Multiple myeloma (MM) is the second most common hematologic malignancy worldwide, characterized by the proliferation of terminally differentiated antibody-producing plasma cells (PCs) ^[1][2][1,2]. Although MM has intrinsic genetic heterogeneity [3], a common feature of malignant clones is the production of elevated amounts of immunoglobulins [1], ultimately leading to organ dysfunctions such as hypercalcemia, renal insufficiency, anemia and bone disease (overall referred to as CRAB criteria) ^[4][5][4,5]. The bone marrow microenvironment (BMM) has been shown to play a significant role in MM pathogenesis triggering PC survival, proliferation and drug resistance ^[1][2][1,2]; moreover, genetic complexity, defined by chromothripsis and hyperdiploidy, along with copy number variations and single-nucleotide polymorphisms, account for the early evolution from asymptomatic stages (e.g., monoclonal gammopathy of undetermined significance (MGUS) and smoldering multiple myeloma (SMM)) to overt disease [6]. Additional alterations, including aberrant DNA or histone methylation [7] and microRNA (miRNA) dysregulation ^[8][9][8,9], may contribute to disease progression.

The ubiquitin–proteasome system (UPS) is crucial homeostatic machinery for protein degradation; it constantly regulates protein turnover, thus affecting various cellular functions, spanning from cell cycle regulation to survival, apoptosis, metabolism and protein quality control [10]. Since PCs secrete high amounts of immunoglobulins, they are strictly dependent on the UPS machinery and are highly reliant on the deregulation of protein degradation. Moreover, due to the constitutive activation of the nuclear factor kappa B (NF-κB) signaling pathway, malignant PCs are more sensitive to PI than healthy ones [11]. In fact, the anti-MM activity of proteasome inhibitors (PIs) was originally ascribed to the inhibition of oncogenic NF-κB through the blockage of the degradation of its negative regulator IκBα [12]. However, additional processes whose targeting contributes to the anti-tumor effects of PIs have subsequently been identified, including the reversal of cell cycle aberrations, apoptosis induction, endoplasmic reticulum stress, angiogenesis and DNA repair, as well as the reactivation of hypermethylated tumor-suppressor genes ^{[12][13][14][15]}[12,13,14,15].

The exquisite sensitivity of MM cells to PIs, along with the design of successful clinical protocols, have led to their approval for the treatment of MM patients, with three compounds currently being used in clinics [16]. PIs are true milestones for the treatment of MM and other hematologic malignancies (e.g., mantle cell lymphoma), and are currently being investigated for other diseases. The first approved PI was bortezomib, a slowly reversible inhibitor of the β5 catalytic proteasomal subunit, followed by carfilzomib, an irreversible inhibitor of the β5 site, and by the first oral PI, ixazomib [2].

Natural products have always played a key role in drug discovery, especially for cancer and infectious diseases. In oncology, several drugs derived from Nature have been approved, and are widely used in clinics, including paclitaxel, romidepsin, vincristine and vinblastine ^{[17][18][19][20]}[17,18,19,20].

Importantly, marine-derived natural compounds isolated from sponges, mollusks, cyanobacteria, corals and tunicates have been clinically approved for MM by the US and Australian FDA as well as the European Medicines Agency (EMA), such as belantamab mafodotin and plitidepsin, while others are under clinical trial or extensive preclinical assessment [21].

Additionally, naturally derived compounds with PI activity have been isolated from various sources, and are currently emerging as potential anti-cancer drugs in a wide variety of cancers, including MM [22].

