Abstract
It has been proposed that the superphylum of Asgard Archaea may represent a historical link between the Archaea and Eukarya. Following the discovery of the Archaea, it was soon appreciated that archaeal ribosomes were more similar to those of Eukarya rather than Bacteria. Coupled with other eukaryotic-like features, it has been suggested that the Asgard Archaea may be directly linked to eukaryotes. However, the genomes of Bacteria and non-Asgard Archaea generally organize ribosome-related genes into clusters that likely function as operons. In contrast, eukaryotes typically do not employ an operon strategy. To gain further insight into conservation of the r-protein genes, the genome order of conserved ribosomal protein (r-protein) coding genes was identified in 17 Asgard genomes (thirteen complete genomes and four genomes with less than 20 contigs) and compared with those found previously in non-Asgard archaeal and bacterial genomes. A universal core of two clusters of 14 and 4 cooccurring r-proteins, respectively, was identified in both the Asgard and non-Asgard Archaea. The equivalent genes in the E. coli version of the cluster are found in the S10 and spc operons. The large cluster of 14 r-protein genes (uS19-uL22-uS3-uL29-uS17 from the S10 operon and uL14-uL24-uL5-uS14-uS8-uL6-uL18-uS5-uL30-uL15 from the spc operon) occurs as a complete set in the genomes of thirteen Asgard genomes (five Lokiarchaeotes, three Heimdallarchaeotes, one Odinarchaeote, and four Thorarchaeotes). Four less conserved clusters with partial bacterial equivalents were found in the Asgard. These were the L30e (str operon in Bacteria) cluster, the L18e (alpha operon in Bacteria) cluster, the S24e-S27ae-rpoE1 cluster, and the L31e, L12..L1 cluster. Finally, a new cluster referred to as L7ae was identified. In many cases, r-protein gene clusters/operons are less conserved in their organization in the Asgard group than in other Archaea. If this is generally true for nonribosomal gene clusters, the results may have implications for the history of genome organization. In particular, there may have been an early transition to or from the operon approach to genome organization. Other nonribosomal cellular features may support different relationships. For this reason, it may be important to consider ribosome features separately.
1. Introduction
Ever since the initial discovery of Archaea by Woese and Fox [1], the number of recognized lineages comprising the Archaea domain has rapidly increased. Currently, five superphyla (and 35-38 phyla) of Archaea are described in the NCBI database [2–4]. Of these, the recently discovered Asgard include several phyla and are the most diverse [5, 6]. To emphasize this, they have been referred to as a “superphylum” [7].
Amongst the three domains, there are approximately 102 recognized ribosomal protein (r-protein) families [8–11]. Essentially universal amongst these are seventeen large subunit and nineteen small subunit r-proteins. Many of these 36 r-proteins likely had an ancient origin possibly before the last universal common ancestor (LUCA) [10–16]. However, this conclusion is dependent on how the gene content of LUCA is identified. At least eight of these genes are found in multiple versions of LUCA (Rivas personal communication). The Archaea and the Eukarya also share eleven large subunit (LSU) r-proteins and twenty small subunit (SSU) r-proteins. However, neither group shares additional r-proteins with the Bacteria. For this reason, it has been hypothesized [17, 18] that the eukaryotic translation system originated from an earlier archaeal version and was subsequently expanded.
In Bacteria, approximately 32 r-proteins from both subunits and two translation-related proteins are grouped into seven well-studied clusters. They are, viz., alpha, beta/L10, L11, S10, S20, str, and the spc clusters [19, 20]. In E. coli and other well-studied organisms, these clusters have been experimentally shown to be operons [21]. In general, we will refer to them as clusters with the implication that many of them could likely be operons. The S10 and the spc operons (clusters) are the largest, each typically encoding ten-twelve r-proteins [21, 22]. Gene expression when known is frequently regulated by one of the r-protein components within the cluster. Besides these clusters, there are several smaller clusters as well. The r-protein clusters are widely conserved amongst bacterial species and remnants, and Bacteria-like clusters are also found in some chloroplasts and mitochondria [23]. However, it is not clear if the clusters seen in mitochondria are functional. If they are not, then their presence or absence is likely a mere reflection of their nonessentiality.
The universality of r-protein gene clusters is often linked with a long history of their association. They also represent ancient/regulatory relationships, with implications for early gene clusters in mini chromosomes. Eukaryotic transcription usually involves a single transcript. Thus, eukaryotic operons/clusters of genes are restricted to few organisms [24, 25]. While eukaryotes do have meaningful gene clusters, these clusters do not include r-protein genes [24, 26–29].
