Discovery of novel 1H-pyrazolo[3,4-d]pyrimidine derivatives as BRK/PTK6 inhibitors

Introduction

Breast tumor kinase (BRK), also known as PTK6, is a non-receptor protein tyrosine kinase containing a catalytic kinase domain and SH2/SH3 domains.(1) It is an oncogenic driver overexpressed in breast, ovarian, prostate, and pancreatic cancers, yet shows minimal expression in normal tissues.(2–9) It is overexpressed in approximately 85% of breast carcinomas—including up to 78% of triple-negative tumors.(10, 11) BRK drives migration, invasion, and metastasis, and correlates with poor outcomes.(12, 13) Although BRK can be detected in non-cancerous tissues, its kinase activity is observed only in the tumor plasma membrane,(14) making it a viable therapeutic target for breast cancer treatment.

BRK kinase activity fuels epithelial-mesenchymal transition (EMT), a phenomenon linked to treatment resistance and metastases.(8, 11, 15) In fact, only catalytically active BRK can drive EMT by suppressing epithelial markers (such as E-cadherin) and boosting mesenchymal regulators (such as N-cadherin and SNAIL).(10, 16) Beyond EMT, BRK phosphorylates crucial substrates—including Stat3/5, FAK, Akt, BKS, paxillin, β-catenin, HIF-1, SNAIL, DOK1, and SMAD4—thereby advancing oncogenic signaling.(1, 10, 16–19) Suppression of BRK levels through shRNA slows metastasis, diminishes migration and invasion, and increases anoikis and these effects are also observed through pharmacological modulation of BRK kinase activity.(10) By degrading tumor-suppressive proteins like SMAD4 and driving malignant phenotypes,(16) it is well justified to impair the tumor-promoting functions of BRK through the development of potent, BRK-specific inhibitors.

While several BRK kinase inhibitors have been identified (see Figure 1), their effects on cell viability by targeting BRK have been inconsistent.(20–23) For example, XMU-MP-2 inhibited the growth of BRK-expressing cells in vitro and in vivo,(21) whereas other BRK inhibitors failed to limit proliferation.(20) This inconsistency is likely caused by the promiscuity of small-molecule inhibitors, which arises due to the highly conserved ATP-binding pocket among kinases. This lack of selectivity contributes to inconsistent findings by confounding the interpretation of the role of BRK in tumor progression, as off-target effects may mask or mimic its function.(24, 25) Although considerable efforts have been made, no BRK inhibitors have progressed to clinical trials, indicating a need for advanced molecules with improved specificity.

Figure 1:

A. Reported BRK inhibitors. The hinge binders are highlighted in green, the back pocket region in red, and the solvent front region in blue. B. Crystal structure of BRK protein. The hinge region is highlighted in green, the back pocket region in red, and the solvent front region in blue.

Herein, we describe the design, discovery, and biological evaluation of 1H-pyrazolo[3,4-d]pyrimidin-4-amine scaffolds as selective BRK inhibitors (see Figure 2). Biochemical assays reveal that these compounds exhibit excellent potency, with IC₅₀ values in the low nanomolar range. Molecular modeling studies suggest that these derivatives function as Type I inhibitors, binding to the DFG-in conformation of the BRK kinase. Furthermore, biological assessments highlight their anti-metastatic potential, making them promising candidates for further development in targeting BRK-driven cancer progression.

Figure 2:

Rational design of BRK inhibitors.

Materials and Methods

Synthesis

The synthesis of tricyclic pyrimido-indole compounds used in this study is outlined in Scheme 1. Briefly, 4-bromo-1-fluoro-2-nitrobenze (1) was reacted with ethyl-2-cyanoacetate (2) to give intermediate 3, which was subsequently reduced with zinc and acetic acid in ethanol to give ethyl 2-(4-bromo-2-nitrosophenyl)-2-cyanoacetate (4). The intermediate 4 was then cyclized in formamide to afford compound 5 followed by chlorination with POCl₃ to obtain 7-bromo-4-chloro-9H-pyrimido[4,5-b]indole (6). Nucleophilic substitution of compound 6 with various amines led to various final compounds.

