Exploration of Pyrido[3,4-d]pyrimidines as Antagonists of the Human Chemokine Receptor CXCR2

Upregulated CXCR2 signalling is found in numerous inflammatory, autoimmune and neurodegenerative diseases, as well as in cancer. Consequently, CXCR2 antagonism is a promising therapeutic strategy for treatment of these disorders. We previously identified, via scaffold hopping, a pyrido[3,4-d]pyrimidine analogue as a promising CXCR2 antagonist with an IC50 value of 0.11 µM in a kinetic fluorescence-based calcium mobilization assay. This study aims at exploring the structure–activity relationship (SAR) and improving the CXCR2 antagonistic potency of this pyrido[3,4-d]pyrimidine via systematic structural modifications of the substitution pattern. Almost all new analogues completely lacked the CXCR2 antagonism, the exception being a 6-furanyl-pyrido[3,4-d]pyrimidine analogue (compound 17b) that is endowed with similar antagonistic potency as the original hit.


Introduction
Upon detection of a pathogen or in response to trauma an acute inflammatory response is mounted by tissue-resident immune cells, such as macrophages, dendritic cells and mast cells. It involves the release of a number of pro-inflammatory chemokines such as CXCL1, CXCL2 and CXCL8, resulting in the chemo-attraction of circulating immune cells [1][2][3]. Neutrophils are among the first to be recruited to the site of inflammation, in order to clear the infection, to release cytotoxic agents and other effector proteins that attract a variety of immune cells [4]. Neutrophils are primarily recruited to the injury site via the CXCL8 and human chemokine receptor CXCR2 axis [5]. Hence, it is clear that both CXCL8 and CXCR2 are essential mediators of the immune defense. For example, it has been demonstrated that CXCL8 or CXCR2 knockout in zebrafish hinders wound healing and reduces neutrophil recruitment [6]. CXCR2 deletion in mice resulted in failure to clear bacteria from the bladder and kidneys due to reduced neutrophil infiltration [7]. CXCR2 upregulation has been linked to a myriad of inflammatory disorders (such as asthma, chronic obstructive pulmonary disease, cystic fibrosis), autoimmune diseases (e.g., rheumatoid arthritis and psoriasis), and even neurodegenerative diseases (such as multiple sclerosis and Alzheimer's disease) [8]. Hence, CXCR2 antagonism is a promising therapeutic strategy for treating these inflammatory illnesses.
Furthermore, elevated expression and signaling of CXCR2 is correlated with aggressive cancer phenotypes and a poor prognosis in various cancers, such as lung cancer, colorectal cancer, pancreatic cancer, prostate cancer, breast cancer, gastric cancer and squamous cell carcinoma [9]. In addition, the glutamate-leucine-arginine positive (ELR+) CXC chemokines, which all interact with CXCR2, directly function as growth factors for tumor cells and stimulate metastases. For example, in melanoma cells, CXCL1 and CXCL8 promote cell growth and metastases, and CXCR2 inhibition has been shown to counteract these for the combination of reparixin and paclitaxel in treatment of metastatic breast cancer. A phase III trial for ladarixin in the treatment of recent onset type 1 diabetes is ongoing. Based on the thiazolo [4,5-d]pyrimidine core of the known CXCR2 antagonist AZD-8309 [34,37,38], we previously reported on a scaffold hopping study in which the central core was replaced by various other pyrimidine fused bicyclic heterocycles, with the aim of identifying CXCR2 antagonists based on underexplored chemotypes ( Figure 2) [39]. The pyrido [3,4-d]pyrimidine analogue 2 emerged as a promising hit, displaying an IC50 Besides the urea and urea isosteres class, several derivatives with an aromatic ring core have been explored, such as SX-682 [33], AZD-8309 [34], AZD-5069 [35], reparixin and ladarixin [36], which all went into clinical trials ( Figure 1). Phase II clinical trials for a combination of SX-682 with nivolumab for the treatment of metastatic colorectal cancer, and a combination of SX-682 with pembrolizumab for metastatic melanoma are ongoing. AZD-8309 was evaluated in a phase I trial for rheumatoid arthritis, COPD and pancreatitis. AZD-8309 was the starting point for the discovery of AZD-5069 via a ring opening strategy [35], which has gone into several clinical trials for COPD, asthma and bronchiectasis treatment, which were all terminated. Additionally, different combination therapies including AZD-5069 have been explored. The combination with MEDI4736 was evaluated for treatment of metastatic pancreatic ductal adenocarcinoma, and with enzalutamide with metastatic castration resistant prostate cancer. Reparixin was evaluated in several phase III clinical trials for preventing graft rejection after organ transplantation, in a phase I trial for treating metastatic breast cancer, and a phase II trial for the combination of reparixin and paclitaxel in treatment of metastatic breast cancer. A phase III trial for ladarixin in the treatment of recent onset type 1 diabetes is ongoing.
In this study, a systematic exploration of the substitution pattern of the pyrido [3,4d]pyrimidine scaffold is pursued in an effort to explore the structure-activity relationship (SAR) of pyrido [3,4-d]pyrimidine 2 and to improve its CXCR2 antagonistic potency.
The Suzuki-coupling of 5-chloropyrido [3,4-d]pyrimidine 9 with phenylboronic acid gave no conversion when standard reaction conditions were applied, i.e., tetrakis(triphenylphosphine)palladium(0) as catalyst and using potassium carbonate as a base in a mixture of water and dioxane. When transitioning to a more catalytically active system [47], using tris(dibenzylideneacetone)dipalladium(0) (Pd 2 dba 3 ), dicyclohexyl-(2 ,4 ,6triisopropylbiphenyl)phosphine (XPhos) and K 2 CO 3 in nBuOH, exclusively hydrolysis of the (R)-alaninol moiety was observed, yielding the 4-oxo-pyrido [3,4-d]pyrimidine derivative 8. The poor outcome of these Suzuki reactions can be ascribed to different factors. Lewis-basic heterocycles or compounds having coordinating groups are known to behave poorly in palladium-catalyzed cross-coupling reactions [47]. Compound 9 has a Lewis-basic aminopyridine core and an (R)-alaninol coordinating group. In addition, since the coupling partner is a heteroaryl chloride, the rate of oxidative addition is low. Moreover, steric hindrance of the chlorine at position 6 of the pyrido [3,4-d]pyrimidine scaffold prevents an efficient Suzuki reaction.
via treatment of 11 with oxalyl chloride, followed by a Schotten-Bauman reaction with aqueous ammonia. This resulted in a concomitant transhalogenation of the 6-bromo group, yielding the 6-chloropyridine analogue 12a. Reduction of the nitro group using iron powder under acidic conditions with ultrasonication (Béchamp reaction circumstances) [48], gave the aniline 13a, which was cyclocondensed with carbon disulfide in the presence of DBU to afford the 6-chloro-2-thioxopyrido [3,4-d]pyrimidine-4-one 14a [46]. Subsequent benzylation gave compound 15a. Chlorination of the lactam moiety with thionyl chloride and a catalytic amount of DMF yielded the 4-chloro intermediate, which was not isolated, but directly subjected to an amination reaction, furnishing compound 16a in good yield. For derivatization at position 6 of the pyrido [3,4-d]pyrimidine scaffold via palladium-catalyzed cross-coupling reactions, the 6-bromo analogue is preferred. Therefore, several attempts to prevent this transhalogenation were made. Performing the Béchamp reduction of the nitro group of compound 11 prior to oxalyl chlorination should greatly reduce the chance of transhalogenation, since the aminopyridine is far less electrophilic than the corresponding nitropyridine [49]. However, purification of the 5bromo-3-aminopyridine-carboxamide proved unsuccessful because of its low solubility. Attempts to prepare the bromo-analogue of 12a via an intermediate acid bromide using phosphorus tribromide, followed by the dropwise addition of an aqueous ammonia solution while cooling to −78 °C, afforded only trace amounts of the desired primary amide. Alternatively, the carbonyl 1,1-carbonyldiimidazole (CDI) mediated coupling with ammonia (alleviating the need of a chlorine source) resulted in concomitant nucleophilic aromatic substitution of the bromide by imidazole [50], giving the 5-imidazolyl-2nitropyridine-4-carboxamide 12b in low yield. Catalytic hydrogenation of the nitro group of compound 12b furnished aminopyridine 13b. Cyclocondensation with carbon disulfide towards 6-imidazolyl-2-thioxopyrido [3,4-d]pyrimidin-4-one 14b was followed by benzylation yielding compound 15b. Next, the one-pot benzotriazol-1yloxytris(dimethylamino)phosphonium hexafluorophosphate (BOP)-mediated amination with (R)-alaninol gave the 4-amino-6-imidazolyl-pyrido [3,4-d]pyrimidine 16b in good yield [51]. In a final attempt to prepare the 6-bromo analogue, a BOP-mediated Scheme 2. Synthesis of 6-substituted pyrido [3,4-d]pyrimidines. Reagents and conditions. (a) K 2 Cr 2 O 7 , For derivatization at position 6 of the pyrido [3,4-d]pyrimidine scaffold via palladiumcatalyzed cross-coupling reactions, the 6-bromo analogue is preferred. Therefore, several attempts to prevent this transhalogenation were made. Performing the Béchamp reduction of the nitro group of compound 11 prior to oxalyl chlorination should greatly reduce the chance of transhalogenation, since the aminopyridine is far less electrophilic than the corresponding nitropyridine [49]. However, purification of the 5-bromo-3-aminopyridinecarboxamide proved unsuccessful because of its low solubility. Attempts to prepare the bromo-analogue of 12a via an intermediate acid bromide using phosphorus tribromide, followed by the dropwise addition of an aqueous ammonia solution while cooling to −78 • C, afforded only trace amounts of the desired primary amide. Alternatively, the carbonyl 1,1-carbonyldiimidazole (CDI) mediated coupling with ammonia (alleviating the need of a chlorine source) resulted in concomitant nucleophilic aromatic substitution of the bromide by imidazole [50], giving the 5-imidazolyl-2-nitropyridine-4-carboxamide 12b in low yield. Catalytic hydrogenation of the nitro group of compound 12b furnished aminopyridine 13b. Cyclocondensation with carbon disulfide towards 6-imidazolyl-2thioxopyrido [3,4-d]pyrimidin-4-one 14b was followed by benzylation yielding compound 15b. Next, the one-pot benzotriazol-1-yloxytris(dimethylamino)phosphonium hexafluorophosphate (BOP)-mediated amination with (R)-alaninol gave the 4-amino-6-imidazolylpyrido [3,4-d]pyrimidine 16b in good yield [51]. In a final attempt to prepare the 6-bromo analogue, a BOP-mediated amidation of the carboxylic acid 11 with ammonia was applied. Unfortunately, this gave a highly complex mixture, from which the desired compound could not be isolated.
Since various attempts to synthesize the 6-bromopyrido [3,4-d]pyrimidine analogue were unsuccessful, the derivatization of the 6-chloropyrido [3,4-d]pyrimidine 16a via palladiumcatalyzed cross-coupling was considered (Scheme 3). The Suzuki coupling with phenylboronic acid, 2-furanylboronic acid and 5-methylfuran-2-boronic acid under the standard conditions gave the desired compounds 17a-c in low to moderate yields. The low yields can partly be explained by challenging purification, since the reactions did not proceed to full conversion. However, for the reaction with 5-methylfuran-2-boronic acid, the low yield resulted from hydrolysis of the (R)-alaninol group towards 15a. For the Suzuki coupling with 2-thienylboronic acid, the more active catalyst system, Pd 2 dba 3 /Xphos [47], was used, giving the desired 6-thienylpyrido [3,4-d]pyrimidine 17d in fair yield.
amidation of the carboxylic acid 11 with ammonia was applied. Unfortunately, this gave a highly complex mixture, from which the desired compound could not be isolated.
Finally, the formal introduction of ammonia at position 6 of the pyrido [3,4d]pyrimidine scaffold was performed using benzophenone imine as ammonia surrogate (Scheme 4) [52]. The pyrimidin-4-one 15a was protected with p-methoxybenzyl chloride which proceeded quantitatively to compound 18. [53] Buchwald-Hartwig amination with benzophenone imine gave the secondary imine 19 in fair yield. Both the benzophenone and p-methoxybenzyl protecting moieties were simultaneously cleaved off by heating in neat trifluoroacetic acid (TFA), giving 6-aminopyrido [3,4-d]pyrimidine-4-one 20 which was isolated by simple filtration. Finally, compound 21 was obtained via a BOP-mediated amination with (R)-alaninol. Finally, the formal introduction of ammonia at position 6 of the pyrido [3,4-d]pyrimidine scaffold was performed using benzophenone imine as ammonia surrogate (Scheme 4) [52]. The pyrimidin-4-one 15a was protected with p-methoxybenzyl chloride which proceeded quantitatively to compound 18. [53] Buchwald-Hartwig amination with benzophenone imine gave the secondary imine 19 in fair yield. Both the benzophenone and p-methoxybenzyl protecting moieties were simultaneously cleaved off by heating in neat trifluoroacetic acid (TFA), giving 6-aminopyrido [3,4-d]pyrimidine-4-one 20 which was isolated by simple filtration. Finally, compound 21 was obtained via a BOP-mediated amination with (R)-alaninol. a highly complex mixture, from which the desired compound could not be isolated.

