Genome Sequencing of Pathogenic Rhodococcus spp.

To the Editor: Infections caused by non-equi Rhodococcus spp. are uncommon but can cause severe pneumonia and bloodstream infections with sepsis (1–4). Increasing prevalence of immune-compromising illnesses and use of immunosuppressive agents might contribute to reemergence of these pathogens (5,6). Rhodococci infections may go undiagnosed or misclassified because of difficulties in laboratory identification, nomenclatural instability, and similarity of signs and symptoms to Mycobacterium tuberculosis infection (e.g., increasing cough, dyspnea, hemoptysis, and weight loss; failure to respond to broad-spectrum antimicrobial drugs; progressive cavitary lesion on repeated chest imaging). 
 
A 73-year-old immunocompetent man with chronic obstructive pulmonary disease who lived on a ranch in the southwestern United States was evaluated for possible malignancy of the left lung. Computed tomography showed expansion of an upper lobe cavitary lesion, and he underwent a partial lobectomy. Rhodococcus spp. infection was suspected on the basis of the finding of the progressive lesion, salmon-pink colony growth on chocolate agar, and gram-positive coccobacilli on Gram stain. The isolate, R1101, could not be further identified by using commercial automated systems in the laboratory and so was subjected to genome study. 
 
Alignment of 16S rRNA gene sequences showed that R1101 was most closely related to R. rhodochrous, with identities of 98.78%, 99.32%, and 99.86% to the 16S rRNA gene for type strains R. rhodochrous DSM43271, R. pyridinivorans PDB9, and R. gordonae W4937 (7), respectively. In contrast, sequence identity with R. equi type strain DSM20307 was 95.57%. Phylogenetic analysis with Rhodococcus 16S rRNA genes suggested that R1101 was a member of R. rhodochrous and evolutionarily separate from R. equi (Figure, panel A). A query against the National Center for Biotechnology Information nonredundant DNA database (http://blast.ncbi.nlm.nih.gov/Blast.cgi) found 60 rhodococcal 16S rRNA sequences that are 99%–100% identical to R1101, which demonstrates worldwide distribution of these closely related strains in diverse environments. 
 
 
 
Figure 
 
Phylogenetic and proteomic comparison of isolate R1101 and Rhodococcus spp. strains. A) 16S rRNA–based phylogenetic analysis. Complete or partial 16S rRNA gene sequences for 29 Rhodococcus spp. type strains were aligned with complete sequence ... 
 
 
 
Optical genome mapping (8) of R1101 generated a consensus full-length map showing a circular chromosome map of ≈4.32 Mb (± 5%) in length. The whole-genome restriction map for R1101 was compared with in silico NcoI restriction maps of 33 complete genome sequences from the family Nocardiaceae, retrieved from GenBank; these sequences included R. equi. No substantial similarity (alignment scores >15%) was detected between R1101 and these genomes. Thus, we determined that R1101 was not an R. equi isolate. 
 
Pyrosequencing of the R1101 genome yielded an average coverage depth of 14-fold. The total length of 1,020 assembled de novo contigs was 4.65 Mb, with a G + C content of 68%. A total of 3,969 putative protein-encoding genes were identified. Whole-genome phylogeny was used to show evolutionary distances between R1101 and the sequenced Rhodococcus species (Figure, panel B). We sampled ≈50.4 kb of sequences, encoding 67 putative proteins shared among all selected rhodococci, to generate the phylogenetic tree. Results confirmed that R1101 is a member of R. rhodochrous and is phylogenetically distant from R. equi. 
 
We used a BLASTP search to determine the similarity of predicted R1101 proteins to homologous proteins from R. equi 103S, R. erythropolis (PR4 and SK121), and R. pyridinivorans (AK37 and BKS6-46) (Figure, panel C). These data indicate that R1101 is highly similar to but distinct from R. pyridinivorans and R. rhodochrous at the protein sequence level and is more distantly related to R. equi. 
 
Each R1101 protein >1 database blast hit with an E-value 92% of the R1101 proteins showed greatest similarity to a protein from R. pyridinivorans, R. rhodochrous, or Rhodococcus sp. R04. However, 120 proteins (3%) have highest homology to their counterparts in the pathogen R. equi; 10 of these are unique to R1101 and R. equi. Furthermore, genes with the greatest similarity to R. equi are not randomly distributed throughout the R1101 genome. 
 
We calculated the probability of observing groups of adjacent R1101 genes that are most similar to R. equi and observed 1 cluster of 5 adjacent genes (p = 2.33 × 10−8), 2 groups of 6 adjacent genes (p = 6.75 × 10−10), and 2 blocks of 7 contiguous genes (p = 1.94 × 10−11). Nucleotide BLAST analysis showed that R1101 sequence contigs 604, 456, 139, and 610 align to nt 4454241–4469589 of the R. equi 103S chromosome, with >97% identity at the DNA level. The 12 proteins wholly or partially encoded within this 15.3-kb region show >99% identity at the amino acid level with their R1101 orthologs. Among them, contig 139 encodes a cluster of 7 consecutive proteins with high homology to R. equi; this portion of the R1101 genome is likely to have been acquired from R. equi through horizontal gene transfer. R1101 may have acquired clusters of virulence-related genes, such as those crucial iron-uptake proteins (9), from the phylogenetically distant species R. equi. 
 
In conclusion, infections caused by non-equi Rhodococcus spp. are rare, especially in immunocompetent patients, but may represent an emerging threat. Specialized diagnostics such as genome sequencing may be needed to accurately identify these pathogens.

