Generation of BT-Amide, a Bone-Targeted Pyk2 Inhibitor, Effective via Oral Administration, for the Prevention of Glucocorticoid-Induced Bone Loss

Abstract

Glucocorticoid-induced osteoporosis (GIOP) is the leading cause of iatrogenic osteoporosis due to the widespread clinical use of glucocorticoids (GC) as immunosuppressants. Previous research identified the proline-rich tyrosine kinase 2, Pyk2, as a critical mediator of GC-induced bone loss, and that blocking Pyk2 could protect the skeleton from adverse GC actions. However, systemic administration of current Pyk2 inhibitors causes harmful side effects, such as skin lesions. To address this, we developed bone-targeted (BT) Pyk2 inhibitors by conjugating them with bisphosphonates (BP), ensuring adherence to the bone matrix and reducing impact on noncalcified tissues. We synthesized BT-Amide by linking a derivative of TAE-226, a Pyk2 inhibitor, with alendronic acid. Oral administration (gavage) of BT-Amide prevented GC-induced bone loss in mice without causing skin lesions, or elevation of any organ toxicity markers. These findings introduce BT-Amide as the first orally effective bone-targeted Pyk2 inhibitor for preventing GC-induced bone loss while minimizing off-target effects.

Graphical Abstract

1. INTRODUCTION

Glucocorticoids (GC) are widely prescribed in the treatment of multiple conditions including rheumatoid arthritis, asthma, inflammatory bowel disease, chronic lung, and liver and skin diseases, as well as for the management of organ transplantation and as components of chemotherapy regimens for cancers. However, one of the most severe adverse side-effects of GC is osteoporosis.^1–3 Notably, iatrogenic GC-induced osteoporosis is the third leading cause overall of osteoporosis in the United States, and the leading iatrogenic inducer of osteoporosis.^4,5

Bone remodeling entails the resorption of old or damaged bone, followed by the deposition of new bone material.⁶ Osteoblasts contribute to bone formation, while osteoclasts cause bone resorption. The cellular process of remodeling begins when osteoclast precursor cells fuse to form a multinucleated osteoclastic cell. In the resorption phase, osteoclasts secrete lysosomal enzymes and hydrogen ions to break down the bone matrix, which comprises an inorganic portion of calcium phosphate crystals (hydroxyapatite) and an organic portion composed of collagen, proteoglycans, and glycoproteins. Upon completion of bone resorption, osteoclasts undergo apoptosis. During the reversal phase, mononuclear cells continue to degrade and deposit organic material while releasing growth factors to initiate the bone deposition phase. In the formation phase, osteoblasts fill the gap by depositing new collagen and minerals.⁷ After bone formation, osteoblasts have three possible fates: undergo apoptosis or transform into either bone-lining cells or osteocytes.⁸

The initial rapid bone loss induced by GC is primarily attributed to exaggerated bone resorption, accompanied by a suppression of bone formation.^9,10 This exaggerated bone resorption is a result of both the prolonged lifespan of preexisting osteoclasts and enhanced osteoclastogenesis (Figure 1).¹¹ GC suppresses bone formation through several mechanisms such as increasing apoptosis of osteoblasts and osteocytes, inhibiting osteoblast differentiation, and reducing the synthetic capacity of osteoblasts.^12–14 However, in contrast to the early, acute phase characterized by increased osteoclast activity and decreased osteoblast activity, the late effects of GC result in reduced numbers of both osteoclasts and osteoblasts. In the long term, GC predominantly inhibits the generation of osteoclasts. This inhibition results from the apoptotic loss of the cells expressing receptor activator of nuclear factor kappa β (NFkB) ligand (RANKL)—a critical ligand needed for binding to the precursor RANK receptor to initiate osteoclastic fusion and differentiation.¹⁵ Consequently, in the later stages, GC suppresses both bone resorption and formation.

Figure 1.

Effects of glucocorticoids on bone cells, as adapted from Sato et al.¹ Glucocorticoids (GCs) induce the apoptosis of osteoblasts and osteocytes, decreases the synthetic capacity of osteoblasts, and prolongs the survival of osteoclasts. GCs also initially increase osteoclastogenesis and with long-term administration led to a low bone turnover state with decreased osteoclastogenesis due to the loss of RANKL expressing cellular populations.

Two antiresorptive/antiremodeling bone therapeutic strategies are the usage of bisphosphonate (BP) agents, which are the most commonly utilized,^16,17 and the usage of anti-RANKL antibody, which has been recently FDA-approved for treatment of glucocorticoid-induced osteoporosis. BPs alter the material properties of bones through three effects: increasing the degree of mineralization, the accumulation of microdamage, and increasing collagen cross-linking.¹⁸ In addition, BPs also affect the normal cellular activities of all the bone cells, especially osteoclasts. First, after entering the cell, BPs can be incorporated into adenosine 5′-O-(2,3-methylenetriphoshate) (AppCH₂p) by class II aminoacyl RNA synthetase. AppCH₂p is a nonhydrolyzable ATP analogue, accumulation of which disrupts normal cellular function and may ultimately cause the death of osteoclasts.^19,20 Also, the second generation of BP agents are distinct from the first generation by bearing nitrogen atoms (N-BPs) that block the mevalonate pathway as an additional cellular target.^21–23 Alternative antiresorptive/antiremodeling strategies involve targeting RANKL, which is a critical osteoclast differentiation factor secreted by osteoblasts, osteocytes, and immune cells.²⁴ Anti-RANKL antibody was recently approved by the USA FDA in 2018 to specifically treat GC-induced osteoporosis.²⁵ Although BPs and anti-RANKL antibody effectively prevent bone loss by inhibiting resorption, these agents are unable to reverse GC-induced suppression of bone formation.^26–28 However, the long-term reduction in bone resorption by these agents reduces bone turnover, which in turn leads to accumulation of microdamage, increased brittleness, reduced toughness, and increased risk of low-energy atypical fractures.²⁹ Therefore, there is a need for novel therapeutic strategies that prevent GC-induced bone loss without these associated long-term skeletal complications.

The survival of bone cells is critically dependent on their attachment and signaling to the extracellular matrix through integrin. Disruption of this interaction causes detachment and apoptosis (termed anoikis) of osteoblasts, osteocytes, and osteoclasts.^30,31 Specifically, GC activates cellular signaling of proline-rich tyrosine kinase 2 (Pyk2, also known as focal adhesion kinases 2 or FAK2), which mediates cellular detachment and apoptosis.^32,33 Our previous studies identified Pyk2 as an essential mediator of anoikis regulated by GC in vivo in bone cells of both lineages: osteoclasts and osteocytes/osteoblasts.³⁴ Inhibition of Pyk2, either by genetic or pharmacological means, prevents GC-induced bone loss by stimulating anoikis of osteoclasts while preventing anoikis of osteoblasts and osteocytes. Moreover, the bone biomechanical material properties are preserved.³⁴ However, systemic administration of Pyk2 inhibitors in mice induces adverse effects in other tissues, such as skin lesions/impaired wound healing, as reported in other studies, which limits their future application for the prevention of GC-induced bone loss.³⁵

Here, we report the novel generation and discovery of (4-(6-((4-((5-chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)amino)-6-oxohexanamido)-1-hydroxybutane-1,1-diyl)bis(phosphonic acid), (BT-Amide) as the first bone-targeting Pyk2 inhibitor effective for the prevention of GC-induced bone loss in vivo, while avoiding adverse side effects in other tissues.

