Cancer continues to be a predominant global health concern, resulting in over 10 million fatalities in 2020, accounting for approximately one-sixth of all deaths globally (Sestilo and Rapposelli, 2019; World Health Organization, 2025). Despite advancements in precision medicine and immunotherapies, cancer persists in demonstrating considerable variability and adaptability (Fruman et al., 2017; Gillet et al., 2013). Variation in mutations, cell origins, and environments gives rise to many cancer subtypes. Characteristics encompass unregulated proliferation, evasion of apoptosis, invasion, metastasis, and resistance to pharmacological agents (Liu et al., 2015; Tandel et al., 2019), frequently associated with the dysregulation of protein kinases (Fruman et al., 2017).

In the past 20 years, protein kinases have become key drug targets for discovery and therapeutic innovation. Small-molecule kinase inhibitors are among the fastest-growing therapies, reflecting their critical function in cellular regulation. By 2025, over 85 kinase inhibitors are expected to gain the U.S. Food and Drug Administration approval (Roskoski R., 2025), with 42 launched from 2018 to 2023, emphasizing their clinical value (Bhullar et al., 2018; Busschots et al., 2012; Latham et al., 2024). ATP-competitive inhibitors targeting Breakpoint Cluster Region-Abelson (BCR-ABL), Epidermal Growth Factor Receptor (EGFR), and the B version of the Raf kinase family discovered in a rapidly accelerated fibrosarcoma (BRAF) have had a significant clinical impact (Anaya et al., 2025). Yet resistance, toxicities, and varied efficacy highlight the need for new kinase targets (Jiao et al., 2018).

The PDPK1 gene encodes 3-phosphoinositide-dependent protein kinase-1 (PDK1), a key serine/threonine kinase in the PI3K/AKT signaling pathway (Chan and Tsichlis, 2001; Jiang et al., 2021). PDK1 acts as a “master kinase,” adding phosphate groups to the activation-loop (T-loop) residues of many AGC family members (Bayascas, 2008; Jiang et al., 2021). When class I PI3Ks are activated, PIP₂ is converted to PIP₃ at the membrane (Falasca and Fyffe, 2013; Maurer et al., 2009). This brings PDK1 and AKT together via their PH domains, enabling PDK1 to phosphorylate AKT at Thr-308 to activate it (with Ser-473 phosphorylated by the mTORC2 complex, which supports cell growth signaling).

Besides AKT, PDK1 also activates other members of the AGC kinase family, a group of related enzymes involved in cell growth, survival, metabolism, and motility. The activity of PDK1 is tightly controlled (Yang et al., 2022). For example, when S6K1 (an AGC kinase, downstream of mTORC1) adds a phosphate to the PH domain of PDK1, it promotes binding to 14-3-3 proteins (a regulatory protein), leading to PDK1 pairing with itself (dimerization) and moving away from cell membrane phospholipid (PIP₃), which reduces AKT activation (Jiang et al., 2022). This type of feedback regulates AKT activities during stimulation by another enzyme complex, mTORC1 (Chan and Tsichlis, 2001; Jiang et al., 2021). Figure 2 shows regulators in the PI3K/Akt pathway, with arrows indicating activation (turning on) or inhibition (turning off). Akt signaling promotes key processes in cancer, such as tumor cell growth, survival, proliferation, changes in immune response, metabolism, and angiogenesis (He et al., 2021).

In cancer, dysregulation of PDK1 signaling is recognized as a cause of oncogenesis (Feldman et al., 2005; Zheng et al., 2023). Research on breast cancer indicates that the gene for PDPK1 is often found in a higher amplified or overexpressed form, frequently alongside upstream oncogenic modifications within the PI3K pathway (Wang et al., 2024). These encompass activating mutations in PIK3CA, the loss of the tumor suppressor PTEN, and the amplification or overexpression of ERBB2/HER2, all of which culminate in increased PDK1 signaling output (Medina, 2013; Park et al., 2023; Piri et al., 2024).

Increased PDK1 expression augments AKT activation, stimulates cellular proliferation, heightens tumor aggressiveness, and fosters resistance to PI3K-targeted therapies (Yang et al., 2017; Yang et al., 2019). Persistent hyperactivation of PDK1, particularly by prolonged phosphorylation at Ser241, is commonly detected in various cancer types and correlates with unfavorable clinical outcomes and aggressive disease characteristics (Bamodu et al., 2020). The genetic deletion of PDK1 in PTEN-deficient tumor models significantly inhibits tumor growth, highlighting its pivotal function as a therapeutic target. Aberrant PI3K–AKT–mTOR signaling, instigated by PTEN loss, activating PI3K mutations, or hyperactive receptor tyrosine kinases (RTKs), constitutes a pivotal oncogenic occurrence, with PDK1 serving as a crucial mediator for relaying oncogenic signals to downstream AGC kinases (Yang et al., 2017; Yang et al., 2019).

In accordance with this model, genetic investigations reveal that a partial deficiency of PDK1 substantially inhibits tumorigenesis in PTEN-deficient environments (Chen et al., 2017; Haddadi et al., 2018), whereas diminished PDK1 expression in PTEN^+/− mice significantly postpones spontaneous tumor formation (Bayascas et al., 2005). In Braf^V600E; Cdkn2a^−/− melanoma models, the deletion of PDK1 significantly hinders tumor growth and metastatic advancement, despite the presence of functional PTEN signaling (Scortegagna et al., 2015). In cerebral tissues, the suppression of PDK1 reduces the hypertrophic reaction induced by PTEN deficiency, thereby underscoring the context-specific carcinogenic functions of PDK1 (Chalhoub et al., 2009).