Scheme 1:

a) K₂CO₃, DMF, RT; b) Zn, AcOH, EtOH, 90 °C; c) Formamide, 140 °C; d) POCl₃, 110 °C; e) pTSA, appropriate amines, EtOH, 90 °C.

Scheme 2 depicts the synthesis of pyrazine derivative 28. 2,6-dichloropyrazine (26) was subjected to a Suzuki reaction to obtain the final compound 3-(6-chloropyrazin-2-yl)phenol (28).

Scheme 2:

a) Pd(PPh₃)₄, K₂CO₃, 4: 1 DMF:H₂O, 90 °C.

The synthesis of 3-(quinoxalin-2-yl)phenol (32) began with cyclization of benzene-1,2-diamine (29) with glyoxylic acid in ethanol, yielding quinoxalin-2-ol (30) as shown in Scheme 3. Following cyclization, intermediate 30 was chlorinated in POCl₃ to obtain 2-chloroquinoxaline (31). Suzuki coupling of intermediate 31 with (3-hydroxyphenyl)boronic acid (27) afforded to final compound 32.

Scheme 3:

a) glyoxylic acid, EtOH, reflux; b) POCl₃, 110 °C; c) 27, Pd(PPh₃)₄, K₂CO₃, 4: 1 DMF:H₂O, 90 °C.

The starting materials 7-bromo-4-chloroquinazoline (33) and 3-aminophenol (34) were refluxed in EtOH in the presence of DIPEA to obtain the final compound 3-((7-bromoquinazolin-4-yl)amino)phenol (35) as shown in Scheme 4.

Scheme 4:

a) DIPEA, EtOH, 90 °C.

Scheme 5 illustrates the synthesis of 1H-pyrazolo[3,4-d]pyrimidin-4-amine derivatives used in this study. 1H-pyrazolo[3,4-d]pyrimidin-4-amine (36) was reacted with N-iodosuccinimide in DMF at 60 °C to obtain 3-iodo-1H-pyrazolo[3,4-d]pyrimidin-4-amine (37). Intermediate 37 was then reacted with various alkyl- and amine- halides to obtain a range of pyrazolo-substituted intermediates. The intermediates were then subjected to Suzuki cross coupling reactions to obtain various final products as shown in Scheme 5.

Scheme 5:

a) N-iodosuccinimide, DMF, 60 °C; b) appropriate alkyl or amine halide, K₂CO₃, DMF, 80 °C; c) appropriate boronic acids or esters, Pd(PPh₃)₄, K₂CO₃, 4: 1 DMF:H₂O, 90 °C.

BRK Biochemical Inhibitory Assay

Enzymatic inhibition of BRK was determined using the ADP-Glo kinase assay (Promega). Briefly, the BRK enzyme was pre-incubated with inhibitors for 10 minutes before adding substrate and the ATP mixture. The kinase reaction (a total of 5 μL) was then incubated for an hour and the ADP Glo reagent was added followed by incubation for 40 minutes. Kinase detecting reagent (KDR) was added and incubated for 45 minutes before measuring luminescence using a microplate reader (BMG CLARIOstar).

BRK Docking Study

Molecular docking studies were performed to better understand the interaction of the ligands with the BRK kinase. The docking studies were performed using Maestro 2018–4 suite. The BRK protein (PDB ID: 5H2U)(26) and ligands were prepared using the protein preparation and ligand preparation wizards, respectively. The ligands were then docked into the grid box, determined based on the crystal structure of the kinase with known ligand, to determine the binding pose. The binding pose was visualized using Maestro.

Results and Discussion

Structure Activity Relationship Studies – Hydroxyl group at the meta position is necessary to maintain BRK inhibition

Hilgeroth et al reported 4-anilino-α-carboline derivatives (tilfrinib or compound 4f) as BRK inhibitors, and are predicted to be Type I kinase inhibitors.(22) However, α-carbolines are known to be mutagenic and carcinogenic.(27) Furthermore, their synthesis is often reported in low yields.(27) Therefore, our primary goal was to design a scaffold with minimal or no mutagenic and carcinogenic potential while ensuring synthetic feasibility. To this end, we modeled kinase inhibitor fragments into a known BRK crystal structure(26) and identified a new ligand – 9H-pyrimido[4,5-b]indole – as a potential BRK inhibitor that creates key intermolecular contacts at the ATP hinge and back pocket similar to 4f (Figure 3A). The pyrimido-indole scaffold accepts a hydrogen bond at the hinge region with Met267 whereas the phenolic OH interacts with the DFG motif by donating a hydrogen bond. We then synthesized pyrimido-indole derivatives by keeping the hinge binding warhead constant while altering substituents that occupy the back pocket to better understand the importance of the phenolic moiety (Figure 3B).