Biological Evaluation
All pyrido [3,4-d]pyrimidines were evaluated as potential CXCR2 antagonists in an in vitro cell-based calcium mobilization assay. Stimulating human glioblastoma U87 cells overexpressing CXCR2 with CXCL8 results in increased intracellular calcium levels, allowing for the determination of functional CXCR2 antagonism. Navarixin, a well-known CXCR2 antagonist, was included as positive control, giving an IC 50 value of 0.0049 µM, in agreement with literature values [55]. Since the original pyrido [3,4-d]pyrimidine lead compound 2 was derived from a thiazolo [4,5-d]pyrimidine scaffold, the 2-amino-thiazolo [4,5d]pyrimidine analogue, endowed with an IC 50 value of 0.079 µM, was also included for comparison purposes.
In a first round of SAR, the substitution pattern of the pyridine moiety of the pyrido [3,4d]pyrimidine scaffold was explored ( Table 1). The rationale of selecting halogens was driven by the fact that they can function as handle for subsequent introduction of structural diversity by palladium-catalyzed cross couplings or nucleophilic aromatic substitutions. The 5-chloro compound 9 was one hundred fold less potent as CXCR2 antagonist (IC 50 = 11 µM) when compared to the original hit 2. Synthesis problems (vide supra) prevented us from further elaboration of this particular position. The 6-chloro analogue 16a was completely devoid of CXCR2 antagonism. Further elaboration at the 6-position by the synthesis of a 6-N-imidazolyl, 6-phenyl-and 6-thienyl-pyrido [3,4-d]pyrimidine analogue (compounds 16b, 17a and 17d) did not have any beneficial effect on CXCR2 antagonism. Remarkably, the 6-furanyl congener 17b (IC 50 = 0.54 µM) showed comparable activity to its unsubstituted congener 2 (IC 50 = 0.11 µM). However, introduction of a methyl group on the furanyl group (compound 17c) led again to complete loss of activity. It is well known that an amino group on the thiazolo moiety of thiazolo [4,5-d]pyrimidine yields potent CXCR2 antagonists. [37] In contrast, within the pyrido [3,4-d]pyrimidine series, the presence of an amino group at position 6 (compound 21) led to a complete loss of CXCR2 antagonistic activity. This suggests that the structure-activity relationship (SAR) in the thiazolo [4,5-d]pyrimidines and pyrido [3,4-d]pyrimidines is different and does not run in parallel. Although compounds 2 and 17b are the most potent CXCR2 antagonists within this series, they are still clearly much less active than navarixin and the reference thiazolo [4,5-d]pyrimidine analogue. Table 1. CXCR2 antagonism of 5-and 6-substituted pyrido [3,4-d]pyrimidines.