Furthermore, the design and synthesis of BT-Amide utilized a strategy to link Pyk2 inhibitors with BPs to enhance bone delivery. The delivery of other drug classes specifically to bone by conjugating them with BPs has been widely studied, with numerous studies reported.^36–42 This is determined by the strong binding affinity of BPs to bone tissue, primarily attributed to the di- and tridentate interactions between the phosphonate groups to hydroxyapatite (HA).^43,44 However, it is important to note that, to the best of our knowledge, no agent-BP conjugate has been reported to be orally active, and no such conjugate has received clinical approval. In this study, we present evidence that BT-Amide is orally active and effective for bone protection in a GC-excess mouse model. This finding represents the first report of an orally active conjugate of BP with another drug class. Our report may also serve as an example to inspire the future designs and developments of next-generation agent-BP conjugates for oral activity.

2. RESULTS AND DISCUSSION

2.1. Design.

2.1.1. Modification of Pyk2 inhibitor.

The overall strategy to design bone-targeted Pyk2 inhibitor is by conjugating a Pyk2 inhibitor with a bone-targeting moiety with high affinity for bone tissue, while maintaining the anti-Pyk2 activity of the whole structure. TAE-226 was identified in 2007 as a focal adhesion kinase (FAK) inhibitor, also possessing potent Pyk2 inhibitory effects.^45,46 TAE-226 is composed of five major components: the pyrimidine core, the aniline ring, the methyl carbamoyl groups, the methoxy aniline ring, and the morpholine group (Figure 2, Panel A).⁴⁷ We used molecular docking to predict the binding mode of TAE-226 and Pyk2 kinase. As predicted, the pyrimidine fragment projects inside the Pyk2 ATP-binding pocket and binds to the hinge domain, while the morpholine fragment is directing outside of the pocket. The binding mode was mainly stabilized by the hydrogen bond between TAE-226, and K457 and Y505; and by aromatic-H between TAE-226 and E503, respectively. (Figure 2, Panel B). The binding conformation was also supported by the crystal structure of FAK kinase in complex with TAE-226 (PDB: 2JKK), because of the high homogeneity between Pyk2 and FAK, the conformation in which adopts in a very similar way.⁴⁸ In both structures, the morpholine group does not seem to have any significant interactions with the kinase, making itself a proper position to structurally modify within. Modification of TAE-226 led to two derivatives, compound T1 with an exposed phenyl-amine, and compound T2 with an exposed hydroxy group (Figure 3). The purpose of the structural modification is to create exposed groups for linkage binding.

Figure 2.

Structure of TAE-226 (A) and its predicted binding with Pyk2 kinase (DFG-in, PDB code: 5TOB) (B).

Figure 3.

modification of TAE-226 and design of BT-Pyk2 inhibitors.

2.1.2. Linkage Type and Linker.

Two linkage types were designed between the linker and TAE-226, amide and ester. As for amide linkage, the morpholine group was modified into a free amine directly for the peptide bond reaction. The strategy for ester linkage is for morpholine to be modified into 4-hydroxypiperidine, generating a free hydroxy group to esterify. Here, we chose adipic acid as the linker to connect the TAE-226 derivative and the bone affinitive moiety.

2.1.3. Selection of Bone Affinitive Moiety.

Concerning the bone-targeting moiety, alendronic acid was chosen. Its bisphosphonate group has high affinity to bones.¹⁷ Here, in our study, two bone-targeting conjugates—BT-Amide and BT-Ester, were designed by attaching alendronic acid to the modified Pyk2 inhibitors (Figure 3).

2.2. Chemistry.

2.2.1. TAE-226.

The synthetic route of TAE-226 is depicted in Scheme 1. 2,4,5-trichloropyrimidin (1) was subjected to a coupling reaction with 2-amino-N-methylbenzamide and another coupling reaction with 2-methoxy-4-morpholinoaniline, leading to the obtain of TAE-226.

Scheme 1. Synthesis of TAE-226^a.

^aReagents and conditions: (a) 2-amino-N-methylbenzamide, DIPEA, isopropanol, reflux, 18h, 92%; (b) 2-methoxy-4-morpholinoaniline, p-toluenesulfonic acid, isopropanol, reflux, overnight, 54%.

2.2.2. BT-Amide.

The synthesis of BT-Amide is depicted in Scheme 2. The amine group of 3-methoxy-4-nitroaniline (3) was first protected by Boc-, forming tert-butyl (3-methoxy-4-nitrophenyl)carbamate (4). Then, the nitro group of 4 was reduced by H₂ to gain tert-butyl (4-amino-3-methoxyphenyl)carbamate (5). The coupling of 5 and 2-((2,5-dichloropyrimidin-4-yl)amino)-N-methylbenzamide yielded tert-butyl (4-((5-chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)carbamate (6). Then the deprotection of the Boc-amine in 6 yielded 2-((2-((4-amino-2-methoxyphenyl)amino)-5-chloropyrimidin-4-yl)amino)-N-methylbenzamide, T1. The following amide condensation reaction, catalyzed by HATU, generated 6-((4-((5-chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)amino)-6-oxohexanoic acid (7). The acid group generated was then activated by NHS to transform into 2,5-dioxopyrrolidin-1-yl 6-((4-((5-chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)amino)-6-oxohexanoate (8). Finally, compound 8 reacted with alendronic acid, generating the final product (4-(6-((4-((5-chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)amino)-6-oxohexanamido)-1-hydroxybutane-1,1-diyl)bis(phosphonic acid) (BT-Amide).

Scheme 2. Synthesis of T1 and BT-Amide^a.

^aReagents and conditions: (a) Boc₂O, DMAP, Et₃N, DCM, overnight, 61%; (b) H₂, Pd/C, MeOH, 5 h, 106%; (c) 2, CH₃COOH, dioxane, 110 °C, sealed bottle, 18 h, 30%; (d) CF₃COOH, DCM, 3 h, 94%; (e) adipic acid, HATU, DIPEA, DMF, 3 h, 74%; (f) NHS, EDCI, DMF, 4h, 28%; (g) alendronic acid, Et₃N, DMF, H₂O, 47%.