Earlier studies of non-Asgard Archaea showed that the universal r-protein genes are in clusters similar or identical to those found in Bacteria [20]. While three archaeal r-protein genes S4e, L32e, and L19e were found associated with the archaeal version of the spc operon, the nonuniversal archaeal r-protein gene L18e was part of the conserved L13 cluster. 17 r-protein genes were part of one of the ten previously unrecognized gene clusters. These clusters were found to be associated with genes involved in the initiation of protein synthesis, transcription, or other cellular processes. Since such associations in the universal clusters could not be found, it was posited that the ribosome had its own independent line of evolution [20].
It has been clear since their initial discovery that archaeal ribosomes more closely resemble those of Eukarya than Bacteria as has been documented in detail [30–32]. However, the non-Asgard Archaea have not been shown to have any strong specific relationship to Eukarya, and the matter was not initially aggressively pursued. Efforts to understand the relationship between the domains of life were rekindled when it was discovered that the then newly discovered Asgard Archaea have many nonribosomal features that are shared with Eukarya [7, 31, 32]. This has fostered the hypothesis that Eukarya may have descended from the Archaea [33]. Asgard Archaea were proposed to represent a historical link or bridge between Archaea and Eukarya [34–41]. A recent report has suggested that the genomic material is condensed and spatially distinct from the riboplasm within certain Loki- and Heimdallarchaeota cells, as further proof of the role of Asgard Archaea in eukaryogenesis [42]. Amongst the Asgard group, the Heimdallarchaeota have been proposed to be the closest to the Eukarya [43, 44]. Other analyses based on multiple gene sequences placed the Asgard between the TACK group of Archaea and Eukarya [7, 36, 40]. In contrast, one branch of the Asgard, namely, the Lokiarchaeota, and their close relatives were proposed to be closer to the Euryarchaeota, than to Eukarya [45].
Genes that are translated together frequently share similar origins, physical interactions, functions, or regulatory mechanism(s) [46]. The arrangement of genes in genomes is thus a window to understand how organisms are related. Assuming there was a universal ancestor of the three domains, it likely had a gene order/arrangement that possibly was a precursor of operons. Its evolutionary origins may have dated back to four billion years ago [47–51]. While some portions of this gene order underwent shuffling over evolutionary timescales [52, 53], the arrangement of some gene clusters continues to be conserved [54]. In particular, the translation machinery and the genes that encode the same are highly conserved. Their association frequently has had a long evolutionary history [55–58]. The origin of the ribosome has been posited to be strongly coupled with the early history and even origins of life [56, 59–63], which is further supported by recent papers by Bose et al. [64] and Bose et al. [65].
In order to gain further insight into the r-protein gene clusters in Asgard Archaea, the genome order of r-protein coding genes was analyzed and compared with non-Asgard archaeal and bacterial genomes.
2. Materials and Methods
2.1. Retrieval of Sequences of Genomes and Protein Coding Sequences
The data used herein are based on the availability of sequences as of July 2023. The feature table and GenBank/RefSeq sequences of the genomes of Asgard Achaea were obtained from the public databases of the National Center for Biotechnology Information (NCBI) [66–68] and the Integrated Microbial Genomes and Microbiomes (IMG/M) system of the DOE’s Joint Genome Institute (JGI) [69, 70]. All the raw genomes and the annotated gene and protein sequences of the same were saved as such and used. When required, r-protein sequences were obtained from the ProteoVision server [71]. Genomes that were not annotated and deposited as raw sequences were excluded.
2.2. Mapping of the r-protein Gene Cluster(s)
The feature table of each genome was first checked for the presence/absence of genes in any given cluster. The closest available homologs of those genes that appeared to be missing in a given Asgard genome were used to ascertain the presence/absence of the same using stand-alone blast [72]. Missing genes, which were found to be misannotated as “hypothetical proteins,” were identified using blastX search and/or gene sequence alignment and included in the cluster map(s) (Figures 1–5).
(a)
(b)
(a)
(b)
(a)
(b)
(a)
(b)
(a)
(b)
3. Results
A total of ~552 Asgard genomes grouped under the various phyla that were available on the NCBI server (https://www.ncbi.nlm.nih.gov/datasets/genome/?taxon=1935183) as of July 2023 were examined [36, 73]. Of all the Asgard genomes, only thirteen occur as a single complete (scaffold) genome(s) (Table 1). They will henceforth be referred to as MKD1, FW102, B-35, bin132, bin108 (Lokiarchaeota), LCB4 (Odinarchaeota), bin27 and bin8 (Thorarchaeota), PM71, PR6, bin6, bin76, and bin272 (Heimdallarchaeota). The next genomes that are of reasonably good quality (assembled into less than 20 contigs) with annotated genes/proteins are those of the Thorarchaeote strains FW25 (twelve contigs) and BC (nineteen contigs) and the Heimdallarchaeote strains AC18 (eleven contigs) and SZ_4 bin2.246 (sixteen contigs). They will henceforth be referred to as FW25, BC, AC18, and bin2.246, respectively. These genomes provide useful information for some of the clusters of interest but typically not all.