Cmpd# IC 50 (µM) a
All pyrido [3,4-d]pyrimidines were evaluated as potential CXCR2 antagonists in an in vitro cell-based calcium mobilization assay. Stimulating human glioblastoma U87 cells overexpressing CXCR2 with CXCL8 results in increased intracellular calcium levels allowing for the determination of functional CXCR2 antagonism. Navarixin, a wellknown CXCR2 antagonist, was included as positive control, giving an IC50 value of 0.0049 µM, in agreement with literature values [55]. Since the original pyrido [3,4-d]pyrimidine lead compound 2 was derived from a thiazolo [4,5-d]pyrimidine scaffold, the 2-aminothiazolo [4,5-d]pyrimidine analogue, endowed with an IC50 value of 0.079 µM, was also included for comparison purposes.
In a first round of SAR, the substitution pattern of the pyridine moiety of the pyrido [3,4-d]pyrimidine scaffold was explored ( Table 1). The rationale of selecting halogens was driven by the fact that they can function as handle for subsequent introduction of structural diversity by palladium-catalyzed cross couplings or nucleophilic aromatic substitutions. The 5-chloro compound 9 was one hundred fold less potent as CXCR2 antagonist (IC50 = 11 µM) when compared to the original hit 2. Synthesis problems (vide supra) prevented us from further elaboration of this particular position The 6-chloro analogue 16a was completely devoid of CXCR2 antagonism. Further elaboration at the 6-position by the synthesis of a 6-N-imidazolyl, 6-phenyl-and 6-thienylpyrido [3,4-d]pyrimidine analogue (compounds 16b, 17a and 17d) did not have any beneficial effect on CXCR2 antagonism. Remarkably, the 6-furanyl congener 17b (IC50 = 0.54 µM) showed comparable activity to its unsubstituted congener 2 (IC50 = 0.11 µM) However, introduction of a methyl group on the furanyl group (compound 17c) led again to complete loss of activity. It is well known that an amino group on the thiazolo moiety of thiazolo [4,5-d]pyrimidine yields potent CXCR2 antagonists. [37] In contrast, within the pyrido [3,4-d]pyrimidine series, the presence of an amino group at position 6 (compound 21) led to a complete loss of CXCR2 antagonistic activity. This suggests that the structureactivity relationship (SAR) in the thiazolo [4,5-d]pyrimidines and pyrido [3,4d]pyrimidines is different and does not run in parallel. Although compounds 2 and 17b are the most potent CXCR2 antagonists within this series, they are still clearly much less active than navarixin and the reference thiazolo [4,5-d]pyrimidine analogue. Table 1. CXCR2 antagonism of 5-and 6-substituted pyrido [3,4-d]pyrimidines.

21)
led to a complete loss of CXCR2 antagonistic activity. This suggests that the structure-activity relationship (SAR) in the thiazolo [4,5-d]pyrimidines and pyrido [3,4d]pyrimidines is different and does not run in parallel. Although compounds 2 and 17b are the most potent CXCR2 antagonists within this series, they are still clearly much less active than navarixin and the reference thiazolo [4,5-d]pyrimidine analogue.