2.2.3. BT-Ester.

Similarly, the synthesis of BT-Ester is depicted in Scheme 3. 4-fluoro-2-methoxy-1-nitrobenzene (9) was first coupled with piperidin-4-ol to gain 1-(3-methoxy-4-nitrophenyl)piperidin-4-ol (10). Then, 10 was reduced and further coupled with 2-((2,5-dichloropyrimidin-4-yl)amino)-N-methylbenzamide to gain 2-((5-chloro-2-((4-(4-hydroxypiperidin-1-yl)-2-methoxyphenyl)amino)pyrimidin-4-yl)amino)-N-methylbenzamide (T2). Catalyzed by DMAP, the hydroxyl group of T2 then reacted with adipic acid anhydride to form the ester, yielding 6-((1-(4-((5-chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)piperidin-4-yl)oxy)-6-oxohexanoic acid (12). The acid group at the other side of the linker, generated by the reaction, was then activated by NHS and reacted with alendronic acid to generate the final product, (4-(6-((1-(4-((5-chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)piperidin-4-yl)oxy)-6-oxohexanamido)-1-hydroxybutane-1,1-diyl)bis(phosphonic acid) (BT-Ester).

Scheme 3. Synthesis of T2 and BT-Ester^a.

^aReagents and conditions: (a) piperidin-4-ol, K₂CO₃, DMSO, overnight, 107%; (b) H₂, Pd/C, MeOH, 5 h, 99%; (c) 2, p-toluenesulfonic acid, isopropanol, reflux, overnight, 67%; (d) oxepane-2,7-dione, DMAP, Et₃N, DCM, overnight, 70%; (e) NHS, EDCI, DMF, 4 h, 37%; (f) alendronic acid, Et₃N, DMF, H₂O, 47%.

2.3. In Vitro Kinase Inhibitory Assay.

We conducted tests to evaluate the Pyk2 kinase inhibitory effects of the synthesized compounds (Table 1). Interestingly, following structural modifications of TAE-226, both T1 and T2 exhibited a significant decrease in kinase inhibitory potency, with the IC₅₀s from 16.95 ± 3.62 to 43.50 ± 12.39 and 101.4 ± 31.0 nM, respectively. This suggests that the morpholine fragment of TAE-226 contributes to stabilizing the ligand binding, despite not appearing to have crucial interactions with Pyk2 kinase based on molecular modeling results. Furthermore, the results of the kinase inhibitory assay indicated that BT-Amide and T1 share very similar IC₅₀ values (44.69 ± 18.86, and 43.50 ± 12.39 nM) for Pyk2, and BT-Ester also exhibits IC₅₀ values similar to those of T2 (118.6 ± 41.0, and 101.4 ± 31.0 nM). This comparison allows us to conclude that the attachment of the linkage and the alendronic acid did not significantly reduce the Pyk2 inhibitory effects of T1 and T2.

Table 1.

Pyk2 Kinase Inhibitory Effects (IC₅₀s) of Compounds of Interest

name	Pyk2 Kinase IC50 (nM)
TAE-226	16.95 ± 3.62
T1	43.50 ± 12.39
BT-Amide	44.69 ± 18.86
T2	101.4 ± 31.0
BT-Ester	118.6 ± 41.0

2.4. Molecular Modeling.

In order to understand the binding modes of the synthesized Pyk2 inhibitors with Pyk2 kinase, we performed molecular docking of T1, BT-Amide, T2, and BT-Ester to Pyk2 kinase in the DFG-in pose (PDB code: 5TOB) (Figure 4). As derivatives of TAE-226, both T1 and T2 adopted similar binding poses as TAE-226 within the Pyk2 kinase domain. They bind to the hinge region within the ATP-binding pocket, forming stabilizing interactions with multiple amino acid residues, including hydrogen bonds with Y505. Comparing BT-Amide to T1, we observed that BT-Amide also adopts a similar pose within the ATP-binding pocket, while the linker and alendronic acid extend outward from the pocket. The alendronic acid of BT-Amide forms hydrogen bonds with R429, I430, and L431 on the kinase surface outside of the ATP-binding pocket, along with salt bridges with R429. In general, the attachment of the linker and the alendronic acid does not hinder the binding of BT-Amide to Pyk2 within the ATP-binding pocket, which aligns with the results of the kinase inhibitory assay we conducted. Likewise, BT-Ester also adopts a similar pose as T2 within the ATP-binding pocket, with the linker and alendronic acid extending outside of the pocket. This structure is stabilized by hydrogen bonds formed between the alendronic acid and amino acid residues on the kinase surface.

Figure 4.

Docking results of Pyk2 inhibitors toward Pyk2 kinase (DFG-in, PDB code: 5TOB). (A) Docking mode of T1 to Pyk2. (B) Docking mode of BT-Amide to Pyk2. (C) Docking mode of T2 to Pyk2. (D) Docking mode of BT-Ester to Pyk2. The ribbons representing for the hinge region are colored in blue. The hydrogen bonds, salt bridges, and pi-cation interactions are depicted as dashed lines in yellow, pink, and green, respectively.

2.5. In Vivo Study.

In our previous studies, we discovered that pharmacological inhibition of Pyk2 by PF-431396 prevents GC-induced bone loss.³⁴ In the above-mentioned kinase inhibitory assay, BT-Amide exhibited moderate inhibitory effects on Pyk2 kinase, while BT-Ester was observed to have low inhibition on Pyk2. In order to examine if Pyk2 inhibitor TAE-226 and its bone-targeting derivatives BT-Amide and BT-Ester exhibit similar protective effects on bones as PF-431396 against GC-induced bone loss, we conducted a mice study. Mice were randomly assigned into different groups for controlled treatments. They were administered either vehicle, TAE-226 (10 mg/kg), BT-Amide (10 mg/kg), or BT-Ester (10 mg/kg) by oral gavage 5 times per week. Alongside, they were treated with either placebo or GC for 2 weeks (Figure S1). After that, the bone mineral density (BMD) of each mouse was assessed (Figure 5). In the groups of mice administered with the vehicle, a noticeable difference was observed between placebo-treated mice and GC-treated mice, although the difference was not statistically significant (P = 0.086). This result indicates that GC alone leads to a decrease in total BMD, consistent with previous reports.¹ Furthermore, in the groups of mice administered with TAE-226 and BT-Amide, no significant difference was observed between the placebo-treatment group and GC-treatment group. This suggests that both TAE-226 and BT-Amide effectively prevent GC-induced bone loss. However, a significant difference of total BMD was observed within the groups of mice administered with BT-Ester between the placebo and GC treatment groups. Additionally, a significant difference was observed within GC-treatment groups between TAE-226 administration and BT-Ester administration. Both the comparisons indicate that unlike TAE-226, BT-Ester lacks protective effects on bones from GC treatments in the mouse model.

Figure 5.