Figure 3:

Identification and design of the pyrimido-indole warhead as BRK inhibitor. A. Molecular docking of pyrimido-indole scaffold as BRK inhibitor (PDB ID: 5H2U). The inhibitor was docked into the DFG-in crystal structure and was found to interact with Met267 at the hinge region and Asp330 at the DFG motif. B. Design of pyrimido-indole derivatives to study structure-activity relationships (SAR). The R substitution was changed to study SAR.

Following synthesis, we evaluated the activity of the 9H-pyrimido[4,5-b]indole compounds at a single point concentration of 2 μM against BRK using tilfrinib (4f) as a positive control. The enzymatic inhibition data indicated that 4f (IC₅₀ = 25 nM) is more potent than our designed inhibitor (7, IC₅₀ = 1 μM, data not shown). Both 4f and compound 7 have a phenol at the meta position, which is predicted to form a key hydrogen bond with Asp330 of the DFG motif. However, phenols are known to be rapidly metabolized through sulfation and glucuronidation, rendering them metabolically unstable.(28) This led us to evaluate whether alternative substitutions at the meta-position could improve potency and metabolic stability while retaining the activity.

To this end, we synthesized various derivatives by substituting phenols with other functional groups (Table 1). Changing the substitution to bulkier groups such as methoxy (8) and trifluoromethoxy (9), resulted in a loss of activity. Electronegative substituents such as halogens (10, 11, 12, 13, and 14), alkyne (15), and cyano (16) groups at the meta-position led to a decrease in activity. Furthermore, removal of phenol (17) or replacing phenol with a secondary alcohol (19) also reduced the activity. Activity of the pyrimido-indole derivatives was not retained when meta-substituted phenols were replaced with para-substituted electronegative functional groups such as 4-(methylsulfonyl)aniline (22). To mimic the acidic nature of the phenolic OH, substitution was changed to amine (23) indole (24), and indolamine (25), but these modifications did not retain activity. These findings suggest that the phenolic group at the meta-position is essential for BRK activity.

Table 1:

Single point inhibition assay of pyrimido-indole derivatives against BRK. The compounds were tested in the ADP-Glo assay at 2 μM.

Name	R	% BRK Inhibition
4f	-	100.0 ± 7.8
7		77.5 ± 4.2
8		26.5 ± 9.5
9		22.0 ± 6.2
10		4.3 ± 10.9
11		0
12		0
13		37.0 ± 5.1
14		5.7 ± 9.9
15		19.3 ± 13.1
16		44.2 ± 0.3
17		17.1 ± 1.4
18		0
19		21.8 ± 4.6
20		41.4 ± 14.7
21		37.5 ± 16.1
22		36.5 ± 1.6
23		10.6 ± 7.1
24		50.9 ± 6.2
25		25.3 ± 11.8

Since 9H-pyrimido[4,5-b]indoles did not exhibit improved BRK inhibition compared to 4f, we progressed to identifying a new scaffold. We modified the nitrogenous warhead region that hydrogen bonds to the BRK hinge while retaining the phenolic group so interactions at the back pocket are retained. To this end, we synthesized commonly used hinge warheads in kinase inhibitors, including pyrazine (26), quinoxaline (32), quinazoline (35), and pyrazolo-pyrimidine (38) while keeping the phenolic OH constant. The inhibitors were initially screened at a single point concentration of 2 μM (Table 2). The single point inhibition data show that the pyrazolo-pyrimidine scaffold had the highest inhibitory effect among all warheads. We then performed IC₅₀ experiments to compare the potency of 38 with 4f, and compound 38 was found to be more potent than 4f exhibiting an IC₅₀ of 0.153 μM compared to 0.196 μM (Figure S1).

