Peptide Inhibitors

Abstract

The proapoptotic cysteine protease caspase-6 participates in the neuropathology of several diseases. Unlike the active dimeric form, the caspase-6 zymogen forms a unique tetramer that can be stabilized by allosteric inhibitors, which prevents caspase-6 activation.

Apoptosis is a programmed form of cell death that is typically executed by caspases or cysteine-dependent aspartate-specific proteases¹. Caspases are activated in a hierarchical manner, wherein the upstream initiator caspases 2, 8, 9 and 10 process the downstream effector caspases 3, 6 and 7, which in turn proteolytically dismantle the cell. Initiator caspases are single-chain monomers that are activated through dimerization following association with their specific adaptor proteins. Effector caspases, in contrast, are obligate dimers that are activated through cleavage of their intersubunit linkers¹. Most efforts to design caspase inhibitors have, until now, targeted their active site cysteines with electrophiles, but achieving selectivity with this approach has been difficult. In this issue, Stanger et al.² demonstrate that procaspase-6 forms a unique tetramer that is resistant to activation. This complex can be stabilized for both the zymogen and active enzyme with peptides that bind an allosteric site²; these findings pave the way for the development of selective inhibitors of caspase-6.

Caspases contain either a long or short prodomain, followed by large and small subunits that are connected by an intersubunit linker (Fig. 1a)¹. Two major caspase cascades are initiated during apoptosis, depending on the type of cellular stress encountered. In the extrinsic pathway, ligation of death receptors such as Fas/CD95 results in the formation of a receptor–Fas-associated protein with death domain (FADD)–caspase-8 complex that promotes caspase-8 activation through dimerization. Similarly, in the intrinsic pathway, stressors such as DNA damage trigger mitochondrial release of cytochrome c and formation of an apoptotic protease-activating factor 1 ‘apoptosome’ complex that recruits and activates caspase-9. Active caspases 8 and 9 then process the downstream effector caspases 3 and 7 (ref. 1). Caspase-3 targets >800 protein substrates in the cell, including caspase-6, and is by far most responsible for dismantling the cell during apoptosis. Caspase-6 has comparatively few substrates, the most notable being lamin A, and is not directly activated by any established initiator caspase. Consequently, caspase-6 has been relatively undervalued and understudied.

Figure 1.

The caspase-6 zymogen forms a tetramer that is stabilized by pep419. (a) In this structure of the caspase-6 dimer, the large and small subunits are shown in red and orange, respectively. The intersubunit linker is shown in magenta; the single β-sheet, containing the ¹⁹⁰TEVD¹⁹³ cleavage site, is inserted into the active site of the enzyme. The region of the large subunit, containing the residues that interact with pep419, is shown in blue. (b) The A-C and B-D chains of two caspase-6 zymogen dimers face one another to form the tetramer, which is stabilized by pep419 (green). Renderings of the dimeric caspase-6 zymogen (Protein Data Bank code 3NR2)⁸ were performed using UCSF Chimera.

Increasing evidence, however, suggests that caspase-6 may have a critical role in the pathogenesis of Alzheimer's disease³ and Huntington's disease^4,5. Indeed, caspase-6–dependent cleavage of huntingtin protein is essential for the development of Huntington's disease in a mouse model, and transgenic mice expressing the cleaved N-terminal fragment of huntingtin protein likewise develop the disease^4,5. Moreover, whereas caspase-6–deficient mice were originally thought to develop normally, they now appear to have learning disabilities, and their neurons are resistant to excitotoxicity and growth factor deprivation⁶.

Despite the accumulating evidence in support of its role in neuropathology, one of the puzzling questions regarding caspase-6 is how it is activated in this context. As noted above, caspase-6 can be directly processed by caspase-3, downstream of either the extrinsic or intrinsic pathways⁷. However, if either of these pathways is activated sufficiently for caspase-3 to activate caspase-6, one would expect caspase-3 to cleave numerous substrates, resulting in immediate cell death. So how is caspase-6 activated in neurons during the much slower process of forming plaques, tangles and protein aggregates? One possible explanation is that, in addition to serving as a classical effector caspase, perhaps caspase-6 can also function as a type of hybrid or combined initiator-effector caspase. Indeed, a recent structure of the caspase-6 zymogen indicates that its intersubunit linker inserts into the active site of caspase-6. This orientation positions the ¹⁹⁰TEVD¹⁹³ cleavage site very near the active site cysteine, so that even a minor change in conformation should allow for catalytic attack of Asp193 (ref. 8). In other words, caspase-6 may self-activate through a highly unusual intramolecular cleavage event.

Stanger et al.² have discovered yet another interesting aspect of caspase-6 biology; namely, that its zymogen is a tetramer at pH 8.0. Given the successful identifcation of allosteric inhibitors of caspase-7 (ref. 9), the authors used phage display in an effort to identify peptides capable of binding the caspase-6 zymogen outside of its active site. One peptide, pep419, bound the caspase-6 zymogen and stabilized the tetrameric complex, ven under low-pH conditions that would otherwise favor dissociation of the tetramer into a dimer^2,10. Subsequent crystallographic studies revealed that pep419 formed a disulfde-linked antiparallel β-strand, sandwiched between each of the A-C and B-D chains of the zymogen, with Glu12 and His9 making key contacts with caspase-6 at His126 and Asp131, respectively (Fig. 1b)². Pep419 also shifed the equilibrium of active caspase-6 from its dimeric to its tetrameric form and correspondingly inhibited enzyme activity through a mixed noncompetitive mechanism, indicating that the peptide could stabilize the tetrameric complex, regardless of the processed state of the enzyme. Remarkably, compared to the zymogen, the peptide-bound complex showed no radical changes in active site conformation. Nevertheless, mutations of Glu12, His126 or both diminished the affinity of pep419 for caspase-6, prevented stabilization of the tetramer and failed to effectively inhibit active caspase-6 in vitro and in cultured cells².

Although the current study has tremendous implications for the treatment of Alzheimer's and Huntington's disease, a number of unanswered questions remain. For example, how is caspase-6 activated in neurons during the progression of Alzheimer's or Huntington's disease? The zymogen may be primed for ‘self-activation’ owing to the placement of its intersubunit linker, but something more is required to trigger cleavage. Also, does caspase-6 inhibition in vivo result primarily from stabilization of the zymogen or the active caspase-6 tetramer, and is the zymogen tetramer resistant to processing by active caspase-3 in the more classical caspase cascade? There are also practical hurdles to overcome in terms of drug design and delivery. Designing small molecules that can selectively stabilize the caspase-6 tetramer and cross the blood-brain barrier will be challenging. Nevertheless, Stanger et al.² provide the first proof of principle that caspase-6 (and perhaps other caspases) can be selectively inhibited through noncovalent binding of molecules to an allosteric site.

This is a commentary on article Stanger K, Steffek M, Zhou L, Pozniak CD, Quan C, Franke Y, Tom J, Tam C, Krylova I, Elliott JM, Lewcock JW, Zhang Y, Murray J, Hannoush RN. Allosteric peptides bind a caspase zymogen and mediate caspase tetramerization. Nat Chem Biol.. 2012;8(7):655-60.