**1. Introduction**

The selective degradation of proteins in eukaryotic cells is essential for the maintenance of physiological homeostasis. Protein degradation is implemented primarily via the ubiquitin-proteasome system [1,2]. The proteasome is huge protein complex (26S), comprised of a 20S core particle (CP) and one or two 19S regulatory particles (RPs). The 20S CP, which has proteolytic activity, is composed of seven homologous α subunits, α1–α7, and seven homologous β subunits, β1–β7, which are assembled into a cylindrical structure with an α1-7β1–7β1–7α1–7 arrangement. The 19S RP is responsible for the collection of ubiquitinated substrates, the opening of the central gating pore of the 20S CP, and the de-ubiquitination and translocation-coupled unfolding of the substrates. Recent structural studies using cryo-electron microscopy have shed light upon the cooperative working mechanisms of this huge proteolytic machinery [3,4].

The correct arrangement of the proteasomal subunits is essential to the proper functioning of eukaryotic proteasomes. There is considerable evidence that the assembly of the eukaryotic 26S proteasome does not proceed spontaneously, but is mediated by several assembly chaperones [5–8]. The formation of the 20S CP is assisted by five proteasome-specific chaperones: PAC1–PAC4 and POMP in humans; Pba1–Pba4 and Ump1 in yeast. Four dedicated chaperones, p27 (Nas2), gankyrin (Nas6), PAAF1 (Rpn14), and S5b (Hsm3) are responsible for the formation of the base subcomplex of the 19S RP. Malfunctions of these assembly chaperones cause the accumulation of imperfectly assembled or mis-assembled complexes of the proteasomal subunits. For example, knock-down experiments involving PAC3 and PAC4 results in the accumulation of abnormal α-subunit oligomers lacking the α3–α7 subunits [9,10].

Because proteasome biogenesis is known to be significantly upregulated in cancer cells [11], the proteasome has potential as a target for therapeutic drugs for cancer treatment [12,13]. Bortezomib (Velcade) has been widely used as proteasome inhibitor for the treatment of patients with multiple myeloma [14]. The chaperones involved in proteasome assembly have also been considered as potential drug targets for anticancer treatments [12]. Selective inhibitors that specifically suppress proteasome biogenesis could be valuable for minimizing the undesired side effects which can occur when using compounds which target mature proteasomes.

Recently reported knock-out experiments indicated that α4, α5, α6, and α7 form a core assembly intermediate as part of the initial process of α-ring assembly, which is supported by PAC3-PAC4 [15]. However, most of the biochemical and structural data about the proteasome-assembly chaperones have been generated mainly from yeast proteins, which have only modest sequence identities with the human counterparts; less than 20%, for PAC3 when compared with Pba3. As with CP-assembly, yeast Pba3 and Pba4 have structural resemblance, and form a heterodimer [16] which functions as a matchmaker mediating the association between α4 and α5 [17]. It remains unclear, however, how the human PAC3-PAC4 complex functions in α-ring assembly through specific, direct interactions with cognate proteasomal subunits, although the crystal structures of human PAC3 and PAC4 have been solved for their homodimeric forms [16,18].

Structural insights into the chaperone-mediated formation of the human proteasome are important for the design and development of low-toxicity anticancer drugs which can inhibit the protein-protein interactions involved in the proteasome-assembly process. We performed a biochemical and biophysical study of the human PAC3-PAC4 heterodimer in order to understand the functional and structural mechanisms of α-ring formation mediated by the proteasome-assembling chaperones.

#### **2. Results and Discussion**