Table 2:

Single point inhibition of various nitrogenous warheads as BRK inhibitors. The hinge binding region is highlighted in red. The compounds were tested in the ADP-Glo assay at 2 μM.

Name	Structure	% BRK Inhibition
26		29.6 ± 22.9
32		48.9 ± 8.9
35		76.1 ± 7.3
38		96.8 ± 0.7

SAR of pyrazolo-pyrimidine inhibitors

Following the identification of 38, we aimed to explore SAR profiles of the scaffold. Specifically, we modified the substitution at the pyrazole ring (R₁) and the phenol ring (R₂) (Table 3). The inhibitors were then screened in a biochemical assay at 2 and 0.2 μM concentrations to assess activity.

Table 3:

BRK inhibition by various pyrazolopyrimidine derivatives in the enzymatic assay and in breast cancer cells. Percent inhibition and IC₅₀ values of select ligands were determined in the ADP-Glo assay. BRK inhibitors were tested for their cytotoxicity via MTT assay in the MDA.MB.231 cell line after 72h treatment.

Name	R1	R2	Biochemical assay			Cellular assay MDA.MB.231
			% Inhibition		IC50 (nM)	% Inhibition 10 μM
			2 μM	0.2 μM
4f	-	-	100	86	25.73 ± 5.05	31.40
38	-CH3		100	91	11.12 ± 10.42	26.00
39			100	100	--	83.98
40			100	100	--	52.99
41			100	100	--	82.01
42			100	87	26.79 ± 4.11	63.01
43			100	25	-	12.94
44			100	22	-	-
45			100	28	-	54.32
46	-CH3		47	-	-	10.31
47			49	-	-	-6.94
48			58	-	-	6.33
49			100	62	-	14.92
50			100	93	8.20 ± 4.04	18.28
51			100	87	3.37 ± 2.19	32.50
52			100	92	30.00 ± 25.15	14.19
53			68	-	-	5.22

Initially, we performed a single point assay at 2 μM to evaluate potency. At this concentration, substitutions at the N-position of the pyrazole ring (R₁) did not affect potency. However, at lower concentrations (0.2 μM), alkyl substitution exhibited improved inhibition compared to amine substitution.

We then turned our focus to modifying the phenolic substitution at the meta-position as phenols are known Phase II conjugation sites(29) and are therefore metabolically unstable. Our goal was to replace the phenol with a bioisostere that preserves pharmacodynamic properties while enhancing metabolic stability. Toward this objective, we first replaced the phenol with a pyrazole moiety (46) but this modification resulted in reduced activity. Subsequently, we replaced the phenolic group with various substituents such as trifluoromethoxy benzene (47), benzodioxole (48), indole (49), and difluoromethyl benzene (53). However, all modifications exhibited a reduction in activity suggesting that the phenolic OH group at the meta-position is essential for the activity. This result is consistent with results observed for pyrimido-indole inhibitors in Table 1.

Since phenolic substitution was detrimental to activity, we focused on modifying substitution within the phenol group, specifically exploring ortho-, meta-, and para-substituted phenols (compounds 50, 51, and 52). Beyond Phase II conjugation, cytochrome P450s (CYP450) are known to oxidize phenols at the para-position.(30–32) Metabolites containing para-oxidized phenols can undergo further oxidation, forming quinone mutagens. Since α-carbolines, such as those in 4f, are known to be mutagenic, we aimed to prevent the bioactivation of the pyrazolopyrimidine scaffold into a mutagenic metabolite. To achieve this, we introduced fluorine into the phenolic system to reduce the likelihood of oxidative metabolism. Additionally, fluorine substitution is a well-established strategy for enhancing ligand binding. Therefore, our approach was designed to simultaneously improve both metabolic stability and ligand efficiency of the inhibitors. IC₅₀ values for various fluorine substituted pyrazolo-pyrimidine derivatives were determined, leading to identification of 51 as a potent BRK inhibitor (IC₅₀ = 3.37 ± 2.19 nM) with para-fluorine substitution on the phenolic system.

