Regioselective Synthesis of New 2,4-(Het)aryl-3H-pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidines Involving Palladium-Catalyzed Cross-Coupling Reactions

The design of some novel di-(het)arylated-3H-pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidine derivatives is reported. The series was developed from 1-aminopyridinium iodide, which afforded the key intermediate bearing two thiomethyl and amide functions, each of them useful for palladium catalyzed cross coupling reactions by alkyl sulfur release and C-O activation, respectively. The two regioselective and successive cross-coupling reactions were first carried out in C-4 by in situ C-O activation and next in C-2 by a methylsulfur release. Process optimization furnished conditions leading to products in high yields. The scope and limitations of the methodologies were evaluated and the final compounds characterized.


Introduction
The exploration of chemical space is a key prior step in the discovery of biologically active molecules. This strategy in heterocyclic chemistry includes the design and the functionalization of polynitrogen structures which are chosen for their similarity with major biomarkers and may contain scaffolds such as indoles, imidazoles, pyrimidine or imidazopyrimidine as well as tricyclic folate cores [1][2][3]. In view of their potential for the design of bioactive molecules, a variety of polynitrogenated skeletons have received considerable attention due to the synthetic challenge that they represent [4]. In this area, tricyclic fused heteroaromatic derivatives with a bridgehead nitrogen have less often been examined [5,6]. There is therefore a need to provide synthetic methods for their functionalization through reproducible and versatile strategies.
Apart from the more classical pyrimidine ring, the currently most promising scaffold appears to be one containing a pyrazolopyridine, which has been used in diverse biologically and pharmaceutically active molecules including anticancer drugs [22], diuretic [23] and antiherpetic [24] drugs, and p38 kinase inhibitors [25]. With a view to increasing molecular diversity, tricyclic In order to build C-2 and C-4 disubstituted pyrido[1′,2′:1,5]pyrazolo [4,3-d]pyrimidine derivatives, we developed a straightforward strategy which included from a versatile pattern 9, a C-4 Suzuki-Miyaura cross coupling reaction using an in situ C-O activation [28,29] followed by a C-2 Liebeskind-Srogl cross-coupling reaction [20,[30][31][32][33]. We report herein the preparation of 9 and its regioselective functionalization, the optimization of the experimental conditions and finally the scope of both reactions on these two identified positions ( Figure 2).

Results and Discussion
To design the platform 9, we modified the available access (Scheme 1). Starting from the commercially available 1-aminopyridinium iodide (1), the 1,3-dipolar condensations of dimethyl acetylenedicarboxylate (DMAD) led to diester 2 in 78% yield (versus 29% in the literature [34]) and after saponification, the di-acid 3 was generated in a quantitative manner (64% in the literature). We next replaced the reported 4-step synthesis (decarboxylation, methylation, nitration, reduction) [27,34] by a shorter 3-step sequence. The discrimination of carboxylic acid functions was performed using an in situ anhydride formation followed by its regioselective opening with methanol to give the mono-methyl ester 4 [11]. This synthetic strategy allowed us to achieve a Curtius rearrangement on the residual acid function to produce, after cleavage of the Boc protective group, the hetaryl amine 6. Finally, the condensation of benzoyl isothiocyanate with 6 followed by cyclization furnished 8 in basic media which was converted to the thiomethyl derivative 9 in satisfactory yield.
In order to tackle the usefulness of 9 as a building block and taking advantage of the presence of the amide group, we began by the regioselective functionalization of the C-4 position using a C-O direct activation involving PyBroP and Et3N to generate the required in situ generated Ophosphonium leaving group [35,36]. The second step required the addition of boronic acid, base, water and catalyst source. Some representative pyrido [1 ,2 :1,5]pyrazolo [3,4-d]pyrimidine derivatives I and pyrido [1 ,2 :1,5]pyrazolo [4,3-d]pyrimidine II, III.

