The first “resident” nucleolar protein to be identified in Dictyostelium was heat shock protein 32 (Hsp32) [26]. Colocalizing with rDNA as beads on a string around the periphery of the nucleolus in unstressed cells, during heat shock it redistributes throughout the nucleolus and nucleoplasm [13,26]. Extended periods of heat shock produce a nucleolus with more pronounced rDNA beads revealing that the structure of the nucleolus in Dictyostelium responds to stress, as it does in other organisms [26,27]. In keeping with this idea, heat shock treatment of Dictyostelium induces a redistribution of Hsp32 from the nucleolar periphery and Snf12 from the nucleoplasm to nucleolus, as discussed below. Attempts to knock out the Hsp32 gene have failed suggesting it may be a critical protein [26].

Due to the presence of a highly acidic region rich in aspartic acid (asp) and glutamic acid (glu) residues, Hsp32 shares sequence similarity to nucleophosmin (NPM1) and nucleolin, both highly conserved nucleolar proteins in mammals [28]. The highly acidic regions are a common feature thought to be responsible for binding to basic ribosomal proteins and to NLSs of other proteins [29,30]. The role of the acid rich glu/asp region is discussed again for NumA1 below. Hsp32 also binds with high affinity to DNA but this association is not involved in its localization [26].

Hsp32 possesses a monopartite and bipartite NLS, thus sharing similarity with nucleolar proteins from other species [26,31,32]. However, an NoLS has not been identified and the means by which Hsp32 localizes to the nucleolus remains to be elucidated.

Eukaryotic translation initiation factor 6, eIF6, was originally identified as a nucleolar protein based on its localization as peripheral patches in DAPI-stained nuclei plus its sensitivity of actinomycin D treatment). [33]. This highly conserved protein is essential to the production of 60S ribosomal subunits serving as a rate-limiting step in the cell cycle [2,34,35,36]. The pathogenesis of two forms of leukemia—inherited Shwachman-Diamond syndrome (SDS) and sporadic SDS—involve a common pathway in 60S-subunit maturation and the functional activation of ribosomes [37,38]. eIF6 is involved in 60S-subunit maturation and thus could play a central role in the disease process. Since eIF6 is shared by eukaryotes and archaea, Dictyostelium serves as a model to detail its function. For example, Weiss et al. [39] used single-particle cryo-EM to dissect the mechanism by which eIF6 gets released from nascent 60S ribosomal subunits in Dictyostelium. As in other species, eIF6 prevents 60S maturation by blocking the binding of essential maturation factors and, thus, must be removed for functional ribosomal formation to occur. This conserved mechanism involving eIF6 release is impaired in both inherited and sporadic leukemias.

The N-terminal region of eIF6 contains both an NLS and NoLS with a second potential NLS in the C-term [33]. Scott et al. [40] published an “Experimentally determined NoLS” and a predicted NoLS for eIF6, the latter generated from a program they compiled. However, based on their information summarized in Table 2 of their publication, they erroneously analyzed the sequence of NumA1 not eIf6, albeit with additional errors. Using their program to analyze eIF6 in fact detects no NoLS for this nucleolar protein (O’Day, unpublished results). So, although the NoLS in eIF6 has been mapped to a subdomain, the precise location of the NoLS for eIF6 remains to be revealed. Deletion of the eIF6 gene is lethal in Dictyostelium as it is in other species [34,35,36]. Examination of the data from Sillo et al. [41], from their study of genes linked to phagocytosis, revealed that eIF6 is upregulated by factors that induce phagocytosis in Dictyostelium.

Tumor necrosis factor receptor-associated protein 1 (TRAP1) is a member of the Hsp90 family [42,43]. It is a multifunctional protein linked to cell cycle progression, cell differentiation, and apoptosis. Found in the outermost layer of the spore coat it is believed to protect these dormant structures from physicochemical stresses [44,45]. While TRAP1 is a questionable nucleolar protein, there is some evidence it localizes there as well as in other cellular locations including mitochondria to where it translocates during early differentiation [45]. It was first observed in intranuclear patches reminiscent of nucleoli but not verified through actinomycin D or other treatments. As expected, TRAP1expression is induced by heat shock and like other members of the Hsp90 family null mutations are lethal [42,43]. More recently, TRAP1 has gained attention as a protein that can help with understanding mitochondrial diseases [46].

