|Home | About | Journals | Submit | Contact Us | Français|
A large collection of sequenced mycobacteriophages capable of infecting a single host strain of Mycobacterium smegmatis shows considerable genomic diversity with dozens of distinctive types (clusters) and extensive variation within those sharing evident nucleotide sequence similarity. Here we profiled the mycobacterial components of a large composting system at the São Paulo zoo.
We isolated and sequenced eight mycobacteriophages using Mycobacterium smegmatis mc2155 as a host. None of these eight phages infected any of mycobacterial strains isolated from the same materials. The phage isolates span considerable genomic diversity, including two phages (Barriga, Nhonho) related to Subcluster A1 phages, two Cluster B phages (Pops, Subcluster B1; Godines, Subcluster B2), three Subcluster F1 phages (Florinda, Girafales, and Quico), and Madruga, a relative of phage Patience with which it constitutes the new Cluster U. Interestingly, the two Subcluster A1 phages and the three Subcluster F1 phages have genomic relationships indicating relatively recent evolution within a geographically isolated niche in the composting system.
We predict that composting systems such as those used to obtain these mycobacteriophages will be a rich source for the isolation of additional phages that will expand our view of bacteriophage diversity and evolution.
The online version of this article (doi:10.1186/s12866-016-0734-3) contains supplementary material, which is available to authorized users.
Bacteriophages (phages) are the most abundant life forms on the planet, and can influence the entire bacterial ecosystem . They are capable of delivering, through horizontal gene transfer, mechanisms of adaptation and resistance, toxins and photosynthesis genes to their hosts. They also contribute to the cycling of organic and inorganic nutrients . According to the International Committee on Taxonomy of Viruses, double-stranded deoxyribonucleic acid (dsDNA) tailed phages constitute one Order (Caudovirales), three Families (Myoviridae, Podoviridae and Siphoviridae) and 39 Genera (http://www.ictvonline.org/virusTaxonomy.asp?taxnode_id=20120366) [3, 4].
Mycobacteriophages infect members of the genus Mycobacterium and were first isolated in the 40’s and 50’s [5, 6]. They were initially used as tools in the study of bacterial genetics and for diagnosis and typing of Mycobacterium tuberculosis [7–11]. More than 800 mycobacteriophage genomes have been sequenced (http://phagesdb.org) a collection that has grown rapidly due in part to the impact of integrated research-education programs [12, 13].
All mycobacteriophages isolated to date are dsDNA tailed phages and morphologically are Siphoviridae or Myoviridae . Mycobacteriophages have been isolated from clinical samples of diseased and healthy people [14, 15], animals , laboratory cultures [17, 18], but primarily from the environment, especially from soil to compost samples [19–21]. Most known mycobacteriophages were isolated using Mycobacterium smegmatis mc2155 as a host, comprising the largest collection of sequenced viruses infecting a common host . However, there is a geographical disequilibrium in mycobacteriophage isolation, and only about 7 % were isolated outside United States of America (USA) (http://phagesdb.org). The analysis of sequenced genomes shows exchange of genome segments among the phage population, revealing mosaicism as the main characteristic of such genomes . Nonetheless, genome comparison allowed the grouping of phages showing significant genomic similarities in clusters and subclusters .
Composting is a controlled process of plant and other living material decomposition by autochthone organisms that results in the production of compost to be used as fertilizer. Macro and microorganisms, as common bacteria, fungi, protozoan, algae, larvae and arthropods can participate in the degradation of organic matter. However, composting processes are typically carried out mainly by complex microbial communities . Such intense microbial activity promotes orchestrated changes in temperature and potential of hydrogen (pH) . The high diversity of organisms present at different times during the composting process offers an important source of environmental microorganisms and phages for study.
The São Paulo Zoo Park Foundation (FPZSP) has an area of 825,000 square meters (m2) and maintains an exhibition fauna of more than 3,200 animals, as well as the free-living fauna living in the neighboring Atlantic rain forest. The park has operated since 2007 as an environmental management program, including an effluent treatment plant, a water treatment plant and an organic compost production unit, which receives four tons/day of organic residues from Zoo activities and solid waste products from the effluent and water treatment stations. The humus-rich compost generated in the unit is used for the production of food for the Zoo animals in the São Paulo Zoo Farm and for fertilization of the park. The Sao Paulo Zoo composting process has been shown to hold a remarkable bacterial diversity, which can fully account for the biomass degradation . The composting operation was started in 2004 and processes 120 ton/month of organic waste for compost production. Its implementation allowed the recovery and preservation of environmental conditions of the park, reduction of fertilization products costs and better control of the food offered to the Zoo animals. Indirectly, the park obtains financial benefits.