The remaining Asgard genomes each consist of at least twenty contigs and frequently many more. Given the incompleteness of these genomes, the r-protein coding genes in a given cluster were often found in multiple contigs. Some of them, as illustrated in Supplementary Figures 1a-b, either end or begin a contig. The genomes (contigs) of many other genomes have not been annotated into ORFs. Hence, their gene order is not available and therefore was not considered when mapping the clusters. In summary, only the complete genomes and those available with less than twenty contigs, as defined above, were used for the comparative analysis of r-protein gene arrangement. Thus, the choice of non-Asgard archaeal genomes for representation in the figures for comparison is based on the four major groups classified under Archaea [74], viz., the Asgard, the DPANN, the TACK, and the Euryarchaeota [2, 7, 34, 36, 40, 41, 75–78]. Furthermore, the phylogenetic relatedness as described in the literature was also taken into account [7, 34, 37, 39, 41, 79]. E. coli and B. subtilis were used as representatives of Bacteria.
3.1. Big Cluster Comprising Genes from S10 and spc
A core of 14 cooccurring genes (uL22-uS3-uL29-uS17-uL14-uL24-uL5-uS14-uS8-uL6-uL18-uS5-uL30-uL15) belonging to the segment of the S10-spc cluster is characterized and established as operons in E. coli and other Bacteria. In the archaeal (both Asgard and non-Asgard) genomes, this cluster core has 4 additional Archaea-Eukarya-specific genes (RNP1 (ribonuclease protein 1), S4e, L32e, and L19e) in Asgard and non-Asgard genomes (Figures 1(a) and 1(b)). The main core (of 14 genes) was identified as conserved in gene order, arrangement, and genome location in Asgard Archaea, non-Asgard Archaea, and Bacteria (Figure 1). The one exception is in the Thaumarchaeote N. aquarius Aq6f (Figure 1(b)). In this genome, a part of this cluster, viz., the segments L19e-L32e and uL18-uS5-uL15, occurs in the immediate neighborhood of the L11-L1-L10-L12 cluster.
In Lokiarchaeote sp. B-35, six genes of this cluster (uS3, RNP1, uL24, uS14, L32e, and uL15) are incorrectly annotated as hypothetical proteins. Nevertheless, this core occurs as a complete set in the complete genomes of Asgard Archaea (listed in Table 1) except four Heimdallarchaeote genomes (bin6, bin272, bin76, and bin2.246) (Figures 1(a) and 1(b)).
In the Thorarchaeotes bin8, bin27, FW25, and BC and the Odinarchaeote LCB, a gene for a heat shock protein occurs as an addition to this cluster (Figures 1(a) and 1(b)). A domain analysis using the Conserved Domain Database (CDD) [80] showed that this protein belongs to the family of HSP20 (small heat shock protein IbpA) family (COG0071) involved in posttranslational modification, protein turnover, and chaperones. Acquisition of this gene in these Asgard Archaea could be an unusual instance of horizontal gene transfer occurring within a highly conserved cluster. Additionally, in Thorarchaeotes bin8 and bin27, uL18 is annotated as a pseudogene.
In the Heimdallarchaeote bin6, the cluster is split into three segments found on different parts of the complete genome. Ten of these genes (uL22..uS14) (segment 1) are in one contiguous section, while uS8-uL6 (segment 2) occurs separately in a second segment. Six genes (L32e-L19e-uL18-uS5-uL30-uL15) (segment 3) are in another contiguous section. In bin272, segments 1 and 2 are similarly found separately on the genome. However, only a partial version of the gene L32e from segment 3 is present, while the rest of segment 3 is missing. In bin76, segment 1 is missing in entirety, while segments 2 and 3 are present. Additionally, in Lokiarchaeote bin108 and Heimdallarchaeotes bin76 and bin272, 10%, 15%, and 21% of the genes are partial. In the case of the Asgard genomes, which are only represented as contigs, the core occurs as a contiguous set on one contig in FW25, BC, and AC18. Only in the case of Heimdallarchaeote SZ_4_bin2.246 (sixteen contigs), segment 1 is on one contig, while segments 2 and 3 are both at different locations on another contig. This pattern is observed even in the Asgard genomes that consist of more than twenty contigs. In these genomes, the S10-spc r-protein cluster organization, despite being found in multiple contigs, is considered indicative of cooccurrence, if the genomes were complete (Supplementary figure 1). In all the non-Asgard Archaea examined, the entire cluster occurs as one contiguous set of genes (with the exception of the Thaumarchaeote N. aquarius AQ6f).