Clinically, heightened PDK1 expression and sustained phosphorylation at Ser241 correlate with unfavorable prognosis in patients with breast, prostate, and glioblastoma cancers, highlighting the importance of PDK1 activation status as both a biomarker for disease severity and a potential predictor of treatment efficacy (Hossen et al., 2015).

PDK1 is crucial in carcinogenesis by enabling tumor initiation and progression through enhancing cell motility, invasion, and adaptation to unfavorable microenvironmental conditions, such as hypoxia and nutrient scarcity (Emmanouilidi and Falasca, 2017; Sestito et al., 2018). Comprehensive experimental evidence indicates that PDK1 governs various stages of the metastatic cascade, encompassing invadopodia development, increased matrix metalloproteinase activity, and dynamic reorganization of the actin cytoskeleton (Baldi et al., 2025; Xie et al., 2006). These activities are facilitated by both kinase-dependent and kinase-independent methods, involving downstream effectors including ROCK1, MRCKα, and PLCγ1 (Garcia-Viloca et al., 2022). In ovarian cancer, PDK1 overexpression is significantly correlated with metastatic spread, chemoresistance, and reduced patient survival (Siu et al., 2020). PDK1 mechanistically promotes cancer cell adhesion, invasion, and angiogenesis by activating the α₅β₁ integrin axis and the downstream JNK/IL-8 signaling pathways (Siu et al., 2020). The findings jointly designate PDK1 as a crucial regulator of tumorigenesis, metastatic potential, and therapeutic resistance, underscoring its importance as a primary target for anti-cancer interventions (Wang et al., 2024).

The justification for focusing on PDK1 is strengthened by the inherent heterogeneity and adaptive potential of cancer cells. Notwithstanding considerable progress in precision oncology, targeted treatments, and immunotherapies, therapeutic resistance persists, fueled by genetic heterogeneity and the constant reconfiguration of intracellular signaling networks (Fruman et al., 2017; Gillet et al., 2013). Deviant kinase signaling perpetuates fundamental characteristics of cancer, including unregulated growth, resilience to metabolic and genotoxic stress, invasion, and metastatic spread (Firnau and Brieger, 2022; Manning and Dibble, 2024). PDK1, as a principal regulator of AGC kinases, serves a pivotal role in these signaling networks, with its dysregulation directly linked to tumor growth and resistance to therapy (Zheng et al., 2023). Thus, PDK1 serves as a compelling and mechanistically validated target for the development of anticancer therapeutics (Emmanouilidi and Falasca, 2017).

Early attempts to block PDK1 with a drug faced significant challenges. Early compounds such as BX-320, BX-795, and BX-912, lacked selectivity in cellular contexts (Dangelmaier et al., 2014; Knight, 2011). A deeper understanding of PDK1’s structure and the new drug design method led to the development of more selective inhibitors, such as GSK2334470, which demonstrates sub-10 nM potency and increased selectivity (Najafov et al., 2011). New approaches that block regions such as the PIF pocket have also been developed. These offer ways to stop PDK1 activity beyond simply blocking its ATP site (Schulze et al., 2016; Wilhelm et al., 2012). These new inhibitor types, along with combination therapies and biomarker use, are revitalizing interest in PDK1 inhibitors for cancer treatment (Chandrasekhar et al., 2018; Schulze et al., 2016; Wilhelm et al., 2012).

Recent evidence underscores the potential benefits and obstacles linked to the targeting of PDK1 (Park et al., 2023). The structural similarities of kinase ATP-binding sites, the context-dependent effects of PDK1 inhibition, and compensatory signaling systems undermine the effectiveness of single-agent inhibitors. The area is progressively transitioning towards combination therapy strategies, the creation of multitarget ligands, and biomarker-driven patient selection methods to address adaptive resistance and enhance clinical outcomes (Sadeghian et al., 2025). Beyond oncology, PDK1 inhibition is being explored for potential applications in allergic conditions and antifungal drug development, indicating its broader therapeutic potential (Sestito and Rapposelli, 2019). Taken together, these recent advancements affirm PDK1’s pivotal role in tumor initiation, progression, and resistance to therapy, underlining its promise as a drug target. This review consolidates existing knowledge of PDK1’s biological functions and offers an updated perspective on its inhibition through innovative medicinal chemistry strategies, including the development of ATP-competitive, allosteric, and substrate-based inhibitors, as well as techniques targeting protein–protein interactions. Additionally, it emphasizes the integration of high-throughput screening, structure-based drug design, polypharmacology, combination therapies, and predictive biomarker strategies, guiding the clinical translation of PDK1-targeted agents and informing the rationale for designing next-generation inhibitors in precision oncology.

PDK1’s multi-domain structure is central to AGC kinase signaling (Sadeghian et al., 2025; Xu et al., 2019). PDK1 contains a bilobal catalytic domain and a C-terminal PH domain. This pH domain binds membranes via phosphoinositides, particularly phosphatidylinositol 3, 4, 5-trisphosphate (PIP₃) (Barile et al., 2012). A conserved PIF (PDK1-Interacting Fragment) pocket connects these domains, docks site substrates, and recognizes AGC kinases (Schulze et al., 2016). This architecture integrates kinase structure with signaling. The glycine-rich loop and Asp205 enhance flexibility, facilitating domain transitions and the dynamic conformation of PDK1 (Xu et al., 2024). These characteristics combined affect activation and regulation. Activation Mechanisms and Regulation of Conformational.