Bone mineral density (BMD) of mice after treatments. (A) The detection of BMD in mice by dual energy X-ray absorptiometry (DEXA). (B) Statistics of mice BMD in different groups of treatments in the context of placebo (PL) or the synthetic glucocorticoid (GC) prednisolone 2.1 mg/kg/d for 2 weeks (two-way ANOVA, Tukey posthoc, N = 9–10). *, P < 0.05 for mice orally administered with BT-Ester and treated by GC, vs mice orally administered with BT-Ester and placebo treated. ^#, P < 0.05 for mice orally administered with TAE-226 and treated by GC vs mice orally administered with BT-Ester and treated by GC.

The different in vivo protective effects of BT-Amide and BT-Ester can be attributed to two potential reasons. First, the Pyk2 inhibitory effects of BT-Ester are significantly lower compared to BT-Amide and TAE-226. This reduced Pyk2 inhibitory potency may limit its in vivo effects. Second, considering the chemical structures, BT-Amide is structurally more stable than BT-Ester. This is because BT-Amide was designed with an amide linkage, whereas the BT-Ester was designed with an ester linkage. Ester linkages are known to be more prone to hydrolysis than amide linkages, potentially resulting in the rapid loss of the bone-affinitive moiety, alendronic acid, in BT-Ester. Consequently, BT-Ester may not accumulate sufficiently in bone tissues. In summary, BT-Ester lacks protective effects on bones, but can serve in comparison with BT-Amide in this study.

Alendronic acid also has protective effects on GIOP;⁴⁹ however, the antiresorptive potency of the bisphosphonate is greatly reduced when the nitrogen is in a nonbasic form.⁵⁰ Although we did not directly compare our compounds with alendronate in the current experiments, we did compare BT-Amide with the ineffective compound BT-Ester, which shares the same amide moiety derived from alendronate. The ineffectiveness of BT-Ester suggests that the amide analogue of alendronate does not contribute to the prevention of GIOP.

Both TAE-226 and BT-Amide exhibited protective effects against GC-induced bone loss. However, it is noteworthy that TAE-226 also had adverse side effects in the mouse model. Among the 20 mice administered with TAE-226, three of them (15%) developed skin lesions (Figure 6). This can be attributed to the distribution of TAE-226 in skin tissues. In contrast, no adverse side effects were observed with BT-Amide administration during this study. This could be explained by the bone-affinitive character of the alendronic acid groups driving BT-Amide distribution toward the target tissue, bone. Although mice treated with BT-Ester also did not develop skin lesions, it was not effective as well for protecting bone loss. Compared to TAE-226, the inhibition of Pyk2 kinase for both BT-Ester and its cleaved product, T2 was significantly reduced (Table 1). In conclusion, BT-Amide displayed protective effects against GC-induced bone loss while avoiding the adverse side effects associated with systemic administration.

Figure 6.

15% (3/20) of mice administered with TAE-226 developed skin lesions.

It should be noted that bisphosphonate drugs are poorly absorbed from the gastrointestinal tract, typically with less than 2% bioavailability.^51–54 It is likely that BT-Amide has not altered this characteristic. However, the enhanced localization provided by the bisphosphonate component may magnify its efficacy. We have not yet compared BT-Amide with TAE-226 using intravenous or intraperitoneal dosing, but this would be an interesting direction for future studies.

2.6. In Vivo Toxicity.

In order to verify if the BT-Amide administration introduced additional toxicity to mice, we euthanized the mice at the end of the study and collected the serum of mice to assess the circulating organ toxicity biomarkers (Figure 7). Biomarkers from untreated mice, mice administered with the vehicle, and mice administered with BT-Ester were used as control groups. Among the tested biomarkers, no significant difference was observed between BT-Amide administration and any of the control groups. This result indicates the absence of organ toxicities in mice treated with BT-Amide.

Figure 7.

Evaluation of BT-Ester and BT-Amide toxicity in vivo by measuring circulating markers of organ toxicities. (N = 9–10; mean ± SD). Normal ranges for the blood marker are shown by dark blue lines as exhibited by nontreated control animals. None of the markers were significantly elevated above control levels.

3. CONCLUSIONS

Our previous research identified the blocking GC activation of the Pyk2 kinase by either genetic loss of Pyk2 function or pharmacologic means is sufficient to prevent bone loss by these immunosuppressant steroids in vivo. More specifically, administration of PF-431396, a pharmacologic Pyk2 inhibitor, prevented bone loss in a murine model of GC excess.³⁴ However, systemic administration of PF-431396 and other commercially available Pyk2 inhibitors are also known to induce adverse side-effects in other tissues, like the induction of skin lesions. In this study, we sought to overcome the harmful off-target actions of Pyk2 inhibition in vivo by the generation of a novel, specific bone-targeting Pyk2 inhibitor. We are the first to synthesis these bone-targeting Pyk2 inhibitors by chemically linking the Pyk2 inhibitor TAE-226 with alendronic acid as reported here. We first made modifications on TAE-226 to generate exposed amine or hydroxyl group for linkage, resulting in TAE-226 derivatives T1 and T2. After that, we attached a linker and alendronic acid to each derivative, leading to BT-Amide and BT-Ester (Figure 3). In in vitro studies, BT-Amide exhibited moderate Pyk2 inhibitory effects, similar as T1 (Table 1). This could be explained by the results of molecular modeling, which indicate that the attachment of the linker and alendronic acid did not significantly influence the binding between T1 and Pyk2 kinase (Figure 4). In in vivo studies, BT-Amide was observed to have protective effects on bones against GC-induced bone loss in the mouse model. Notably, BT-Amide was found to be effective via oral administration by gavage (Figure 5). Moreover, BT-Amide avoids adverse effects such as skin lesions observed after TAE-226 administration, which were caused by the distribution of Pyk2 inhibitors in skin tissues (Figure 6). Furthermore, circulating biomarkers such as albumin, calcium, glucose, etc. after BT-Amide administration remain within the same level as control, indicating the absence of organ toxicities (Figure 7). In conclusion, BT-Amide is the first bone-targeting Pyk2 inhibitor shown to effectively prevent GC-induced bone loss, while avoiding the adverse side effect of skin lesion development.

The strategy of delivering other drug classes specifically to the mineralized bone matrix by conjugation with BPs has been widely applied in numerous studies.^55–58 However, to the best of our knowledge, none of the agents in these studies was reported to be effective via an oral route of administration (by gavage). One possible reason for this is that most previously reported agent-BP conjugates were designed to be prodrugs. In such cases, after delivery into bone tissues, the linkage between the BP and the agent needs to be cleaved to release the active agent. Consequently, the designed linkage is typically not too stable and can be susceptible to digestive enzymes. However, BT-Amide has an amide bond as linkage, which is strong enough to resist immediate hydrolysis in the digestive enzymes, allowing itself to survive in the digestive system before entering the internal environment. This could explain the effective oral route of administration with usage of BT-Amide. Furthermore, to investigate whether BT-Amide works as a prodrug, we collected mouse serum 24 h after BT-Amide administration and measured the concentration of possible active metabolites for Pyk2 inhibition using LC-MS/MS, particularly T1, which is the most likely metabolite. Interestingly, no active metabolite was detected to have substantial concentration. Therefore, BT-Amide may not function as a prodrug; instead, it may act as a Pyk2 inhibitor itself. This is a plausible characteristic of BT-Amide, given the chemical properties of the amide bond, which is also supported by the extensive reports in the evergreen field of peptide-based drugs and the emerging field of bifunctional small molecules highlighting the stability of amide bonds.^59–66 Taking together, we are the first to generate and report the effective oral administration of a novel BT-Amide Pyk2 inhibitor in a preclinical model of GIOP, which overall highlights the innovative BP conjugate strategy as a potential avenue for generating new lines of therapeutics with lower off-target risks.