The objectives of this work were to evaluate if compost from São Paulo Zoo Park is a suitable source for isolation of mycobacteria and mycobacteriophages, to understand the dynamics of mycobacterial and phage isolation during the compost processing and to compare isolated phages with mycobacteriophages from other countries. We report the genome sequences of eight newly isolated phages providing insights into viral diversity and the first evidence for mycobacteriophage evolution within a geographically isolated niche.
Several colonies of acid-fast bacteria were isolated from the São Paulo Zoo Park composting chambers 1, 2 and 3 at each time point (see Table 1), ranging from 2 to 20 per plate (data not shown). This finding is consistent with results obtained in a previous study of metagenomics with two composting chambers from the FPZSP, not related to the ones studied here. The study revealed that Domain Bacteria predominated during the whole process, comprising approximately 90 % of the analyzed sequences and the order Actinomycetales, which includes the Family Mycobacteriaceae, was included among the 10 most abundant orders . Moreover, Kitamura et al. showed that food waste composts contained relatively high amounts of Actinobacteria  and Partanen et al. confirmed that Phyllum Actinobacteria was predominantly isolated in the thermophilic phase of the composting process .
Identification of 38 randomly selected isolates from the first two collections by polymerase chain reaction (PCR) restriction enzyme analysis (PRA) of the heat shock protein 65 gene (PRA-hsp65) showed that most isolates belonged to the Mycobacterium fortuitum group with the following PRA-hsp65 patterns: M. fortuitum 1: BstE II [base pairs (bp)] (235,120,85) and Hae III [bp] (145,120,60,55), M. fortuitum 2: BstE II [bp] (235,120,85) and Hae III [bp] (140,120,60,55), Mycobacterium peregrinum 1 BstE II [bp] (235,210) and Hae III [bp] (145,140,100,50), M. peregrinum 3: BstEI I [bp] (235,130,85) and Hae III [bp] (145,140,100,60) and Mycobacterium septicum 1: BstE II [bp] (235,210) and Hae III [bp] (140,125,100,50). Albeit very similar, these PRA-hsp65 patterns could be clearly distinguished by electrophoresis in 3 % agarose gels (Fig. 1). Other detected PRA-hsp65 patterns were Mycobacterium kumamotonense 1: BstE II [bp] (320,115) Hae III [bp] (130,110,70); Mycobacterium terrae 1: BstE II [bp] (320,115) Hae III [bp] (180,130) and a novel profile: BstE II [bp] (235,120,100) Hae III [bp] (140,90,60).
One or two isolates showing each PRA-hsp65 pattern were submitted to sequencing of the gene that encodes the beta subunit of bacterial ribonucleic acid polymerase (rpoB) and of the hsp65 gene (Table 2). The sequences from isolates showing the PRA-hsp65 patterns of M. fortuitum 1, M. fortuitum 2, M. peregrinum 1, M. peregrinum 3 and M. septicum 1 showed highest similarity with sequences from species that belong to the M. fortuitum group. This group is composed of M. fortuitum, M. peregrinum, Mycobacterium senegalense, Mycobacterium conceptionense, Mycobacterium alvei, Mycobacterium mucogenicum, Mycobacterium setense and the 3rd biovariant sorbitol positive (Mycobacterium brisbanense, Mycobacterium houstonense and Mycobacterium mageritense) and sorbitol negative (Mycobacterium bonickei, Mycobacterium neworleansense, Mycobacterium porcinum and M. septicum) [29–32]. The separation of these species is problematic because they share most phenotypic characteristics, the similarity of deoxyribonucleic acid (DNA) sequences is high and the PRA-hsp65 patterns are very similar, some of them being shared by more than one species: the patterns M. peregrinum 2, M. porcinum 1 and M. septicum 1 are identical, as well as the patterns of M. senegalense 1, M. conceptionenese 1, M. houstonense 1 and M. neworleansense 1. A precise definitive identification was only reached for M. fortuitum and M. septicum, which showed concordant identification results by PRA-hsp65, rpoB and hsp65 sequencing. Nevertheless, considering these isolates in conjunct, the M. fortuitum group was the most prevalent in compost in this study, corresponding to 76.3 % (29/38) of the obtained isolates.
The isolate showing the PRA-hsp65 pattern of M. terrae 1 was identified as Mycobacterium senuense by rpoB sequencing and as Mycobacterium algericum by hsp65 sequencing. These species, and M. kumamotonense, whose identification was confirmed by rpoB and hsp65 sequencing, belong to the M. terrae complex [33, 34], which was the second group most prevalent in compost samples, corresponding to 21 % (8/38) of all isolates.