3.2. Small Cluster of Genes Normally Found in the S10 Operon
A second smaller cluster comprising the homologs of the most conserved genes of the bacterial S10 operon/cluster, namely, uS10-uL3-uL4-uL23-uL2, occurs independently on the Asgard genomes, separate from the rest. This feature is shared by some non-Asgard Archaea as well (Figures 1(a) and 1(b)). Part of the small cluster, uL3-uL4-uL23, always cooccurs. In contrast, uS10 is always found to occur independently from this cluster in both the Asgard and non-Asgard Archaea. In bin8, bin27, FW25, and BC (Thorarchaeotes) and PR6, PM71, bin6, bin272, AC18, and SZ_4_bin2.246 (Heimdallarchaeotes), uL3-uL4-uL23-uL2 occurs as one whole segment. In MKD1, FW102, and B-35 (Lokiarchaeotes), uL2 occurs independently, separate from the remainder of the small cluster. However, uL3-uL4-uL23 is entirely missing in bin132 and bin108 (Lokiarchaeotes).
In the bacterial genomes, E. coli and B. subtilis, the S10 and spc operons (clusters) cooccur as one contiguous cluster of twenty-one genes, including this smaller cluster of five genes. This is then followed by the smaller cluster of bL36-uS13-uS11-uS4-rpoA-bL17, which is part of another cluster, namely, the Alpha operon/L18e cluster, described below. Thus, in the bacterial genomes, the S10-spc operon/cluster is contiguous with a part of the S13…S9 cluster. L18e being Archaea-Eukarya-specific is not found in this segment.
3.3. Str/L30e Cluster
The core of this cluster is comprised of nine contiguous genes, which is observed to cooccur with the L11 cluster in B. subtilis (Bacteria). In the E. coli genome, the cluster is split into two sections. One section comprising rpoB-rpoC cooccurs with the L11 cluster, while the other section is S12..tufA. Each section is found in different parts of the genome (Figures 2(a) and 2(b)). The equivalent of the str/L30e cluster in non-Asgard Archaea (A. pernix K1 (TACK group; Crenarchaeota), M. jannaschii DSM2661 (Euryarchaeota)) and the DPANN group Archaea (Figure 2) has 9-10 genes, viz., rpoH-rpoB-rpoA-L30e-nusA-S12-S7-(fusA)-EF(1a)-S10. In the Thaumarchaeote N. aquarius AQ6f and DPANN group archaeon LC1Nh, EF(1a)-S10fusA occurs independent of this section. In the genomes of the Lokiarchaeotes FW102, B-35, bin132, bin108, and bin8; the Thorarchaeotes FW25 and BC; the Odinarchaeote LCB; and the Heimdallarchaeotes (bin6, bin272, bin76, and bin2.246), a part of this cluster, viz., rpoA1-rpoA2-L30e-nusA-S12, occurs as a conserved core. In the other Asgard genomes, this core is split into two parts, comprised of rpoA1-rpoA2-L30e-nusA (five genes) and S12-S7 (two genes), respectively, which are found on different regions of the genome(s). Notable are the complete genomes of the Heimdallarchaeotes PR6 and PM71, in which the entire cluster is dispersed with no conservation of the gene order (Figure 2(a)). Additionally, in the Thorarchaeote genomes FW 25 and BC, the str/L30e cluster cooccurs with a portion each of the L7ae (a new cluster described below) and the L18e clusters. Despite these two genomes occurring in fourteen and nineteen contigs, respectively, the cluster is fairly contiguous in one contig. The arrangement of the str/L30e cluster in the Asgard Thorarchaeota most closely resembles the arrangement found in the non-Asgard Archaea.
3.4. Alpha Operon/L18e Cluster
This cluster is an operon (S13-S11-S4-rpoA-L17) in bacterial genomes and is found immediately downstream of the S10-spc operon (Figure 3). In the non-Asgard Archaea, A. pernix K1 (TACK group) and M. jannaschii DSM 2661 (Euryarchaeota), this cluster is comprised of a contiguous core of eight genes, viz., S13-S4-S11-rpoD-L18e-L13-S9-rpoN (Figure 3(b)). Only in M. jannaschii DSM 2661, this cluster is in the genomic neighborhood of the uL3-uL4-uL23-uL2 section of the S10 cluster.
In the Asgard genomes, the L18e cluster is split into two main sections. The first is comprised of S13-S4-S11, which, in five Asgard genomes, is found cooccurring with L34e-L14e-Cbf5 in the genomic neighborhood of the big S10-spc cluster (Figures 3(a) and 3(b)). Cbf5 encodes an RNA-guided pseudouridylation complex pseudouridine synthase subunit. In three Asgard genomes, the S13-S4-S11 group is in the immediate genomic neighborhood of uL3-uL4-uL23-uL2.
