Introduction

Species of the genus Mycoplasma have extremely small genomes, likely contributing to the need of the species to gain resources from host cells, and while Mycoplasma form a variety of relationships with hosts, many are pathogenic in vertebrates [1]. In North American tortoises, an upper respiratory tract disease is associated with both Mycoplasma testudineum and its close relative, Mycoplasma agassizii [2,3,4,5]. North American tortoise populations are in decline, with infectious disease as a possible agent in these declines [6,7,8], though importantly, our knowledge of the mechanisms of disease progression and its impacts on populations is lacking [9, 10]. To understand URTD, we must improve our understanding of the pathogens associated with the disease. By sequencing the genome of M. testudineum , we may gain insight into proteins associated with its pathogenicity and virulence.

Until now, DNA sequence data available for this species in GenBank was limited to ribosomal RNA genes and the associated intergenic spacer region, as well as the RNA polymerase beta subunit gene. To obtain genomic data on the species, we extracted DNA from a culture of the type-strain, BH29T, which was collected from the upper respiratory tract of a wild Mojave desert tortoise, Gopherus agassizii [3]. This sequencing work is part of a larger project addressing mycoplasmal variation among host species.

Organism information

Classification and features

M. testudineum infects the upper respiratory tracts of tortoises causing upper respiratory tract disease [3, 4]; however, recent investigations in wild tortoises suggest it may be present in the host without pathogenicity [11]. This microbe has been found in five tortoise species inhabiting North America— G. agassizii , G. morafkai, G. evgoodei, G. berlandieri, and G. polyphemus [3, 11,12,13]—and its presence has yet to be investigated in the sixth tortoise congener, G. flavomarginatus (located in north-central Mexico). From wild samples, there is some indication that M. testudineum may have a facilitative relationship with M. agassizii in tortoise hosts, but interactions with other community members are unknown [11].

M. testudineum is a sugar-fermenting, coccoid Mycoplasma , which is very similar in phenotype to the closely-related M. agassizii [3] (Table 1, Fig. 1). M. testudineum grows in culture at 22–30°C, with an optimal growth at 30°C [3] (Table 1). These temperatures are frequently experienced in their hosts during the seasons when tortoises are found to be most active [14, 15], though tortoise body temperatures can fluctuate well above or below these temperatures within a day and over the seasons [14,15,16].

Table 1 Classification and general features of Mycoplasma testudineum strain BH29T
Full size table
Fig. 1
figure 1

Transmission electron micrograph of thin section of Mycoplasma testudineum strain BH29T. Image from ref. [3], reproduced with permission from the publisher

Full size image

To determine the placement of M. testudineum in the mycoplasmal phylogeny, all 16S rRNA gene sequences from the type strains of Mycoplasma species were obtained from the SILVA database [17] and aligned using MUSCLE 3.8.31 [18], and a phylogenetic tree was constructed using the maximum likelihood method implemented in MEGA7 [19] (Fig. 2). M. agassizii is a sister group of M. testudineum in the resultant tree, and the M. testudineum / M. agassizii clade is a sister group of Mycoplasma pulmonis —the agent of murine respiratory mycoplasmosis, which also seems to be present in humans who are in contact with rodents [20]. All three species fall within the hominis group of Mycoplasma (see ref. [21] for group definitions). The M. testudineum 16S rRNA gene sequence is 93.1 and 89.2% identical to those of M. agassizii and M. pulmonis , respectively. Remarkably, these species are not closely related to Mycoplasma testudinis , isolated from the cloaca of a spur-thighed tortoise ( Testudo graeca ) in the UK [22], which are placed in the pneumoniae group. A previous taxonomic analysis placed M. testudinis within the pneumoniae group (in agreement with our results), but placed M. testudineum and M. agassizii in different hominis subgroups: the hyorhinis and the fermentans groups, respectively [23]. Our result is, however, in agreement with that by Volokhov et al. [24], which was also based on 16S rRNA data.