The isolate showing a novel PRA-hsp65 pattern not present in the PRASITE database was identified as Mycobacterium insubricum by rpoB and hsp65 sequencing. This recently described species was isolated from sputum samples but was not considered clinically significant  and from fish . This is the first description of its isolation from environmental samples and PRA-hsp65 pattern.
Mycobacteriophages were isolated from chambers 1 and 2 using M. smegmatis mc2155 as host. Crude extracts from the first collection (Mar 11, 2011) yielded 627 lytic plaques with material from chamber 1 and 484 with material from chamber 2. In the second collection (Mar 25, 2011) the number of lytic plaques was reduced to 14 in chamber 1 and 74 in chamber 2. Additional crude extracts collected at intervals of 15 days, were used for phage isolation, but after the fourth collection no lytic plaques were detected, showing that mycobacteriophages able to infect M. smegmatis were more prevalent in the initial phases of the process. The same experiment was repeated using two fresh samples from a different chamber, collected at 15-day intervals in the initial phase of the composting process, and 150 and 290 lytic plaques were obtained, respectively. Eight mycobacteriophages randomly selected were purified and stored.
We concluded that compost is a good source of mycobacteriophages capable of infecting M. smegmatis mc2155. To investigate if mycobacterial isolates from compost could also be used as hosts for isolation of mycobacteriophages from compost we used eight isolates representing the different observed PRA-hsp65 patterns (M. fortuitum 1, M. fortuitum 2, M. peregrinum 1, M. peregrinum 3, M. septicum 1, M. kumamotonense 1, M. terrae 1 and M. insubricum) as hosts. Each isolate was infected with the stored material from chambers 1 and 2 and fresh material from two additional chambers. No lytic plaques were obtained in these experiments. Control infections using M. smegmatis mc2155 as host in the same experiments were successful in generating lytic plaques (data not shown). Moreover, the eight purified phages previously isolated from chambers 1 and 2 were not able to infect the mycobacterial isolates from compost. The inability of phages to infect co-existing strains has been also witnessed in other environments, such as seawater cyanophages .
Mycobacteria isolated in this work may not include the natural hosts of mycobacteriophages from compost. Less than 1 % of environmental prokaryotes can grow in controlled conditions  and some mycobacteria can be difficult to recover from the environment. Moreover, few known phages infect the mycobacterial species identified in this work. Phage F-phi WJ-1 is the only described mycobacteriophage isolated from M. fortuitum . Additionally, the protocol used for phage isolation, obtained at http://phagesdb.org/workflow/ included an initial step of filtration of the composting material in 0.22 micrometer (μm) membranes in order to eliminate bacteria and other organisms. This implies that the isolated phages were present as free particles in the material. Occasional phages inside mycobacteria, either free or integrated in the genome, were most probably eliminated in the filtration step. Interestingly, the metagenomic analysis of two composting cells from the FPZSP, distinct from the ones analyzed here, revealed a low amount of viral sequences, corresponding to 0.05 and 0.25 % of the obtained sequences in each chamber .
Two sequencing runs yielded 689,550 and 518,987 reads respectively, with 615 nucleaotides (nt) mean length, totaling 424.3 mega base-pairs (Mbp) and 319.4 Mbp, respectively. The reads were assembled generating contigs between 49,117 bp and 69,377 bp. The number of reads in each contig ranged from 2,573 to 161,255. Final assembly revealed that genome sizes varied from 51,355 bp to 69,377 bp. Three phages were shown to have circularly permuted genomes and five had defined termini with 3’ single strand extensions of 10 bases in length (Table 3).
The eight phages sequenced here were compared to metagenomic reads obtained from a sample that included material from five different days during the composting process of a single chamber at FPZSP (unpublished work). Some hits were observed, particularly with phage Florinda, in sequences of 583 bp, 460 bp and 352 bp, all with at least 99 % identity. The evidence presented suggests that mycobacteriophage sequences can be at least partially recovered by metagenomics.
By comparison with sequences available in the Mycobacteriophages Database at http://phagesdb.org/ and based on the classification in clusters and subclusters proposed by Hatfull et al. , two phages (Nhonho and Barriga) were grouped in Subcluster A1, two phages (Pops and Godines) in Cluster B, one (Pops) from subcluster B1 and the other (Godines) in B2 and three (Florinda, Girafales and Quico) were grouped in Subcluster F1. One phage (Madruga) has high DNA sequence similarity with singleton phage Patience , and these two phages were grouped to form the new Cluster U. Functional assignments from database matches, HHPred results, and syntenic placement can be made for 25.6 % of the genes in these genomes. Across the eight genomes, ten new orphams have been identified, four in Subcluster A1 phages Barriga and Nhonho, and six in Subcluster F1 phages, Florinda, Giraffes, and Quico. The genomic features of each of these groups will be discussed in further detail.