4. EXPERIMENTAL SECTION

4.1. Chemistry.

All the starting materials were obtained from commercial suppliers and used without further purification. Thermo Finnigan LCQ Deca with Thermo Surveyor LCMS System at variable wavelengths of 254 and 214 nm was used to monitor the reaction and test the purity of the compounds. The purity of all the final compounds is >95%. The water–methanol gradient buffered with 0.1% formic acid was used as the mobile phase for the HPLC system. NMR spectra was completed on a Varian 400 MHz instrument. The ¹H NMR spectra and ¹³C spectra were recorded at 400 and 101 MHz, respectively. All final compounds were purified using Silica gel (0.035–0.070 mm, 60 Å) flash chromatography, unless otherwise noted. Microwave assisted reactions were completed in sealed vessels using a Biotage Initiator microwave synthesizer (Biotage, Uppsala, Sweden).

4.1.1. Synthesis of 2-((2,5-Dichloropyrimidin-4-yl)amino)-N-methylbenzamide, (2).

Compound 2 was synthesized according to the literature.⁶⁷ Compound 1 (13.00 g, 70.88 mmol), 2-amino-N-methylbenzamide (12.77 g, 85.05 mmol) and N,N-diisopropylethylamine (14.82 mL, 85.05 mmol) were dissolved in 260 mL isopropanol. The resulting mixture was stirred at 85 °C, reflux for 18 h. After completion of reaction, the reaction mixture was cooled down and filtrated. The resulting solid was washed with isopropanol and dried under decreased pressure yielded product compound 2 as white solid (19.33 g, 91.78%). MS m/z [M + 1] 298.5. ¹H NMR (400 MHz, DMSO-d₆) δ 12.19 (s, 1H), 8.88 (d, J = 4.1 Hz, 1H), 8.52 (d, J = 8.4 Hz, 1H), 8.49 (s, 1H), 7.80 (d, J = 7.8 Hz, 1H), 7.61 (t, J = 7.9 Hz, 1H), 7.23 (t, J = 7.6 Hz, 1H), 2.81 (d, J = 4.5 Hz, 3H).

4.1.2. Synthesis of 2-((5-Chloro-2-((2-methoxy-4-morpholinophenyl)amino)pyrimidin-4-yl)amino)-N-methylbenzamide, (TAE-226).

TAE-226 was synthesized according to the literature.⁶⁸ Compound 2 (500 mg, 1.68 mmol), 2-methoxy-4-morpholinoaniline (350 mg, 1.68 mmol) and p-toluenesulfonic acid (580 mg, 3.37 mmol) were dissolved in 5 mL isopropanol. The resulting mixture was stirred at 80 °C for 16 h. After completion of reaction, the reaction mixture was cooled down and filtrated. The resulting solid was washed with isopropanol and dried under decreased pressure. Flash chromatography C18 (MeOH/H₂O 10–70%) yielded TAE-226 as a white solid (427 mg, 54.12%). MS m/z [M + 1] 469.3. ¹H NMR (400 MHz, DMSO-d₆) δ 11.60 (s, 1H), 8.72 (d, J = 4.6 Hz, 1H), 8.59 (s, 1H), 8.15 (s, 1H), 8.10 (s, 1H), 7.71 (dd, J = 7.9, 1.4 Hz, 1H), 7.43 (d, J = 8.7 Hz, 1H), 7.36–7.28 (m, 1H), 7.10–7.03 (m, 1H), 6.66 (d, J = 2.5 Hz, 1H), 6.49 (dd, J = 8.7, 2.5 Hz, 1H), 3.81–3.72 (m, 5H), 3.15–3.10 (m, 3H), 2.79 (d, J = 4.5 Hz, 2H).

4.1.3. Synthesis of tert-Butyl (3-Methoxy-4-nitrophenyl)carbamate, (4).

Compound 3 (3.00 g, 17.84 mmol), ditert-butyl dicarbonate (4.67 g, 21.41 mmol), triethylamine (4.97 mL, 35.68 mmol), and 4-dimethylaminopyridine (217.97 mg, 1.78 mmol) were dissolved in 50 mL DCM. The resulting mixture was stirred at room temperature overnight. After completion of reaction, the reaction mixture was concentrated in vacuo. Flash chromatography silica (EA/Hexane 0–15%) yielded 4 as a yellow solid (2.94 g, 61.43%). MS m/z [M + 1] 269.1. ¹H NMR (400 MHz, DMSO-d₆) δ 7.90 (d, J = 8.6 Hz, 1H), 7.35 (d, J = 1.4 Hz, 1H), 6.99 (dd, J = 8.6, 1.7 Hz, 1H), 3.89 (s, 3H), 1.40 (s, 9H). The structural data corresponds with the previous literature.⁶⁹

4.1.4. Synthesis of tert-Butyl (4-Amino-3-methoxyphenyl)carbamate, (5).

We used a different method from the previous literature to prepare compound 5.⁷⁰ Compound 4 (2.89 g, 10.77 mmol) and Pd/C (0.28 g), was dissolved in 30 mL MeOH. The resulting mixture was stirred at room temperature for 5 h under H₂ atmosphere. After completion of reaction, the reaction mixture was filtrated twice and concentrated in vacuo. The resulting solid was dried under decreased pressure yielded 5 as a purple solid (2.73 g, 106.35%). MS m/z [M + 1] 239.8. ¹H NMR (400 MHz, DMSO-d₆) δ 6.57 (s, 1H), 6.55 (s, 1H), 6.45 (dd, J = 8.2, 1.5 Hz, 1H), 4.80 (s, 2H), 3.72 (s, 3H), 1.38 (s, 9H).

