5′-Nor-3-Deaza-1′,6′-Isoneplanocin, the Synthesis and Antiviral Study

The arbocyclic nucleosides aristeromycin and neplanocin have been studied as a source for new antiviral agents. A convenient synthesis of C-5′-truncated 3-deaza-1′,6′-isoneplanocin, which combines the features of antiviral candidates 5′-noraristeromycin and 3-deaza-1′,6′-isoneplanocin is reported from (−)-cyclopentenone to give the two C-4′ epimers of 5′-nor-3-deaza isoneplanocin. Antiviral assays showed activity against the JC virus (EC50 = 1.12 µM for (4′R)-8; EC50 = 59.14 µM for (4′S)-7) and inactivity of both compounds against several DNA and RNA viruses. Both compounds lacked cytotoxicity.


Introduction
Emerging and reemerging viral infectious diseases are continuously posing huge threats to global public health and have had a substantial socioeconomic impact. For example, a total of 28,616 confirmed and suspected cases with 11,310 deaths were reported during the 2014-2016 Ebola outbreak [1]. At the end of 2019, a novel coronavirus, named SARS-CoV-2, emerged and has infected 12,970,605 people in 188 countries/regions with 570,220 deaths (as of 13 July 2020 [2]) and continues to increase.
In the search for antiviral countermeasures, repurposed or newly designed nucleosides and nucleotide analogues are serving as a resource for the frontline defense, especially in those urgent situations [3,4]. For instance, BCX 4430 (Galidesivir, a) and GS-5734 (Remdesivir, b) ( Figure 1) were developed during the 2014-2016 Ebola outbreak [5]. Because of its activity towards SARS-CoV-2, Remdesivir is being repurposed for treatment in this current pandemic.

Introduction
Emerging and reemerging viral infectious diseases are continuously posing huge threats to global public health and have had a substantial socioeconomic impact. For example, a total of 28,616 confirmed and suspected cases with 11,310 deaths were reported during the 2014-2016 Ebola outbreak [1]. At the end of 2019, a novel coronavirus, named SARS-CoV-2, emerged and has infected 12,970,605 people in 188 countries/regions with 570,220 deaths (as of 13 July 2020 [2]) and continues to increase.
In the search for antiviral countermeasures, repurposed or newly designed nucleosides and nucleotide analogues are serving as a resource for the frontline defense, especially in those urgent situations [3,4]. For instance, BCX 4430 (Galidesivir, a) and GS-5734 (Remdesivir, b) ( Figure 1) were developed during the 2014-2016 Ebola outbreak [5]. Because of its activity towards SARS-CoV-2, Remdesivir is being repurposed for treatment in this current pandemic. Galidesivir (a) and Remdesivir (b) are C-nucleosides with the glycosidic linkage replaced by a more stable C-C bond and, hence, are metabolically stable to hydrolytic and phosphorlytic breakdown, Molecules 2020, 25, 3865 2 of 5 a relevant feature for nucleoside-based therapeutic candidates [6]. A similar property is seen with carbocyclic nucleosides, such as the naturally occurring aristeromycin (1) and neplanocin A (2), ( Figure 2) which possess antibacterial, -parasitic, -viral and -cancer properties [3,7], due, principally, to the non-selective inhibition of S-adenosylhomocysteine hydrolase (SAHase). The therapeutic of 1 and 2 is limited by their cytotoxicity as a result of biomolecular inference by their 5 -phosphate metabolites.
Galidesivir (a) and Remdesivir (b) are C-nucleosides with the glycosidic linkage replaced by a more stable C-C bond and, hence, are metabolically stable to hydrolytic and phosphorlytic breakdown, a relevant feature for nucleoside-based therapeutic candidates [6]. A similar property is seen with carbocyclic nucleosides, such as the naturally occurring aristeromycin (1) and neplanocin A (2), (Figure 2) which possess antibacterial, -parasitic, -viral and -cancer properties [3,7], due, principally, to the non-selective inhibition of S-adenosylhomocysteine hydrolase (SAHase). The therapeutic of 1 and 2 is limited by their cytotoxicity as a result of biomolecular inference by their 5′phosphate metabolites.
To address this undesirable feature, the C-4′ truncated variations (3 and 4) were prepared and found to be effective against a number of viruses and to be non-cytotoxic [8]. A similar modification on neplanocin A (that is, 5) is, however, unlikely due to its enolic structure (red structure in Figure  2). Another carbocyclic nucleoside structural modification developed in our labs has been the 1′,6′isoneplanocin series (herein designated as isoneplanocin and represented by the 3-deaza analogue, 6) that displays a broad-based, non-cyctotoxic antiviral profile [9]. We have recently desired to combine the features of 3 with 6 and, thus, set 7 and 8 as targets. (Figure 3) These results are reported here.