Results and Discussion
To design the platform 9, we modified the available access (Scheme 1). Starting from the commercially available 1-aminopyridinium iodide (1), the 1,3-dipolar condensations of dimethyl acetylenedicarboxylate (DMAD) led to diester 2 in 78% yield (versus 29% in the literature [34]) and after saponification, the di-acid 3 was generated in a quantitative manner (64% in the literature). We next replaced the reported 4-step synthesis (decarboxylation, methylation, nitration, reduction) [27,34] by a shorter 3-step sequence. The discrimination of carboxylic acid functions was performed using an in situ anhydride formation followed by its regioselective opening with methanol to give the mono-methyl ester 4 [11]. This synthetic strategy allowed us to achieve a Curtius rearrangement on the residual acid function to produce, after cleavage of the Boc protective group, the hetaryl amine 6. Finally, the condensation of benzoyl isothiocyanate with 6 followed by cyclization furnished 8 in basic media which was converted to the thiomethyl derivative 9 in satisfactory yield.
In order to tackle the usefulness of 9 as a building block and taking advantage of the presence of the amide group, we began by the regioselective functionalization of the C-4 position using a C-O direct activation involving PyBroP and Et3N to generate the required in situ generated Ophosphonium leaving group [35,36]. The second step required the addition of boronic acid, base, water and catalyst source.

Results and Discussion
To design the platform 9, we modified the available access (Scheme 1). Starting from the commercially available 1-aminopyridinium iodide (1), the 1,3-dipolar condensations of dimethyl acetylenedicarboxylate (DMAD) led to diester 2 in 78% yield (versus 29% in the literature [34]) and after saponification, the di-acid 3 was generated in a quantitative manner (64% in the literature). We next replaced the reported 4-step synthesis (decarboxylation, methylation, nitration, reduction) [27,34] by a shorter 3-step sequence. The discrimination of carboxylic acid functions was performed using an in situ anhydride formation followed by its regioselective opening with methanol to give the mono-methyl ester 4 [11]. This synthetic strategy allowed us to achieve a Curtius rearrangement on the residual acid function to produce, after cleavage of the Boc protective group, the hetaryl amine 6. Finally, the condensation of benzoyl isothiocyanate with 6 followed by cyclization furnished 8 in basic media which was converted to the thiomethyl derivative 9 in satisfactory yield.
In order to tackle the usefulness of 9 as a building block and taking advantage of the presence of the amide group, we began by the regioselective functionalization of the C-4 position using a C-O direct activation involving PyBroP and Et 3 N to generate the required in situ generated O-phosphonium leaving group [35,36]. The second step required the addition of boronic acid, base, water and catalyst source. We used conditions exploited in our previous research, which involved Na2CO3 and PdCl2(dppf).CH2Cl2 under thermal conditions. Under these conditions which have proved to be useful in diverse heterocyclic series, the reaction was successfully achieved with para-tolylboronic acid to afford 10 in a good 85% yield (Table 1, entry 1).
To evaluate the effect of boronic acid substituents, we first studied the reaction using other electron-donating groups such as methoxy in ortho, meta or para positions (entries 2-4). In para position, no alteration in the reaction efficiency was observed and compound 11 was isolated in 87% yield (entry 2). In contrast, the ortho orientation of the OMe group induced steric hindrance and consequently, a dramatic decrease in yield was observed. An intermediate yield was obtained with the meta-oriented methyl ester boronic acid (entries 3, 4) as the meta-methoxy group induces an inductive withdrawing effect. With an electron-withdrawing substituent on the phenylboronic acid, efficiency depends on the nature and strength of the electronic effects. With a moderate inductive effect, reactivity was maintained and 14 was isolated in a 90% yield. In the presence of strong electronwithdrawing substituents such as CF3 or CN the yields decreased to 70% and 58%, respectively (entries 6, 7). The same behaviour was observed with heteroaryl boronic acids. Electron-rich heterocycles such as 2-furan were introduced in good yield (74%, entry 9) whereas no reaction occurred with π-deficient heterocycles such as 4-pyridylboronic acid (entry 10). Finally, we investigated the interference of the OH group in basic media, which led to a moderate yield compared to its methyl ether equivalent (entry 8 vs. 2). In conclusion, the method is suitable with a wide variety of boronic acids, and limits concern only strongly deactivated (het)aryl boronic acids.