Cmpd# IC50 (µM) a
Navarixin see Figure 1 0.0049 ± 0.0033 Since the 5-and 6-substituted derivatives displayed strongly reduced activity compared to the unsubstituted pyrido [3,4-d]pyrimidine 2, structural modifications of the substituents at positions 2 and 4 were explored. Subtle structural variations of the (R)alaninol group of hit compound 2 (Table 2), such as elongation of the methyl to an ethy group (compound 24a), insertion of an additional methyl group yielding the gemdimethyl analogue 24b, introducing a hydroxyl functionality (compound 24c) or remova of the methyl group (compound 24d), afforded pyrido [4,3-d]pyrimidines that were completely devoid of CXCR2 antagonism, as evidenced by IC50 values exceeding 30 µM Similarly, an ethyleneglycol linker at position 4 gave compound 24e also lacking CXCR2 antagonism. This situation is different from the thiazolo [4,5-d]pyrimidine series, since for this latter scaffold, CXCR2 antagonism tolerates structural variation at this position [37]. 0.079 ± 0.058 2 0.11 ± 0.019 a IC50: the compound concentration that inhibits CXCL8 induced intracellular calcium flux by 50% Data are expressed as average ± SD (in µM) of at least three experiments.
Since the 5-and 6-substituted derivatives displayed strongly reduced activity compared to the unsubstituted pyrido [3,4-d]pyrimidine 2, structural modifications of the substituents at positions 2 and 4 were explored. Subtle structural variations of the (R)alaninol group of hit compound 2 (Table 2), such as elongation of the methyl to an ethy group (compound 24a), insertion of an additional methyl group yielding the gemdimethyl analogue 24b, introducing a hydroxyl functionality (compound 24c) or remova of the methyl group (compound 24d), afforded pyrido [4,3-d]pyrimidines that were completely devoid of CXCR2 antagonism, as evidenced by IC50 values exceeding 30 µM Similarly, an ethyleneglycol linker at position 4 gave compound 24e also lacking CXCR2 antagonism. This situation is different from the thiazolo [4,5-d]pyrimidine series, since for this latter scaffold, CXCR2 antagonism tolerates structural variation at this position [37]. 0.079 ± 0.058 2 0.11 ± 0.019 a IC50: the compound concentration that inhibits CXCL8 induced intracellular calcium flux by 50% Data are expressed as average ± SD (in µM) of at least three experiments.
Since the 5-and 6-substituted derivatives displayed strongly reduced activity compared to the unsubstituted pyrido [3,4-d]pyrimidine 2, structural modifications of the substituents at positions 2 and 4 were explored. Subtle structural variations of the (R)alaninol group of hit compound 2 (Table 2), such as elongation of the methyl to an ethyl group (compound 24a), insertion of an additional methyl group yielding the gemdimethyl analogue 24b, introducing a hydroxyl functionality (compound 24c) or removal of the methyl group (compound 24d), afforded pyrido [4,3-d]pyrimidines that were completely devoid of CXCR2 antagonism, as evidenced by IC50 values exceeding 30 µM Similarly, an ethyleneglycol linker at position 4 gave compound 24e also lacking CXCR2 antagonism. This situation is different from the thiazolo [4,5-d]pyrimidine series, since for this latter scaffold, CXCR2 antagonism tolerates structural variation at this position [37]. 0.079 ± 0.058 2 0.11 ± 0.019 a IC50: the compound concentration that inhibits CXCL8 induced intracellular calcium flux by 50% Data are expressed as average ± SD (in µM) of at least three experiments.
Since the 5-and 6-substituted derivatives displayed strongly reduced activity compared to the unsubstituted pyrido [3,4-d]pyrimidine 2, structural modifications of the substituents at positions 2 and 4 were explored. Subtle structural variations of the (R) alaninol group of hit compound 2 (Table 2), such as elongation of the methyl to an ethy group (compound 24a), insertion of an additional methyl group yielding the gem dimethyl analogue 24b, introducing a hydroxyl functionality (compound 24c) or remova of the methyl group (compound 24d), afforded pyrido [4,3-d]pyrimidines that were completely devoid of CXCR2 antagonism, as evidenced by IC50 values exceeding 30 µM Similarly, an ethyleneglycol linker at position 4 gave compound 24e also lacking CXCR2 antagonism. This situation is different from the thiazolo [4,5-d]pyrimidine series, since for this latter scaffold, CXCR2 antagonism tolerates structural variation at this position [37].
0.079 ± 0.058 2 0.11 ± 0.019 a IC50: the compound concentration that inhibits CXCL8 induced intracellular calcium flux by 50% Data are expressed as average ± SD (in µM) of at least three experiments.
Since the 5-and 6-substituted derivatives displayed strongly reduced activity compared to the unsubstituted pyrido [3,4-d]pyrimidine 2, structural modifications of the substituents at positions 2 and 4 were explored. Subtle structural variations of the (R)alaninol group of hit compound 2 (Table 2), such as elongation of the methyl to an ethy group (compound 24a), insertion of an additional methyl group yielding the gemdimethyl analogue 24b, introducing a hydroxyl functionality (compound 24c) or remova of the methyl group (compound 24d), afforded pyrido [4,3-d]pyrimidines that were completely devoid of CXCR2 antagonism, as evidenced by IC50 values exceeding 30 µM Similarly, an ethyleneglycol linker at position 4 gave compound 24e also lacking CXCR2 antagonism. This situation is different from the thiazolo [4,5-d]pyrimidine series, since for this latter scaffold, CXCR2 antagonism tolerates structural variation at this position [37]. 0.079 ± 0.058 2 0.11 ± 0.019 a IC50: the compound concentration that inhibits CXCL8 induced intracellular calcium flux by 50% Data are expressed as average ± SD (in µM) of at least three experiments.
Since the 5-and 6-substituted derivatives displayed strongly reduced activity compared to the unsubstituted pyrido [3,4-d]pyrimidine 2, structural modifications of the substituents at positions 2 and 4 were explored. Subtle structural variations of the (R) alaninol group of hit compound 2 (Table 2), such as elongation of the methyl to an ethy group (compound 24a), insertion of an additional methyl group yielding the gem dimethyl analogue 24b, introducing a hydroxyl functionality (compound 24c) or remova of the methyl group (compound 24d), afforded pyrido [4,3-d]pyrimidines that were completely devoid of CXCR2 antagonism, as evidenced by IC50 values exceeding 30 µM Similarly, an ethyleneglycol linker at position 4 gave compound 24e also lacking CXCR2 antagonism. This situation is different from the thiazolo [4,5-d]pyrimidine series, since for this latter scaffold, CXCR2 antagonism tolerates structural variation at this position [37]. 0.079 ± 0.058 2 0.11 ± 0.019 a IC50: the compound concentration that inhibits CXCL8 induced intracellular calcium flux by 50% Data are expressed as average ± SD (in µM) of at least three experiments.
