Intracellular Signaling by Akt: Bound to Be Specific

Science Signaling  17 Jun 2008:
Vol. 1, Issue 24, pp. pe29
DOI: 10.1126/scisignal.124pe29


Over the past decade, the serine/threonine kinase Akt (also known as protein kinase B) has emerged as a critical signaling molecule within eukaryotic cells. In addition to the research required for the clarification of its regulation by upstream kinases and phosphatases, progress has been made in the identification of Akt-binding partners that modulate its activation, regulate its kinase activity, and define its impact on downstream biological responses. Studies of Akt-binding molecules have highlighted novel mechanisms involved in the regulation of signaling downstream of activated phosphoinositide 3-kinase. Akt-interacting molecules may have important roles in Akt signal transduction both under physiological and pathological conditions.

A key component for translating extracellular information into downstream biological responses is the serine/threonine kinase Akt [also called protein kinase B (PKB)]. The Akt signaling module, including its upstream regulators and downstream effectors, is evolutionarily conserved among invertebrate and vertebrate species. As a target of phosphoinositide 3-kinase (PI3K) (1), Akt regulates a wide range of biological responses that include cell motility, growth, proliferation, and survival, as well as transcription, protein synthesis, and nutrient metabolism (2, 3). In mammals, three independent genes encode three isoforms of Akt, each of which is present, although in varying amounts, in all mammalian cells. The Akt isoforms share similar structures, namely (i) an N-terminal regulatory domain, which includes a pleckstrin homology (PH) domain; (ii) a hinge region, which connects the PH domain to (iii) a kinase domain with serine/threonine specificity; and (iv) a C-terminal region necessary for the induction and maintenance of kinase activity (2). The structural homology between the Akt isoforms extends to their mechanisms of activation (Fig. 1). All Akt isoforms contain PH domains with similar specificities for the D3-phosphorylated phosphoinositide products of PI3K. They are also substrates for phosphoinositide-dependent kinase 1 (PDK1), a PH domain–containing kinase downstream of PI3K that phosphorylates Akt isoforms and other AGC kinase family members on a critical threonine residue in the activation loop (4).

Fig. 1.

Regulation of Akt signaling complexity by Akt-interacting molecules. Tyrosine autophosphorylation of receptor tyrosine kinases (RTKs) induces the recruitment of the p85 regulatory subunit leading to phosphatidylinositol 3-kinase (PI3K) activation. Once activated, the p110 catalytic subunits α and β phosphorylate plasma membrane–bound phosphoinositides (PI-4-P and PI-4,5-P2) on the D3-position of their inositol rings (2, 3). The second messengers resulting from this PI3K-dependent reaction are PI-3,4-P2 and PI-3,4,5-P3 (also called PIP3). PIP3, in turn, is the substrate for the phosphoinositide 3-phosphatase and tensin homolog (mutated in multiple advanced cancers 1) (PTEN), an endogenous inhibitor of PI3K signaling. The phosphoinositide products of PI3K form high-affinity binding sites for the pleckstrin homology (PH) domains of intracellular molecules. Phosphoinositide-dependent kinase 1 (PDK1) and Akt are two of the many targets of PI3K products. After localization of cytoplasmic inactive Akt to the plasma membrane, phosphoinositide-binding Akt is phosphorylated by PDK1 on a critical threonine residue in its kinase domain. The mammalian target of rapamycin complex 2 (mTORC2) constitutes a second kinase activity that, through phosphorylation of a serine residue in the C-terminal hydrophobic motif (HM) of Akt, locks the enzyme into an active conformation (35). Studies examining Akt-interacting molecules suggest that this canonical and evolutionarily conserved pathway of Akt regulation may be modulated in a cell type– or context-specific manner. In addition to proteins that bind to the Akt kinase domain to increase Akt phosphorylation and activity [such as BTBD10 (21), APE (15), and APPL (20); see text for details], TRB3 also binds to the kinase domain but instead inhibits Akt (19). In the case of APPL, it has been postulated that changes in Akt phosphorylation are due to the inhibition of Akt phosphatases (20). CTMP binds to the C terminus of Akt (14), where it inhibits a critical phosphorylation event in the hydrophobic motif. The proto-oncogene product Tcl-1 and its family members bind to the PH domain to facilitate Akt activation (22). In contrast, CKIP-1 (26) and Ca2+·CaM (27) interfere with the binding of the PI3K product PIP3 to the PH domain and, as a result, block Akt activation. Inhibition of PIP3 binding has also been reported for the Ins(1,4,5)P3 kinase (InsP3K) product Ins(1,3,4,5)P4 (30). It is unknown whether phosphatases regulate the abundance of Ins(1,3,4,5)P4 analogously to the interplay between PI3K and PTEN in regulating PIP3 levels.

