Synthesis and In vitro Activity of N -sulfonylamidine-derived Pyrimidine Analogues

T HIS PAPER IS DEDICATED TO PROF . M LADEN Ž INIĆ ON THE OCCASION OF HIS 70 TH BIRTHDAY Abstract: Two novel series of N -sulfonylamidino pyrimidine derivatives were synthesized via Cu-catalyzed three-component reaction of propargylated nucleobases with different benzenesulfonyl azides and amines. In this way 4 - acetamido, 4 - methyl and 4 -carboxybenzenesulfonyl amidine products 15–26 in the uracil series and 4 -acetamidobenzenesulfonyl amidine derivatives 27–29 in the cytosine series were prepared in 34 − 69 % yields. Attempts to prepare N - sulfonylamidino cytosine derivatives in reaction with 4 -methylbenzenesulfonyl azide were unsuccessful. The cytosine derivatives 32 and 33 were prepared from the N -sulfonylamidino uracil derivatives via the C4 triazole intermediates. The prepared N -sulfonylamidino pyrimidine derivatives 1–28 were tested for the antiproliferative activity on a panel of seven tumor cell lines of different histological origin (HeLa, Caco - 2, NCI - H358, Raji, HuT78, K562, Jurkat) and on normal MDCK I cells. Most of the synthesized compounds showed antiproliferative activity on the tested cell lines.


INTRODUCTION
N the last two decades, our group has been intensively involved in the design and synthesis of a new series of N-sulfonyl nucleobase derivatives that exhibit antitumor activity. [1][2][3][4] We have shown that N-1-sulfonylpyrimidine derivatives have strong antiproliferative activity on human tumor cell lines and an ability to induce apoptosis in the treated tumor cells. [5][6][7][8][9][10] In vivo experiments showed that some N-sulfonylcytosine derivatives had strong antitumor activity against mouse mammary carcinoma. [9,[11][12][13][14] Recently, we reported an efficient multicomponent synthesis of the new N-sulfonylamidino thymine derivatives 1−14 using Cu(I) catalyzed three component reactions of 1propargyl thymine, selected benzenesulfonyl azides, and amines or ammonium salts (Table 1). [15] We have shown that this one-pot three component reaction appears to be advantageous for the preparation of variously substituted N-sulfonylamidino thymine derivatives in moderate to good yields and opens the way for the preparation of the libraries of other nucleobase N-sulfonylamidino derivatives as possible biologically active molecules.
Tested compounds were dissolved in dimethyl sulfoxide as a 1 × 10 -2 mol dm -3 stock solution. Working dilutions were prepared in high pure water at a concentration range 10 -4 -10 -7 mol dm -3 .
For the MTT test, the adherent cells, MDCK1, HeLa, and Caco-2, were seeded in 96 micro-well plates at concentration of 2 × 10 4 cells / cm 3 and allowed to attach overnight in the CO2 incubator. After 72 hours of the exposure to tested compounds, medium was replaced with 5 mg cm -3 MTT solution and the resulting formazane crystals were dissolved in DMSO.
Leukemia cells (1 × 10 5 cells ⁄ cm 3 ), were plated onto 96 micro-well plates and after 72 hours of incubation, 5 mg cm -3 MTT solution was added to each well and incubated 4 hours in the CO2 incubator. To each well, 10 % SDS with 0.01 mol dm -3 HCl was added to dissolve waterinsoluble MTT-formazane crystals. The microplate reader (iMark, BIO RAD, Hercules, CA, USA) was used for measurement of the absorbance at 595 nm. All experiments were performed three times in triplicates.
The GI50 value, defined as the concentration of compound achieving 50 % of cell growth inhibition was calculated and used to compare cytotoxicity among the compounds. Calculation of GI50 value curves and QC analysis is performed by using the Excel tools and GraphPadPrism software (La Jolla, CA), v. 5.03. Briefly, individual concentration effect curves are generated by plotting the logarithm of the concentration of tested compounds (X) vs. corresponding percent inhibition values (Y) using least squares fit. The best fit GI50 values are calculated using Log (inhibitor) versus normalized response -Variable slope equation, where Y = 100 / (1 + 10^((log GI50-X) * HillSlope)). QC criteria parameters (Z0, S:B, R2, HillSlope) were checked for every GI50 curve.