The nucleolar localization of a group of well-established calmodulin-binding proteins from other organisms such as calcineurin, CaM kinase II, and myosin light chain kinase supports a role for calmodulin in the nucleolus [47,48]. In keeping with this, this calcium-sensor and -effector has also been shown to bind and localize in mammalian nucleoli [49,50]. Nucleomorphin isoform NumA1 represents the only verified nucleolar calmodulin-binding protein in Dictyostelium. Acting as a regulator of nuclear number and interacting with calmodulin in a Ca²⁺-dependent manner, it is predominantly a nucleolar protein with secondary nucleoplasmic localization [51,52,53]. This pattern of localization is a common feature of nucleolar proteins (e.g., nucleophosmin, adenosine deaminases, and murine double minute 2) that shuttle between the nucleolus and nucleoplasm [54,55].

Full length nucleomorphin contains a breast cancer carboxy-terminus domain (BRCT) that is found in cell cycle checkpoint proteins in other organisms [51,56]. The presence of a highly acidic, glu/asp domain is a common feature of nucleolar proteins including nucleophosmin, nucleoplasmin, nucleolin, and Hsp32 [26,28]. Overexpression of GFP-NumA1 lacking its palindromic glu/asp or DEED domain results in multinuclearity fitting with NumA1’s involvement in cell cycle regulation [51]. In contrast to the regulation of nuclear number in human cells where over 100 proteins appear to be involved, this is the only protein so far linked to this function in Dictyostelium [20].

A large number of attributes of NumA1 suggest it is a functional equivalent of the mammalian nucleolar protein nucleophosmin (NPM1) [51]. NPM1 has diverse functions including a role in DNA repair, centrosome duplication and cell proliferation. Mutations in NPM1 are a major cause of acute myeloid leukemia (AML) being present in 20–30% of the cases [57]. The acidic glu/asp domain of NPM1 is involved in histone binding but this function has not been studied in NumA1. The DEED domain of NumA1 is sufficient to target FITC to the nucleus thus acting as an unconventional NLS [51]. Furthermore, NPM1 can act as both a proto-oncogene and as a tumor suppressor [58]. The study of NumA1 thus has the possibility of offering additional insight into the mode of action of NPM1. For example, as covered in the next section, its DEED domain binds Cbp4a in a calcium-dependent manner suggesting NumA1 could be involved in recruiting other proteins to the nucleolus, a mechanism apparently not yet studied for NPM1.

Yeast two hybrid studies revealed that NumA1 interacts with the calcium-binding protein Cbp4a and Zn²⁺-metallopeptidase puromycin-sensitive aminopeptidase A (PsaA) which, in other species, is associated with cell cycle progression and several human diseases including Huntington’s and Alzheimer’s disease [59]. Dictyostelium PsaA is similar to PSA from Drosophila, mouse, and human. In contrast to Cbp4a (see below), PsaA does not localize to the nucleolus but colocalizes with NumA1 in the nucleoplasm independent of Ca²⁺/Calmodulin [60]. The functional relationship between Cbp4a and NumA1 is strengthened by their apparent co-regulation by developmental morphogens where differentiation factor-1 (DIF-1) upregulates them while cyclic AMP and ammonia leads to their downregulation [61].