Phages Nhonho and Barriga belong to Subcluster A1 and share extensive nucleotide sequence similarity across most of their genomes to other A1 phages, including phages from geographically dispersed locations (including China, South Africa and USA) (Fig. 2, Additional file 1: Table S1). Cluster A is the largest of all mycobacteriophage clusters, and includes more than 300 members, almost 40 % of the sequenced mycobacteriophage population. Phages Bxb1 and L5 – members of Subcluster A1 and A2 respectively – were among the first mycobacteriophages to be fully sequenced [42, 43], and Nhonho and Barriga share many genomic features with them (Fig. 2). The structure and assembly genes coding for terminase, portal, protease, scaffolding, capsid, major tail subunit, tail assembly chaperones, tape measure and minor tail proteins have the canonical syntenic organization, and the lysis cassette is located between the structural genes and the genome left end. The integration cassette that includes a serine-integrase is positioned near the center of the genome and separates the rightwards-transcribed structural genes of the left arm from the leftwards-transcribed genes in the right arm. The repressor genes (Nhonho 71 and Barriga 82) are closely related to the Bxb1 gp69 repressor (98 and 100 % aa identity respectively) and presumably are homoimmune to Bxb1  and the other Subcluster A1 phages.
Although Nhonho and Barriga have similar genomic architectures, there are several informative differences between them (Fig. 2). For example, in the Barriga right arm there are eight open reading frames of unknown function (72 – 79) that are absent from Nhonho, but present in a small subset of other Subcluster A1 phages. We also note that Barriga 84 is replaced by gene 73 in Nhonho, an orphan with no close relatives in the mycobacteriophage database. Finally, we note that Barriga gp52, which is a homologue of the Bxb1 Recombination Directionality Factor (RDF) gp47  lacks an intein that is present in the homologue (gene 52) in Nhonho.
Comparison of the gene contents of Barriga and Nhonho with those of other Cluster A phages illustrates that these two phages are more closely related to other Subcluster A1 genomes than those in other subclusters within Cluster A (Fig. 3). The Subcluster A1 phages are numerous and diverse, but Barriga and Nhonho are nonetheless quite closely related to each other, suggesting the possibility that they have evolved together relatively recently within the geographical region from which they were isolated.
Phages Pops and Godines belong to Subclusters B1 and B2, respectively, with close similarity to phages isolated elsewhere, notably within the USA (Additional file 1: Table S1) . Pops is most closely related to the Subcluster B1 phages Numberten and Suffolk, but differs from both in several genomic features (Fig. 4). Several of these differences suggest genes that are unlikely to be essential for lytic growth as they are present in some but not all of these genomes: First, Pops and Numberten share a gene (gene 54) of unknown function that is absent in Suffolk; second, Pops lacks a small gene (50 codons) present in both Suffolk and Numberten (genes 61 and 62 respectively), third, Pops 79 and Suffolk 78 are nearly identical, but the corresponding gene in Numberten (80) has a central deletion removing about 50 % of the gene (Fig. 4). Finally, Pops gene 92 is closely related to Numberten gene 95, but is deleted in Suffolk.
The Subcluster B2 phage Godines is most closely related to phages Ares and Arbiter (Fig. 4). Godines shares the common Subcluster B2 features such as the lack of Lysin B gene [47–49]. Another characteristic of this group is the presence of six genes between the gene for terminase and the left genome end (Fig. 4), whose products are predicted to be involved in the synthesis of queuosine from guanosine-5’-triphosphate (GTP).
Informative differences between Godines and other B2 genomes include the insertion of a small gene (56) of unknown function between the putative primase/helicase and DNA polymerase genes, which is has only one other relative present in the mycobacteriophages, in the Subcluster B5 phage Baee (Fig. 4). Also, a gene present in both Ares and Arbiter (77 and 74 respectively) is absent from Godines, and a Godines gene (89) is present in Ares (gene 90) but is absent from Arbiter; presumably these genes are not required for lytic growth.
Pops and Godines share only low DNA sequence similarly to each other and share fewer than 25 % of their genes through amino acid sequence comparisons. They are each more closely related to other phages than to each and thus there is no evidence to suggest they have evolved in evolutionary isolation in the geographic area from where they were isolated.