Fig. 2
figure 2

Phylogenetic tree of the Mycoplasma genus based on 16S rRNA gene sequences showing the phylogenetic position of M. testudineum BH29T (●). All 16S sequences from the Mycoplasma genus were obtained from the SILVA database [17]. Only sequences in the ‘The All-Species Living Tree’ Project (LTP), release 128, were retained. This dataset only contains sequences from type strains, designated with a superscripted “T”. Clostridium botulinum strain ATCC 25763 was also included in the dataset as outgroup. Sequences were aligned using MUSCLE 3.8.31 [18]. A phylogenetic tree was obtained using the maximum likelihood method implemented in MEGA7 [19], with 1000 bootstrap replicates. Species with available genomes at the NCBI Genomes database or the Genomes Online Database are represented in bold face. GenBank accession numbers are shown in parentheses. Bootstrap support values above 50% are represented. The scale bar represents a divergence of 0.05 nucleotide substitutions per nucleotide position

Full size image

Genome sequencing information

Genome project history

The type strain of M. testudineum , strain BH29T, was selected for sequencing. This strain was isolated from a nasal flush of the choana of a Mojave desert tortoise, which was filtered through a 0.45 μm filter and then grown in SP4 broth [2, 3]. Sequencing was conducted in October 2016. The Whole Genome Shotgun project was deposited at DDBJ/ENA/GenBank under the accession number NNCE00000000. The version described in this paper is the first version, NNCE01000000. A summary of the project information in compliance with MIGS version 2.0 [25] is shown in Table 2.

Table 2 Project information
Full size table

Growth conditions and genomic DNA preparation

Freeze-dried M. testudineum , strain BH29T, was obtained from the ATCC in November 2014 (ATCC 700618T) and had been cultured by the ATCC on Spiroplasma SP4 medium at 30°C in aerobic conditions. Genomic DNA was extracted using the Qiagen DNeasy Blood and Tissue Kit protocol for Gram-negative bacteria and eluted with ultra-pure water. Extracted DNA was quantified on a Qiagen QIAxpert system and with Picogreen analysis.

Genome sequencing and assembly

Genome sequencing was conducted using the Illumina Nextera XT DNA Library Preparation Kit (Illumina, Inc., San Diego, USA) with the Illumina NextSeq500 platform (150 bp, paired-end) and 2 ng of starting genomic DNA at the Nevada Genomics Center (University of Nevada, Reno). Sequencing was performed in multiplex with multiple samples, using dual index sequences from the Illumina Nextera XT Index Kit, v2 (index 1, N701; index 2, S502). A total of 455,422 read pairs were obtained. Using Trimmomatic, version 0.36 [26], reads were trimmed to remove Nextera adapter sequences and low quality nucleotides from either end (average Phred score Q ≤ 5, four bp sliding window), and sequences trimmed to < 35 bp were removed. After trimming, 412,763 read pairs and 36,907 single-reads (the pairs of which were removed) remained. De novo genome assembly was performed using SPAdes 3.10.1 [27], using as inputs the trimmed paired reads, and the trimmed single reads (assembly k-mer sizes 21, 33, 55, and 77; with read error-correction enabled and ‘--careful’ mode mismatch correction). After removing scaffolds of less than 500 bp, the final assembly consisted of 25 scaffolds with a total length of 960,895 bp, an average length of 38,435 bp, and an N50 of 130,815 bp. The coverage was 64×.

Genome annotation

Gene prediction was carried out using the NCBI Prokaryotic Genome Annotation Pipeline (PGAP) 4.2 [28]. For each predicted protein, (i) families were identified using the Pfam 31.0 [29] batch search tool (“gathering threshold” option), (ii) COG categories were assigned using eggNOG-mapper [30] based on eggNOG 4.5.1 data [31], (iii) signal peptides were identified using the SignalP server 4.1 [32], and (iv) transmembrane helices were inferred using the TMHMM server v. 2.0 [33]. CRISPR repeats were identified using PGAP and CRISPRFinder [34].

Genome properties

The properties of the draft genome are summarized in Table 3. The final assembly consisted of 25 scaffolds, with a total length of 960,895 bp and a G + C content of 27.54%. The small genome size and low G + C content is consistent with those of other Mycoplasma genomes sequenced [35, 36]. PGAP [28] identified a total of 788 protein-coding genes, 6 pseudogenes, and 35 RNA genes. The identified RNA genes include 3 rRNAs (one 5S, one 16S and one 23S), 3 ncRNAs and 29 tRNAs. PGAP identified 4 CRISPR repeats, and CRISPRFinder [34] identified 4 “confirmed” repeats, and another 3 that were flagged as “questionable” by the server. The numbers of protein-coding genes in each COG category [37] are summarized in Table 4.