4.1.5. Synthesis of tert-Butyl (4-((5-chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)carbamate, (6).

Compound 5 (2.38 g, 9.99 mmol), 2 (1.98 g, 6.66 mmol) and acetic acid (1.20 g, 19.98 mmol) was dissolved in 30 mL dioxane. The resulting mixture was stirred at 110 °C overnight in a sealed bottle. After completion of reaction, the reaction mixture was concentrated in vacuo. Then, water (20 mL) and EA (20 mL × 3) were added, and the organic layers were collected, washed with brine (20 mL), dried over anhydrous Na₂SO₄, and concentrated in vacuo. Flash chromatography silica (EA/Hexane 0–40%) yielded 6 as a yellow solid (2.94 g, 61.43%). MS m/z [M + 1] 499.1. ¹H NMR (400 MHz, DMSO-d₆) δ 11.61 (s, 1H), 9.31 (s, 1H), 8.74 (d, J = 4.5 Hz, 1H), 8.57 (d, J = 7.5 Hz, 1H), 8.21 (s, 1H), 8.12 (s, 1H), 7.71 (d, J = 7.8 Hz, 1H), 7.49 (d, J = 8.6 Hz, 1H), 7.34 (s, 1H), 7.32 (s, 1H), 7.07 (t, J = 7.5 Hz, 1H), 6.95 (d, J = 8.5 Hz, 1H), 3.73 (s, 3H), 2.79 (d, J = 4.3 Hz, 3H), 1.49 (s, 9H).

4.1.6. Synthesis of 2-((2-((4-Amino-2-methoxyphenyl)amino)-5-chloropyrimidin-4-yl)amino)-N-methylbenzamide, (T1).

Compound 6 (0.60 g, 1.20 mmol) and trifluoroacetic acid (1.37 g, 12.02 mmol) was dissolved in 6 mL DCM. The resulting mixture was stirred for 3 h at room temperature. After completion of reaction, the excessive CF₃COOH was neutralized by triethylamine dropwise. The reaction mixture was concentrated in vacuo. Flash chromatography silica (EA(1‰NH₃·H₂O)/Hexane 0–80%) yielded T1 as a yellow solid (0.45 g, 93.83%). MS m/z [M + 1] 399.1. ¹H NMR (400 MHz, DMSO-d₆) δ 11.58 (s, 1H), 8.72 (d, J = 4.3 Hz, 1H), 8.61 (s, 1H), 8.05 (s, 2H), 7.70 (d, J = 7.9 Hz, 1H), 7.28 (s, 1H), 7.12–6.99 (m, 2H), 6.32 (s, 1H), 6.15 (d, J = 8.3 Hz, 1H), 5.02 (s, 2H), 3.65 (s, 3H), 2.79 (d, J = 4.4 Hz, 3H). ¹³C NMR (101 MHz, DMSO-d₆) δ 169.00, 159.66, 154.82, 154.71, 153.85, 147.28, 139.81, 131.41, 127.81, 127.17, 121.27, 120.89, 119.83, 116.69, 105.16, 103.37, 97.81, 54.97, 26.34.

4.1.7. Synthesis of 6-((4-((5-Chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)amino)-6-oxohexanoic Acid, (7).

Compound T1 (0.98 g, 2.45 mmol), adipic acid (537 mg, 3.69 mmol), HATU (1.40 g, 3.69 mmol) and N,N-diisopropylethylamine (1.28 mL, 7.37 mmol) was dissolved in 20 mL DMF. The resulting mixture was stirred for 3 h at room temperature. After completion of reaction, the reaction mixture was concentrated in vacuo directly. Flash chromatography silica (MeOH/DCM 0–8%) yielded 7 as a yellow solid (0.96 g, 74.36%). MS m/z [M + 1] 527.1. ¹H NMR (400 MHz, DMSO-d₆) δ 12.09 (s, 1H), 9.99 (s, 1H), 9.03 (s, 1H), 8.83 (d, J = 4.3 Hz, 1H), 8.51 (s, 1H), 8.31 (s, 1H), 8.22 (s, 1H), 7.76 (d, J = 7.8 Hz, 1H), 7.52 (s, 1H), 7.49 (d, J = 8.6 Hz, 1H), 7.42–7.28 (m, 1H), 7.23–7.08 (m, 2H), 3.75 (s, 3H), 2.80 (d, J = 4.4 Hz, 3H), 2.33 (t, J = 6.9 Hz, 2H), 2.26 (t, J = 7.1 Hz, 2H), 1.71–1.44 (m, 4H).

4.1.8. Synthesis of 2,5-Dioxopyrrolidin-1-yl 6-((4-((5-chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)amino)-6-oxohexanoate, (8).

Compound 7 (0.96 g, 1.82 mmol), 1-hydroxypyrrolidine-2,5-dione (566 mg, 4.92 mmol) and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide (764 mg, 4.92 mmol) was dissolved in 10 mL DMF. The resulting mixture was stirred for 4h at room temperature. After completion of reaction, the reaction mixture was concentrated in vacuo directly. Then, water (20 mL) and EA (20 mL × 3) were added, and the organic layers were collected, dried over anhydrous Na₂SO₄, and concentrated in vacuo. Flash chromatography silica (EA/Hexane 0–100%) yielded 8 as a yellow solid (320 mg, 28.15%). MS m/z [M + 1] 624.2.

4.1.9. Synthesis of (4-(6-((4-((5-Chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)amino)-6-oxohexanamido)-1-hydroxybutane-1,1-diyl)bis(phosphonic Acid) (BT-Amide).

A solution of alendronic acid triethylammonium salt (pH ~ 10) was prepared by mixing alendronic acid (511 mg, 2.05 mmol, 4 equiv), water (3 mL), DMF (5 mL), and triethylamine (0.8 mL). To a solution of 8 (320 mg, 513 μmol, 1 equiv) in DMF (1.5 mL) was added the previously prepared solution of alendronic acid/Et₃N. The resulting mixture was stirred for 1h at room temperature. After completion of reaction, the reaction mixture was concentrated in vacuo and washed with EA. The resulting solid was collected and H₂O was added. The resulting mixture was filtrated and the solution was collected and concentrated in vacuo. Finally, the resulting solid was mixed with methanol and filtrated; the solution was collected and concentrated in vacuo. Preparation HPLC C18 (MeOH/H₂O 30–100%, 1‰ trifluoroacetic acid) yielded BT-Amide as a colorless solid (180 mg, 47.30%). MS m/z [M + 1] 758.1. ¹H NMR (400 MHz, CD₃OD) δ 8.50 (d, J = 7.9 Hz, 1H), 8.04 (s, 1H), 7.90 (d, J = 8.8 Hz, 1H), 7.65 (d, J = 7.8 Hz, 1H), 7.44 (d, J = 13.0 Hz, 2H), 7.17–7.09 (m, 1H), 7.05 (d, J = 8.9 Hz, 1H), 3.87 (s, 2H), 3.65 (s, 1H), 3.34 (s, 3H), 2.91 (s, 2H), 2.66 (s, 3H), 2.54 (d, J = 11.1 Hz, 2H), 2.41–2.32 (m, 2H), 2.24 (s, 2H), 1.92 (s, 2H), 1.71 (s, 2H).