We next investigated the reactivity of the C-2 position. Methyl sulfur release was achieved using a Liebeskind-Srogl reaction which furnished the desired original 2,4-disubstituted compounds. The reactivity of the thioether 10 was explored using various (het)aryl boronic acids in the presence of Pd(PPh3)4 and copper(I) thiophene-2-carboxylate (CuTc) in THF at 100 °C under microwave irradiation ( Table 2). As observed during the first cross coupling procedure, the use of electron-rich phenyl boronic acids was well tolerated and furnished the 4-tolyl-or 4-methoxyphenyl-substituted derivatives 20 and 21 in excellent yields (entries 1-2). The compartmental similarity between the two reactions is even clearer in the following results.
Firstly, the OMe position switch to the 2-or 3-methoxyphenyl boronic acids afforded the desired products in moderate yields (entries 3 and 4). Secondly, the presence of a strong electronwithdrawing CF3 group on the phenylboronic acid lowered the efficiency of the reaction. Nevertheless, the CF3 vs F atom exchange fortunately restored the reactivity and led to 24 in a satisfactory 85% yield (entry 6 vs. 5) indicating that decreasing the electron-withdrawing character had a major impact on the reaction rate. This result was confirmed with (het)aryl boronic acids as Scheme 1. Synthetic pathway to generate 9.
We used conditions exploited in our previous research, which involved Na 2 CO 3 and PdCl 2 (dppf).CH 2 Cl 2 under thermal conditions. Under these conditions which have proved to be useful in diverse heterocyclic series, the reaction was successfully achieved with para-tolylboronic acid to afford 10 in a good 85% yield ( Table 1, entry 1).
To evaluate the effect of boronic acid substituents, we first studied the reaction using other electron-donating groups such as methoxy in ortho, meta or para positions (entries 2-4). In para position, no alteration in the reaction efficiency was observed and compound 11 was isolated in 87% yield (entry 2). In contrast, the ortho orientation of the OMe group induced steric hindrance and consequently, a dramatic decrease in yield was observed. An intermediate yield was obtained with the meta-oriented methyl ester boronic acid (entries 3, 4) as the meta-methoxy group induces an inductive withdrawing effect. With an electron-withdrawing substituent on the phenylboronic acid, efficiency depends on the nature and strength of the electronic effects. With a moderate inductive effect, reactivity was maintained and 14 was isolated in a 90% yield. In the presence of strong electron-withdrawing substituents such as CF 3 or CN the yields decreased to 70% and 58%, respectively (entries 6, 7). The same behaviour was observed with heteroaryl boronic acids. Electron-rich heterocycles such as 2-furan were introduced in good yield (74%, entry 9) whereas no reaction occurred with π-deficient heterocycles such as 4-pyridylboronic acid (entry 10). Finally, we investigated the interference of the OH group in basic media, which led to a moderate yield compared to its methyl ether equivalent (entry 8 vs. 2). In conclusion, the method is suitable with a wide variety of boronic acids, and limits concern only strongly deactivated (het)aryl boronic acids.
We next investigated the reactivity of the C-2 position. Methyl sulfur release was achieved using a Liebeskind-Srogl reaction which furnished the desired original 2,4-disubstituted compounds. The reactivity of the thioether 10 was explored using various (het)aryl boronic acids in the presence of Pd(PPh 3 ) 4 and copper(I) thiophene-2-carboxylate (CuTc) in THF at 100 • C under microwave irradiation ( Table 2). As observed during the first cross coupling procedure, the use of electron-rich phenyl boronic acids was well tolerated and furnished the 4-tolyl-or 4-methoxyphenyl-substituted derivatives 20 and 21 in excellent yields (entries 1-2). The compartmental similarity between the two reactions is even clearer in the following results.