Results
Ullmann coupling of a vinyl iodide with an adenine moiety is well established in our lab as a powerful synthetic tool for the preparation of 1′,6′-isoneplanocin analogues [9]. For the purposes of this investigation, vinyl halide 11 was foreseen as the requisite building block. Its synthesis (Scheme To address this undesirable feature, the C-4 truncated variations (3 and 4) were prepared and found to be effective against a number of viruses and to be non-cytotoxic [8]. A similar modification on neplanocin A (that is, 5) is, however, unlikely due to its enolic structure (red structure in Figure 2).
Another carbocyclic nucleoside structural modification developed in our labs has been the 1 ,6 -isoneplanocin series (herein designated as isoneplanocin and represented by the 3-deaza analogue, 6) that displays a broad-based, non-cyctotoxic antiviral profile [9]. We have recently desired to combine the features of 3 with 6 and, thus, set 7 and 8 as targets. (Figure 3) These results are reported here. Galidesivir (a) and Remdesivir (b) are C-nucleosides with the glycosidic linkage replaced by a more stable C-C bond and, hence, are metabolically stable to hydrolytic and phosphorlytic breakdown, a relevant feature for nucleoside-based therapeutic candidates [6]. A similar property is seen with carbocyclic nucleosides, such as the naturally occurring aristeromycin (1) and neplanocin A (2), (Figure 2) which possess antibacterial, -parasitic, -viral and -cancer properties [3,7], due, principally, to the non-selective inhibition of S-adenosylhomocysteine hydrolase (SAHase). The therapeutic of 1 and 2 is limited by their cytotoxicity as a result of biomolecular inference by their 5′phosphate metabolites.
To address this undesirable feature, the C-4′ truncated variations (3 and 4) were prepared and found to be effective against a number of viruses and to be non-cytotoxic [8]. A similar modification on neplanocin A (that is, 5) is, however, unlikely due to its enolic structure (red structure in Figure  2). Another carbocyclic nucleoside structural modification developed in our labs has been the 1′,6′isoneplanocin series (herein designated as isoneplanocin and represented by the 3-deaza analogue, 6) that displays a broad-based, non-cyctotoxic antiviral profile [9]. We have recently desired to combine the features of 3 with 6 and, thus, set 7 and 8 as targets. (Figure 3) These results are reported here.

Results
Ullmann coupling of a vinyl iodide with an adenine moiety is well established in our lab as a powerful synthetic tool for the preparation of 1′,6′-isoneplanocin analogues [9]. For the purposes of this investigation, vinyl halide 11 was foreseen as the requisite building block. Its synthesis (Scheme
Our attention turned to employing the Ullmann coupling of 11 and 3-deazaadenine. This succeeded in giving 14 (Scheme 2) in a moderate yield in contrast to 12, suggesting a hydroxyl substituent adjacent to the vinyl coupling site was necessary for the Ullmann to succeed. Acid deprotection of 14 availed the desired (4′R)-8. In addition to NMR data, the structure of 8 was confirmed by X-ray crystallography (CCDC 2018731), which served to confirm the regiochemistry of the cyclopentenyl and the 3-deaza base of 8 (Supplementary Meterials).
To achieve epimer 7, acid catalyzed isopropylidene rearrangement of 14 to 13 was followed by a Mitsunobu C-4′ inversion to 15. Basic removal of the benzoate of 15 to 16 and subsequent acid deprotection yielded (4′S)-7.
Our attention turned to employing the Ullmann coupling of 11 and 3-deazaadenine. This succeeded in giving 14 (Scheme 2) in a moderate yield in contrast to 12, suggesting a hydroxyl substituent adjacent to the vinyl coupling site was necessary for the Ullmann to succeed. Acid deprotection of 14 availed the desired (4 R)-8. In addition to NMR data, the structure of 8 was confirmed by X-ray crystallography (CCDC 2018731), which served to confirm the regiochemistry of the cyclopentenyl and the 3-deaza base of 8 (Supplementary Meterials).
Our attention turned to employing the Ullmann coupling of 11 and 3-deazaadenine. This succeeded in giving 14 (Scheme 2) in a moderate yield in contrast to 12, suggesting a hydroxyl substituent adjacent to the vinyl coupling site was necessary for the Ullmann to succeed. Acid deprotection of 14 availed the desired (4′R)-8. In addition to NMR data, the structure of 8 was confirmed by X-ray crystallography (CCDC 2018731), which served to confirm the regiochemistry of the cyclopentenyl and the 3-deaza base of 8 (Supplementary Meterials).
To achieve epimer 7, acid catalyzed isopropylidene rearrangement of 14 to 13 was followed by a Mitsunobu C-4′ inversion to 15. Basic removal of the benzoate of 15 to 16 and subsequent acid deprotection yielded (4′S)-7.
To achieve epimer 7, acid catalyzed isopropylidene rearrangement of 14 to 13 was followed by a Mitsunobu C-4 inversion to 15. Basic removal of the benzoate of 15 to 16 and subsequent acid deprotection yielded (4 S)-7.

Discussion
Compounds 7 and 8 were subjected to antiviral assays [14]. Compound 8 displayed potent activity (EC 50 = 1.12 µM) against the JC virus, a polyomavirus. Compound 7 had much lower activity (EC 50 = 59.14 µM) against the JC virus. Both epimers showed no cytotoxicity (CC 50 > 150 µM) towards the host COS7 cell-line. There was no activity for either compound against human cytomegalovirus, adenovirus, vaccinia virus, Epstein-Barr virus and human norovirus. No cytotoxicity was found as a result of these assays.
Further studies will consider variations of 8 for improving its JC antiviral potential, correlating its enzymatic effects (for example, towards SAHase) with the parent 6, and its usefulness for developing novel C-4 hydroxyl-based analogues within the 3-deazaisoneplanocin series.