The second section is comprised of rpoD-L18e-L13-S9 (Figures 3(a) and 3(b)) and is found independent of S13-S4-S11 in the genomes of Lokiarchaeota, Heimdallarchaeota, and three Thorarchaeotes. The gene for the DNA-directed RNA polymerase subunit N (rpoN), when not found in the main cluster, is often associated with the L7ae cluster. In two Asgard Thorarchaeote genomes, part of the L18e cluster cooccurs with the str/L30e cluster. In all the Asgard Thorarchaeote genomes examined, a partial portion of the L7ae cluster cooccurs with either a part of or the whole of the rpoD-L18e-L13-S9 segment.
3.5. S24e-S27ae Cluster
This cluster has a core of eight genes (gcp-pNP-S15-S3ae-S27ae-S24e-rpoE2-rpoE1). It cooccurs and is contiguous with the L7ae and the Alpha-L18e clusters in the Asgard Odinarchaeota (Figure 4(b)). gcp and pNP encode a bifunctional tRNA threonylcarbamoyladenosine biosynthesis protein and a noncanonical purine NTP pyrophosphatase, respectively.
Based on how it is found in the rest of the archaeal genomes, the S24e cluster can be split into two sections, one containing the three genes, pNP-S15-S3ae, and the other with five genes, rpoE1-rpoE2-S24e-S27ae-gcp. In all five complete Lokiarchaeote genomes, the pNP-S15-S3ae section occurs independently. In three of the Lokiarchaeote genomes, the second segment cooccurs with a part of the L7ae cluster. A section comprising S27ae-S24e-rpoE2-rpoE1 cluster cooccurs with the S6e-eIF2g-utp24 section of L7ae cluster in all the Asgard Thorarchaeota examined. In the two complete Asgard Heimdallarchaeota (PR6 and PM71), the two sections of the S24e cluster “sandwich” a small segment of the L7ae cluster, viz., S6e and Utp24-eIF2g. In four Heimdallarchaeotes, the pNP-S15-S3ae section is found in the immediate neighborhood of one section of the spc cluster. Amongst the non-Asgard Archaea, the S24e-S27ae-rpoE1 cluster is contiguous with the L7ae cluster, the Alpha-L18e, and the str/L30e clusters in the Desulfurococcales (Crenarchaeota) Archaea, namely, D. amylolyticus 1221n and S. hellenicus DSM 12710 (Supplementary Figure 2).
3.6. Identification of a New Cluster of r-protein Coding Genes
A new cluster L7ae comprised of eight genes (L7ae-S28e-L24e-ndk-infB-S6e-eIF2g-Utp24) was identified in the Asgard Odinarchaeota (Figure 4). The genes ndk, infB, eIF2g, and Utp24 encode a nucleoside diphosphate kinase, a translation initiation factor IF-2, the gamma subunit of translation initiation factor 2, and a 30S proteasome protein, respectively. In the Odinarchaeote genome, this cluster co-occurs with the entire S24e-S27ae cluster, in the following order: S24e-S27ae cluster, L7ae cluster, archaeal mevalonate pathway genes [81] and a partial L18e cluster. Such a contiguous arrangement of multiple clusters was also found in the genomes of D. amylolyticus 1221n and S. hellenicus DSM 12710 (Desulfurococcales (Crenarchaeota)) (Supplementary Figure 2). In three Lokiarchaeote genomes, L7ae..infB (5 genes, since ndk could not be found in these genomes using a BLAST search) cooccur with a portion of the S24e-S27ae cluster. In the two complete genomes of Heimdallarchaeotes PR6 and PM71, 3 genes, S6e-eIF2g-utp24, are sandwiched in between two parts of the S24e-S27ae cluster (Figure 4). In the two complete Asgard Thorarchaeote genomes, this 3-gene section S6e-eIF2g-utp24 cooccurs with a five-gene segment of the S24e-S27ae cluster. In the other two Thorarchaeote genomes, a partial portion (5 genes) of this new cluster is found sandwiched between the str/L30e and a partial section (6 genes) of the L18e clusters.
3.7. L31e and L11-L1-L10-L12 Clusters
The genes for the Archaea unique r-proteins, namely, L39e, L31e, S19e, and LXa, have been described to be part of a cluster that includes the conserved segment of 4 genes, viz., S19e-COG2118-COG2117-L39e [20]. This 4-gene conserved segment occurs as such only in the non-Asgard Archaea M. jannaschii. COG2117 is missing in all the other non-Asgard and Asgard Archaea examined (Figure 5). In the Asgard genomes of Odinarchaeota LCB4 and the three Thorarchaeote strains, this segment is part of a larger contiguous segment that includes the four gene L11-L1-L10-L12 cluster, which is described as an operon [82, 83]. The genes in this cluster include those encoding RNP (ribonuclease P protein component 4), RBP (an RNA binding protein), Nep1 (ribosomal RNA small subunit methyltransferase), Spt5 (transcription elongation factor), SecE (gamma subunit of the protein translocase SEC61 complex), RbsK (a carbohydrate kinase), FtsZ (a cell division protein), FtsY (a signal recognition particle-docking protein), and DtdA (D-tyrosyl-tRNA(Tyr) deacylase).