An attribute of some nucleolar proteins is the presence of multiple NLSs. NumA1 contains four identified NLSs [51,62]. Three of them (NLS-1, -2, and -4) reside within N-terminal residues 1-120: NLS1, ³¹PKSKKKF³⁷ and NLS-1 and -4 within ⁴⁸KKSYQDPEIIAHSRPRK⁶⁴ (NLS-4 is underlined). NLS-3 is found in the C-terminus (²⁴⁶PTKKRSL²⁵²). Multiple NLSs may function to modulate the amount of the nuclear-localized protein, as seen for human TBX5, nucleolar ribosomal protein L7a and transcription factor Nrf2 [63,64,65]. FITC-peptide constructs showed each of these sequences, including the terminal RPRK sequence of the bipartite NLS (48–64), localize to nucleoli revealing the peptide sequences are all NoLS/NLS (e.g., Figure 1) [60]. As with the localization of GFP-NumA1 the nucleolar localization was abolished in the presence of actinomycin D but unaffected by treatments with calcium chelators or calmodulin antagonists. Hence these sequences serve as joint NLS/NoLS for the targeting of NumA1. NoLSs, like NLSs, are typically rich in basic residues but, in the absence of a nucleolar envelope, rather than serving as transport signals they appear to function more as retention signals [66]. Human proteins containing NLS/NoLSs include human NF-κB-inducing kinase, the novel human nucleolar protein phosphatidylinositol 4-kinase and others [66,67].

In addition to NumA1, mitotic translocations have only been studied for four other proteins—Cbp4a, Src1, Snf12 and FhkA—as discussed in the following sections (Figure 2). During prophase nucleoli become indistinct with NumA1 appearing in smaller accumulations adjacent to the inner nuclear envelope as well as being associated with the centrosome. By metaphase these inner membrane accumulations disappear and, for the rest of mitosis, NumA1 appears throughout the cytoplasm with a border of protein adjacent the outer nuclear envelope plus continued centrosomal localization.

Studies using yeast two hybrid and coimmunoprecipitation, identified calcium binding protein 4a, (Cbp4a) as a nucleolar Ca²⁺-dependent, NumA1-binding partner [52]. This binding occurs via the glu/asp or DEED domain. Addition of actinomycin D leads to a loss of nucleolar CBP4a as does calcium-chelation with BAPTA-AM, supporting Cbp4a as a nucleolar protein that requires calcium for its localization. Fitting with its association with the cell cycle protein NumA1, CBP4a has a putative forkhead-associated domain that is present in numerous cell cycle proteins. Dictyostelium possesses 13 calcium-binding proteins (CBPs) including CaM, a major Ca²⁺ effector in all eukaryotes [68,69,70]. In contrast to CaM, calfumirin, and CBP3, the function of the other CBPs remains to be researched [53].

Cbp4a residues ⁴⁰KKCK⁴³ have been verified as a true NLS but not an NoLS since FITC-bound peptides show nuclear but not nucleolar localization. In total the results indicate that Cbp4a localizes to the nucleolus not via an NoLS but via the calcium-dependent binding to the DEED domain of NumA [60]. This would suggest that Cbp4a diffuses into the nucleolus to be held there through NumA1-binding. The relationship and behavioral differences between these two proteins was revealed during the events of mitosis where they were the first two nucleolar proteins to be studied during mitosis in Dictyostelium.

During mitosis CBP4a reveals a unique distribution that suggests the presence of previously undetected intranuclear subdomains that persist throughout the mitotic stages (Figure 2) [71]. During prophase, nucleolar dissolution is accompanied by the accumulation of CBP4a as multiple, discrete nucleoplasmic accumulations called “CBP4a islands”. In addition to these multiple smaller accumulations, during metaphase two larger islands localize to the metaphase plate region. Through anaphase and telophase, these accumulations migrate to the inner membrane as if in anticipation of reforming post-mitotic nucleoli. To date, no other nucleolar or nuclear protein has shown this sequence of events during mitosis. If the nucleolar binding of Cbp4a is dependent on diffusion, then retaining this protein within the nucleus may be essential to ensuring there is enough protein present when the nucleolus reforms during telophase.

SWI/SNF is a nucleosome remodeling complex, composed of 9–12 proteins called BAFs (Barrier-to-autointegration factor), highly conserve proteins that regulate gene transcription [73]. The complex mediates multiple other processes including cell proliferation, differentiation and DNA repair. It serves as a tumor suppressor by regulating the p53-mediated transcription of cell cycle genes. BAF60a (Snf12 in yeast) mediates its interaction with p53. Dictyostelium Snf12 is a predominately nucleoplasmic protein that localizes to nucleoli in ~20% of cells, as seen in mice [74,75]. Appropriately, it possesses conserved SWIB and COG domains found in BAF proteins, but these are not involved in nucleolar positioning. Instead an experimentally defined NLS/NoLS (³⁷²KRKR³⁷⁵) defines both its nuclear and nucleolar localization.