The activity of PDK1 is regulated by a comprehensive framework that includes interactions with the autoinhibitory domain, phosphoinositide binding, and phosphorylation of the activation loop, which collectively influence kinase conformation, localization, and signaling output (Meng et al., 2025). The involvement of PIP₃-rich membranes alleviates autoinhibition, facilitates PDK1 dimerization and trans-autophosphorylation at Ser241, and dynamically modifies substrate and adaptor contacts to enhance AGC kinase signaling (Shaw and Burke, 2025). Prolonged lipid-dependent activation and continuous Ser241 phosphorylation, rather than mere PDPK1 expression, characterize pathological PDK1 addiction in cancer and establish a molecular foundation for its prognostic and therapeutic significance (Levina et al., 2022; Sacerdoti et al., 2023; Shaw and Burke, 2025; Xu et al., 2024).

PDK1 activity is rigorously regulated by an inherent autoinhibitory mechanism, wherein interactions among the pleckstrin homology (PH) domain, the kinase domain, and the intervening regulatory linker maintain an inactive cytosolic conformation and restrict substrate accessibility. Structural and computational analyses demonstrate that this closed conformation inhibits productive PDK1 dimerization and restrains activation-loop phosphorylation until suitable upstream signals are activated, thereby linking structural rearrangement to functional activation instead of allowing constitutive signaling (Mulpuri et al., 2025; Park et al., 2023). The phosphorylation of C-terminal regulatory residues, including Ser393, has been shown to reduce autoinhibitory contacts and promote conformational release; nevertheless, these modifications act as modulators rather than primary activators (Xu et al., 2024).

The binding of phosphoinositides formed by PI3K—mainly phosphatidylinositol-(3,4,5)-trisphosphate (PIP₃) and PI(3,4)P₂ to the PH domain facilitates the recruitment of PDK1 to the plasma membrane, inducing a change from an autoinhibited monomer to an active, face-to-face dimeric form (Levina et al., 2022; Xu et al., 2024). The lipid-induced dimerization significantly enhances trans-autophosphorylation and reveals the PIF pocket, promoting effective binding to AGC kinases devoid of PH domains, including S6K, SGK, and RSK (Sacerdoti et al., 2023; Wu et al., 2009). Phosphoinositide binding serves mostly as a signal-amplifying spatial organizer, improving substrate colocalization and route flow instead of directly augmenting intrinsic catalytic turnover (Heras-Martínez et al., 2019).

The structure and conformational changes of PDK1 identify significant targets for drug design, including the residues Lys111, Ala162, Ser160, Met120, and the PIF-pocket, all essential for inhibitor interaction (Sadeghian et al., 2025). These characteristics facilitate the systematic design of ATP-competitive, allosteric, and dual-site inhibitors (Garcia-Viloca et al., 2022; Nagashima et al., 2011; Sadowsky et al., 2011).

The adaptability of the glycine-rich loop and PH domain allows for the targeting of various PDK1 conformations, enhancing inhibitor selectivity and aiding in efficient drug design (Kumar et al., 2025). This conformational plasticity facilitates the creation of compounds that more accurately inhibit PDK1 activity, reducing compensatory signaling and improving target selectivity (Yao et al., 2020). These structural features therefore enhance rational drug design strategies for PDK1 in oncology and other diseases, establishing a direct link to therapeutic applications (Xu et al., 2024).

PDK1 is primarily known for activating AKT by phosphorylating Thr308 (Yang et al., 2022). This modification is essential for full AKT activity (Lu et al., 2010). Activated AKT promotes cell survival, glucose metabolism, and angiogenesis. Cancer frequently exploits these processes (Irazoqui et al., 2023). PDK1 also phosphorylates other AGC kinases, including S6K, SGK, and various PKC isoforms (Levina et al., 2022). This action broadens its influence to cell growth, cell cycle progression, migration, invasion, and metabolic adaptation (Tsai et al., 2022; Wang et al., 2021).

PDK1 is often elevated in cancers with PTEN loss or PI3K hyperactivation, leading to sustained AKT signaling and downstream pathway activation. Genetic models show their essential role in PTEN-loss–driven tumorigenesis. These findings highlight PDK1 as a key component of oncogenic signaling networks (Chen et al., 2017; Gagliardi et al., 2018).

PDK1 directly contributes to tumor progression by facilitating epithelial–mesenchymal transition (EMT), stem cell–like traits, and therapy resistance (Chen et al., 2018; Chen et al., 2017). In breast cancer, elevated PDK1 activity is associated with resistance to trastuzumab and tamoxifen (Wang et al., 2022). In glioblastoma, inhibiting both PDK1 and Aurora-A reduces tumor cell survival. This phenomenon underscores its involvement in adaptive resistance mechanisms (Emmanouilidi and Falasca, 2017; Falasca and Fyffe, 2013). Furthermore, PDK1 facilitates carcinogenic characteristics via AKT-independent pathways (Vasudevan et al., 2009). In BCR–ABL–driven hematological malignancies, pharmacological or genetic suppression of PDK1 diminishes tumor cell survival with negligible alterations in AKT activity, suggesting dependence on alternate downstream effectors. These data indicate that PDK1 maintains cancer cell viability and proliferation via non-AKT AGC kinase signaling pathways. These data collectively affirm the role of PDK1 as a significant promoter of tumor growth, invasion, and adaptive survival in response to therapeutic pressure (Castel et al., 2016; Di Blasio et al., 2017; Emmanouilidi and Falasca, 2017; Manning and Toker, 2017; Raimondi et al., 2012).