In the Asgard genomes examined in this study, except four Asgard genomes, all the other genomes show a conserved 5-gene core cluster, S19e-COG2118-L39e-L31e (Figures 5(a) and 5(b)). The remaining genes of the L31e and L11-L1-L10-L12 clusters show significant rearrangements in the Asgard Archaea. Overall, it can be divided into four tentative segments, viz., nep1-RNP-RBP-S19e-COG2118-L39e-L31e-eIF6, LXa-pfdA-ftsY, ftsZ-secE-spt5, and L11-L1-L10-L12.
In the Asgard Lokiarchaeotes MKD1 and FW102, part of the large segment, S19e..eIF6, is contiguous with LXa…secE, and in B-35, bin132, and bin108, nep1…secE occurs as a whole unit. The L11-L1-L10-L12 segment is on a different location in these genomes. In all the Asgard Heimdallarchaeotes examined, L11 is split from its operon/cluster (L11-L1-L10-L12) and found in other parts of the cluster. In the two complete Heimdallarchaeotes PR6 and PM71, the remainder of the operon L12-L10-L1 while on the negative strand is contiguous with the section containing S19e..eIF6 on the positive strand, followed by RBP..ftsZ on the negative strand. The three-gene segment ftsY-pfdA-LXa occurs independently. This pattern of rearrangement is identical in these two Asgard Heimdallarchaeotes, with the only difference being the strand orientation of all these segments is the opposite relative to each other. In the incomplete Heimdallarchaeote bin2.246 genome (16 contigs) and three Heimdallarchaeote genomes (bin6, bin272, and bin76), parts of these segments occur as L11, contiguous with L39e-L31e-eIF6-LXa-ftsY, followed by L10-spt5-secE. These three genomes also show truncated versions of the S10-spc cluster (Figure 1).
Additionally, an annotation error was observed in the Asgard Thorarchaeote BC genome. A hypothetical protein coding gene (GenBank accession number: TXT56565.1) is downstream of rbsK (GenBank accession number: TXT56564.1, which is also incorrectly annotated as a hypothetical protein). This gene is likely an artifact of erroneous annotation, because the coordinates of the same (gene for TXT56565.1: 125555-125674) overlap the coordinates of the gene ftsZ immediately downstream by 28 bases (gene for TXT56566.1: 125646-126779). Overall, the S19e-COG2118-L39e order is conserved in all the genomes with the exception of the non-Asgard Archaea Thaumarchaeota (Figure 5).
4. Discussion
Classical operons are found to occur in both Bacteria and Archaea [84, 85]. The conservation of the arrangement of the r-protein gene cluster is in congruence with how they are conserved in sequence [21, 22, 55]. A large chunk of 14 genes from the most highly conserved S10-spc cluster of r-protein coding genes was thus found to be similarly conserved in arrangement, whether it was Bacteria, Asgard, or non-Asgard Archaea. The arrangement of the four r-protein coding gene clusters as observed in the 17 Asgard archaeal genomes (Lokiarchaeota, Odinarchaeota, Heimdallarchaeota, and Thorachaeota) suggests that apart from the big cluster (S10-spc), the arrangement/order of the rest of the r-protein clusters is not conserved and is dispersed. In the context of one section of the Alpha operon found in the vicinity of the S10-spc cluster, the conservation of the location resembles what is observed in bacterial genomes. Assuming these Asgard genomes have been correctly assembled, a portion of the Alpha operon (L18e) cluster associated with the highly conserved and universal S10-spc cluster is likely a reflection of their association despite the phylogenetic distance between Bacteria and Asgard Archaea.
It has been suggested that the Heimdallarchaeota group is the most likely sister group of Eukarya [7, 36, 40, 43, 44]. The dispersive nature of the str/L30e cluster order and the unique arrangement of the S24e-S27ae, the L7ae, the L31e, and the L12-L10-L11-L1 clusters are particularly evident in the case of the Heimdallarchaeote genomes. The unique arrangement of the r-protein clusters, as observed in the case of the Heimdallarchaeota members of the Asgard in this study, may corroborate this. This could possibly be a marker for evolutionary divergence within the Asgard genomes. Additional complete Asgard genomes, when available, may help understand such divergences better. Examination of the plastid genome of A. thaliana and the gene order of (S10-spc cluster)-L36-S11 shows similarities with the order in Bacteria (E. coli, B. subtilis), though a section of this cluster is missing from the same (data not shown).
The dispersion of gene clusters in eukaryotes has been suggested to be directly correlated with higher rates of chromosomal/genome evolution [26, 86]. It remains to be seen if Asgard Archaea show a higher rate of genome evolution as compared to the non-Asgard Archaea.