In studies examining Akt signaling as a correlate of upstream signaling or downstream biological outcomes, a common laboratory technique is the measurement of Akt activity by performing kinase assays and blotting with phosphorylation-specific antibodies. A plethora of inputs induce quite similar Akt signaling in vitro when comparing different growth factors, extracellular signals, and matrix components, even after training of rodents in behavioral paradigms (2, 3). Often unrelated inputs increase Akt activity to a similar extent with comparable time courses. This apparent redundancy in Akt activation in response to varying inputs lends itself to questioning whether there is any specificity of Akt signaling under physiological conditions and, if so, how it is achieved mechanistically.

Conversely, there are marked differences regarding the involvement of different Akt isoforms in human disease. For example, whereas a mutation in Akt2 has been associated with familial diabetes (5), coding variations in Akt1 have been observed in schizophrenia (6), thus suggesting that different Akt isoforms may be associated with distinct disturbances in downstream biological processes. Furthermore, the selective disruption of Akt genes in the mouse germ line results in isoform-deficient mice with specific phenotypes: Akt1-deficient mice show placental hypotrophy, accompanied by retardation of growth and reduction of body weight; Akt2-deficient mice become hyperinsulinemic and hyperglycemic; and Akt3-deficient mice exhibit reduced brain size (7). This lack of functional compensation suggested that not all Akt isoforms perform equally, which is corroborated by studies of mouse embryonic fibroblasts from the different strains of Akt-deficient mice, in which the lack of specific Akt isoforms yields distinct phenotypes (8).

These physiological findings of nonredundancy suggest that additional mechanisms are involved in fine-tuning the involvement of Akt in its diverse biological functions. One growing focus of research with potential to provide insights into possible mechanisms explaining the context-dependent regulation of Akt activity concerns Akt-interacting proteins. Mainly through yeast two-hybrid analyses, several proteins have been identified that interact with one or more of the functional domains of Akt, and with one or more of its isoforms (9, 10). Several enzymes are found among these Akt-interacting proteins, including PDK1 (4). Other such enzymes include inosine-5′ monophosphate dehydrogenase (IMPDH) (11), the histone H3 methyltransferase SETDB1 (12), and PH domain leucine-rich repeat protein phosphatase (PHLPP) (13).

Other Akt-interacting proteins lack discernible intrinsic enzymatic activities. This shared property suggests that their interactions with one or more regions of different Akt isoforms affect the kinase activity of Akt, access to upstream regulatory components, or its intracellular localization. Akt-interacting proteins bind to one or more of its functional domains. C-terminal modulator protein (CTMP) (14), Akt phosphorylation enhancer (APE) (15), Arg-binding protein 2γ (ArgBP2γ) (16), and prohibitin 2 [PHB2, also known as repressor of estrogen activity (REA)] (17) interact with the C terminus. Heat shock protein 90 (Hsp90) (18), tribbles homolog 3 (TRB3) (19), adaptor protein containing PH domain, PTB domain, and leucine zipper motif (APPL) (20), and the bric-a-brac/poxvirus and zinc finger proteins (BTB/POZ)–homologous domain–containing protein (BTBD10) (21) interact with the kinase domain. Finally, PH domain–interacting proteins include the Tcl-1 oncoprotein and its related family members (22), c-Jun N-terminal kinase (JNK)–interacting protein 1 (JIP1) (23), growth factor receptor–binding protein–10 (Grb10) (24), the guanosine triphosphatase (GTPase)–activating protein RasGAP (25), casein kinase 2–interacting protein-1 (CKIP-1) (26), and Ca2+/calmodulin (Ca2+/CaM) (27). The consequences of the binding of any of these proteins to Akt isoforms are diverse. Thus, binding may increase or decrease Akt phosphorylation, activity, or both, and in some cases this depends on the cell types and experimental conditions involved (28). However, the identification of Akt-binding partners may still provide a unique opportunity to better understand the mechanisms that modulate Akt signaling.