Clinically, PDK1 activity is linked to poor outcomes in many cancer types. Elevated phosphorylation at Ser241 correlates with a worse prognosis in breast, prostate, glioblastoma, and pancreatic cancers. This highlights its potential as a prognostic biomarker (Huang et al., 2024). Its role in therapy resistance supports targeting PDK1 in clinical settings. Inhibiting PDK1 may enhance sensitivity to chemotherapy, targeted therapies, or combination regimens. In summary, PDK1’s key role in oncogenic signaling, tumor progression, and drug resistance makes it a significant therapeutic target and predictor of clinical outcomes (Falasca and Fyffe, 2013). These results indicate that PDK1 is a central signaling hub linking PI3K activation to downstream AGC kinases. These kinases are important for tumor growth, metabolism, and treatment resistance. Phosphorylation at Ser241 is a key marker of activation and may indicate how well a cancer will respond to treatment (Han et al., 2023).

ATP-competitive inhibitors bind the conserved ATP-binding pocket of PDK1 and directly compete with ATP at the catalytic site (Jug and Ilaš, 2025). Early compounds such as BX-320, BX-795, and BX-912 showed high potency, with IC₅₀ values ranging from 11 to 30 nM (Feldman et al., 2005; Islam et al., 2007). However, these inhibitors lacked selectivity. They often affected other kinases, including TBK1 and IKKε. This reduced their therapeutic potential and highlighted the challenge of targeting a conserved active site (Knight, 2011; Peifer and Alessi, 2008).

The introduction of GSK2334470 marked a significant advancement. This ATP-competitive inhibitor shows exceptional selectivity (Yang et al., 2017). With an IC₅₀ of about 10 nM, GSK2334470 is roughly 1000-fold more selective across a panel of 93 kinases (Najafov et al., 2011). It effectively inhibits phosphorylation of downstream substrates such as SGK and S6K. Its effects on AKT are context-dependent and reflect the complex substrate-specific interactions of PDK1 (Knight, 2011; Medina, 2013). Despite these improvements, challenges remain. Compensatory signaling pathways and potential toxicity from residual off-target effects continue to pose problems. The situation emphasizes the need for alternative or combination strategies.

A second generation of 7H-pyrrolopyrimidine analogues was made to explore how changing the size of the substituent influences how the PDK1 catalytic site works (O’Brien et al., 2014). These refined analogues showed that they were more effective than first-generation drugs. Compound 9j (4-methoxy) was the most promising lead, showing better kinase inhibition The research underscored the importance of the additional ring nitrogen in the purine-like scaffold, particularly for the most active compounds (9f and 9j), indicating the need for future structure–activity investigations (O'Brien et al., 2014).

Allosteric inhibitors target regions outside the ATP-binding site, including the PIF pocket, which is essential for substrate docking. These compounds deliver greater selectivity and minimize off-target effects compared to ATP-competitive inhibitors (Mingione et al., 2023).

To address pathway redundancy and drug resistance, multi-target strategies have been developed (Haynes and Manogaran, 2025). These strategies inhibit PDK1 along with complementary kinases, such as PI3K or Aurora-A. Pyridonyl compounds targeting the DFG-out conformation have shown efficacy against glioblastoma stem cells (Sestito et al., 2016). Heterocyclic inhibitors from companies like Sunesis and Biogen attack both ATP and PIF sites at the same time. These multi-target agents are efficacious against resistant tumors and frequently function independently of AKT signaling (Daniele et al., 2017; Park et al., 2023).

The multi-target approach reflects a growing understanding that single-target inhibition is often inadequate. The adaptability of cancer signaling networks undermines such approaches. Combining structural and mechanistic knowledge helps scientists come up with smart ways to make powerful, selective next-generation inhibitors. These strategies offer promise for overcoming resistance and improving clinical outcomes (Mulpuri et al., 2025; Xu et al., 2019).

The chemical diversity presented in Table 1 and Figure 4 provide a robust basis for the ongoing enhancement of PDK1-targeted therapeutics and demonstrates the progression of medicinal chemistry approaches in the domains. Initial ATP-competitive inhibitors like BX-795 and BX-912 exhibited significant biochemical potency but were constrained by extensive polypharmacology and inadequate translational applicability, while subsequent compounds such as GSK2334470 attained enhanced selectivity via structure-guided optimization, though they lacked sufficient in vivo efficacy (Jug and Ilaš, 2025). Concurrently, allosteric PIF-pocket inhibitors like RS1 and the dicarboxylate-derived PS210/PS48 family demonstrated the potential for selectively obstructing substrate docking while also exposing ongoing issues concerning permeability and metabolic stability (Busschots et al., 2012). Recent advancements in dual-site inhibitors, such as pyridonyl and hybrid chemotypes, combine ATP-site interaction with PIF-pocket inhibition to diminish both catalytic function and compensatory substrate acquisition (Jug and Ilaš, 2025). The scaffold classes collectively illustrate that a singular binding mode is inadequate for completely inhibiting PDK1 signaling in various biological contexts; instead, the comparative insights from Table 1 underscore the necessity for mechanism-informed, multi-modal design strategies that harmonize potency, selectivity, and drug-like characteristics to address pathway redundancy and enhance clinical translatability (Mansi et al., 2020).