Firstly, the OMe position switch to the 2-or 3-methoxyphenyl boronic acids afforded the desired products in moderate yields (entries 3 and 4). Secondly, the presence of a strong electron-withdrawing CF 3 group on the phenylboronic acid lowered the efficiency of the reaction. Nevertheless, the CF 3 vs. F atom exchange fortunately restored the reactivity and led to 24 in a satisfactory 85% yield (entry 6 vs. 5) indicating that decreasing the electron-withdrawing character had a major impact on the reaction rate. This result was confirmed with (het)aryl boronic acids as electron-rich aromatics such as furan provided the desired product 27 in good yield whereas attempts with 4-pyridineboronic acid totally failed (entries 8,9). Table 1. Scope of (het)arylation of 9 under C-O amide activation.  To assess whether the nature of the C-4 residue really interferes with the C-2 reactivity, we varied the nature of 4-arylated tricyclic derivatives and fixed p-tolylboronic acid as sole arylation partner. When the pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidine derivative was substituted by a singly enriched 4-MeO-phenyl group at the C-4 position, the reaction gave derivative 29 in very good yield (entry 10), higher than that observed with 10 (entry 1). The inversion of the electronic effect in C-4 reduced the reactivity of the tricyclic system (entry 11) as was confirmed with the assay conducted starting from 14 that led to 30 in only 53% yield. In conclusion, the C-2 (het)arylation followed the same behavior as the C-4 cross coupling reaction with an additional favorable parameter i.e., the presence of electron rich (het)aryl moieties in C-4. electron-rich aromatics such as furan provided the desired product 27 in good yield whereas attempts with 4-pyridineboronic acid totally failed (entries 8,9). To assess whether the nature of the C-4 residue really interferes with the C-2 reactivity, we varied the nature of 4-arylated tricyclic derivatives and fixed p-tolylboronic acid as sole arylation partner. When the pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidine derivative was substituted by a singly enriched 4-MeO-phenyl group at the C-4 position, the reaction gave derivative 29 in very good yield (entry 10), higher than that observed with 10 (entry 1). The inversion of the electronic effect in C-4 reduced the reactivity of the tricyclic system (entry 11) as was confirmed with the assay conducted starting from 14 that led to 30 in only 53% yield. In conclusion, the C-2 (het)arylation followed the same behavior as the C-4 cross coupling reaction with an additional favorable parameter i.e., the presence of electron rich (het)aryl moieties in C-4.  To assess whether the nature of the C-4 residue really interferes with the C-2 reactivity, we varied the nature of 4-arylated tricyclic derivatives and fixed p-tolylboronic acid as sole arylation partner. When the pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidine derivative was substituted by a singly enriched 4-MeO-phenyl group at the C-4 position, the reaction gave derivative 29 in very good yield (entry 10), higher than that observed with 10 (entry 1). The inversion of the electronic effect in C-4 reduced the reactivity of the tricyclic system (entry 11) as was confirmed with the assay conducted starting from 14 that led to 30 in only 53% yield. In conclusion, the C-2 (het)arylation followed the same behavior as the C-4 cross coupling reaction with an additional favorable parameter i.e., the presence of electron rich (het)aryl moieties in C-4.  To assess whether the nature of the C-4 residue really interferes with the C-2 reactivity, we varied the nature of 4-arylated tricyclic derivatives and fixed p-tolylboronic acid as sole arylation partner. When the pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidine derivative was substituted by a singly enriched 4-MeO-phenyl group at the C-4 position, the reaction gave derivative 29 in very good yield (entry 10), higher than that observed with 10 (entry 1). The inversion of the electronic effect in C-4 reduced the reactivity of the tricyclic system (entry 11) as was confirmed with the assay conducted starting from 14 that led to 30 in only 53% yield. In conclusion, the C-2 (het)arylation followed the same behavior as the C-4 cross coupling reaction with an additional favorable parameter i.e., the presence of electron rich (het)aryl moieties in C-4.  To assess whether the nature of the C-4 residue really interferes with the C-2 reactivity, we varied the nature of 4-arylated tricyclic derivatives and fixed p-tolylboronic acid as sole arylation partner. When the pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidine derivative was substituted by a singly enriched 4-MeO-phenyl group at the C-4 position, the reaction gave derivative 29 in very good yield (entry 10), higher than that observed with 10 (entry 1). The inversion of the electronic effect in C-4 reduced the reactivity of the tricyclic system (entry 11) as was confirmed with the assay conducted starting from 14 that led to 30 in only 53% yield. In conclusion, the C-2 (het)arylation followed the same behavior as the C-4 cross coupling reaction with an additional favorable parameter i.e., the presence of electron rich (het)aryl moieties in C-4.  