Since the 5-and 6-substituted derivatives displayed strongly reduced activity compared to the unsubstituted pyrido [3,4-d]pyrimidine 2, structural modifications of the substituents at positions 2 and 4 were explored. Subtle structural variations of the (R)alaninol group of hit compound 2 (Table 2), such as elongation of the methyl to an ethy group (compound 24a), insertion of an additional methyl group yielding the gemdimethyl analogue 24b, introducing a hydroxyl functionality (compound 24c) or remova of the methyl group (compound 24d), afforded pyrido [4,3-d]pyrimidines that were completely devoid of CXCR2 antagonism, as evidenced by IC50 values exceeding 30 µM Similarly, an ethyleneglycol linker at position 4 gave compound 24e also lacking CXCR2 antagonism. This situation is different from the thiazolo [4,5-d]pyrimidine series, since for this latter scaffold, CXCR2 antagonism tolerates structural variation at this position [37]. Since the 5-and 6-substituted derivatives displayed strongly reduced activity compared to the unsubstituted pyrido [3,4-d]pyrimidine 2, structural modifications of the substituents at positions 2 and 4 were explored. Subtle structural variations of the (R)-alaninol group of hit compound 2 (Table 2), such as elongation of the methyl to an ethyl group (compound 24a), insertion of an additional methyl group yielding the gem-dimethyl analogue 24b, introducing a hydroxyl functionality (compound 24c) or removal of the methyl group (compound 24d), afforded pyrido [4,3-d]pyrimidines that were completely devoid of CXCR2 antagonism, as evidenced by IC 50 values exceeding 30 µM. Similarly, an ethyleneglycol linker at position 4 gave compound 24e also lacking CXCR2 antagonism. This situation is different from the thiazolo [4,5-d]pyrimidine series, since for this latter scaffold, CXCR2 antagonism tolerates structural variation at this position [37].
To probe the importance of the sulfur linker for CXCR2 antagonism, the corres ing amino and oxygen linkers were prepared, yielding compounds 26a and 26b, r tively, that were both devoid of CXCR2 antagonistic activity. Since a sulfur was fo be important for imparting CXCR2 antagonism, a limited number of thiols was app to position 2 of the pyrido[3,4-d]pyrimidine scaffold. Shortening (compound 26c) o gation (compound 26d) of the linker between the sulfur and the phenyl moiety fur analogues that were inactive as CXCR2 antagonists. Since it has been shown befo
To probe the importance of the sulfur linker for CXCR2 antagonism, the c ing amino and oxygen linkers were prepared, yielding compounds 26a and 2 tively, that were both devoid of CXCR2 antagonistic activity. Since a sulfur wa be important for imparting CXCR2 antagonism, a limited number of thiols was to position 2 of the pyrido [3,4-d]pyrimidine scaffold. Shortening (compound 2 gation (compound 26d) of the linker between the sulfur and the phenyl moiety analogues that were inactive as CXCR2 antagonists. Since it has been shown in the thiazolo [4,5-d]pyrimidine series, an aliphatic side chain is tolerated for tagonism, an n-hexyl linker was installed (compound 26e), which was inactive antagonist, with an IC50 value exceeding 30 µM (Table 3).  To probe the importance of the sulfur linker for CXCR2 antagonism, the c ing amino and oxygen linkers were prepared, yielding compounds 26a and 2 tively, that were both devoid of CXCR2 antagonistic activity. Since a sulfur w be important for imparting CXCR2 antagonism, a limited number of thiols was to position 2 of the pyrido [3,4-d]pyrimidine scaffold. Shortening (compound 2 gation (compound 26d) of the linker between the sulfur and the phenyl moiety analogues that were inactive as CXCR2 antagonists. Since it has been shown in the thiazolo [4,5-d]pyrimidine series, an aliphatic side chain is tolerated for tagonism, an n-hexyl linker was installed (compound 26e), which was inactive antagonist, with an IC50 value exceeding 30 µM (Table 3).  To probe the importance of the sulfur linker for CXCR2 antagonism, the co ing amino and oxygen linkers were prepared, yielding compounds 26a and 26 tively, that were both devoid of CXCR2 antagonistic activity. Since a sulfur wa be important for imparting CXCR2 antagonism, a limited number of thiols was to position 2 of the pyrido [3,4-d]pyrimidine scaffold. Shortening (compound 26 gation (compound 26d) of the linker between the sulfur and the phenyl moiety analogues that were inactive as CXCR2 antagonists. Since it has been shown b in the thiazolo [4,5-d]pyrimidine series, an aliphatic side chain is tolerated for C tagonism, an n-hexyl linker was installed (compound 26e), which was inactive antagonist, with an IC50 value exceeding 30 µM (Table 3).  To probe the importance of the sulfur linker for CXCR2 antagonism, the co ing amino and oxygen linkers were prepared, yielding compounds 26a and 26 tively, that were both devoid of CXCR2 antagonistic activity. Since a sulfur wa be important for imparting CXCR2 antagonism, a limited number of thiols was to position 2 of the pyrido [3,4-d]pyrimidine scaffold. Shortening (compound 26 gation (compound 26d) of the linker between the sulfur and the phenyl moiety analogues that were inactive as CXCR2 antagonists. Since it has been shown b in the thiazolo [4,5-d]pyrimidine series, an aliphatic side chain is tolerated for C tagonism, an n-hexyl linker was installed (compound 26e), which was inactive antagonist, with an IC50 value exceeding 30 µM (Table 3).  To probe the importance of the sulfur linker for CXCR2 antagonism, the co ing amino and oxygen linkers were prepared, yielding compounds 26a and 26 tively, that were both devoid of CXCR2 antagonistic activity. Since a sulfur wa be important for imparting CXCR2 antagonism, a limited number of thiols was to position 2 of the pyrido [3,4-d]pyrimidine scaffold. Shortening (compound 26 gation (compound 26d) of the linker between the sulfur and the phenyl moiety analogues that were inactive as CXCR2 antagonists. Since it has been shown b in the thiazolo [4,5-d]pyrimidine series, an aliphatic side chain is tolerated for C tagonism, an n-hexyl linker was installed (compound 26e), which was inactive antagonist, with an IC50 value exceeding 30 µM (Table 3).  To probe the importance of the sulfur linker for CXCR2 antagonism, the co ing amino and oxygen linkers were prepared, yielding compounds 26a and 26 tively, that were both devoid of CXCR2 antagonistic activity. Since a sulfur wa be important for imparting CXCR2 antagonism, a limited number of thiols was to position 2 of the pyrido [3,4-d]pyrimidine scaffold. Shortening (compound 26 gation (compound 26d) of the linker between the sulfur and the phenyl moiety analogues that were inactive as CXCR2 antagonists. Since it has been shown b in the thiazolo [4,5-d]pyrimidine series, an aliphatic side chain is tolerated for C tagonism, an n-hexyl linker was installed (compound 26e), which was inactive antagonist, with an IC50 value exceeding 30 µM (Table 3). To probe the importance of the sulfur linker for CXCR2 antagonism, the corresponding amino and oxygen linkers were prepared, yielding compounds 26a and 26b, respectively, that were both devoid of CXCR2 antagonistic activity. Since a sulfur was found to be important for imparting CXCR2 antagonism, a limited number of thiols was appended to position 2 of the pyrido [3,4-d]pyrimidine scaffold. Shortening (compound 26c) or elongation (compound 26d) of the linker between the sulfur and the phenyl moiety furnished analogues that were inactive as CXCR2 antagonists. Since it has been shown before that in the thiazolo [4,5-d]pyrimidine series, an aliphatic side chain is tolerated for CXCR2 antagonism, an n-hexyl linker was installed (compound 26e), which was inactive as CXCR2 antagonist, with an IC 50 value exceeding 30 µM (Table 3).