Synthesis
Cu(I)-catalyzed 1,3-dipolar cycloadditions of azides with terminal alkynes (CuAAC) afford a 1,4-disubstituted 1,2,3triazole. This powerful, widely used reaction [17,18] is the most representative example of click chemistry. Employment of electron-deficient phosphoryl or sulfonyl azides led to a path change in the copper-catalyzed reaction with 1-alkynes. [19] Proposed key intermediate ketenimine, which is generated in situ upon ring-opening of the corresponding copper-triazole intermediate, undergoes addition reactions with various nucleophiles such as amines, alcohols, water or heterocycle compounds to give amidines, imidates, amides and other coupled compounds. [20] These three component reactions allow access to biologically interesting compounds that are typically otherwise prepared by multistep functional group transformations.
The synthesis of target N-sulfonylamidino uracils 15−26 was started by the preparation of 1-propargyl uracil. [21,22] First, the uracil was silylated with N,Obis(trimethylsilyl)acetamide (BSA) in acetonitrile and then the silylated intermediate was treated with propargyl bromide, giving the N-1 alkylated product in 64 % yield.
The copper-catalyzed reactions of 1-propargyl uracil with three different commercially available sulfonyl azides (4-acetamidobenzenesulfonyl, 4-methylbenzenesulfonyl and 4-carboxybenzenesulfonyl) and amines or amine salts were performed in THF or dichloromethane at room temperature affording N-sulfonylamidino uracils 15−26 in the 34−69 % yields (Table 1). All reactions included 20 % molar excess of sulfonyl azide and amine, except the reaction with 4-carboxybenzenesulfonyl azide (entry 12-14), where a 100 % molar excess of amine was required. In the case of amine/ammonium salts (entry 10 and 11), additional triethylamine base in 50 % molar excess was used. [15] Table 1 provides a comparison between the previously reported [15] N-sulfonylamidino thymine derivatives 1−14 and newly synthesized compounds from the uracil series 15−26. Compared to the propargyl thymine, in all reactions, a slightly lower reactivity of propargyl uracil is apparent, with the exception of compound 24 (entry 12, Table 1). Among used azides, reaction with 4-acetamidobenzenesulfonyl azide afforded products with highest yields.
As expected reactions of 1-propargyl uracil with 4-acetamidobenzenesulfonyl azide or 4-methylbenzenesulfonyl azide (tosyl azide) and with secondary amines (Table 1, compounds 15, 16 and 22) gave better yields compared to those with primary amines (Table 1,  compounds 17, 18 and 23) and aromatic amine (Table 1, compound 19). The reactions with 4-carboxybenzenesulfonyl azide afforded products 24-26 in lower yields (Table 1, entry 12-14). A similar trend has been noticed for thymine series. A significant difference between the reactivity of propargyl uracil and propargyl thymine was observed since the preparations of the uracil analogs of compounds 10 and 11, failed. Although identical conditions were used, with the same reagents, there was no sign that the reaction occurred even at elevated temperatures, and the unreacted 1-propargyl uracil was isolated from the reaction mixture (> 90 %).
In the next step we decided to examine the conditions for the three-component coupling reactions with 1-propargyl cytosine, which was synthesized by the known condensation method of N 4 -acetylcytosine with propargyl bromide. [23] The cytosine amino group was acetylated with acetic anhydride in pyridine [24] and obtained N 4 -acetylcytosine was activated with K2CO3 in DMF and condensed with propargyl bromide. In the reaction mixture N1 and O2 regioisomers were obtained in the 96:4 ratio. N1 isomer was isolated by recrystallization from water and the acetyl group was readily removed by methanolic ammonia to give the desired N-1-propargyl cytosine.
Next, using the same conditions for Cu-catalyzed three-component reaction (Method A), 1-propargyl cytosine was reacted with 4-acetamidobenzenesulfonyl azide and secondary or primary amines giving desired products 27-29 in 45-62 % yields (Scheme 1). Cu-catalyzed three-component reactions of 1-propargyl cytosine and ammonium chloride (Method B) or 4-methylbenzenesulfonyl azide (Method A) failed to give any N-sulfonylamidine product and the starting material, 1-propargyl cytosine, was completely recovered.

In vitro Antiproliferative Screening
We have shown before that N-1-sulfonylpyrimidine derivatives have strong and selective antiproliferative activity on different human tumor cell lines in vitro and on xenograft model in vivo. [5][6][7][8][9][10] Amidines are important units in chemistry for the synthesis of heterocycles [29,30] and widely used in bioactive chemicals and drug molecular design. [31][32][33] We assumed the improved antiproliferative capacity of novel hybrid compounds could be obtained if Table 2. Inhibitory effects of N-sulfonylamidine pyrimidine derivatives on the growth of normal and human tumor cells.

Normal cells
As indicated in Table 2, two large groups of prepared N-sulfonylamidine-derived pyrimidine analogues showed great variation in the antiproliferative effect on tumor cell lines, depending on the cell line and structure of the tested compounds.

CONCLUSIONS
In conclusion, the series of new aliphatic N-sulfonylamidino pyrimidine derivatives incorporating nucleobase, Nsulfonyl and amidine pharmacophores in the structure were synthesized by Cu(I)-catalyzed three-component coupling of 1-propargyl nucleobase, benzenesulfonyl azides and amines. New N-sulfonylamidino pyrimidine derivatives possess good inhibitory potential against tested tumor cells. These results stimulate further studies directed to investigate their mechanisms of action.