Table 3 Genome statistics
Full size table
Table 4 Number of genes associated with general COG functional categories
Full size table

Insights from the genome sequence

Brown et al. [3] sequenced most of the 16S rRNA gene of M. testudineum strain BH29T (GenBank ID: AY366210). They had previously sequenced the homologous region for M. testudineum strain H3110, which differed only in one nucleotide position (GenBank ID: U19768, ref. [23]). Comparison of their BH29T sequence and that obtained by us revealed 5 point differences and an indel of 14 nucleotides (present in Brown et al.’s sequence but not in ours) (Fig. 3). Remarkably, 4 of the 5 point differences were located toward the ends of Brown et al.’s sequence, and thus may represent sequencing errors. The other differences probably represent mutations accumulated since the isolation of the strain in 1995. Our 16S rRNA gene sequence is identical to that generated by Volokhov et al. [24], with the exception of the first nucleotide of Volokhov et al.’s sequence. Nevertheless, the placement of M. testudineum in the tree (Fig. 2) is not affected by the particular sequence used.

Fig. 3
figure 3

Comparison of the 16S rRNA gene sequences generated by Brown et al. [3], by Volokhov et al. [24], and in our study. All three sequences correspond to M. testudineum BH29T. Asterisks represent identical sites

Full size image

In general, Mycoplasma cells need to adhere to mucosal epithelial cells of the hosts as a pre-requisite for pathogenesis. The mechanisms of adhesion are relatively well understood in Mycoplasma pneumoniae and its close relatives, but much less so in other Mycoplasma groups [38]. We used BLASTP and TBLASTN (E < 10− 5; low-complexity regions filtered out) to search for homologs of M. pneumoniae cytadhesins P1, P30, P65, P40 and P90 —proteins involved in adhesion— and cytadhesin accessory proteins Hmw1, Hmw2 and Hmw3 in all available Mycoplasma genomic data (nr database). We only found homologs in species closely related to M. pneumoniae ( Mycoplasma genitalium , Mycoplasma gallisepticum , Mycoplasma pirum , Mycoplasma alvi , Mycoplasma imitans , and M. testudinis ), as previously noted [38, 39]. Searches against the M. testudineum BH29T proteome detected no hits, and none of the 788 predicted M. testudineum proteins contained any of the Pfam domains present in the M. pneumoniae cytadhesins and accessory proteins (domains “CytadhesinP1”, “Adhesin_P1”, “Cytadhesin_P30”, “MgpC” and “EAGR_box”). These observations may have at least three alternative explanations: (i) the adhesion proteins used by M. pneumoniae may be specific to its group, (ii) adhesion proteins evolve very fast, perhaps due to co-evolutionary races, thus hindering the detection of distant homologs, or (iii) M. testudineum may exhibit limited adhesion capabilities. In support of the first possibility, M. pulmonis , the most closely related species to the M. testudineum / M. agassizii clade (Fig. 2), is known to have adhesion mechanisms different from M. pneumoniae : M. pneumoniae exhibits a specialized attachment organelle, whereas M. pulmonis adhesion takes place by generalized interaction of the pathogen and the host cell membranes [40]. The adhesins of M. pulmonis are unknown. In support of the second scenario, putative cytadhesins identified in M. pirum and M. gallisepticum are only 26–29% identical at the amino acid level to those of M. pneumoniae [41, 42].

To extend our search, we obtained a list of known Mycoplasma adhesins from the UniProt database [43] (search: “ Mycoplasma adhesin”). Again, BLASTP and TBLASTN searches (E < 10− 5; low-complexity regions filtered out) against the M. testudineum BH29T proteome/genome did not identify any significant hits. M. pneumoniae proteins GAPDH and EF-Tu and M. hominis protein OppA have been reported to be adhesins in addition to their traditional functions [44,45,46]. We found homologs of all three proteins in M. testudineum . It should be noted, however, that this does not guarantee that these proteins act as adhesins in M. testudineum . For instance, whereas M. pneumoniae EF-Tu binds fibronectin [45], M. genitalium EF-Tu, which is 96% identical, does not [47]. The M. testudineum protein is only 70% identical to that of M. pneumoniae , and serine 343, proline 345, and threonine 357 (replacement of which significantly reduces the fibronectin binding of EF-Tu in M. pneumoniae ; ref. [47]) are not conserved in M. testudineum . Additional work will be required to understand the mechanisms of adhesion in M. testudineum and its close relatives.

Conclusions

We have obtained a draft genome sequence of M. testudineum BH29T isolated from the upper respiratory tract of a desert tortoise with URTD in the Mojave Desert. Our analysis revealed some features typical of Mycoplasma genomes: a very small size and low G + C content. The new genome will enable comparative genomic studies to help understand the molecular bases of the pathogenicity of this and other Mycoplasma species.