Kinase Selectivity Profile

Small-molecule kinase inhibitors are often associated with kinase promiscuity. Indeed, the Protein Data Bank (PDB) contains 471 crystal structures that feature pyrazolopyrimidine derivatives as kinase inhibitors, with 50% of these structures featuring 1H-pyrazolo[3,4-d]pyrimidines.(33) Therefore, it is crucial to determine the selectivity to ensure that the biological activity is due to BRK inhibition rather than off-target effects. To assess the selectivity of our lead compound, 51, the compound was screened at a concentration of 30 nM using the Eurofins KINOMEscan^® panel (Figure 4A). The compound demonstrated excellent selectivity, with an S(35) value of 0.012 at 30 nM, indicating that only 1.2% of the 468 kinases tested were significantly inhibited at this concentration. Furthermore, we determined the dissociation constant (K_d) value of the lead compound against BRK, which was found to be 44 nM (Figure 4B). The low Kd value indicates a strong binding affinity of 51 further supporting its potential as a selective and potent BRK inhibitor.

Figure 4:

A) Kinome-wide selectivity profiling of 51 (30 nM) against 468 kinase targets. TREEspot^™ Interaction Maps of 51 are presented. Red circles indicate kinases that were inhibited >70% by 51. B. Binding affinity (Kd) of 51 against BRK.

Cell Viability Assays – Pyrazolo-pyrimidine derivatives have no significant effect on cellular viability

Following the discovery of novel pyrazolo-pyrimidine BRK inhibitors and confirming the selectivity of the lead compound (51), we sought to evaluate the effect of the synthesized compounds on cell viability in the TNBC cell line MDA-MB-231 cell line. We found that the inhibitors had no significant effect on cell viability at a concentration of 10 μM (Table 3). The results indicate a lack of correlation between the biochemical inhibition of BRK and cell viability suggesting BRK inhibition is not cytotoxic. This is supported by other research groups that have shown that BRK inhibition is not particularly cytotoxic and has minimal effects on cell viability.(20, 21, 23) Furthermore, in the case where BRK inhibition is cytotoxic,(21) the reduction in cell viability could be attributed to off-target effects of the inhibitors. The inhibitors were screened in other TNBC and non-TNBC cell lines with no significant effect on cell viability suggesting the inhibitors have limited off-target effects (Table S1).

To further evaluate the anti-proliferative effects of the inhibitors, we performed a colony formation assay in the TNBC cell lines MDA.MB.231, BT-20, and HCC1806 (Figure 5 and S2). The compounds inhibited colony formation in a dose-dependent manner in MDA.MB.231 and BT-20 cells and at 10 μM in HCC1806. The results indicate that BRK inhibition may not be an effective monotherapy. Interestingly, previous studies indicate that in vitro cell proliferation is independent of BRK kinase activity. For instance, Lange et al. reported no significant difference in proliferation between cells expressing constitutively active BRK and those expressing kinase-dead BRK.(34) Furthermore, in vivo studies showed no significant difference in the tumor volume between NGS mice injected into the mammary gland with BRK-overexpressing cells and those with BRK-knockout cells.(35) Notably, mice injected with BRK-overexpressing cells exhibited a 2- to 3-fold increase in metastatic burden to the lungs compared to those injected with BRK-knockout cells. These findings suggest that while BRK inhibition does not significantly affect cell viability, it plays a critical role in suppressing metastasis.

Figure 5:

BRK inhibitors reduce the colony formation in TNBC cells. MDA.MB.21 and BT-20 cells were treated with varying concentrations of compounds 4f and 51, while HCC1806 was treated with a single concentration of 10 μM.

These findings are further supported by our viability studies with the BRK control compound tilfrinib (4f), which demonstrated minimal cytotoxicity, reinforcing the notion that BRK inhibition does not directly impact proliferation. Instead, BRK appears to be more closely linked to the metastatic process, making it a key target for disrupting cancer dissemination. By focusing on metastasis rather than tumor growth, BRK inhibition may represent a distinct and valuable approach to limiting cancer progression.

