Specific defects in the process of ribosome production lead to a heterogeneous group of human disorders that are well known today as ribosomopathies [1
]. The term ribosomopathy, with its correspondent meaning, was first suggested [3
] for the skin and bone marrow failure syndrome X-linked dyskeratosis congenita (X-DC), in which the human pseudouridine synthase dyskerin is mutated [4
]. Among its functions, dyskerin mediates the modification of approximately 100 specific uridine residues to pseudouridines in rRNA, an essential step of ribosomal biogenesis. Soon after, it was discovered that approximately one fourth of Diamond Blackfan anemia (DBA) patients harbor a mutation in the gene encoding the ribosomal protein (RP) S19 (or eS19 according to the new nomenclature [5
], suggesting that the term ribosomopathy could be shared by more than a single disease. The list of recognized ribosomopathies then grew rapidly to include Schwachman-Diamond syndrome (SDS), cartilage hair hypoplasia (CHH), and Treacher Collins syndrome (TCS) [7
]. Ever since the earliest classification attempts, these five disorders have been considered examples of known or suspected inherited ribosomopathies [10
]. Their number continued to grow further, coming to include a list of other less characterized inherited disorders, as well as acquired conditions such as the 5q− myelodysplastic syndrome [11
] and cancer [12
]. To understand the molecular mechanism underlying most of these disorders, it may be helpful to quickly review the fundamental steps in ribosome production in human cells.
Ribosomes are ribonucleoprotein complexes dedicated to messenger RNA translation and protein synthesis. Human cytoplasmic ribosomes are made of four ribosomal RNA molecules and approximately 80 proteins divided into two subunits: the large subunit (60S) and the small subunit (40S). The first accounts for three rRNAs, 5S, 5.8S, and 28S together with 47 ribosomal proteins (RPs). The small subunit is made up of 18S rRNA and 33 RPs.
Ribosome biogenesis is an intricate and coordinated process (reviewed in [14
]) that occurs in the nucleolus and later in the cytoplasm (Figure 1
); it involves more than 200 trans-acting factors [16
] that are required during the numerous steps of ribosomal subunits maturation. All three RNA polymerases are required in this process: RNA polymerase I is needed for the transcription of 28S, 5.8S, and 18S rRNAs [17
], RNA pol II produces the mRNAs of the 80 ribosomal proteins and numerous ribosomal processing factors, and RNA pol III synthetizes 5S rRNA [18
]. In the nucleolus, RNA polymerase I, complexed with different transcription initiation factors, synthetizes a large polycistronic transcript (47S) from the rDNA genes present in hundreds of copies within the cell [19
]. The pre-rRNA 47S obtained is made up of two external transcribed spacers (ETS) positioned at 5’ and 3′ of the molecule, while the sequence of 18S, 5.8S, and 28S are separated by two internal transcribed spacers (ITS1 and ITS2).