Chemistry
All chemicals were purchased from Acros Organics (Geel, Belgium), Merck (Darmstadt, Germany), Alfa Aesar (Kandel, Germany), Fluorochem (Hadfield, UK) and TCI Europe (Zwijndrecht, Belgium) and used as received. Moisture sensitive reactions were carried out under nitrogen or argon-atmosphere, using flame dried glassware. Reactions were stirred magnetically. Reaction conversion was monitored via TLC analysis using MilliporeSigma™ Silica Gel 60 F254 Coated Aluminum-Backed TLC Sheets or Macherey-Nagel SILPre-coated ALUGRAM ® Xtra SIL G/UV254 TLC sheets. Compounds were visualized via UV irradiation (254 nm), iodine coated silica or KMnO4-staining. Column chromatography was performed via standard column chromatography or using an MPLC apparatus. For column chromatography, 70-230 mesh silica 60 (Acros, Geel, Belgium) was used as the stationary phase. MPLC was performed using a CombiFlash EZ prep appa-

2
Molecules 2023, 28, x FOR PEER REVIEW 1 Navarixin See Figure 1 0.0049 ± 0.0033 26e >30 a IC50: the compound concentration that inhibits CXCL8 induced intracellular calcium flux b Data are expressed as average ± SD (in µM) of at least three experiments.