Three phages, Florinda, Girafales, and Quico are related to Cluster F phages, and all belong in Subcluster F1 (Fig. 5). Nucleotide similarity searches using Nucleotide Basic Local Alignment Search Tool (BLASTN) suggest that the three phages are somewhat more similar to each other than to other Subcluster F1 phages, and this is reflected in a phylogenetic representation based on gene content (Fig. 6). This presents compelling evidence that these three phages have evolved with some degree of isolation from the other Subcluster F1 phages, most of which were isolated within the United States, as suggested for Nhonho and Barriga above. Interestingly, there is no close general correlation between geological location and genome type for the mycobacteriophage phage collection as a whole, and the relatedness between Florinda, Girafales and Quico may indicate localization within the specific zoo location from which they were isolated. Comparison of the genome maps (Fig. 5) shows that the gene content relationships are likely dominated by a series of 12 small open reading frames (29–86 codons; Girafales genes 38–50 and their homologues) located immediately to the left of the integrase gene. These small open reading frames (ORFs) are nearly identical in Florinda, Girfales and Quico but not found in any other mycobacteriophage genome (Fig. 5).
Other notable features of these phages include the 7–8 genes – including the terminase, portal, and capsid subunit genes – at the extreme left end of the genomes (Fig. 5). The Cluster F genomes have two distinct types of these genes that have unrelated DNA sequences but encode similar functions: Florinda has one distinct type (related to genes in phages such as Daenerys and Llij), whereas Girafales and Quico have the alternative type (related to phages Boomer and Ramsey). Secondly, there is a small group of genes in Girafales (58, 59, 62) that substitute for unrelated genes in Florinda and Quico (Fig. 5). We also note that a small gene (115, 55 codons) present at the right end of the Florinda genome is absent from both Quico and Girafales, although found in otherwise unrelated genomes in Subclusters A1, D1 and D2. Its function is unknown (Fig. 5).
Phage Madruga is very closely related to the singleton mycobacteriophage Patience, with similarity spanning >95 % of their genome lengths (Table 3) , and as such, these two phages constitute a newly assigned Cluster U (Fig. 7). Patience has an unusually low percent of guanine and cytosine (GC%) content (50.3 %) and its proteome has been defined by mass spectrometry; it was isolated in 2009 in Durban, South Africa . Madruga therefore offers new perspectives from the comparative genomic analysis.
First, Madruga shares the low GC% content (50.4 %) to that of Patience, and the genomes are of similar lengths (69,377 bp and 70,506 bp in Madruga and Patience respectively). The length difference is primarily accounted for by three genes that are present in Patience but absent from Madruga: Patience genes 5, 7 and 104 (Fig. 7). Patience 5 is a putative histidine-asparagine-histidine motif (HNH) endonuclease and there are distantly related copies in Cluster C phages Myrna and Dandelion (39 % and 33 % aa identity respectively); the protein is expressed both early and late in lytic growth . Patience 7 is related to the Holliday Junction resolvase endonuclease VII which is expressed late in lytic growth  and although this function is likely required for DNA packaging, we note that Patience and Madruga both encode another Holliday Junction resolvase (RusA-like; genes 91 and 93 respectively) elsewhere in the genome that is expressed late in lytic growth and could provide functional redundancy (Fig. 7). Finally, Patience gene 104 is absent from Madruga and was shown to be expressed late in Patience lytic growth but is presumably not required; its function is unknown.
Finally, we note that putative tails proteins in both Madruga and Patience (gp34 and gp36 respectively) contain a C-terminal 150-residue domain that is also present at the C-termini of putative tail proteins encoded by other phages, including Pops (and other Subcluster B1 phages), and phages in Subcluster D1, D2, H1, H2, and Cluster R (Fig. 7). HHPred analysis strongly indicates that this domain has a carbohydrate binding function, perhaps with Endo-1,4-beta-xylanase activity. Presumably this is associated with host surface recognition or capsule degradation to facilitate the early stages of infection.
In conclusion, this work provides new insights into the genetic diversity of mycobacteriophages, and presents supporting evidence for local evolution of groups of phages (Florinda, Quico and Girafales; Barriga and Nhonho) in relative isolation from other mycobacteriophages. Although the database of sequenced mycobacteriophage genomes is quite large, this is the first clear evidence for this phenomenon, and may reflect in part the richness and growth of the microbial communities within the zoo composting systems from which they were isolated. Interestingly, none of the phages isolated here appear to infect any of the mycobacterial strains isolated from the same locations, and thus the natural host for these phages is unclear. We predict that composting systems such as those used to obtain these mycobacteriophages will be a rich source for the isolation of additional phages that will expand our view of bacteriophage diversity and evolution.
Composting material from 8 cubic meters (m3) chambers was collected for this work. Two chambers (1 and 2) were followed and four collections were done from March 11 to April 28, 2011. After this period the partially composted material from both cells was reunited in a third chamber, which was followed until June 10, with three additional collections. This is the regular procedure of compost production carried out at the FPZSP. At each collection, the temperature of five points located 1.5 meters (m) bellow the cell surface was checked and the mean temperature of the five points was calculated. The pH of the collected material was measured using a pH measure strip (EM Science, Germany). The composition, mean temperature and pH of each chamber and the air temperature at each collection are shown in Table 1. At each time point, material from five different points from each chamber was collected and pooled as previously described . The material was aliquoted in two 50 mililitre (mL) tubes, one used for mycobacteria and mycobacteriophage isolation and the other for storage at −20 °C.