Effect of BRK Inhibitors on Cell Migration and Invasion

TNBC is an aggressive form of breast cancer, with its high metastatic potential being the primary cause of breast cancer-related mortality. Several studies have shown that high expression of BRK is associated with poor prognosis in TNBC.(11, 36) In vitro and in vivo studies demonstrate that constitutively active BRK increases migration and invasion but does not significantly alter baseline cell viability.(1, 35) To assess the anti-metastatic potential of the lead compound 51, we performed wound healing and cell invasion assays in BT-20 and MDA.MB.231 cells. We tested 51 in the wound healing assay to evaluate its effect on cell migration, and the results show that it effectively suppressed the migration of both cell lines, emphasizing its potential as an anti-metastatic agent (Figure 6A and 6B). To further support the anti-metastatic properties of compound 51, we performed an invasion assay in the highly invasive TNBC cell line MDA.MB.231. Treating the cells with 51 at concentrations 10 μM and 20 μM (Figure 6C) resulted in significant reduction in cell invasion compared to the control (p<0.0001).

Figure 6:

Compound 51 inhibits migration and invasion of TNBC cells. A) Wound healing assay in MDA.MB.231 cells treated with compound 51 at 10 μM and 20 μM for 24h and 48h showed a reduction in wound healing. B) Wound healing assay in BT-20 cells treated with compound 51 at 10 μM for 48h showed a reduction in wound healing. The blue vertical dashed lines in the images define the open wound areas. C) Invasion assay in MDA.MB.231 cells treated with compound 51 at 10 μM and 20 μM for 24h showed a reduction in cell invasion. A one-way ANOVA with Dunnett’s multiple comparison test was performed to compare the results to the DMSO control. ****p<0.0001, n = 3. D) Western blot in MDA.MB.231 cells treated with compound 51 at 10 μM for 24h results in reduced MMP-9 protein levels compared to the DMSO control.

Matrix metalloproteinase-9 (MMP-9) is known to play a crucial role in extracellular matrix remodeling and the promotion of metastasis.(37, 38) Furthermore, MMP-9 has been implicated in invasiveness of TNBC cells.(39–41) For instance, a study reported by Mehner et al. demonstrated that MMP9 knockdown in MDA.MB.231 cells completely blocked pulmonary metastasis in vivo compared to the control.(42) Therefore, investigating the effect of compound 51 on MMP-9 expression provides a valuable insight into its potential anti-metastatic properties. To this end, we performed a Western blot to compare the MMP-9 protein expression in treated and untreated cells. Reduction in MMP-9 protein levels in cells treated with 51, compared to the control, likely explains the reduction in invasion seen in MDA.MB.231 cells. The reduction in both wound closure and invasion was achieved without significant cytotoxicity, supporting the hypothesis that kinase-active BRK is primarily a driver of metastasis. Additionally, BRK has been implicated in conferring resistance to chemotherapy in several cell lines including MCF-7 and T47D cells.(12, 43) When combined with traditional chemotherapies, BRK inhibitors could help overcome BRK-mediated resistance, enhance efficacy of chemotherapies, and reduce the likelihood of metastatic relapse. Therefore, by specifically targeting BRK-driven metastatic pathways, BRK inhibitors such as 51 can be used as an adjunct therapy to combat the aggressive spread of TNBC.(44)

Effect of BRK Inhibitors on Pro-migratory Signaling Pathways

BRK has been demonstrated the ability to facilitate migration and metastasis of tumor cells, in part through downstream activation of AKT contributing to a migratory phenotype.(45) To further elucidate the effects of BRK inhibition, we performed western blot analysis to understand how BRK inhibitors impact downstream targets. BRK has been shown to directly phosphorylate key signaling molecules, including AKT, thereby promoting migration and metastasis.(8, 46, 47) Preliminary western blot data suggests that BRK ligands (40, 51, and 52) can suppress AKT phosphorylation compared to the control in BT-20 cells (Figure S3). Paradoxically, tilfrinib (4f) exhibited an increase in AKT phosphorylation suggesting the inhibitor may be non-selective with activity against multiple cellular targets.