2. Proteasome Structure and Function

By ruling protein degradation, the UPS represents a fundamental intracellular system whose components act in a highly coordinated manner through different steps, such as the polyubiquitylation, deubiquitylation and degradation of the target protein ^[16][23][16,23]. Thanks to the coordinated activities of this system, all the damaged or misfolded proteins, or the proteins which are no longer needed, are correctly destroyed [16]. The proteasome is the central core of this system. The eukaryotic 26S proteasome is a large (1500–2000 kDa) multi-subunit complex that degrades most cellular proteins under physiologic conditions. It is a multicatalytic proteinase complex able to bind, deubiquitylate and unfold its substrates prior to completing their degradation. The proteasome is predominately located in the cytosol and nucleus, and its functions are not limited to the maintenance of proteostasis—it is involved in a wide array of biological processes, including cell differentiation and proliferation, DNA repair and apoptosis, regulation of gene expression and response to stress ^[24][25][24,25]. Alternative forms of the proteasome, such as the immunoproteasome and thymoproteasome, are involved in routine proteolytic functions, antigen processing and T-cell selection [16]. The 26S proteasome is a dynamic ATP-dependent macromolecular machine, comprising two subcomplexes: the 19S regulatory particle (RP) and the 20S core particle (CP). Either one or two 19S RPs can attach to a single 20S core particle to form the 26S proteasome (19S-20S) (Figure 1A) or the 30S proteasome (19S-20S-19S), respectively. However, both conformations in the literature are referred to as “26S” (Figure 1B) ^[26][27][28][26,27,28]. The core of the RP is formed by a heterohexameric ATPase associated with various cellular activities (AAA + ATPase), which is the driver of the large-scale conformational dynamics of the RP. The core of the RP prepares substrates for degradation in coordination with at least three ubiquitin receptors (26S proteasome non-ATPase regulatory subunit 1, Rpn1, Rpn10 and Rpn13, and a deubiquitylating subunit, Rpn11). The other subunits mainly have structural roles, such as holding the CP and RP together [29]. The proteasome regulatory subunits bind to each end of the 20S proteasome (Figure 1) and mediate deubiquitylation. These accessory subunits unfold substrates and feed them into the 20S catalytic complex, whilst removing the attached ubiquitin molecules. Once the target proteins have been deubiquitylated, they undergo degradation via the 20S proteasome particle [31], which removes misfolded and damaged proteins, and digests foreign proteins as part of the adaptive immune system [32]. The eukaryotic 20S core particle is the catalytic core of this complex, constituted by a cylindrical structure consisting of four heptameric rings containing α and β subunits (Figure 2). Seven α subunits form the two outer rings and serve as the gates through which proteasome substrates enter, meanwhile the seven β subunits form the inner rings. The passage through the gates formed by the N-termini of the α subunits is the rate-limiting step, and prevents unregulated protein degradation [34]. The αN-terminal tails are highly conserved, and contain a tyrosine-aspartate-arginine (YDR) motif that forms salt bridges with neighboring tails that obstruct the 13 Å entry pore. The β subunits contain the N-terminal threonine that provides the nucleophile that attacks the carbonyl group of the peptide bond in the target proteins. The Thr residue is part of a conserved Thr-Lys-Asp catalytic triad, which functions similarly in both processes [25] (Figure 3). Other important residues for the active site are glycine 129 (Gly129), aspartate 167 (Asp167) and tyrosine 169 (Tyr169). The Thr1, Asp17 and Lys33 are the most important residues for the proteolytic mechanism meanwhile the other residues are important for the active site integrity [35]. The proteolytically active β subunits, three on each β ring, exhibit different substrate preferences. The proteasome is associated to at least three distinct proteolytic activities, based upon preference to cleave a peptide bond after a particular amino acid residue: the postglutamyl peptidyl hydrolytic-like, the tryptic like and the chymotryptic-like. These proteolytic activities of the proteasome are carried out by the β1, β2 and β5 subunits, respectively [36]. The chymotrypsin-like (β5) activity cleaves after hydrophobic residues, the trypsin-like (β2) sites preferentially cleave after hydrophobic and basic residues, while the caspase-like site (β1) cleaves after the acidic residues [36]. The multiple catalytic sites with varying specificities advantageously allow for the rapid and processive degradation of cellular proteins. As a result, the proteasome is not simply a complex of independent proteases, but a unique multicatalytic enzyme properly functioning when fully intact.