M. smegmatis mc2155 was cultivated on solid Middlebrook 7H10 medium (Becton Dickinson and Company, Franklin Lakes, NJ, USA) supplemented with 10 % OADC (oleic acid, albumin, dextrose and catalase - Becton Dickinson and Co). Isolated colonies were then transferred to liquid Middlebrook 7H9 medium (Becton Dickinson and Co) supplemented with 10 % OADC, Tween 80® 0.05 % and CaCl2 1 mM (Merck, Darmstadt, HE, Germany) and cultivated at 37 °C for 24 h. The culture was diluted 1:20 in the same medium without Tween 80 until 0.5<D.O.600nanometers(nm)<0.7 (log phase) was reached.
Isolation of mycobacteria was carried out as described . The collected material [10 gram (g)] was diluted in 40 mL of sterile water and maintained in a shaker for 1 h (h) at room temperature (RT). The solution was centrifuged at 600 force of gravity (g) for 5 min (min) at 4 ° C for precipitation of large particles and the supernatant was transferred to sterile tubes and centrifuged at 8000g for 15 min at 4 ° C. Sodium dodecyl sulfate (SDS) 3 %/NaOH 4 % were added to the pellet, which was incubated for 15 min at RT. The sample was centrifuged again under the same conditions and the pellet was treated with a solution of cetrimide (Henrifarma Produtos Químicos e Farmacêuticos, São Paulo, Brazil) 2 % for 5 min. After centrifugation under the same conditions, the pellet was washed and resuspended in 0.5 mL of deionized water. An aliquot of 0.1 mL was distributed in duplicates on Löwenstein-Jensen (Probac do Brasil, São Paulo, Brazil) slants and on 7H10-OADC plates supplemented with PANTA (antibiotic solution containing final concentrations of 40 U/mL polymyxin, 4 μg/mL amphotericin B, 16 μg/mL nalidixic acid, 4 μg/mL trimethoprim and 4 μg/mL azlocillin) . The cultures were incubated at 30 ° C and 37 ° C and examined every 2 days in the first 10 days and once a week for a total of 90 days.
Isolated colonies were visualized by microscopy after Ziehl-Neelsen staining. Acid-fast bacilli were identified by PRA-hsp65. DNA was extracted by boiling a loop full of acid-fast bacilli taken from solid plates in 300 microlitres (μL) of TET [Tris–HCl 10 milimolar (mM), EDTA 1 mM, pH 8.0, Tween 20® 0.05 %] for 10 min. A 441 bp fragment from the hsp65 gene was amplified using primers Tb11 (5’-ACCAACGATGGTGTGTCCAT-3') and Tb12 (5’-CTTGTCGAACCGCATACCCT-3') as previously described . Amplified products were separately digested with BstE II at 60 ° C and Hae III at 37 ° C for 90 min. Digestion products were separated by electrophoresis in 3 % agarose gels stained with ethidium bromide and the resulting bands were analyzed by comparison with restriction profiles at the PRASITE (http://app.chuv.ch/prasite/index.html).
The identification of 11 selected isolates was confirmed by DNA sequencing. Partial sequences of rpoB and hsp65 genes were obtained. The primers used for amplification and sequencing of rpoB were MycoF (5'-GGCAAGGTCACCCCGAAGGG-3') and MycoR (5'-AGCGGCTGCTGGGTGATCATC-3')  and for hsp65 were hsp667-forward (5'-GGCCAAGACAATTGCGTACG-3') and hsp667-reverse (5'-GGAGCTGACCAGCAGGATG-3') . The amplicons were purified using QIAquick PCR purification Kit (Qiagen, Germany). Dideoxy sequencing was performed using BigDye® Terminator v 3.1 Cycle Sequencing Kit (Applied Biosystems, USA) and run in ABI PRISM 3100 DNA Analyzer (Applied Biosystems). The sequences were analysed by comparison with sequences deposited in the GenBank using the Basic Local Alignment Search Tool (BLAST: http://blast.ncbi.nlm.nih.gov/Blast.cgi).