To the Editor: Infections caused by non-equi Rhodococcus spp. are uncommon but can cause severe pneumonia and bloodstream infections with sepsis (1)(2)(3)(4). Increasing prevalence of immune-compromising illnesses and use of immunosuppressive agents might contribute to reemergence of these pathogens (5,6). Rhodococci infections may go undiagnosed or misclassifi ed because of diffi culties in laboratory identifi cation, nomenclatural instability, and similarity of signs and symptoms to Mycobacterium tuberculosis infection (e.g., increasing cough, dyspnea, hemoptysis, and weight loss; failure to respond to broad-spectrum antimicrobial drugs; progressive cavitary lesion on repeated chest imaging).
A 73-year-old immunocompetent man with chronic obstructive pulmonary disease who lived on a ranch in the southwestern United States was evaluated for possible malignancy of the left lung. Computed tomography showed expansion of an upper lobe cavitary lesion, and he underwent a partial lobectomy. Rhodococcus spp. infection was suspected on the basis of the fi nding of the progressive lesion, salmon-pink colony growth on chocolate agar, and gram-positive coccobacilli on Gram stain. The isolate, R1101, could not be further identifi ed by using commercial automated systems in the laboratory and so was subjected to genome study.
Alignment of 16S rRNA gene sequences showed that R1101 was most closely related to R. rhodochrous, with identities of 98.78%, 99.32%, and 99.86% to the 16S rRNA gene for type strains R. rhodochrous DSM43271, R. pyridinivorans PDB9, and R. gordonae W4937 (7), respectively. In contrast, sequence identity with R. equi type strain DSM20307 was 95.57%. Phylogenetic analysis with Rhodococcus 16S rRNA genes suggested that R1101 was a member of R. rhodochrous and evolutionarily separate from R. equi ( Figure, panel A). A query against the National Center for Biotechnology Information nonredundant DNA database (http://blast.ncbi. nlm.nih.gov/Blast.cgi) found 60 rhodococcal 16S rRNA sequences that are 99%-100% identical to R1101, which demonstrates worldwide distribution of these closely related strains in diverse environments.
Optical genome mapping (8) of R1101 generated a consensus fulllength map showing a circular chromosome map of »4.32 Mb (± 5%) in length. The whole-genome restriction map for R1101 was compared with in silico NcoI restriction maps of 33 complete genome sequences from the family Nocardiaceae, retrieved from GenBank; these sequences included R. equi. No substantial similarity (alignment scores >15%) was detected between R1101 and these genomes. Thus, we determined that R1101 was not an R. equi isolate.
Pyrosequencing of the R1101 genome yielded an average coverage depth of 14-fold. The total length of 1,020 assembled de novo contigs was 4.65 Mb, with a G + C content of 68%. A total of 3,969 putative protein-encoding genes were identifi ed. Wholegenome phylogeny was used to show evolutionary distances between R1101 and the sequenced Rhodococcus species ( Figure, panel B). We sampled ≈50.4 kb of sequences, encoding 67 putative proteins shared among all selected rhodococci, to generate the phylogenetic tree. Results confi rmed that R1101 is a member of R. rhodochrous and is phylogenetically distant from R. equi.
We used a BLASTP search to determine the similarity of predicted R1101 proteins to homologous proteins from R. equi 103S, R. erythropolis (PR4 and SK121), and R. pyridinivorans (AK37 and BKS6-46) (Figure, panel C, Appendix, wwwnc.cdc.gov/ EID/article/18/11/12-0818-F1.htm). These data indicate that R1101 is highly similar to but distinct from R. pyridinivorans and R. rhodochrous at the protein sequence level and is more distantly related to R. equi.
Each R1101 protein >1 database blast hit with an E-value <1 × 10 −10 was assigned a best match to another species on the basis of its bit score. Consistent with the 16S rRNA-based phylogenetic analysis, >92% of the R1101 proteins showed greatest similarity to a protein from R. pyridinivorans, R. rhodochrous, or Rhodococcus sp. R04. However, 120 proteins (3%) have highest homology to their counterparts in the pathogen R. equi; 10 of these are unique to R1101 and R. equi. Furthermore, genes with the greatest similarity to R. equi are not randomly distributed throughout the R1101 genome.
We calculated the probability of observing groups of adjacent R1101 genes that are most similar to R. equi and observed 1 cluster of 5 adjacent genes (p = 2.33 × 10 -8 ), 2 groups of 6 adjacent genes (p = 6.75 × 10 −10 ), and 2 blocks of 7 contiguous genes (p = 1.94 × 10 −11 ). Nucleotide BLAST analysis showed that R1101 sequence contigs 604, 456, 139, and 610 align to nt 4454241-4469589 of the R. equi 103S chromosome, with >97% identity at the DNA level. The 12 proteins wholly or partially encoded within this 15.3-kb region show >99% identity at the amino acid level with their R1101 orthologs. Among them, contig 139 encodes a cluster of 7 consecutive proteins with high homology to R. equi; this portion of the R1101 ge-nome is likely to have been acquired from R. equi through horizontal gene transfer. R1101 may have acquired clusters of virulence-related genes, such as those crucial iron-uptake proteins (9), from the phylogenetically distant species R. equi.
In conclusion, infections caused by non-equi Rhodococcus spp. are rare, especially in immunocompetent patients, but may represent an emerging threat. Specialized diagnostics such as genome sequencing may be needed to accurately identify these pathogens. 1 These authors contributed equally to this article.

Letters
Letters commenting on recent articles as well as letters reporting cases, outbreaks, or original research are welcome. Letters commenting on articles should contain no more than 300 words and 5 references; they are more likely to be published if submitted within 4 weeks of the original article's publication. Letters reporting cases, outbreaks, or original research should contain no more than 800 words and 10 references. They may have 1 Figure or Table and should not be divided into sections. All letters should contain material not previously published and include a word count.