Unexpectedly, treatment of cells with actinomycin D increases the levels of nucleolar Snf12 which leads to an outward bulging of nucleoli followed by the cytoplasmic accumulation of Snf12-rich vesicles. Heat shock treatment also leads to a major increase in the nucleolar localization of Snf12. The rapid increases in nucleolar localization after heat shock and actinomycin D treatment, suggests Snf12 may function in the stress response. As mentioned above, heat shock treatment of Dictyostelium induces a redistribution of Hsp32 from the nucleolar periphery [26]. This is not surprising since the nucleolus is the central hub for coordinating the response to cell stress in other species where the composition of the nucleolus is stress-dependent [76]. In keeping with the results summarized here, heat shock and AM-D treatment both cause nucleolar accumulation of specific human proteins [77].

Snf12 undergoes several translocations during mitosis (Figure 2). With nucleolar dissolution during prometaphase, it first shifts from its nucleolar locale to take up a predominantly nucleoplasmic location with some localization in the cytoplasm. During metaphase and through anaphase it then exhibits a relatively uniform cellular distribution before reacquiring its nuclear/nucleolar localization during telophase.

Rad53 (CHK2 in humans) is a tumor suppressor protein involved in DNA damage (genotoxic) stress response [78]. It is recognized as a nuclear protein that possesses a C-terminal bipartite NLS [79]. The Dictyostelium Rad53 homologue forkhead-associated kinase protein A, FhkA, is a nucleolar protein. Immunolocalization shows is resides at the periphery of the nucleolar patches (i.e., NoSC3) being more concentrated adjacent to the nuclear envelope. Like Snf12, actinomycin D treatment leads to its nuclear expulsion as nucleolar protein-containing vesicles that end up in the cytoplasm. Its mitotic dynamics are also fitting for a nucleolar protein, yet its specific function there remains to be elucidated. During mitosis FhkA redistributes throughout the cell with an enhanced level of localization evident adjacent to the nuclear envelope from prometaphase through telophase (Figure 2). Like NumA1, FhkA also localizes within the spindle fiber region.

As one of the last nucleolar proteins to be identified so far, less is known about Bud31 than any of the others. A comparative study of spliceosomal genes in Dictyostelium discoideum identified Bud31 but no further analysis of the protein was carried out [80]. In yeast where the protein was first identified, Bud31 is involved in cell cycle regulation, specifically functioning at the G1/S regulatory or start point [81]. A search of dictyBase.org indicates Bud31 is a putative RNA splicing factor or transcription factor. While the gene has been identified in humans and other species, its function in Dictyostelium has not been studied. Selected as a nuclear protein for comparison, it was shown that Bud31 localized throughout the Dictyostelium nucleolus along with NumA1 and eIF6 [18]. However, its location during mitosis and the presence of an NoLS have not been assessed.

Dictyostelium Src1 is homolog of the helix-extension-helix family that localizes adjacent to the inner nuclear membrane [72]. Because of its interaction with the major nuclear lamina protein NE81, its juxtaposition to the inner nuclear membrane and its unchanging location during the cell cycle, Src1 has attributes of an inner nuclear membrane protein that is involved in nuclear lamina formation. In other species, Src1 is also implicated in nucleolar organization through its ability to stabilize repetitive rDNA sequences [82]. GFP-Src1 and immune-transmission electron microscopy reveal that Src1 is an inner nuclear membrane protein that is tightly linked to the positioning of nucleoli in Dictyostelium [72]. Future work will have to be done to verify the significance of this relationship and whether Src1 fits the description of a true nucleolar protein. Since Src1 retains its localization throughout mitosis, this puts it in a position to serve as part of the reformation points for nucleolar reassembly during telophase (Figure 2).