Approximately 30 g of collected material were diluted 1:1 weight/volume percent (w/v) in phage buffer (Tris 10 mM pH 7.5, MgSO4 10 mM, NaCl 68 mM, CaCl2 0.1 mM). After homogenization and incubation at RT for 30 min, the samples were centrifuged at 3,500g for 15 min (Sorvall, Newtown, CT, USA) and the liquid phase was filtered through a 0.22 μm pore membrane (Techno Plastic Products AG-TPP, Trasadingen, SH, Switzerland). A volume of 10 μL of the filtered material was incubated with 0.5 mL of log-phase cultures of M. smegmatis mc2155 or mycobacterial isolates from compost at RT for 30 min and added to 2.5 mL of Middlebrook top agar (MBTA)/CaCl2 (7H9 liquid medium with agar 0.7 % and CaCl2 0.1 mM) and 2 mL of 7H9-OADC. The final volume was homogenized and poured over solid 7H10-OADC/CaCl2 plates, which were incubated for 16 h at 37 ° C. Negative controls (no addition of compost filtered material) and positive controls (phage D29) were included in each experiment. After the incubation period, putative lytic plaques showing different sizes and morphologies were collected by removing the agar regions using micropipet tips and transferred to microcentrifuge tubes containing 100 μL of phage buffer. Several rounds of purification were performed by infection of M. smegmatis mc2155 until all lytic plaques in each sample showed the same size and morphology. Each purified phage was then amplified and titered. The amplified phage samples were harvested, filtered through a 0.22 μm membrane and ultracentrifuged at 65,000g at 4 °C for 2 h (Beckman-Coulter Optima XL 100 K, Pasadena, CA, US). The pellet was suspended in 1 mL of ammonium acetate 0.1 molar (M) (Synth, Brazil), pH 7.5 or in phage buffer.
DNA from purified phages was submitted to pyrosequencing using standard Roche 454 GS FLX Titanium protocols (Roche Applied Science), at the Center for Advanced Technologies in Genomics (CATG), Instituto de Quimica, Universidade de Sao Paulo. Barcoded (tagged) shotgun libraries for each DNA sample were constructed using GS Titanium Rapid Library Prep Kit, pooled and submitted to two sequencing runs. Sequencing reads were separated according to their individual tags, quality-filtered and individually assembled using 454 Newbler assembler software version 2.5.3. Automatic annotation was performed using DNA Master software v. 5.0.2 (JG Lawrence, http://cobamide2.bio.pitt.edu/computer.htm), followed by manual curation. Hypothetical protein genes were analyzed for the presence of conserved domains using online tools CDD-NCBI  and InterPro-EMBL-EBI . Genome maps were generated using Phamerator  with the database ‘Actinobacteriophage_554’. This database contains 554 phage genomes of phages infecting Actinobacterial hosts, consisting of 56,572 genes, organized into a total of 8032 phamilies (phams) composed of sequence-related proteins. Parameters for pham construction were as described previously . Network relationships were derived from the Phamerator Pham table, and represented using Splitstree selecting only the Cluster F1 phages.
BLAST, basic local alignment search tool; BLASTN, nucleotide basic local alignment search tool; bp, base-pair; CATG, center for advanced technologies in genomics; DNA, deoxyribonucleic acid; dsDNA, double-stranded deoxyribonucleic acid; FPZSP, Sao Paulo Zoo Park Foundation; g, force of gravity; g, gram; GC%, percent of guanine and cytosine; GTP, guanosine-5’-triphosphate; h, hour; HNH, histidine-asparagine-histidine motif; PCR, polymerase chain reaction; hsp65, heat shock protein 65 gene; m, meter; M, molar; M., Mycobacterium; m2, square meter; m3, cubic meter; Mbp, mega base-pair; MBTA, Middlebrook top agar; min, minute; mL, mililitre; mM, milimolar; nm, nanometer; nt, nucleotide; OADC, oleic acid, albumin, catalase and dextrose supplement; ORF, open reading frame; PANTA, antibiotic solution containing final concentrations of 40 U/mL polymyxin, 4 μg/mL amphotericin B, 16 μg/mL nalidixic acid, 4 μg/mL trimethoprim and 4 μg/mL azlocillin; PCR, polymerase chain reaction; pH, potential of hydrogen; PRA-hsp65, polymerase chain reaction restriction enzyme analysis of the hsp65 gene; rpoB, gene that encodes the beta subunit of bacterial ribonucleic acid polymerase; RT, room temperature; SDS, sodium dodecyl sulfate; USA, United States of America; w/v, weight/volume percent; WGS, whole genome sequencing; μL, microlitre; μm, micrometer
We acknowledge the staff from FPZSP for helping with the collections. We are very grateful to Drs Joao Batista da Cruz and Paulo Bressan, respectively Scientific Director and President of the São Paulo Zoo Park Foundation (FPZSP) for their continued support to this project. Prof. Mario Hiroyuki Hirata is acknowledged for providing phage D29 for this study. We thank Charles Bowman and Steve Cresawn for assistance in database construction.
