Proteomic characterization of cellular and molecular processes that enable the Nanoarchaeum equitans--Ignicoccus hospitalis relationship

Abstract

Nanoarchaeum equitans, the only cultured representative of the Nanoarchaeota, is dependent on direct physical contact with its host, the hyperthermophile Ignicoccus hospitalis. The molecular mechanisms that enable this relationship are unknown. Using whole-cell proteomics, differences in the relative abundance of >75% of predicted protein-coding genes from both Archaea were measured to identify the specific response of I. hospitalis to the presence of N. equitans on its surface. A purified N. equitans sample was also analyzed for evidence of interspecies protein transfer. The depth of cellular proteome coverage achieved here is amongst the highest reported for any organism. Based on changes in the proteome under the specific conditions of this study, I. hospitalis reacts to N. equitans by curtailing genetic information processing (replication, transcription) in lieu of intensifying its energetic, protein processing and cellular membrane functions. We found no evidence of significant Ignicoccus biosynthetic enzymes being transported to N. equitans. These results suggest that, under laboratory conditions, N. equitans diverts some of its host's metabolism and cell cycle control to compensate for its own metabolic shortcomings, thus appearing to be entirely dependent on small, transferable metabolites and energetic precursors from I. hospitalis.

Figure 1. Analysis of protein size and membrane association effects on the number of assigned spectra.

Each circle represents an individual detected protein. Each panel represents an individual sample. Proteins with predicted transmembrane domains (TMD) are in red. The average adjusted spectral counts for proteins containing TMDs or not were calculated for each of the four biological samples. There is no correlation between protein size and the number of detected spectra, however, membrane proteins generated on average approximately three times less spectra than non-TMD proteins.

Figure 2. Proteomic coverage of I. hospitalis and N. equitans functional gene categories (arCOGs).

(A) Summary of arCOG proteins detection (% of total for each individual arCOG class). The values represent the combined samples (single culture/purified and co-culture) for I. hospitalis (blue) and N. equitans (red). The diamonds indicate classes with low representation in the genome (<10 predicted proteins). (B) Number of predicted genes/proteins assigned to the arCOG classes in the two genomes. (C) Percentage of identified spectra assigned to the individual arCOG classes relative to the total proteome from each of the four samples. (D) Average spectral representation assigned to individual proteins depending on the arCOG class (% of total spectra).

Figure 3. Changes in I. hospitalis relative protein abundance between the pure culture and the co-culture with N. equitans.

Proteins up-regulated in the co-culture are in dark green (>3 fold) or light green (2–3 fold). Down regulated proteins are in red (>3 fold) or orange (2–3 fold). Proteins that were only detected in one of the sample types are represented on either axis and an arbitrary spectral count (RSpC) threshold >10 was chosen for coloring. The scatter plot uses the sum of spectral counts for each protein between the three independent measurements for each sample.

Figure 4. Updated reconstruction of I. hospitalis-N. equitans metabolism and interaction (modified from [4]), incorporating results of recent physiological and ultrastructural studies , .

The localization of most I. hospitalis membrane complexes (outer vs. inner membrane) is unknown and arbitrarily depicted here as spanning both membranes. I. hospitalis proteins detected by proteomics are indicated by yellow diamonds (<1.5 fold change between pure culture and co-culture) or blue or red arrows (>1.5 fold increase or decrease, respectively in co-culture versus single culture). Black diamonds indicate proteins that were not detected. In N. equitans, the variation between the purified sample and the co-culture values are not being represented, the yellow/black diamonds only indicate detection/non-detection.

Figure 5. The I. hospitalis proteome and the effect of co-culture with N. equitans at the genome level.

Genes are arranged in rows with the coordinates on the left corresponding to the nucleotide position in the genome. Blue indicate RNA genes. In black are genes encoding proteins that were not detected. Green and red correspond to proteins that were up- or down-regulated, respectively, by 1.5 fold or more between the co-culture versus the independent culture samples. Grey indicates proteins with <1.5 fold variation between the samples. The direction of the slanted rectangle indicates the direction of transcription. Not all consecutive genes transcribed in the same direction are predicted to be in the same transcriptional unit (operon).

Figure 6. I. hospitalis proteins detected in the purified N. equitans sample.

The graph shows the normalized spectral count of detected proteins versus the enrichment factor, φ. The enrichment factor takes into account the difference in proteome abundance and complexity between the samples. A value of φ = 1 corresponds to no enrichment and can explain protein presence based on carry over/contamination. A region with φ>2 and NSpC>20 was shadowed as comprising of the most likely proteins transfer candidates or otherwise specifically associated with Nanoarchaeum. Below NSpC of 20 the low abundance may affect accurate measurements and those proteins were not taken into account. The insert cartoons depict possible explanations of detecting Ignicoccus proteins (green dots) in the Nanoarchaeum sample: (1) real inter-species transfer; (2) adherence of Ignicoccus membrane fragments to Nanoarchaeum, potentially mediated by a distinct “interactome” (red) or (3) sample contamination with Ignicoccus cellular debris.