Chemistry
All chemicals were purchased from Acros Organics (Geel, Belgium), Merck (Darmstadt, Germany), Alfa Aesar (Kandel, Germany), Fluorochem (Hadfield, UK) and TCI Europe (Zwijndrecht, Belgium) and used as received. Moisture sensitive reactions were carried out under nitrogen or argon-atmosphere, using flame dried glassware. Reactions were stirred magnetically. Reaction conversion was monitored via TLC analysis using Milli-poreSigma™ Silica Gel 60 F254 Coated Aluminum-Backed TLC Sheets or Macherey-Nagel SILPre-coated ALUGRAM ® Xtra SIL G/UV 254 TLC sheets. Compounds were visualized via UV irradiation (254 nm), iodine coated silica or KMnO 4 -staining. Column chromatography was performed via standard column chromatography or using an MPLC apparatus. For column chromatography, 70-230 mesh silica 60 (Acros, Geel, Belgium) was used as the stationary phase. MPLC was performed using a CombiFlash EZ prep apparatus with BGB Scorpius Silica 60Å Irregular-50 µm cartridges. 1 H NMR spectra were recorded on a Bruker Avance 300 (300 MHz working frequency) or Bruker Avance 400 (400 MHz working frequency). Proton-decoupled 13 C NMR spectra were recorded on a Bruker Avance 400 (101 MHz working frequency). 19 F NMR spectra were recorded on a Bruker Avance 400 (377 MHz working frequency). Samples were dissolved in d 6 -DMSO or CDCl 3 , and chemical shifts (δ) were reported in parts per million (ppm) referenced to tetramethylsilane ( 1 H), or the internal (NMR) solvent signal ( 1 H and 13 C) as internal standards [56]. All NMR spectra are presented in Supplementary Materials. High-resolution mass spectra were acquired on a quadrupole orthogonal acceleration time-of-flight mass spectrometer (Synapt G2 HDMS, Waters, Milford, MA, USA). Samples were infused at 3 µL/min and spectra were obtained in positive ionization mode with a resolution of 15,000 (FWHM-full width at half maximum) using leucine enkephalin as a lock mass. Melting points were determined using a Reichert Thermovar apparatus and are uncorrected.

Biology
Recombinant human CXCL8 was obtained from Peprotech (Rocky Hill, NJ, USA).

3-Amino-5-chloropyridine-4-carbonitrile (5)
A mixture of 3,5-dichloropyridine-4-carbonitrile 3 (13.76 g, 80.0 mmol, 1.00 eq.) and NaN 3 (5.72 g, 8.80 mmol, 1.10 eq.) in DMF (100 mL) was heated to 80 • C for 19 h. After cooling to room temperature, the reaction mixture was transferred to a separatory funnel together with H 2 O (200 mL) and EtOAc (200 mL). After phase separation the aqueous phase was extracted with EtOAc (2 × 200 mL). The combined organic phases were washed with water (2 × 200 mL) and brine (2 × 200 mL), dried over Na 2 SO 4 and coated on Celite. Filtration over silica using 30% EtOAc/PE gave the crude intermediated product 4a, which was used without further purification in the next step. The obtained solid was dissolved ACN (600 mL) and NaI (107 g, 720 mmol, 9.00 eq.) was added. The suspension was cooled in an ice bath and FeCl 3 (19.4 g, 120 mmol, 1.50 eq.) was added. After stirring for 10 min at 0 • C, the mixture was stirred for 30 min at room temperature, when full conversion was observed by TLC. The reaction mixture was transferred to a separatory funnel along with EtOAc (200 mL) and water (100 mL) and quenched with Na 2 S 2 O 3 . After phase separation, the aqueous phase was extracted with EtOAc (2 × 200 mL). The combined organic phases were washed with H 2 O (2 × 100 mL) and brine (2 × 100 mL), dried over Na 2 SO 4 and filtered over silica. The obtained brown solid was recrystallized twice from methanol to afford the title compound (11.041 g, 71.89 mmol, 90%) as an off-white solid.

2-Bromo-5-nitropyridine-4-carboxylic acid (11)
2-Bromo-4-methyl-5-nitropyridine 10 (25.0 g, 115 mmol, 1.00 eq.) was dissolved in concentrated H 2 SO 4 (100 mL). The mixture was cooled in an ice bath, and K 2 Cr 2 O 7 (50.83 g, 173 mmol, 1.5 eq.) was added portion-wise. Following complete addition, the mixture was stirred in an ice bath for 1 h and at room temperature for 24 h. The resulting green solution was slowly poured onto 1L of ice. The obtained precipitate was filtered, washed with water until the eluent was no longer green, and dried under vacuum to afford the title compound as white solid (22.

2-Chloro-5-nitropyridine-4-carboxamide (12a)
Oxalyl chloride (20 mL) was added to an ice-cold solution of 2-bromo-5-nitropyridine-4-carboxylic acid 11 (10.00 g, 40.49 mmol) in dry DCM (200 mL) at 0 • C and under N 2atmosphere. After stirring for 5 min, dry DMF (0.2 mL) was added. The reaction mixture was stirred at 0 • C for 1 h and at room temperature overnight. The resulting solution was concentrated in vacuo. The obtained residue was dissolved in dry DCM (50 mL) and added dropwise to a heavily stirred, ice cold 25% aqueous ammonia solution (300 mL). After stirring for 15 min, the formed precipitate was filtered, washed with saturated aqueous NaHCO 3 and water, and dried under vacuum overnight to afford the title compound (7.498 g, 34.71 mmol, 86%) as a white solid that was used without further purification.