This study received financial support from São Paulo Research Foundation (FAPESP) (grants 2011/18326-4 and 2011/50870-6), and from the Howard Hughes Medical Institute grant 54308198 to GFH. JDL-Jr and GEM received fellowships from FAPESP (2012/04911-5 and 2014/01825-6). DVCO received undergraduate fellowship from National Council for Scientific and Technological Development (CNPq). JM-Jr is supported by a fellowship from Coordination for the Improvement of Higher Education Personnel (CAPES). LAD, AMDS, JCS and SCL are partially supported by Research Fellowship Awards from National Council for Scientific and Technological Development (CNPq). Funding agencies had no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.
The mycobacterial isolates and mycobacteriophages were deposited in the collection of the Sao Paulo Zoo Park Foundation (FPZSP) and in the collection of the Laboratory of Mycobacteria, Departamento de Microbiologia, Imunologia e Parasitologia, Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, Brazil. The material is available for non-commercial purposes upon request.
JDL-Jr carried out mycobacteriophage isolation, identification of mycobacteria and helped to draft the manuscript, CV-N carried out mycobacterial isolation and helped to draft the manuscript, DVCO carried out identification of mycobacteria and experiments of isolation of phages using mycobacteria from compost, GEM helped in the identification of mycobacteria, MCSR helped to standardize the experiments of phage isolation in the laboratory, JM-Jr participated in sequence assembly, LFM carried out sequencing of the mycobacteriophages, LAD participated in sequence assembly, AMDS participated in coordination of phage sequencing and helped to draft the manuscript, JCS coordinated phage assembly and sequence analysis and helped to draft the manuscript, DAR carried out sequence review and analysis, DJ-S carried out annotation reviews and prepared figures, WHP reviewed the analysis of the mycobacteriophages, GFH coordinated the review of the analysis and helped to draft the manuscript, SCL conceived the study and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.
JDL-Jr, DVCO, MCSR and GEM were students in the Laboratory of Mycobacteria, Universidade Federal de São Paulo; CV-N and SCL are Associate Professors in the same University and have a long-lasting collaboration in projects related to Molecular Biology of Mycobacteria; JM-Jr, LFM, LAD, AMDS and JCS belong to a group that work at Instituto de Quimica, Universidade de São Paulo in projects related to WGS and Bioinformatics. DAR, DJ-S, WHP and GFH are members of the Pittsburgh Bacteriophage Institute and have great expertise in mycobacteriophage characterization and analysis.
The authors declare that they have no competing interests.
This study has been approved by the Ethics Committee from Universidade Federal de Sao Paulo in May 11, 2012, under number 0068/12.
Complete genome sequences for eight phages obtained in this work were deposited in GenBank under accession numbers: [GenBank:KR997929, GenBank:KR997930, GenBank:KR997931, GenBank:KR997932, GenBank:KR997933, GenBank:KR997934, GenBank:KR997967 and GenBank:KR997968].
Additional file 1: Table S1.(50K, xlsx) Similarity of sequenced phages with phages deposited in the GenBank. The phage genomes sequences were compared with sequences deposited in the GenBank using the Basic Local Alignment Search Tool (BLAST: http://blast.ncbi.nlm.nih.gov/Blast.cgi). (XLSX 49 kb)
Similarity of sequenced phages with phages deposited in the GenBank. The phage genomes sequences were compared with sequences deposited in the GenBank using the Basic Local Alignment Search Tool (BLAST: http://blast.ncbi.nlm.nih.gov/Blast.cgi). (XLSX 49 kb)
James Daltro Lima-Junior, Email: moc.liamg@ortladsmj.
Cristina Viana-Niero, Email: email@example.com.
Daniel V. Conde Oliveira, Email: moc.liamg@ednocvleinad.
Gabriel Esquitini Machado, Email: firstname.lastname@example.org.
Michelle Cristiane da Silva Rabello, Email: email@example.com.
Joaquim Martins-Junior, Email: firstname.lastname@example.org.
Layla Farage Martins, Email: rb.psu.qi@alyal.
Luciano Antonio Digiampietri, Email: rb.psu@irteipmaigid.
Aline Maria da Silva, Email: rb.psu.qi@avlismla.
João Carlos Setubal, Email: rb.psu.qi@labutes.
Daniel A. Russell, Email: ude.ttip@87rad.
Deborah Jacobs-Sera, Email: ude.ttip@sjd.
Welkin H. Pope, Email: ude.ttip@niklew.
Graham F. Hatfull, Email: ude.ttip@hfg.
Sylvia Cardoso Leão, Email: email@example.com.