To assess whether the nature of the C-4 residue really interferes with the C-2 reactivity, we varied the nature of 4-arylated tricyclic derivatives and fixed p-tolylboronic acid as sole arylation partner. When the pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidine derivative was substituted by a singly enriched 4-MeO-phenyl group at the C-4 position, the reaction gave derivative 29 in very good yield (entry 10), higher than that observed with 10 (entry 1). The inversion of the electronic effect in C-4 reduced the reactivity of the tricyclic system (entry 11) as was confirmed with the assay conducted starting from 14 that led to 30 in only 53% yield. In conclusion, the C-2 (het)arylation followed the same behavior as the C-4 cross coupling reaction with an additional favorable parameter i.e., the presence of electron rich (het)aryl moieties in C-4.  To assess whether the nature of the C-4 residue really interferes with the C-2 reactivity, we varied the nature of 4-arylated tricyclic derivatives and fixed p-tolylboronic acid as sole arylation partner. When the pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidine derivative was substituted by a singly enriched 4-MeO-phenyl group at the C-4 position, the reaction gave derivative 29 in very good yield (entry 10), higher than that observed with 10 (entry 1). The inversion of the electronic effect in C-4 reduced the reactivity of the tricyclic system (entry 11) as was confirmed with the assay conducted starting from 14 that led to 30 in only 53% yield. In conclusion, the C-2 (het)arylation followed the same behavior as the C-4 cross coupling reaction with an additional favorable parameter i.e., the presence of electron rich (het)aryl moieties in C-4.  To assess whether the nature of the C-4 residue really interferes with the C-2 reactivity, we varied the nature of 4-arylated tricyclic derivatives and fixed p-tolylboronic acid as sole arylation partner. When the pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidine derivative was substituted by a singly enriched 4-MeO-phenyl group at the C-4 position, the reaction gave derivative 29 in very good yield (entry 10), higher than that observed with 10 (entry 1). The inversion of the electronic effect in C-4 reduced the reactivity of the tricyclic system (entry 11) as was confirmed with the assay conducted starting from 14 that led to 30 in only 53% yield. In conclusion, the C-2 (het)arylation followed the same behavior as the C-4 cross coupling reaction with an additional favorable parameter i.e., the presence of electron rich (het)aryl moieties in C-4. 13 20 9 Molecules 2018, 23, x 4 of 14 electron-rich aromatics such as furan provided the desired product 27 in good yield whereas attempts with 4-pyridineboronic acid totally failed (entries 8,9). To assess whether the nature of the C-4 residue really interferes with the C-2 reactivity, we varied the nature of 4-arylated tricyclic derivatives and fixed p-tolylboronic acid as sole arylation partner. When the pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidine derivative was substituted by a singly enriched 4-MeO-phenyl group at the C-4 position, the reaction gave derivative 29 in very good yield (entry 10), higher than that observed with 10 (entry 1). The inversion of the electronic effect in C-4 reduced the reactivity of the tricyclic system (entry 11) as was confirmed with the assay conducted starting from 14 that led to 30 in only 53% yield. In conclusion, the C-2 (het)arylation followed the same behavior as the C-4 cross coupling reaction with an additional favorable parameter i.e., the presence of electron rich (het)aryl moieties in C-4.  To assess whether the nature of the C-4 residue really interferes with the C-2 reactivity, we varied the nature of 4-arylated tricyclic derivatives and fixed p-tolylboronic acid as sole arylation partner. When the pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidine derivative was substituted by a singly enriched 4-MeO-phenyl group at the C-4 position, the reaction gave derivative 29 in very good yield (entry 10), higher than that observed with 10 (entry 1). The inversion of the electronic effect in C-4 reduced the reactivity of the tricyclic system (entry 11) as was confirmed with the assay conducted starting from 14 that led to 30 in only 53% yield. In conclusion, the C-2 (het)arylation followed the same behavior as the C-4 cross coupling reaction with an additional favorable parameter i.e., the presence of electron rich (het)aryl moieties in C-4.  To assess whether the nature of the C-4 residue really interferes with the C-2 reactivity, we varied the nature of 4-arylated tricyclic derivatives and fixed p-tolylboronic acid as sole arylation partner. When the pyrido[1′,2′:1,5]pyrazolo[4,3-d]pyrimidine derivative was substituted by a singly enriched 4-MeO-phenyl group at the C-4 position, the reaction gave derivative 29 in very good yield (entry 10), higher than that observed with 10 (entry 1). The inversion of the electronic effect in C-4 reduced the reactivity of the tricyclic system (entry 11) as was confirmed with the assay conducted starting from 14 that led to 30 in only 53% yield. In conclusion, the C-2 (het)arylation followed the same behavior as the C-4 cross coupling reaction with an additional favorable parameter i.e., the presence of electron rich (het)aryl moieties in C-4. 19 traces a Yield is indicated as isolated compound.