Genome organization seldom remains in stasis with the phenomenon of lateral gene transfer constantly changing the dynamics of the same. It has been hypothesized that the Asgard proteome, particularly the eukaryotic signature protein coding genes, evolved through extensive horizontal gene transfer [36]. It is not clear from the few available complete Asgard genomes if the arrangement of the r-protein clusters is a result of this phenomenon. Nevertheless, the roles of gene order or arrangement in fashioning cellular physiology and genome evolution remain to be explored further [87].
It should be noted that the arrangement of the r-protein coding gene clusters in the Asgard could be in some cases an artifact of poor genome completion. The occurrence of several partial genes, a universal gene annotated as a pseudogene, and several universal gene segments missing in the Asgard genomes in this study (Figures 1–5) suggests that the genomes in which they are found may have been incorrectly assembled. The noncontiguous/disjointed occurrence of the otherwise conserved segment of 14 cooccurring genes of the S10-spc cluster in the three Heimdallarchaeotes (bin6, bin272, and bin76) (Figure 1(a)) and the absence of three of the most universal and highly conserved genes uL3-uL4-uL23 from the genomes of Lokiarchaeotes bin132 and bin108 (Figure 1) suggest that the assembly of these genomes is likely not accurate. The incompleteness or misassembly of genomic contigs has been observed previously [88–92]. In fact, binning errors were implicated in the absence of r-protein genes from metagenome-assembled genomes (MAGs), and hence, analysis of genomics and phylogeny of uncultivated microbes, using MAGs, may be fraught with erroneous interpretations [93, 94].
However, we did observe one interesting feature in a non-Asgard archaeal MAG. Of the three non-Asgard archaeal genomes assembled from metagenomes that are available in the NCBI database [88], only the genome of DPANN archaeon GW2011_AR10 (GenBank accession: CP010424) is complete as a single scaffold/chromosome. Examining the order/arrangement of the r-protein gene clusters in this genome showed that four of the five clusters reported in this paper cooccur contiguously. The arrangement of these clusters in this genome is as follows: str-L30e cluster (rpoH-rpoB-HP-rpoA1-rpoA2-L30e-nusA-S12-S7-fusA-EF(1a)-S10)-5HPs-S10-spc cluster (L3-L4-L23-L2-S19-L22-S3-L29-IF(Sui)-RNP1-S17-L14-L24-S4e-L5-S14-S8-L6-L32e-L19e-L18-S5-L30-L15-secY)-HP-L34e-L14e-L18e (Alpha operon) cluster (S13-S4-S11-rpoD-tRNA(leu)-L18e-L13-S9-rpoN-S2)-HP-gltX-HP-L7ae cluster (L7ae-S28e-L24e-ndk-dUTP diphosphatase-infB-S6e-eIF2g-Utp24-rpoE1-rpoE2-S24e-S27ae-10 ORFS-gcp-S15-HP-S3ae). The L31e and L11-L1-L10-L12 clusters are not part of this section and occur separately as two distinct sections elsewhere in the genome. Only in B. subtilis (Bacteria), a partially similar feature is observed with the L11 cluster found immediately upstream of the S10-spc cluster.
The second non-Asgard MAG, that of DPANN Candidatus Micrarchaeum acidiphilum ARMAN-2 (GenBank assembly: GCA_009387755.1) is in 8 contigs, with the r-protein gene clusters split and not conserved as reported for some of the Asgard genomes in the paper. The third non-Asgard MAG, that of DPANN Candidatus Forterrea multitransposorum archaeon (GenBank accession no.: CP045477), has not been annotated into ORFs (open reading frames)/genes etc. Thus, these sequences derived from metagenome-assembled genomes may not be sound.
It is not clear if the cooccurrence of four r-protein clusters in the MAG of DPANN archaeon GW2011_AR10 is a result of the accuracy of the assembly and annotation/curation of the genome. If it is indeed an accurate assemble, then the assembly of other genomes in the database may need to be revisited for improving better quality which may help in an improved understanding of arrangement of these clusters. Thus, in the absence of good quality genomes, ascertaining the order of highly conserved genes of r-proteins can be challenging, and thus, well-curated genomes are vital towards understanding genome organization. The importance of a more robust curation of MAGS as well publicly available Asgard genomes [6, 95] cannot be emphasized enough to get a better picture of the genome organization of the same.
5. Conclusions
The discovery of the Asgard Archaea has redefined our understanding of the phylogenetic positions of the various groups that comprise the Archaea. The genomic arrangement of the r-protein coding genes in the Asgard superphylum within the domain Archaea suggests that the order may follow the phylogeny. The closer an Asgard is to eukaryotes, the more dispersed is the arrangement of the r-protein gene clusters. This feature suggests a possible ancient gain or loss of the operon strategy of gene regulation in the early history of life. Given that very few Asgard genomes are currently complete, new data and better means of genome assembly will facilitate a better understanding of genome order in the Asgard Archaea in the future.