The structural and mechanistic diversity of PDK1 inhibitors demonstrates the advancement of multi-target design strategies (Arter et al., 2022). Figure 4 Representative inhibitors bind to various regulatory pockets, including the ATP-binding site, PIF pocket, PH domain, and DFG-out conformation, showing the viability of orthosteric, allosteric, and dual-site targeting. ATP-competitive ligands like GSK2334470 set standards for potency (Castel et al., 2016). Molecules such as 2-O-Bn-InsP₅ and RS1 demonstrate that non-ATP binding domains can achieve selectivity and modulated protein interactions (Barile et al., 2012). DFG-out inhibitors and hybrid scaffolds exemplify new “dual-geometry” strategies that stabilize inactive conformations by interacting with accessory domains (Xu et al., 2024). These approaches could overcome pathway redundancy and drug resistance. Together, these structures highlight how rational polypharmacology through multi-target engagement or hybrid binding broadens the therapeutic landscape for PDK1 inhibition beyond traditional single-site models.

Structure–activity relationship (SAR) studies of BX-series compounds (BX-320, BX-795, BX-912, BX-517) illustrate that core-molecule design is key to overcoming these issues (Abeyrathna and Su, 2015). Specifically, all inhibitors in this series share an N-phenylpyrimidin-2-amine core and, with 5-urea groups, form critical hydrogen bonds with Lys111 and Thr222. Notably, BX-517 is more potent due to stronger interactions with these residues (Feldman et al., 2005).

Subsequent systematic modifications of BX-517, including pyrrole C-4′ substitutions, generated derivatives (7b, 7d, 16, 17) with enhanced solubility and improved pharmacokinetics. These changes also led to an extended half-life. Nevertheless, some compounds continued to display limited microsomal stability. This incident illustrates the trade-offs involved in optimizing multiple pharmacological properties at once (Islam et al., 2007b).

Beyond the BX-series, other compound classes use distinct structural features for better selectivity and stability. For example, Sunesis’ pyridinyl compounds carry a difluorobenzyl carboxamide group, which enhances interactions with Ala162 and Ser160. Similarly, quinazolin-12-one compounds, such as 3f, cross the blood–brain barrier, bind less to plasma protein, and show low toxicity. Together, these features demonstrate how rational scaffold design can optimize drug properties (Hossen et al., 2015; Sestito et al., 2016).

Achieving selectivity for PDK1 inhibitors remains a major challenge. Molecules must distinguish among similar kinase active sites while minimizing off-target effects. BX-series analogs are effective but are linked to higher toxicity in vitro. In contrast, quinazolinones and sulfonamides such as RS1 offer better selectivity and oral bioavailability (Rettenmaier et al., 2014; Xu et al., 2019).

In addition to selectivity, optimizing pharmacokinetics requires detailed ADMET profiling. For example, while carboxylate compounds like PS210 need ester prodrugs to improve membrane permeability, these prodrugs are often rapidly hydrolyzed, thereby reducing plasma half-life (Wilhelm et al., 2012). In contrast, sulfonamides have favorable ionization properties, which aid absorption and systemic exposure. However, dicarboxylates remain problematic due to poor stability and short plasma residence times (Sadeghian et al., 2025; Wang et al., 2024). Thus, medicinal chemistry must optimize potency, metabolic stability, solubility, permeability, and in vivo durability to create clinically relevant candidates.

In the development of PDK1 inhibitors, both efficacy and safety are very important. Lipinski, Veber, Ghose, and Brenk’s classical guidelines examine molecular size, hydrogen bonding, lipophilicity, and structural alerts associated with safety risks. Modern design strategies, such as altering the core structure, targeting multiple sites, and using AI, are accelerating the search for better inhibitors. Inhibiting PDK1 at both the ATP-binding site and the PIF pocket simultaneously increases potency and reduces off-target toxicity. This dual blockade stabilizes inactive conformations and disrupts substrate docking, making the pathway more selective and safer for drug targeting (Chen et al., 2017; Irazoqui et al., 2023; Peifer and Alessi, 2008). Table 2 presents the main medicinal chemistry parameters and (SAR) trends that have led to improvements in PDK1 inhibitors. It emphasizes the need to balance molecular weight, polarity, lipophilicity, and ionization to achieve optimal permeability, selectivity, and metabolic stability. These principles are crucial for converting potent scaffolds into clinically viable, safe drug candidates.

Multiple therapeutically relevant signaling networks converge on phosphoinositide-dependent kinase 1 (PDK1), positioning it as a pivotal target in cancer treatment. PDK1 serves as a vital signal integrator in the PI3K–AKT–mTOR pathway, operating downstream of phosphoinositide production. The aberrant activation of this axis—caused by PIK3CA mutations, PTEN loss, or hyperactive receptor tyrosine kinases—constitutes one of the most prevalent oncogenic occurrences across various malignancies, and PDK1 inhibition provides a method to diminish pathway output regardless of specific upstream alterations (Arsenic, 2014; Castel et al., 2016; Maurer et al., 2009). Recent findings indicate that PDK1 dependence signifies pathway flux rather than static genetic alterations, thereby endorsing its targeting in PI3K-addicted cancers that exhibit resistance to AKT-centric therapies (Peng et al., 2025).

In addition to traditional PI3K signaling, PDK1 interacts with the RAS–MAPK pathway, especially in tumors with RAS and BRAF mutations. Genetic or pharmacological inhibition of PDK1 hinders tumor development, invasion, and metastasis in melanoma and pancreatic cancer models, highlighting its essential involvement even in MAPK-dominant environments (Ferro, 2014). This functional crosstalk is supported by findings that PDK1 maintains ERK-driven oncogenic activity via AKT-independent AGC kinases, such as RSK and SGK, thus facilitating signaling plasticity and adaptive resistance (Castel et al., 2016).