The maturation of this polycistronic transcript occurs in the nucleolus and starts with the action of a series of small nucleolar ribonucleoprotein complexes, namely C/D and H/ACA box snoRNPs [20
]. C/D box snoRNPs are made up of fibrillarin and accessory proteins such as Nop56, Nop58, 15.5K/NHPX, and the small nucleolar RNAs (snoRNA) characterized by box C and D. This complex, guided by the complementary hybridization with the sequence of the snoRNA, catalyzes the site-specific 2′-O
-methylation of the ribose of rRNA [21
]. Moreover, the complex is also involved in rRNA processing and folding [20
]. Recently, a new class of specialized C/D box snoRNA, which are able to guide cytosine rRNA acetylation has been reported [21
]. On the other hand, the H/ACA snoRNP complex is made up of dyskerin, Nhp2, Nop10 and Gar1, and the H/ACA snoRNA component. This complex acts similarly to the C/D box snoRNPs: the H/ACA box snoRNA guides the complex to a specific uridine in rRNA, while dyskerin catalyzes the conversion of this uridine into pseudouridine [22
]. These modifications can be found in functionally important regions of the ribosome, some of them being essential to regulate translational efficiency [23
] and fidelity [24
]. In addition, numerous C/D boxes or H/ACA boxes snoRNA not involved in ribose 2′-O
-methylation or in rRNA pseudouridylation, like U3, U8, U14, U22, U17, and some long non-coding RNAs like RMRP are involved in rRNA processing and maturation [25
The ribosomal RNA also undergoes a series of processes and assembly events that give rise to ribosome subunits. Some alternative processing pathways are described, although the most common one starts with the cleavage at the 5′ETS end (at the so-called site 01), removal of 3′ETS and, subsequentially, cleavage at site 2 of ITS1 (as extensively described in [15
]). This last cleavage by RNAse MRP has the important function of separating the maturation of the two ribosome subunits: the small subunit containing 18S rRNA, and the large subunit containing 5.8S and 28S rRNA. After exonucleolytic and endonucleolytic cleavages at the 3′ of pre-18S rRNA, this RNA and the ribosomal proteins of the small subunit constituting pre-40S particles are exported to the cytoplasm to complete their maturation. On the other hand, the maturation of the large subunit rRNA continues in the nucleolus. It has been reported that, in mammals, two different forms of 5.8S are present: a short (5.8Ss) and a long (5.8Sl) form both originating after the cleavage at site 2 of ITS1 operated by RNase MRP, as is the case with yeast. However, the alternative pathway leading to the formation of 5.8Sl is still unclear, while the trimming of the 5′ end of 5.8S sequence operated by XRN2 leads to the formation of the short form. A second endonucleolytic cleavage occurs in ITS2 and leads to the maturation of 28S rRNA. After the cytoplasmic assembly and nucleolar import of 5S RNP, pre-60S subunits also containing ribosomal proteins can be exported to the cytoplasm to fully complete their maturation process. After this important step, the missing ribosomal proteins are added to the two subunits while completing their maturation [26
]. The activation for translation of the nascent 60S subunit occurs when the anti-association factor eIF6 is removed from the large subunit thanks to the activity of EFL1 (elongation factor-like GTPase1) and its cofactor SBDS (Shwachman Bodian Diamond Syndrome) [27
]. The correct dissociation of the different assembly factors from both 60S and 40S subunits consist of the final structural quality control and allows the formation of the complete ribosome in presence of a messenger RNA and the translation initiation complex [27
A few years ago, in an attempt to provide a definition, De Keersmaecker, Sulima and Dinman suggested that a ribosomopathy is “any disease associated with a mutation in a ribosomal protein or biogenesis factor impairing ribosome biogenesis in which a defect in ribosome biogenesis or function can be clearly linked to disease causality” [29
]. The intention of the Authors was to provide a conservative definition that would not include disorders in which ribosomal defects were not causative. Still, this definition, in addition to “pure” ribosomal disorders, leaves room for disorders in which the defect in ribosome biogenesis only concurs to the pathophysiology of the disease. According to this definition, ribosomopathies can be further classified as disorders whose pathogenesis could be fully ascribed to the defect in ribosome biogenesis and/or ribosomal functions on the one hand, and disorders deriving from defects in multiple cellular functions including ribosome biogenesis on the other. Table 1
shows examples of ribosomopathies following this classification.
