Most of the sequences of Pollicipes analyzed in the present study, except the pseudogenes found in P. pollicipes, might be functional genes because they possess the necessary elements for gene expression, viz the presence of control elements in the NTS, the poly-T tail at the 3' end of the transcribing region, and conserved ICRs that function as internal promoters of the gene. Until recently it was thought that the NTS had no function, but studies of deletion mutants have shown that upstream control elements are required for the expression of 5S rDNA genes . The NTS minimun average size was 83 bp. This size agrees with Martins and Galetti , who proposed that an NTS of 60-80 bp can represent the minimum size for the organization of this rDNA in the genome. Although in general 5S displays a high degree of conservation among species and variants, we found some nucleotide substitutions in Pollicipes spp. In the comparisons with D. melanogaster, the ICR I and ICR II regions were the most conserved. Furthermore, the proportion of conserved nucleotide positions in Pollicipes spp. is higher than those obtained for razor clams  and mussels  which is not surprising since Pollicipes barnacles and D. melanogaster belong to the Arthropoda. The degree of conservation of internal control elements in the 27 crustacean sequences was 10/16 matches within ICR I, 6/8 matches within ICR II, 8/14 matches within ICR III, and 19/21 matches within ICR IV. Many nucleotide substitutions in ICR III were unique for Artemia spp. The highest degree of conservation was in Pollicipes spp. (11/16, 6/8, 12/14 and 21/21 matches respectively) (see Additional File 4, Figure S6). The poly-T tail transcription termination signal of 5S rDNA has been studied in several organisms and seems to be quite conserved. It is part of a transcribed 15-16 nucleotide segment specific to the 5S rRNA precursor. The 135-nucleotide primary transcript was identified in D. melanogaster by in vitro transcription and 3'-processed to yield the approximately 126-nucleotide pre' 5S species and the 120-nucleotide mature-size 5S rRNA . The analysis of upstream sequences of 5S from genus Pollicipes revealed a putative regulatory region, a TATA - like control element, located around positions -30 to -25 as observed in several fish species  and in razor clams . This region, together with RNA pol II-like transcriptional factors, may be involved in RNA pol III transcription . The high degree of conservation of TATA-like sequence positions in all organisms examined to date (e.g.. elasmobranch fishes, ) suggests a shared structural pattern. However, in our case, many sequences did not show the TATA-like motif. The fact that certain conserved regions are associated with a specific variant could be related to a differential expression throughout development as seen in Xenopus . Furthermore, we found a TTC sequence, as previously observed in the silkworm Bombyx mori .
The predicted secondary structure of all 5S sequences analysed in this work consists of five helices, two hairpin loops, two internal loops and a hinge region. This structure is consistent with the general eukaryotic 5S rRNA structure [24, 30] and with that obtained for A. salina . According to Smirnov , helix I is potentially important for RNA-protein recognition and helix III seems to be associated with the integration of 5S rRNA into the large ribosome subunit. Helix IV and the terminal loop are responsible for the interaction of 5S rRNA with 23S rRNA and are involved in the integration of the large subunit RNA component. Helix IV was conserved in all the predicted consensus secondary structures. However, although helix II was conserved in the comparison with the one from A. salina, a nucleotide substitution (G/A) in position 61 was found in some sequences. The ability of the sequence to adopt a correct consensus secondary structure can be used to discriminate between genes and pseudogenes . In this way, the putative pseudogenes of 123 bp did not fold.
In recent studies of molecular organization and evolution of 5S rDNA, several classes of 5S rDNA have been described, for example, in several species of fish , in razor shells , and in mussel species . The number of these different classes of 5S rDNA is low compared with the eight different types (per species) found in filamentous fungi . In this study we obtained several different classes of 5S rDNA which cluster into seven types (this number being the maximum of different variants found in animals). In these studies cited above, the main difference between classes of 5S rDNA is the length of the NTS types. In some cases the nucleotide sequence of the transcribing region also varies. In the case of Pollicipes there is no association between NTS and 5S.
In related species there is greater similarity among repeat units within the same cluster than among repeat units of different clusters; furthermore, similarity among repeat units within the same clusters from different related species is higher than among different clusters in the same species. According to this, some authors as Martins and Galetti  have suggested that different 5S rDNA loci evolve independently. We have found sequences that show a greater similarity in 5S rDNA units within a specific type between two species than between two types in the same species. Other studies have reported that the two types of 5S rDNA are not in separate clusters, since different variants have been found in tandem in the same clone [25, 36]. Similarily, we found that sequences belonging to different types were organized in tandem. We sequenced 9 dimers E-E, 2 dimers F-F, one dimer D-E, another dimer F-G, one trimer E-D-E and another 2 dimers that consisted of C type monomers linked to putative pseudogenes. By gene conversion, dimers and trimers, i.e tandem repeat units, should be composed of the same variants. However, we observed that this is not always the case. This tandem organization might therefore be caused by one or both of two reasons: 1) a variant may have recently been transposed, or 2) the unit of homogenization consists of dimers or trimers. Although these variants are in tandem, the repeats could be dispersed throughout the genome. At this moment, the attempts to locate these loci on metaphase chromosomes, by fluorescent in situ hybridization (FISH), have been unsuccessful.
The evolution of ribosomal gene families has traditionally been explained by the model of concerted evolution, which proposes that all members of a gene family are assumed to evolve in a concerted manner rather than independently, and a mutation occurring in a repeat spreads through all the member genes by repeated occurrence of unequal crossover or gene conversion . Therefore, sequence similarity is greater within a species than among related species . However, previous studies have shown that multigene families could be evolving under the birth-and-death model. Under this model new genes are created by gene duplication, and some duplicated genes are maintained in the genome for a long time, whereas others are deleted or become non-functional through deleterious mutations . Thus, according to the data (Figure 1) B and C variants could have originated after the colonization of the Atlantic ocean by P. pollicipes, whereas variants of the A, D, E, F, and G types are maintained in the three species so that their origin variant may have been present in the species' common ancestor. The case of ribosomal DNA could be more complex and involve a combined effect of concerted and birth-and-death evolution [3, 34]. Our data did not reveal a clustering by species. There were no fixed differences among species and low levels of nucleotide variation within the 5S region. However, divergence was observed among NTSs from different units. Taken together, these observations highlight the importance of purifying selection over the functional regions.
We found two putative pseudogenes in P. pollicipes. The presence of 5S rDNA truncated pseudogenes has also been described in other species, including humans , fishes , and filamentous fungi . As pointed out by Rooney and Ward , the truncated sequences are believed to be pseudogenes because their lack of an intact transcribing sequence effectively destroys the secondary structure of the 5S rRNA molecule that they would have otherwise encoded. The presence of pseudogenes in a multigene family strongly suggests that the family evolves under a birth-and-death process [39–41]. According to this model, a multigene family can expand as a consequence of gene duplication and contract because of gene loss (e.g. as a result of unequal crossover). Eventually, distinct gene copies accumulate differences, leading some of them to degenerate into pseudogenes . Under a birth-and-death process, the 5S rDNA multigene family is expected to show several variants, and the phylogenetic analyses of the genes of several closely related species will not show a within-species clustering pattern, but they should cluster according to their sequence similarities [3, 5, 7]. This agrees with the pattern obtained in the phylogenies and networks where sequences of 5S rDNA belonging to different Pollicipes species clustered together. In some species, 5S rDNA are dispersed throughout the genome, as in Schizosaccharomyces pombe . The dispersed gene organization apparently facilitates birth-and-death evolution wherein rRNA genes diverge from one another, some being unique to a given species, others shared among species .