4.1.10. Synthesis of 1-(3-Methoxy-4-nitrophenyl)piperidin-4-ol (10).

Compound 10 was synthesized according to the literature.⁷¹ Compound 9 (5.00 g, 29.22 mmol), piperidin-4-ol (3.55 g, 35.06 mmol) and K₂CO₃ (4.85 g, 35.06 mmol) was dissolved in 50 mL DMSO. The resulting mixture was stirred at room temperature overnight. After completion of reaction, water (50 mL) and EA (50 mL × 3) were added, and the organic layers were collected, dried over anhydrous Na₂SO₄, and concentrated in vacuo. Compound 10 was yielded as a yellow solid (7.90 g, 107.18%). MS m/z [M + 1] 253.1. ¹H NMR (400 MHz, DMSO-d₆) δ 7.87 (d, J = 9.4 Hz, 1H), 6.58 (dd, J = 9.5, 2.1 Hz, 1H), 6.49 (d, J = 2.1 Hz, 1H), 4.78 (d, J = 3.9 Hz, 1H), 3.90 (s, 3H), 3.81 (dt, 2H), 3.74 (td, J = 8.6, 4.4 Hz, 1H), 3.18 (ddd, J = 13.0, 9.7, 2.9 Hz, 2H), 1.87–1.75 (m, 2H), 1.47–1.35 (m, 2H).

4.1.11. Synthesis of 1-(4-Amino-3-methoxyphenyl)piperidin-4-ol (11).

Compound 11 was synthesized according to the literature.⁷² Compound 10 (4.90 g, 19.42 mmol) and Pd/C(0.49 g) was dissolved in 50 mL methanol. The resulting mixture was stirred at room temperature under H₂ atmosphere for 5h. After completion of reaction, the reaction mixture was filtrated twice and concentrated in vacuo. The resulting solid was dried under decreased pressure yielded 11 as a black solid (4.28 g, 99.13%). MS m/z [M + 1] 223.1. ¹H NMR (400 MHz, DMSO-d₆) δ 6.49 (d, J = 8.5 Hz, 1H), 6.28 (dd, J = 8.3, 2.3 Hz, 1H), 4.62 (d, J = 4.2 Hz, 1H), 4.19 (s, 2H), 3.73 (s, 3H), 3.53 (qd, J = 8.4, 4.0 Hz, 1H), 3.30–3.20 (m, 2H), 2.68–2.55 (m, 2H), 1.85–1.74 (m, 2H), 1.57–1.42 (m, 2H).

4.1.12. Synthesis of 2-((5-Chloro-2-((4-(4-hydroxypiperidin-1-yl)-2-methoxyphenyl)amino)pyrimidin-4-yl)amino)-N-methylbenzamide (T2).

Compound 2 (5.02 g, 16.91 mmol), compound 11 (3.76 g, 16.91 mmol) and p-toluenesulfonic acid (5.82 g, 33.82 mmol) were dissolved in 100 mL isopropanol. The resulting mixture was stirred at 82 °C reflux for 16 h. After completion of reaction, the reaction mixture was concentrated in vacuo. The resulting solid was washed with isopropanol and dried under decreased pressure. Flash chromatography C18 (MeOH/H₂O 10–60%) yielded T2 as a yellow solid (5.44 g, 66.61%). MS m/z [M + 1] 483.2. ¹H NMR (400 MHz, DMSO-d₆) δ 10.97 (s, 1H), 8.69 (d, J = 8.2 Hz, 1H), 8.09 (d, J = 9.3 Hz, 2H), 7.50–7.43 (m, 2H), 7.25 (s, 1H), 7.08 (td, J = 7.6, 1.1 Hz, 1H), 6.56 (d, J = 2.5 Hz, 1H), 6.51 (dd, J = 8.8, 2.5 Hz, 1H), 6.19 (d, J = 4.7 Hz, 1H), 3.87 (s, 3H), 3.51–3.44 (m, 3H), 3.03 (d, J = 4.9 Hz, 3H), 2.89 (ddd, J = 12.6, 9.9, 3.0 Hz, 2H), 2.08–2.00 (m, 2H), 1.74 (dtd, J = 13.0, 9.6, 3.8 Hz, 2H). ¹³C NMR (101 MHz, DMSO-d₆) δ 168.96, 159.06, 154.88, 154.70, 152.80, 149.27, 139.61, 131.43, 127.85, 125.56, 121.51, 120.94, 120.07, 119.59, 107.06, 103.91, 100.40, 66.11, 55.37, 47.18 (2C), 33.99 (2C), 26.34. The structural data corresponds with the previous literature.⁷³

4.1.13. Synthesis of 6-((1-(4-((5-Chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)piperidin-4-yl)oxy)-6-oxohexanoic Acid (12).

Compound T2 (500 mg, 1.04 mmol), oxepane-2,7-dione (398 mg, 3.11 mmol), 4-dimethylaminopyridine (25 mg, 207.05 ummol) and triethylamine (1.01 mL, 7.25 mmol) were dissolved in 10 mL anhydrous DCM. The resulting mixture was stirred under N₂ atmosphere at room temperature overnight. After completion of reaction, the reaction mixture was concentrated in vacuo. Flash chromatography silica (MeOH/DCM 0–3%) yielded 12 as a yellow solid (crude, 440 mg, 69.55%). MS m/z [M + 1] 611.1.

4.1.14. Synthesis of 1-(4-((5-Chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)piperidin-4-yl (2,5-dioxopyrrolidin-1-yl) Adipate (13).

Compound 12 (440 mg, 720.02 μmol), 1-hydroxypyrrolidine-2,5-dione (249 mg, 2.16 mmol) and 1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide (335 mg, 2.16 mmol) was dissolved in 10 mL DMF. The resulting mixture was stirred for 4h at room temperature. After completion of reaction, water (20 mL) and EA (20 mL × 3) were added, and the organic layers were collected, washed with brine water (20 mL), dried over anhydrous Na₂SO₄, and concentrated in vacuo. Flash chromatography silica (EA/Hexane 0–100%) yielded 13 as a yellow solid (190 mg, 37.26%). MS m/z [M + 1] 708.2. ¹H NMR (400 MHz, CDCl₃) δ 10.99 (s, 1H), 8.66 (d, J = 8.3 Hz, 1H), 8.11–7.99 (m, 2H), 7.52–7.40 (m, 2H), 7.29 (s, 1H), 7.11–7.01 (m, 1H), 6.56 (d, J = 2.5 Hz, 1H), 6.50 (dd, J = 8.8, 2.5 Hz, 1H), 6.25 (d, J = 4.6 Hz, 1H), 4.95 (ddd, J = 12.0, 8.0, 3.9 Hz, 1H), 3.87 (s, 3H), 3.41–3.31 (m, 2H), 3.09–2.95 (m, 5H), 2.79 (s, 4H), 2.68–2.61 (m, 2H), 2.38 (t, J = 6.8 Hz, 2H), 2.04–1.99 (m, 2H), 1.90–1.72 (m, 6H).