In proof-of-principle experiments, studies from the Naguchi group have demonstrated the utility of Tcl-1–derived peptides as suitable and specific inhibitors of Akt signaling (29). In support of the physiological importance of endogenous competitors for PI3K second-messenger binding to the Akt PH domain, Jia et al. have shown that inositol 1,3,4,5-tetrakisphosphate [Ins(1,3,4,5)P4] is a specific inhibitor of PI3K-dependent signaling in neutrophils (30). In this paradigm, signaling specificity is achieved through the regulation of Ins(1,4,5)P3 kinase, which is increased in activity following chemoattractant stimulation and impairs plasma membrane binding of Akt (and other PI3K targets). As a different example of PH domain–interacting inhibitors of Akt signaling, Dong et al. have shown that Ca2+/CaM competes with PI3K products for interaction with the PH domain of Akt in vitro (27).

The importance of Akt-binding proteins has been most clearly defined under pathological conditions. For example, overexpression of Tcl-1 in human T cell prolymphocytic leukemia contributes to pathogenesis by facilitating Akt activation and altering its intracellular localization (22). Conversely, in other tumor types, the reduced abundance of mRNAs for CTMP (31) and CKIP-1 (26) has been reported. It may be that related physiological mechanisms exist that regulate Akt activation by altering the abundance of Akt-interacting proteins—for example, during development and differentiation. The abundance and tissue distribution of Tcl-1 during early development substantially differ from that during adulthood (22).

The idea that Akt-interacting proteins may determine and guide Akt activity under physiological conditions, or in response to stress or injury, is appealing. In its support, Fritzius and Moelling have shown that the abundance of the Akt-interacting protein WD-repeat propeller-FYVE (ProF) is transiently increased during adipocyte differentiation, at which point it specifically enhances phosphorylation of the forkhead-related transcription factor, Foxo1, by interacting with Akt (32). Schenck et al. have shown that endosomal partitioning of Akt by the APPL family member Appl1 regulates downstream signaling specificity during zebrafish development (33). Additional complexity is introduced by the observation that not all Akt-interacting proteins are able to bind to all Akt isoforms; JIP1 preferentially binds to Akt1 but not to Akt2 (23). Conversely, PHB2 preferentially binds to Akt2 but not to Akt1 (34), whereas the Tcl-1 family members Tcl-1b and MTCP1 bind to Akt1 and Akt2, but not to Akt3 (22). Thus, specificity of Akt signaling may be controlled by interacting proteins, depending on their relative abundance and isoform-specific binding patterns.

Faced with the task of finding specificity in the commonality of Akt activation, it is conceivable that differential regulation of Akt’s binding partners will determine cell type– and context-specific signaling by Akt, possibly involving isoform-specific interactions. Future studies will elucidate the physiological roles of Akt’s binding partners by applying tissue-specific knockout models or through studying their regulation of specific Akt isoforms in relevant biological settings. Most of the known Akt-interacting proteins have been studied separately for their abilities to stimulate or inhibit Akt activity. However, it is possible that multiple binding partners for Akt are present concomitantly and compete in combination for their effects on Akt.

At a time when Akt signaling has been firmly established as a common eukaryotic signal transduction pathway, studies will be needed that address the complex mechanisms by which this ubiquitious signaling module contributes to context-specific functions. The identification of Ins(1,3,4,5)P4 as an endogenous inhibitor of PH domain–containing PI3K targets such as Akt during neutrophil responses or Appl1 as a mediator of specific substrate interaction during zebrafish development are important first steps (30, 33). It is a safe bet that the identification of other binding partners capable of context-dependent regulation of Akt signaling will follow.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
View Abstract

Navigate This Article