2,4-Dichloro-pyrido[3,4-d]pyrimidine (22)
This compound was prepared following a reported procedure [54]. A mixture of 3-aminoisonicotinic acid (10.00 g, 72.4 mmol, 1.00 eq.) and urea (21.74 g, 181.0 mmol) in a screw-capped sealed 80 mL reaction tube was heated to 210 • C for 1 h. After cooling to room temperature, a solution of NaOH (14.2 g, 360 mmol) in 60 mL water was added and the mixture was heated to 100 • C for 1 h. Next, the reaction mixture was cooled down to room temperature, the solid was filtered and washed with water. The obtained crude product was suspended in AcOH (60 mL) and heated to 100 • C for 1 h. After cooling to room temperature, the reaction mixture was filtered and the obtained solid was washed with a large amount of water, and a small amount of methanol and diethyl ether. After drying in vacuo (40 • C) pyrido [3,4-d]pyrimidine-2,4-dione (9.948 g, 61.0 mmol, 84%) was obtained as a white solid.
Pyrido [3,4-d]pyrimidine-2,4-dione was first finely ground with a pestle and mortar. To a flame dried two-necked 500 mL round bottom flask, equipped with an Ar-balloon, septum and stir bar, pyrido [3,4-d]pyrimidine-2,4-dione (9.800 g, 60.07 mmol, 1.00 eq.), dry toluene (100 mL) and POCl 3 (55 mL, 92 mmol, 10 eq.) were added. While cooling to 0 • C in an ice bath, DIPEA (21.5 mL, 15.9 mmol, 123.1 eq.) was added dropwise. After stirring at 0 • C for 30 min, the reaction mixture was stirred at room temperature for 24 h and concentrated in vacuo. To the obtained oil, diethyl ether and ice water were added. The mixture was neutralized with saturated aqueous NaHCO 3 (pH = 7) while cooling in an ice bath. Due to extensive tar formation, phase separation of the aqueous and organic layers did not occur. This was solved by pouring the mixture into a large beaker (2L) and adding a lot of paper towel to adsorb the tar. This mixture was extracted 5 times with 300 mL diethyl ether, by manually stirring the contents of the beaker with a large spatula and decanting off the organic phase. The combined organic phases were washed with water (2 × 50 mL) and brine (2 × 50 mL), decolorized with activated charcoal and dried over Na 2 SO 4 . After concentrating to dryness in vacuo, while keeping the heating bath at room temperature, and drying in vacuo (30 • C), the title compound (6.544 g, 32.72 mmol, 55%) was obtained as a beige-brown solid which was used without further purification.

General procedure A: Nucleophilic aromatic substitution of 2,4-dichloro-pyrido[3,4-d] pyrimidine with amines
To a flame dried, N 2 -flushed 8 mL reaction tube equipped with a stirring bar 2,4dichloro-pyrido [3,4-d]pyrimidine 22 (200 mg, 1.00 mmol, 1.00 eq.) and dry ACN (4.0 mL) were added. While cooling to 0 • C in an ice bath, Et 3 N (0.21 mL, 1.5 mmol, 1.5 eq.) and an amine nucleophile (1.5 mmol, 1.5 eq.) were added dropwise. After stirring for 10 min at 0 • C, the reaction mixture was stirred at room temperature for the specified time. During and after the addition of Et 3 N and nucleophile, generally, the formation of a yellow precipitate occurred. Next, the reaction mixture was transferred to a 100 mL round bottom flask with methanol; the entire sample was dissolved under gentle heating, and coated on Celite. Flash chromatography using 0-20% MeOH/DCM afforded the title compounds as white or beige solids.

General procedure B: Nucleophilic aromatic substitution of 2,4-dichloro-pyrido[3,4-d] pyrimidine with alcohols
NaH (60% dispersion) was added to a solution of alcohol nucleophile (1.1 eq.) in dry acetonitrile (2.0 mL), in a flame dried N 2 -flushed 8 mL screw capped reaction tube, and the mixture was sonicated for 30 min. In a separate flame dried, N 2 -flushed 8 mL reaction tube equipped with a stir bar, 2,4-dichloro-pyrido [3,4-d]pyrimidine 22 (200 mg, 1.00 mmol, 1.00 eq.) and dry ACN (3.0 mL) were added. While cooling to 0 • C in an ice bath, the solution of alcohol sodium salt was added dropwise to the 2,4-dichloro-pyrido [3,4d]pyrimidine solution. After stirring for 10 min at 0 • C, the reaction mixture was stirred at room temperature for the specified time. Next, the reaction mixture was transferred to a 100 mL round bottom flask with methanol; the entire sample was dissolved under gentle heating, and coated on Celite. Flash chromatography using 0-100% EtOAc/DCM, followed by trituration with pentane and drying under vacuum (40 • C), afforded the title compounds as white or beige solids.

General procedure C: Nucleophilic aromatic substitution of 2-chloropyrido[3,4-d] pyrimidines with thiols
To a flame dried, N 2 -flused 8 mL reaction tube equipped with a stir bar, the appropriate 2-chloropyrido [3,4-d]pyrimidine (0.25 mmol, 1.00 eq.), thiol nucleophile (0.50 mmol, 2.00 eq.) and dry 1,4-dioxane (1 mL) were combined. NaH 60% dispersion in mineral oil (20 mg, 0.50 mmol, 0.50 eq.) was added in one portion. After stirring at room temperature overnight, methanol (1.0 mL) was added and the mixture was heated to 60 • C for 30 min. After cooling to room temperature, the reaction mixture was transferred to a 100 mL round bottom flask with methanol; the entire sample was dissolved under gentle heating, and coated on Celite. Flash chromatography using 0-20 MeOH/DCM, followed by trituration with pentane and drying under vacuum (40 • C), afforded the title compounds as white solids.
The following compounds were made according to these procedures:

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.