To evaluate the impact of BRK inhibition on pro-migratory signaling pathways, particularly the AKT pathway, we stimulated MDA.MB.231 and BT-20 cells with IGF-1, a growth factor that activates PI3K/AKT signaling and enhances BRK tyrosine phosphorylation(48) and is upregulated in the breast cancer microenvironment.(49–51) As shown in Figure 7, compound 51 was able to reduce IGF-1 induced AKT phosphorylation as compared to control in both TNBC cells. By demonstrating that our compound can inhibit BRK even under conditions of IGF-1 stimulation, we provide strong evidence of its anti-metastatic potential, highlighting its potential use in reducing metastasis in aggressive cancers like TNBC. Furthermore, the inhibitors provide anti-metastatic properties at non-cytotoxic concentrations, which is desirable when used in combination with chemotherapy.

Figure 7:

Compound 51 reduces Akt phosphorylation in TNBC cells. Cells were treated with 10 μM of compounds 51 and 4f for 6h and 24h in IGF-1 stimulated MDA.MB.231 cells (A) and BT-20 cells (B). The cells were treated with 100 ng/ml IGF-1 for 30 minutes.

AKT, a key signaling molecule in PI3K/AKT pathway, not only regulates cell survival and proliferation but also motility and EMT.(52–54) AKT activation promotes cytoskeletal rearrangements and increased cell motility facilitating migration and invasion.(55) Additionally, AKT drives the process of EMT enabling cancer cells to disseminate from the primary tumor.(53, 56) By reducing AKT phosphorylation when stimulated by IGF-1 in a pro-migratory environment, BRK inhibition disrupts these critical pro-metastatic pathways, thereby impairing migratory and invasion phenotypes. This is consistent with our findings in wound healing and invasion assays demonstrating that BRK inhibition with compound 51 suppresses both migration and invasion in TNBC cells.

Molecular modeling of compound 51 in BRK

Pyrazolopyrimidine-based BRK inhibitors are predicted to be Type I inhibitors by binding the active DFG-in conformation of the kinase. Because compound 51 was found to be the most potent inhibitor, we used this compound for molecular modeling studies. Compound 51 was docked into the active crystal structure of BRK (PDB ID: 5H2U). The modeling study suggests that the nitrogen in the pyrimidine ring forms a hydrogen bond with Met267 and the amine forms a hydrogen bond with Glu265 at the hinge region (Figure 8). The phenol ring occupies the back pocket while the methyl group orients towards solvent front. The benzene ring in the phenol moiety forms a pi-cation interaction with catalytic Lys219 and the phenolic OH forms a hydrogen bond with Asp330 of the DFG motif.

Figure 8:

Binding pose of 51 docked into BRK (PDB ID: 5H2U). At the hinge region, the pyrimidine -NH₂ forms a hydrogen bond with Glu265, and the pyrimidine ring forms a hydrogen bond with Met267. The phenolic hydroxy group forms a hydrogen bond with Asp330, and the benzene ring is involved in a pi-cation interaction with Lys219.

Conclusion

This study presents the discovery and optimization of novel 1H-pyrazolo[3,4-d]pyrimidine derivatives as selective inhibitors of BRK/PTK6. These compounds exhibit potent biochemical inhibition of BRK with minimal cytotoxicity, a key distinction given BRK’s emerging role in promoting metastasis rather than proliferation. Through systematic structure–activity relationship studies, we identified compound 51 as a lead candidate with nanomolar potency, kinase selectivity, and significant anti-migratory and anti-invasive activity in TNBC cell lines.

Compound 51 effectively suppressed wound closure and reduced cell invasion in vitro, while maintaining a non-cytotoxic profile at therapeutic concentrations. It also lowered MMP-9 expression and inhibited AKT phosphorylation following IGF-1 stimulation, reinforcing its impact on metastatic signaling pathways rather than cell viability. These results align with the growing consensus that BRK primarily facilitates metastatic progression and demonstrate that compound 51 can selectively disrupt this function.

Given the role of BRK in driving EMT, chemoresistance, and metastatic dissemination, inhibitors like 51 offer a compelling therapeutic strategy—particularly in combination with cytotoxic agents where BRK activity undermines chemotherapy efficacy. By selectively targeting BRK’s pro-metastatic functions while preserving cell viability, these inhibitors represent a new class of metastasis-focused therapies that may improve outcomes in aggressive cancers such as TNBC. Future in vivo studies are warranted to validate the anti-metastatic efficacy of compound 51 and to explore its utility in combination regimens.