PDK1 regulates oncogenic transcriptional programs, particularly through the PDK1–PLK1–MYC axis, which stabilizes MYC and maintains proliferative and metabolic characteristics linked to aggressive illness and therapeutic resistance (Tan et al., 2013). Furthermore, PDK1 signaling influences not only tumor-intrinsic effects but also modulates the tumor microenvironment by regulating endothelial function, immune cell activation, and inflammatory signaling, with emerging evidence indicating its roles in immune metabolism and response to immunotherapy (Balkwill et al., 2012; Quail and Joyce, 2013). These actions collectively establish PDK1 as a systems-level regulator, whose targeting may provide distinct therapeutic advantages across genetically and microenvironmentally varied malignancies, as can be seen in Figure 3 (Gagliardi et al., 2018).

Preclinical studies provide compelling evidence for the therapeutic potential of PDK1 inhibition in cancer. In PTEN-deficient tumors, PDK1 inhibition prevents AKT activation and hinders anchorage-independent growth (Lu et al., 2010). These are characteristics typical of aggressive cancer phenotypes (Keenan Natesan et al., 2023). In breast cancer, PKD1 overexpression is linked to resistance to trastuzumab, tamoxifen, and PI3K inhibitors. Pharmacological inhibition of PDK1 restores sensitivity to these treatments and decreases tumor cell viability (Emmanouilidi and Falasca, 2017).

Furthermore, PDK1’s function encompasses neuroblastoma, where its inhibition, alongside AKT inhibitors like MK-2206, amplifies cytotoxicity, highlighting PDK1’s involvement in survival signaling pathways. In glioblastoma, the concurrent inhibition of PDK1 and Aurora-A yields superior antiproliferative effects compared to monotherapy This study illustrates the prospective advantages of concurrently targeting many signaling pathways to surmount compensatory mechanisms and improve therapeutic efficacy (Sestito and Rapposelli, 2019).

The development of combination and biomarker-driven therapies represents a pivotal strategy for PDK1-targeted interventions. PDK1 inhibitors have been found to synergize with various agents, including ATP-competitive AKT inhibitors, mTOR inhibitors, and PI3K inhibitors, thus suppressing oncogenic signaling and mitigating the activation of compensatory pathways. Recent research efforts have focused on dual targeting the PDK1 and SGK1 pathways to address challenges posed by PI3K-resistant tumors, as outlined in patent literature (Najafov et al., 2011).

Biomarkers such as phosphorylated Ser241 (p-Ser241), which indicate PDK1 activation, may help identify patient populations that would benefit from PDK1-targeted therapies. Furthermore, inhibiting PDK1 has been shown to sensitize cancer stem cells to standard treatments, potentially reducing cancer recurrence and metastasis (Nagashima et al., 2011; Park et al., 2023). These findings underscore the need to integrate PDK1 inhibitors into biomarker-driven, rational combination regimens to maximize therapeutic benefits.

In conclusion, summarized data across diverse cancer models (Table 3) indicate that both pharmacological and genetic inhibition of PDK1 attenuates malignant phenotypes. Specifically, in PTEN-deficient breast cancer models, PDK1 inhibitors (e.g., BX-320, GSK2334470) have been shown to reduce anchorage-independent growth and induce apoptosis while concurrently decreasing p-AKT (Thr308) and p-S6K signaling pathways. In neuroblastoma, PDK1 inhibition not only reduces cell viability but also resensitizes cells to AKT inhibitors, underscoring its role in overcoming therapeutic resistance. Furthermore, in glioblastoma stem-like cells, DFG-out PDK1 inhibitors (e.g., SA16) diminish neurosphere formation and invasion, particularly in combination regimens, highlighting the promising translational potential of targeting PDK1 in aggressive and resistant tumor contexts.

PDK1 inhibition is currently being investigated beyond oncology, expanding its therapeutic applications. Research indicates that siRNA-mediated silencing of PDK1 can effectively diminish responses in ocular inflammation and allergy models, pointing towards PDK1’s role in immune regulation (Feldman et al., 2005). The structural distinctions observed between human and fungal PDK1 offer potential for creating selective antifungal drugs that minimize toxicity to host cells. Additionally, CNS-penetrant PDK1 inhibitors, particularly quinazolinone derivatives, present therapeutic potential for brain tumors and possibly neurodegenerative diseases by modulating PDK1-dependent survival pathways (Sestito and Rapposelli, 2019). These developments underscore the PDK1 utility as a pharmacological target across a range of diseases.

Furthermore, Table 4 furnishes a comprehensive overview of various PDK1 inhibitors explored in both preclinical and clinical realms, detailing their chemical classifications, mechanisms of action, and stages of development. It reveals that the majority of these compounds are still in preclinical optimization, with only a limited number progressing to early translational or combination studies, highlighting ongoing challenges in achieving selectivity and favorable pharmacokinetic properties. The table encapsulates the progression from early ATP-competitive inhibitors to contemporary dual-site and allosteric modulators, thereby illustrating the balance between the therapeutic potential of targeting PDK1 and the translational hurdles that remain. This analysis serves as a succinct connection between discovery research and clinical application.

