%), BPA (>98%) and 10 B-BPA (>98%) which came from Katchem spol. s r. o. The reagents were used as provided without further purification, with NMR analysis confirming an acceptable degree of purity and correct structural identity.
Purification via silica gel column chromatography was performed on Davisil 40-63-micron silica gel.
Thin layer chromatography (t.l.c.) was performed on aluminium sheets coated with 60 F254 silica by Merck and visualised using UVG-11 Compact UV lamp (254 nm) or stained with the cerium molybdate stain (12.0 g ammonium molybdate, 0.5 g ceric ammonium molybdate in 15 mL concentrated sulfuric acid and 235 mL distilled water).
Nuclear Magnetic Resonance (NMR) spectra were recorded on Bruker Ascend TM 400 in deuterated chloroform (CDCl 3 ). Chemical shifts (δ) are quoted in ppm and coupling constants (J) in Hz. Residual signals from the CDCl 3 (7.26 ppm for 1 H-NMR and 77.16 ppm for 13 C-NMR) were used as an internal reference [84].
Infrared spectroscopy (IR) spectra were obtained on a PerkinElmer Spectrum Two Spectrometer and on a PerkinElmer Spectrum 2 with UATR. Only characteristic peaks are quoted and in units of cm −1 .
High-resolution mass spectrometry (HRMS) spectra were obtained from samples suspended in acetonitrile (1 mL with 0.1% formic acid at a concentration of~1 mg/mL, before being further diluted to~10 ng/µL in 50% acetonitrile/water containing 0.1% formic acid). Samples were infused directly into the HESI source of a Thermo Scientific Q Exactive™ Plus Hybrid Quadrupole-Orbitrap™ Mass Spectrometer using an on-board syringe pump at 5 µL/min. Data were acquired on the QE+ in both positive and negative ion mode at a target resolution of 70,000 at 200 m/z. The predominant ions were manually selected for MS/MS fragmentation (collision energies were altered for each compound to obtain sufficient fragmentation). Data analysis of each sample was performed manually using Thermo Qualbrowser whilst the Isotopic Patterns of predicted chemical formula were modelled using Bruker Compass Isotope Pattern.
Crystallographic data were collected on an Oxford Diffraction Gemini CCD diffractometer employing either graphite-monochromated Mo-Kα radiation (0.71073 Å) or Cu-Kα (1.54184 Å). The sample was cooled to 190 K with and Oxford Cryosystems Desktop Cooler. Data reduction and empirical absorption corrections were performed with Oxford Diffraction CrysAlisPro software. Structures were solved by direct methods and refined with SHELXL [85]. All non-H atoms were refined with anisotropic thermal parameters. The crystal of ortho 8 was a non-merohedral twin which was refined using the HKLF 5 mode in SHELX. Molecular structure diagrams were produced with Mercury [86]. The data in CIF format were deposited at the Cambridge Crystallographic Data Centre (CCDC 215189 and 2151899).

Conclusions
We reported an expedited synthesis to a small library of novel borylated 2,3,4,5tetrachlorophthalimides and 2,3,4,5-tetrachlorobenzamides. Biological assays against glycosidase enzymes and cancer cell lines highlighted a good inhibitor for bovine liver β-galactosidase and three potent growth inhibitors and, of these, one selective growth inhibitor for cancer versus healthy cell lines in the cancer assay. These drugs are set for further derivatisations and utilisation in BNCT.