Cryptophytes are important organisms for several reasons. In terms of cell biology, their complex compartmentalization is of major interest, because several plasmas and genomes coexist in these organisms, which can be traced back to either a prokaryote or a eukaryote . One of the hallmarks of cryptophytes is the remnant of a second nucleus, which originated by the reduction of the cell nucleus of an engulfed phototrophic eukaryote by another eukaryotic cell . This compartment, the nucleomorph, is minimized in its coding capacity and expresses – in the case of Guillardia theta – only approximately a tenth of that of the E. coli K12 genome . The reduced coding capacity leads to the impression that the genes are still present in the nucleomorph may encode important functions. Thus, we are interested in addressing the functions of proteins encoded by the nucleomorph. However, due to the lack of a method of transfecting cryptophytes, we are studying homologs of the nucleomorph genes and their encoded proteins in genetically accessible organisms in order to identify the functions of the cryptophytic proteins indirectly. One of the best-studied and genetically accessible cyanobacterium is Synechocystis sp. PCC 6803.
Orf222 is one of the uncharacterized nucleomorph-specific open reading frames, for which homologs are present in many cyanobacteria. Analysis of the contribution of this gene within different organisms indicated that a clear correlation between orf222-homolog genes and phycobiliproteins is present, because at least one orf222 homolog is encoded in all organisms expressing phycobiliproteins, including red alga. Phylogenetic studies demonstrated that homologs of the orf222 gene can be classified into the following four groups (Fig. 1): Slr1649-like a, Slr1649-like b, CpeT-like a and CpeT-like b. Because the method for network construction as well as sampling in our studies is different from that of a recently presented phylogeny , it is not surprising that slightly different affiliations are resolved. However, our network corrects erroneous affiliations and indicates uncertainties of the basal grouping. This may be seen in the position of the bacteriophage sequence, which is in the network presented here in the neighbourhood of the bacteria they infect and not in the same branch as the cryptophyte sequence.
Despite the high degree of homology, the members of Slr1649-like and CpeT-like groups differ in the genomic context of the corresponding genes (Table 1). Members of the CpeT group are predominantly localized in the phycoerythrin associated linker protein operon [30, 31] next to the cpeS gene. In some cases, even the cpeR gene is localized directly downstream of cpeT. Because operon structures connect functionally related genes in many cases, CpeT could be an either structurally or functionally moiety of the phycobilisome and it has been shown to be responsible for the attachment of a bilin group to a specific site from β-phycocyanin . It is remarkable that a congruent distribution of members of the Slr1649-groups is not visible, because the genes seem to be localized randomly throughout different genomes. Interestingly, slr1649-homologs exist in some higher plants such as Oryza sativa and Arabidopsis thaliana (Fig. 1). The encoded proteins of these land plants are characterized by a DUF1001 domain as well, but obviously have paralogous functions, since the Arabidopsis thaliana homolog seems to be required for plastid division . It is also suggested to play an important role in cell differentiation and the regulation of the cell division plane in plants . The same could be true for the copy of the bacteriophage S-PM2, but seems to be unlikely since this phage infects different Synechococcus strains and its resource of the homolog may be the result of a selective advantage.
The homozygous knock-out mutant Δslr1649 in Synechocystis sp. PCC 6803 showed features identical to a cpcT knock-out mutant from Synechococcus sp. PCC 7002 described in Shen et al. . Here, the same pale green phenotype and a reduced phycocyanin content, resulting from a missing bilin group in phycobilisomes, was created by knock-out of cpcT, homolog to slr1649 homolog in this cyanobacterium. This indicates that the lyase function of the homologous proteins of Synechococcus sp. PCC 7002 and Synechocystis sp. PCC 6803 is comparable.
Nevertheless, we obtained one additional, not described feature in the Synechocystis sp. PCC 6803 knock-out mutant. Two linker proteins, CpcC2 and CpcD, were missing from the phycobilisomes in the knock-out mutant Δslr1649. CpcD is a small linker (10 kDa) located at the distal tip of rods, possibly functioning as a rod terminating factor . The CpcC2 rod linker (30 kDa) connects the most distal located phycocyanin discs . Both genes are located in the phycocyanin operon from which they are co-transcribed with the phycocyanin subunits and the cpcC1 linker gene . A transcriptional effect causing the loss of the linker proteins appears to be very unlikely, because the α-subunit, the β-subunit and CpcC1 linker are present, although the CpcD and the CpcC2 linker are simultaneously absent. This is indicative of our finding that the cpcC2 gene is indeed transcribed in the mutant as indicated by reverse transcription experiments (data not shown). Therefore, the deficit of the two linker proteins in mutant phycobilisomes is a post-transcriptional effect. However, we can not rule out that a decreased stability of phycobilisomes caused by the altered β-phycocyanin may be the reason for the lack of the two linker proteins in our preparations. In any case, the lack of the linker proteins is a molecular marker for the loss of lyase function, which may be interested to be studied in Synechococcus sp. PCC 7002  as well.
Guillardia theta, the cryptophyte on which we are primarily focusing expresses a homolog of slr1649 in association with phycoerythrin. Phycobiliproteins are located in the thylakoid lumen and apparently not organized in phycobilisomes in cryptophytes. Because the cryptophyte Guillardia theta uses phycoerythrin and not phycocyanin as an accessory pigment for photosynthesis, one might not expect that the putative cryptophytic lyase is able to complement the one of Synechocystis sp. PCC 6803. Surprisingly, the complemented strain showed wild type phycobilisomes structures as shown by the correct attachment of chromophore groups and the linker protein spectrum. Thus, Orf222 from the cryptophyte is able to complement the loss-of-function of Slr1649, indicating that the cryptophytic phycoerythrin lyase has still retained the capacity to couple a bilin group to β-phycocyanin, even after the progenitor of both classes of proteins evolved into apparently paralogous ones. However, a pleiotropic function of a biliprotein lyase with a specificity for phycobilin:cysteine-84 was recently shown in vitro for CpeS1 from Anabaena PCC 7120 , implicating that a multiplicity of proteins like the cryptophytic phycobilin:cysteine-β155 lyase has the capacity to couple bilins to homologous positions in a variety of phycobiliproteins.