The clinical translation of PDK1-targeted therapies has been impeded by a confluence of medicinal chemistry, biological, and translational challenges. The principal challenges include insufficient selectivity and suboptimal pharmacokinetic properties of initial inhibitors, prevalent pathway redundancy and adaptive resistance within the PI3K–AGC kinase network, and the absence of validated prognostic biomarkers for efficient patient stratification (Baldi et al., 2025; Fu et al., 2025). The on-target toxicity associated with the critical physiological functions of PDK1 has limited treatment windows, highlighting the necessity for mechanism-aligned, biomarker-driven, and multi-modal approaches to realize the clinical potential of PDK1 inhibition (Domrachev et al., 2021).

Future strategies emphasize rational design that unites selectivity, potency, and redundancy management. Polypharmacology targeting PDK1 with PI3K, mTOR, or Aurora-A offers a promising solution for compensatory signaling and can improve outcomes.

Structural insights enable targeting unique residues, such as Ala162 in the hinge. Stabilizing autoinhibited conformations using a linker mimetic provides an additional approach for modulating kinases. Biomarker-guided patient selection may enhance clinical efficacy, especially in tumors with elevated Ser241 phosphorylation. These developments could enable personalized therapy. AI-driven, structure-based design can accelerate the discovery of compounds with improved pharmacokinetics, selectivity, and potency (Mulpuri et al., 2025; Peifer and Alessi, 2008).

Recent breakthroughs in computational chemistry and artificial intelligence have significantly enhanced kinase drug discovery by enabling the efficient investigation of vast chemical space and structure-based optimization (Yasuda et al., 2025). Structure-based drug design (SBDD), including molecular docking and molecular dynamics simulations, is extensively utilized to investigate kinase binding pockets, evaluate ligand-target interactions, and enhance lead compounds prior to their synthesis and biological testing. This method saves time and money compared to traditional methods (Wei and McCammon, 2024).

Machine learning–based virtual screening methods, such as deep learning and generative scoring functions, can accelerate the identification of potential inhibitors by predicting binding affinity and chemical features associated with activity. When used with multiconformational protein ensembles, these methods can also improve hit rates against dynamic targets (Loeffler et al., 2024). Another promising approach to suggesting new scaffolds that fit specific kinase pockets is to use chemical language models (CLMs) and reinforcement-learning frameworks to generate new designs from scratch. This combines structural and chemical constraints (Yin et al., 2024).

For PDK1, structure-based pharmacophore modeling and protein–protein docking have recently been used to identify allosteric pharmacophoric features in PDK1’s PIF pocket and to support virtual screening of different libraries. Met dynamics and free-energy analysis have confirmed possible binders (Vennila and Elango, 2024). Integrated workflows that combine docking with molecular mechanics have also become more common. Born surface area (MM-GBSA) rescoring and dynamics can analyze the stability and specificity of potential PDK1 ligands in silico (Kirubakaran et al., 2012; Sadeghian et al., 2025). While there have not yet been any reports of fully AI-generated, experimentally validated PDK1 inhibitors, the combination of generative design, physics-based screening, and machine learning-enhanced optimization provides a strong framework for accelerating the search for selective, potent PDK1 inhibitors in future studies.

PDK1 inhibitors will likely impact clinical practice as part of biomarker-guided combination therapies. Patients with PTEN loss or hyperactive PI3K signaling may respond to AKT inhibitors. Brain-penetrant scaffolds might help in CNS malignancies. Yet, clinical progress is slow (Raju et al., 2023). This situation is due to selectivity challenges, incomplete biomarker validation, and poor in vivo efficacy of current compounds.

Closing the gap between preclinical promise and clinical utility requires validated pharmacodynamic markers, such as p-Ser241 or substrate phosphorylation signatures, alongside proof-of-concept studies showing synergy in combination regimens. Without this, clinical integration of PDK1-targeted strategies will remain challenging.

Despite a robust biological foundation, the clinical application of PDK1-targeted therapeutics has been hindered by a confluence of chemical, biological, and translational obstacles, necessitating a comprehensive evaluation of the factors that have impeded success to date. While PDK1 (PDPK1) is infrequently altered, numerous extensive cancer genomics studies highlight its clinical significance via gene amplification and pathway reliance rather than traditional oncogenic mutation. The TCGA dataset indicates that PDPK1 copy-number gain or amplification occurs in approximately 5%–15% of breast, ovarian, lung, and head and neck squamous cell carcinomas, which is associated with heightened PI3K pathway activity and a poor prognosis (TCGA Research Network, 2020; Lui et al., 2013; He et al., 2021; Qi et al., 2025). Conversely, activating point mutations in PDPK1 is infrequent, constraining mutation-based patient classification and partially elucidating the unsuccessful outcomes of previous therapeutic studies that depended on expression rather than functional reliance (Zhang et al., 2017). Recent research indicates that indicators of pathway flux may more accurately predict sensitivity to PDK1 inhibition than static genetic changes. These factors encompass PTEN loss, PI3KCA mutation, increased PIP₃ levels, and reliance on AKT-independent PDK1 substrates such as SGK3 or RSK, especially in cancers exhibiting diminished AKT signaling or metabolic reprogramming (Prabhu et al., 2025; Scortegagna et al., 2014; Scortegagna et al., 2015; Zhou et al., 2024). Recent mechanistic investigations suggest that phosphoinositide-mediated PDK1 conformational activation and persistent Ser241 phosphorylation, rather than overall protein levels, may characterize functional dependence on PDK1 signaling (Huang et al., 2024; Levina et al., 2022; Sacerdoti et al., 2023). The data collectively suggest that PDK1 represents a context-dependent vulnerability, necessitating sensible patient selection based on integrated lipid signaling, PTEN status, and downstream substrate activation measures, rather than only on PDPK1 expression or mutation.