4. Recently Identified Ribosomopathies
In recent years, there has been an increase in the number of diseases identified as novel congenital ribosomopathies. These extremely rare diseases are characterized by mutations in ribosomal proteins or in factors involved in ribosome biogenesis, but further studies are necessary to fully understand the contribution of altered ribosome production in their pathophysiology. These ribosomopathies are heterogeneous diseases showing generalized multisystemic symptoms or, alternatively, more specific manifestations selective for one tissue or organ. An example of multisystemic disease is Bowen-Conradi syndrome, a rare autosomal recessive disorder first described by Bowen and Conradi [114
] in the Hutterite population, which is characterized by mental retardation, microcephaly, micrognathia, prominent nose, rocker bottom feet, and flexion contractures of the joints [115
]. This severe disease is the cause of early death in children; in fact, the average death age is 13 months. Recently, it has been reported that the cause of Bowen-Conradi syndrome is a mutation in the gene EMG1
, coding for a ribosome assembly protein, thus including this disease in the list of ribosomopathies [116
]. The mutation in EMG1 causes 18S rRNA processing delay, with the result of reduced cell proliferation rates and G2/M arrest [117
A more limited and specific effect is that caused by mutations in a ribosomal protein of the small subunit, RPS20. The mutation in RPS20
has been associated with a subtype of hereditary colorectal cancer (CRC) called Familial colorectal cancer type X (FCCTX), in which no mutation in mismatch repair genes was reported, but several pathogenic variants of predisposing genes were observed [118
]. Two different heterozygous mutations of RPS20
were reported in a CRC-affected family and in a patient with hereditary CRC. The experiment conducted on samples from the CRC-affected family showed that the haploinsufficiency of RPS20 caused a reduced ribosome biogenesis and consequent stabilization of p53, which is probably responsible for the selection of cells that escape p53 regulation [118
]. Since, for the two examples mentioned above at present, no additional extra-ribosomal functions of the products of the genes involved have been reported, they may be considered putative pure ribosomopathies although further research is necessary to confirm this definition.
A further example of the tissue-specific effect of mutations in ribosomal protein is represented by the outcome of mutations in RPL10. In fact, it has been reported that missense mutations causing an alteration in protein sequence can lead to a rare form of autism [120
] or microcephaly [121
]. A defective nervous system development can be caused by a decreased translational capacity of the cell coupled with an increased apoptosis due to the activation of ribosomal stress response. For RPL10, however, extra-ribosomal functions have also been reported [123
]. Ribosomal protein L10 in mitochondria serves as a regulator for the ROS level in pancreatic cancer cells [123
]; therefore, pending a more detailed characterization of the molecular pathogenesis of this disorder, it may be considered a putative mixed ribosomopathy.
In addition to the above-mentioned disorders, other recently identified ribosomopathies have been described. These conditions are listed in Table 2
5. Acquired Ribosomopathies
In addition to inherited ribosomopathies, a defect in the gene encoding for a ribosomal protein is also an underlying factor in an acquired myelodysplastic disorder termed 5q deletion (or 5q−) syndrome. This disorder, which is more frequently found in women over 75 years of age, is due to the somatic deletion of the short arm of chromosome 5, leading to macrocytic anemia and erythroid hypoplasia, which may subsequently progress to AML in some patients [133
]. The haploinsufficiency of the RPS14
gene has been identified by means of an RNA interference-based screening as the predominant cause of the myelodysplastic phenotype in 5q− syndrome [11
], indicating that the alteration in the ribosome biogenesis process may also be at the root of acquired disorders. Since for RPS14 no additional extra-ribosomal functions have been reported, 5q− syndrome may be considered a pure acquired ribosomopathy.
It has long been known that the process of ribosome biogenesis is highly deregulated in cancer [12
], suggesting that a subset of human tumors may also be considered acquired ribosomopathies. Mutations of NPM1
gene encoding the multifunctional ribosome processing factor nucleophosmin have been described as the most frequent mutation in acute myeloid leukemia [133
]. Whereas mutations of genes encoding for ribosomal proteins have been reported for the first time in pediatric acute lymphoblastic leukemia, where recurrent mutations of RPL10
genes have been found in approximately 10% of all cases [135
]. Very interestingly, a large-scale study on more than 10,000 genomes from human tumors of different origins indicated that the hemizygous deletions of ribosomal protein genes occur in more than 40% of the cases [136
]. In addition, a growing amount of data has become available on snoRNA mutations and expression alterations in human multiple cancer types [137
]. All these studies indicate that as ribosome biogenesis deregulation is a frequent feature in cancer. In many cases, cancer itself may be considered, at least to some extent, an acquired ribosomopathy. The exact role of most of these ribosome biogenesis alterations, however, still remains to be determined.
A list of acquired ribosomopathies is also shown in Table 3