Approved and Investigational PIs for MM Treatment

PIs are classified according to their chemical structure and their mechanism of action. Covalent inhibitors are generally electrophilic and react with the catalytic gamma-hydroxyl of Thr1 in the active sites to reversibly or irreversibly inhibit the proteasome depending on the strength of the chemical bond. Based on their chemical structure, PIs can be classified in the following classes: peptide aldehydes, peptide boronates, epoxomicin and epoxyketones, lactacystin, β-lactone and vinyl sulfones. Similar to most protease inhibitors, several PIs are short peptides designed to fit into the substrate binding site of the catalytic subunit. Although the proteasome has three types of catalytic sites, full inhibition of all of them is not required to significantly affect protein degradation. In fact, while β5 inhibition leads to significantly reduced protein breakdown, specific β1 or β2 inhibition does not have any significant effect. Consequently, most PIs act by targeting the β5 site, although they can often have some lesser activity against β1 and/or β2 [37]. To date, three FDA-approved proteasome inhibitors, namely bortezomib (Velcade^®), carfilzomib (Kyprolis^®) and ixazomib (Ninlaro^®), are in clinical use, and several drugs are under development (Figure 4).

• Bortezomib

Bortezomib (Velcade^®) (Figure 4), the first reversible PI approved in 2003 by the US Food and Drug Administration (FDA), rapidly became a mainstay in the treatment of relapsed or refractory (and, subsequently, of previously untreated) MM patients [38]. Chemically, bortezomib ([3-methyl-1-(3-phenyl-2- pyrazin-2-ylcarbonylamino-propanoyl) amino-butyl] boronic acid) is a dipeptide boronic acid derivative which contains pyrazinoic acid, phenylalanine and leucine with boronic acid in its structure [39]. Bortezomib targets the chymotrypsin- and caspase-like active sites, with minimal effect on trypsin-like activity (Table 1) [40]. Its usage has been expanded for newly diagnosed MM patients as well as for the treatment of mantle cell lymphoma.

• Carfilzomib

The tetrapeptide epoxyketone carfilzomib (Kyprolis^®) (Figure 4) is the first irreversible PI approved by the FDA for the treatment of relapsed and refractory MM [39]. Carfilzomib differs from bortezomib in structure, activity and the irreversibility of its binding mode, showing potent and selective inhibition of chymotrypsin-like activity with lower affinity for the trypsin- and caspase-like proteases (Table 1). Carfilzomib is active in vitro against bortezomib-resistant MM cell lines, as well as in vivo in patients with bortezomib-resistant MM.

• Ixazomib

Ixazomib (Ninlaro^®) (Figure 4) is an orally bioavailable, reversible PI developed by Takeda Oncology, approved by the FDA in 2015. It is made up of an alanine leucine dipeptide core, and it acts as a reversible inhibitor of the chymotrypsin-like β5 subunit of the 20S proteasome [41].

• Marizomib

Marizomib (salinosporamide A) is a marine-derived extract from Actinomycete strains, identified as a member of the Micromonospora family at Scripps Institution of Oceanography (SIO) in the late 1980s. With the help of DNA sequencing (16S rRNA gene sequencing), these strains were found to be distinctive from all Micromonospora species, and were named as a new genus, Salinospora [42]. Chemically, marizomib is a bicyclic β-lactone γ-lactam acting as a broad-spectrum PI able to irreversibly bind the three major catalytic β5, β1 and β2 sites (Figure 4–Table 1), demonstrating efficacy in relapsed and/or refractory MM patients becoming resistant to bortezomib, carfilzomib and ixazomib regimens ^[43][44][43,44].

• Oprozomib

Oprozomib (ONX 0912) (Figure 4) is an oral analog of carfilzomib, belonging to the epoxyketone PIs. It inhibits the β5 subunit like carfilzomib, demonstrating an equivalent anti-tumor activity to carfilzomib in in vitro and in vivo models.

• Delanzomib

Delanzomib (CEP-18 770) (Figure 4) is related to bortezomib, with a peptide-like backbone and a boronate warhead. It is an oral and reversible β5 subunit proteasome inhibitor [43].