Subsequent research in PDK1-targeted therapy must emphasize the creation of next-generation allosteric and dual-site inhibitors to enhance selectivity, maintain prolonged pathway suppression, and expand therapeutic windows. A thorough pharmacological assessment in genetically and phenotypically pertinent patient-derived models will be crucial for forecasting clinical efficacy and reducing translational failure. The early incorporation of comprehensive biomarker research initiatives in drug discovery pipelines is crucial for reasonable patient classification and the reduction of on-target toxicity in early-phase clinical evaluations. Collectively, these initiatives facilitate a significant shift from non-selective, ATP-competitive suppression to precision-focused, mechanism-aligned, and multi-modal therapeutic approaches (Roskoski, 2025; Wang et al., 2021).

In the forthcoming decade, rational polypharmacology strategies are anticipated to become a primary emphasis, integrating PDK1 regulation with inhibitors of PI3K, AKT, or analogous signaling pathways to mitigate route redundancy and adaptive signaling plasticity. Structural and signaling investigations increasingly endorse dual-site targeting paradigms, wherein simultaneous interaction with the ATP-binding cleft and rupture of the PIF-pocket stabilize inactive kinase conformations and inhibit the docking of downstream AGC substrates. These “dual-geometry” inhibitors possess the capability to inhibit both catalytic activity and compensatory substrate reprogramming, a prevalent resistance mechanism linked to single-site inhibitors. Concurrently, tailored protein degradation techniques, such as PDK1-directed PROTACs, signify a revolutionary approach by facilitating the removal of both catalytic and non-enzymatic scaffold activities that uphold signaling robustness and are unattainable by traditional small-molecule inhibitors.

Effective clinical translation will ultimately rely on the integration of sophisticated treatment modalities with biomarker-driven patient selection frameworks that emphasize functional pathway reliance rather than solely PDPK1 expression. Genomic factors, including PTEN loss, PIK3CA mutation, or PDPK1 amplification, along with lipid-dependent activation states evidenced by persistent Ser241 phosphorylation and synchronized activation of downstream AGC kinases, may function as dynamic markers of PDK1 dependency and therapeutic susceptibility (Emmanouilidi et al., 2019). Pan-cancer proteogenomic and functional dependency analyses indicate that pathway activation signatures, rather than singular genomic alterations, more precisely reflect PDK1 dependence, with CRISPR and RNAi screens revealing tumor subsets addicted to PI3K or amplified in RTK where PDPK1 is critical for survival. Expanding the exploration of PDK1 beyond oncology to include inflammatory, metabolic, and neurodegenerative diseases where PI3K–PDK1–AGC signaling regulates immune activation, metabolic balance, and neuronal survival highlights the kinase’s extensive pharmacological significance and advocates for its reclassification as a context-specific, multi-disease therapeutic target rather than a universal growth kinase.

Next-generation targeted protein degradation techniques such as proteolysis-targeting chimeras (PROTACs), molecular glues, and novel degrader–antibody conjugates (DACs) present an effective approach to overcome persistent challenges associated with kinase inhibition, especially for signaling hubs like PDK1. By eradicating the target protein instead of merely transiently inhibiting catalytic activity, these strategies possess the capacity to surmount conformational plasticity, disrupt non-enzymatic scaffold functions, and mitigate compensatory signaling networks that often undermine small-molecule inhibitors. The integration of PDK1-binding scaffolds into degrader platforms may facilitate enhanced and prolonged pathway inhibition in tumors exhibiting hyperactive PI3K–AKT signaling, consistent with recent clinical progress in kinase degradation noted for BTK, BRAF, and CDK family members (Békés et al., 2022; Li and Crews, 2022; Mullard, 2025).

It is essential to differentiate between direct degradation of PDK1 and indirect regulation of pathways through the degradation of upstream regulators. The newly identified SMART1 chemical does not act as a PROTAC that directly destroys PDK1; instead, it serves as a highly selective degrader of Smurf1, an E3 ubiquitin ligase that regulates PDK1–AKT signaling via post-translational mechanisms. SMART1 was discovered via the examination of a structurally varied PROTAC library and consists of a ligand that binds to Smurf1, targeting the HECT domain, an optimized linker, and a cereblon-recruiting component, leading to effective, CRBN- and proteasome-mediated degradation of Smurf1 in KRAS-mutant colorectal cancer models (Zhang et al., 2017). This distinction is crucial to prevent confounding pathway-level regulation with direct target engagement in the developing degradation literature.

Mechanistically, the loss of Smurf1 impairs PDK1 signaling indirectly by inhibiting PDK1 neddylation at Lys163, a change essential for the recruitment of SETDB1 and the assembly of the cytoplasmic cCOMPASS complex, which enables AKT membrane localization and Thr308 phosphorylation (Peng et al., 2025). The loss of this regulatory axis diminishes PDK1 Ser241 phosphorylation and weakens downstream PI3K–AKT pathway output without direct interaction with PDK1, indicating an extra layer of upstream regulation over PDK1 functional activation. SMART1 exhibits nanomolar antiproliferative activity in KRAS-mutant colorectal and pancreatic cancer cell lines, significant antitumor efficacy in both cell-derived and patient-derived xenograft models, and favorable tolerability, while demonstrating synergy with ATP-competitive PDK1 inhibitors. These findings collectively affirm the viability of targeting the PDK1 pathway indirectly through E3 ligase degradation as a therapeutic strategy, while highlighting that genuine PDK1-directed degraders are still in the nascent stages of preclinical development, with only recent studies initiating exploration of this area (Peng et al., 2025).