Abbreviations
| MKD1: | Lokiarchaeote strain Candidatus Prometheoarchaeum syntrophicum MK-D1 |
| FW102: | Lokiarchaeote strain Harpocratesius repetitus FW102 |
| B-35: | Lokiarchaeote strain B-35 |
| bin132: | Lokiarchaeote strain bin132 |
| bin108: | Lokiarchaeote strain bin108 |
| LCB4: | Odinarchaeote strain LCB_4 |
| bin27: | Thorarchaeote bin27 |
| bin8: | Thorarchaeote bin8 |
| FW25: | Thorarchaeote strain FW25 |
| BC: | Thorarchaeote strain BC |
| PM71: | Heimdallarchaeote strain PM71 |
| PR6: | Heimdallarchaeote strain PR6 |
| bin6: | Heimdallarchaeote bin6 |
| bin76: | Heimdallarchaeote bin76 |
| bin272: | Heimdallarchaeote bin272 |
| AC18: | Heimdallarchaeote AC18 |
| bin2.246: | Heimdallarchaeote SZ_4_bin2.246 |
| LUCA: | Last universal common ancestor |
| LSU: | Large subunit |
| SSU: | Small subunit |
| r-protein: | Ribosomal protein |
| mtDNA: | Mitochondrial DNA |
| NCBI: | National Center for Biotechnology Information |
| rRNA: | Ribosomal RNA |
| MAGs: | Metagenome-assembled genomes |
| IMG/M: | Integrated Microbial Genomes and Microbiomes |
| DOE: | Department of Energy |
| JGI: | Joint Genome Institute. |
Data Availability
The datasets used and analysed within the current study are available from the NCBI website as referenced in the paper.
Disclosure
A preprint has previously been published [96].
Conflicts of Interest
All authors declare they have no competing interests. LAK and ELS were student volunteers on this work. LAK was a fresh graduate from Clements High School, and ELS had just completed his 10th grade at Obra D. Tompkins High School, who volunteered with the group of Dr. George E. Fox at the University of Houston.
Authors’ Contributions
MRT and GEF conceived and designed the study. MRT obtained the sequences. MRT, RVS, LAK, and ELS compared the organization of the ribosomal protein coding genes in the genomes. RVS, LAK, and ELS prepared the comparative chart for the gene clusters. MRT prepared the figures. MRT and GEF prepared the manuscript paper that was finalized with the help from all the authors. All authors read and approved the final manuscript.
Acknowledgments
This work was supported in part by NASA (Contract 80NSSC18K1139) under the Center for Origin of Life, Georgia Institute of Technology, to George E. Fox and partly by the NSF (NSF-MCB-EAGER 2227347) to Madhan R. Tirumalai.
Supplementary Materials
Supplementary Figure 1: (a) the S10-spc cluster distributed in multiple contigs in Asgard Archaea. In Odinarchaeota LCB_4, arrows represent different locations on a single contig; adjacent arrows do not indicate their order on the genome; genes within an arrow are contiguous; the direction of the arrow denotes the strand orientation. Begins or ends a contig. Hsp20: heat shock protein 20 (a, partial gene; d, pseudogene). Red diamond represents gene(s) absent in the corresponding genome/location; arrows within dashed red boxes are contiguous. HP: hypothetical protein; RNP1: ribonuclease P protein component 1; NCG: noncluster genes. #ORF(s) annotated as HP(s). Hsp20: heat shock protein 20. Blank box represents the entire gene set missing in corresponding genome location. (b) The S10-spc cluster distributed in multiple contigs in Asgard Archaea. Adjacent arrows do not indicate their order on the genome; genes within an arrow are contiguous; the direction of the arrow denotes the strand orientation. Begins or ends a contig. Hsp20: heat shock protein 20 (a, partial gene; d, pseudogene). Red diamond represents gene(s) absent in the corresponding genome/location; arrows within dashed red boxes are contiguous. HP: hypothetical protein; RNP1: ribonuclease P protein component 1; NCG: noncluster gene(s). #ORF(s) annotated as HP(s). Hsp20: heat shock protein 20. Blank box represents the entire gene set missing in corresponding genome location. Supplementary Figure 2: the S24e-S27ae-rpoE1 cluster is contiguous with L7ae…Utp24, the Alpha-L18e, and the str/L30e clusters in the Desulfurococcales (Crenarchaeota) Archaea, namely, D. amylolyticus 1221n and S. hellenicus DSM 12710. Genes within an arrow are contiguous; red diamond represents gene(s) absent in the corresponding genome/location. ORFs: open reading frames. Arrows within dashed red boxes are contiguous. HP: hypothetical protein. (Supplementary Materials)