4.1.15. Synthesis of (4-(6-((1-(4-((5-Chloro-4-((2-(methylcarbamoyl)phenyl)amino)pyrimidin-2-yl)amino)-3-methoxyphenyl)piperidin-4-yl)oxy)-6-oxohexanamido)-1-hydroxybutane-1,1-diyl)bis(phosphonic Acid), (BT-Ester).

A solution of alendronic acid triethylammonium salt (pH ~ 10) was prepared by mixing alendronic acid (295 mg, 1.19 mmol, 4 equiv), water (2 mL), DMF (3 mL), and triethylamine (0.5 mL). To a solution of 13 (210 mg, 297 μmol, 1 equiv) in DMF (1 mL) was added the previously prepared solution of alendronic acid/Et₃N. The resulting mixture was stirred for 1h at room temperature. After completion of reaction, the reaction mixture was concentrated in vacuo and washed with EA. The resulting solid was collected and H₂O was added. The resulting mixture was filtrated, and the solution was collected and concentrated in vacuo. Finally, the resulting solid was mixed with methanol and filtrated; the solution was collected and concentrated in vacuo. Preparation HPLC C18 (MeOH/H₂O 30–100%, 1‰ trifluoroacetic acid) yielded BT-Ester as a colorless solid (120 mg, 47.04%). MS m/z [M + 1] 842.3. ¹H NMR (400 MHz, CD₃OD) δ 8.49 (d, J = 8.3 Hz, 1H), 7.96 (s, 1H), 7.63 (dd, J = 22.8, 8.3 Hz, 2H), 7.33 (s, 1H), 7.06 (s, 1H), 6.64 (s, 1H), 6.50 (d, J = 8.3 Hz, 1H), 3.79 (s, 1H), 3.60 (s, 2H), 3.46 (s, 1H), 3.24 (s, 3H), 3.12 (s, 4H), 2.93 (s, 2H), 2.61 (s, 3H), 2.49 (dd, J = 11.6, 6.1 Hz, 2H), 2.31 (dd, J = 16.6, 6.6 Hz, 2H), 2.13 (s, 2H), 2.06–1.91 (m, 6H), 1.86 (s, 2H), 1.56 (s, 2H).

4.2. Pyk2 Kinase Activity Assay.

Kinase assays was conducted using the bioluminescent ADP-Glo kinase assay (Promega), following the manufacturer’s instructions. Assay was performed with the test compounds at 8-point half log dilutions (1 uM to 0.316 nM). Luminescence signal was measured on a spectrophotometer (BioTek), and IC₅₀ values reported are based on the dose–response curve fitted in GraphPad Prism V9.0 using [inhibitor] vs normalized response model. Each compound was tested at least 3 times.

4.3. Molecular Docking.

The molecular docking studies were completed using Maestro 11.8 (released 2018–4). The ligand was prepared by the LigPrep tool. The crystal structure of Pyk2 bound with an inhibitor (PDB ID: 5TOB) was used as template. The protein preparation was done in Maestro using the protein preparation wizard tool. A grid box around the ATP binding site was created. The ligands were docked in this grid using Ligand Docking. The results were visualized and analyzed with the Maestro suite.

4.4. In Vivo Study.

All animal experiments performed were conducted in compliance with the guidelines as defined by Institutional Animal Care and Use Committee (IACUC ID: IPRO-TO202200000010). Skeletally mature 4-month-old C57BL/6 female mice (five per cage) were fed with a regular diet (Teklad Global 18% Protein Extruded Rodent Diet Sterilizable, catalog no. 2018SX, Harlan/ENVIGO, Indianapolis, IN), received water ad libitum, and were maintained on a 12-h light/dark cycle in polycarbonate cages. Mice were implanted subcutaneously with slow-release pellets delivering placebo, or 2.1 mg/kg/d (GC) prednisolone (Innovative Research of America, Sarasota, FL) for 4 weeks. Previous studies showed that these doses reproduce in the mouse the hallmarks of GC-induced osteoporosis and are equivalent to medium and high therapeutic glucocorticoid doses in humans.^74,75 C57BL/6 4-month-old female mice were administered by oral gavage with either vehicle, 10 mg/kg of TAE-226, 10 mg/kg of BT-Amide, or 10 mg/kg of BT-Ester 5 times per week, as previously published,⁷⁶ initiating 3 days prior to placebo or GC pellet implantation. Mice were euthanized by, first, sedation with 2% isoflurane (Abbott Laboratories, Chicago, IL) administered by a Drager 19.1 Anesthetic Vaporizer, which was then confirmed by the secondary means of cervical dislocation. Analysis was performed in a blinded fashion.

4.5. BMD Measurement.

Bone mineral density (BMD) of the total body excluding the head and the tail were measured by dual-energy X-ray absorptiometry (DEXA) scanning by using a PIXImus II densitometer (GE Medical Systems, Lunar Division, Madison, WI). DXA measurements were performed 2 to 4 days before to determine baseline BMD levels for randomization purposes and 2 weeks after pellet implantation.^74,77 Mice were randomized to the experimental groups based on initial BMD. Briefly, mice were sorted by BMD starting from the highest BMD, and then randomly distributed to the treatment groups. After randomization, no statistical differences on BMD were found between experimental groups prior to the administration of any treatment agent or pellet implantation.

4.6. In Vivo Toxicity Study.

To monitor the function of organs after the exposure with the compounds, 14 parameters of the Comprehensive Diagnostic Profile (Abaxis)—including alanine aminotransferase, albumin, alkaline phosphatase, amylase, blood urea nitrogen, calcium, creatinine, globulin, glucose, phosphorus, potassium, sodium, total bilirubin, and total protein were analyzed in plasma samples to quantitatively assess functions of the liver, kidney, heart, intestine, pancreas, and other organs that could be affected by the compounds (Figure 7). No toxicity was observed as compared to nontreated control animals.

Supplementary Material

Supporting Information

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jmedchem.4c02539.

Docking of TAE-226 (PDB)

Docking of T1 (PDB)

Docking of T2 (PDB)

Docking of BT-Amide (PDB)

Docking of BT-Ester (PDB)

Table of molecular formula strings (CSV)

Experimental design for in vivo study; purity, ¹H and ¹³C NMR spectra for compounds (PDF)

ABBREVIATIONS USED

ATP

adenosine triphosphate

BMD

bone mineral density

BP

bisphosphonate

BT

bone-targeted

DEXA

dual energy X-ray absorptiometry

DFG

aspartic acid-phenylalanine-glycine

FAK

focal adhesion kinase

GC

glucocorticoid

GIOP

glucocorticoid-induced osteoporosis

HA

hydroxyapatite

LC-MS/MS

liquid chromatography with tandem mass spectrometry

NFkB

nuclear factor kappa β

PL

placebo

Pyk2

proline-rich tyrosine kinase 2

RANKL

receptor activator of nuclear factor kappa β ligand

USFDA

U.S. Food And Drug Administration