To assess whether the nature of the C-4 residue really interferes with the C-2 reactivity, we varied the nature of 4-arylated tricyclic derivatives and fixed p-tolylboronic acid as sole arylation partner. When the pyrido[1 ,2 :1,5]pyrazolo[4,3-d]pyrimidine derivative was substituted by a singly enriched 4-MeO-phenyl group at the C-4 position, the reaction gave derivative 29 in very good yield (entry 10), higher than that observed with 10 (entry 1). The inversion of the electronic effect in C-4 reduced the reactivity of the tricyclic system (entry 11) as was confirmed with the assay conducted starting from 14 that led to 30 in only 53% yield. In conclusion, the C-2 (het)arylation followed the same behavior as the C-4 cross coupling reaction with an additional favorable parameter i.e., the presence of electron rich (het)aryl moieties in C-4.

General Methods
Reactions were monitored by thin-layer chromatography (TLC) using silica gel (60 F254) plates, and the compounds were visualized by UV irradiation. Flash column chromatography was performed with silica gel 60 (230-400.13 mesh, 0.040 × 0.063 mm). The melting points were measured with samples in open capillary tubes. The infrared spectra of compounds were recorded with a Nicolet iS10 instrument (Thermo Scientific, Waltham, MA, USA). The 1 H-and 13 C-NMR spectra were recorded with DPX 250 ( 13 C, 62 MHz), Avance II 250 ( 13 C, 63 MHz), Avance 400 ( 13 C, 101 MHz), or Avance III HD Nanobay 400 ( 13 C, 101 MHz) spectrometers (Bruker, Billerica, MA, USA). The chemical shifts are given in ppm from tetramethylsilane as an internal standard. The coupling constants (J) are reported in Hz. High-resolution mass spectra (HRMS) were recorded with a Bruker Maxis 4G instrument Microwave irradiation was carried out in sealed vessels placed in a Initiator or Initiator + system (400 W maximum power, Biotage, Uppsala, Sweden). The temperatures were measured externally by IR. Pressure was measured by a noninvasive sensor integrated in the cavity lid.

Conclusions
In summary, we have described in this work a synthetic pathway for the preparation of an original pyrido[1 ,2 :1,5]pyrazolo [4,3-d]pyrimidine platform, and then, have developed several arylations at its C-2 and C-4 positions. First, the amide function in C-4 position was reacted in a direct C-O activation with PyBroP as activator followed by a Suzuki-Miyaura cross coupling reaction to generate a library of C-4 (het)arylated derivatives and next a Liebeskind-Srogl reaction furnished the desired di-(het)arylated derivatives. The scope of the two reactions was investigated and showed a strong influence of electronic effect and steric hindrance. In both cases electron enrichment of the systems, i.e., the tricyclic core as well as the boronic partner, improved efficiency. This route will offer the opportunity to explore other metal catalyzed cross coupling reactions and to open a new chemical space area to generate new bioactive compounds containing polyfunctionalized pyrido[1 ,2 :1,5]pyrazolo [4,3-d]pyrimidines.