First report of cis-1,4-polyisoprene degradation by Gordonia paraffinivorans

Abstract

The use of rubber has increased over the years, leading to a series of environmental problems due to its indefinite decomposition time. Bioremediation employing microorganisms have drawn an increasing interest and originated several studies of microbial rubber degradation. Genome sequencing and in silico analysis demonstrated that G. paraffinivorans MTZ041 isolate encodes the lcp gene (Latex Clearing Protein), responsible for expressing an enzyme that performs the first step in the assimilation of synthetic and natural rubber. Growth curves and scanning electron microscopy (SEM) were conducted for MTZ041 in natural (NR) and synthetic rubber (IR) as sole carbon source during 11 weeks. After 80 days, robust growth was observed and SEM analysis revealed the presence of bacilli and the formation of biofilm-like structures on natural and synthetic rubber. This is the first report of a G. paraffinivorans rubber degrader. Given the complexity of this substrate and the relative small number of microorganisms with this ability, the description and characterization of MTZ041 is of great importance on bioremediation processes of rubber products.

Introduction

Rubber is a polymer with elastic properties, also known as elastomer, which has a natural and synthetic origin. It is one of the most important raw materials developed by humankind, used in the manufacturing of over 40,000 products [1, 2]. Natural latex is produced by more than 2500 plant species, with Hevea brasiliensis being the most important commercially. Latex extracted from this tree is coagulated and goes through several industrial processes in order to become natural rubber (NR) [3]. Due to the large consumption of this material, rubber tree forests are not able to meet the demand, which forced the industry to seek an alternative polymer similar to natural rubber. Therefore, in 1909, Fritz Hoffmann developed the first synthetic rubber, known as methyl rubber. Since then, more than 500 types of synthetic rubbers have been developed; among them, isoprene rubber (IR) is the closest one to NR. Synthetic isoprene rubber is obtained from thermal cracking of naphtha. This technique consists in using pressure and temperature to break large hydrocarbon molecules into smaller ones, followed by polymerization of the material, which brings together the molecules of 2-methyl-1,3-butadiene, also known as isoprene [4]. NR and IR are both composed of C₅H₈ units (isoprene), each containing a double bond in cis configuration. In NR, these polymers correspond to 94% of its composition, whereas in IR, the percentage of cis-1,4-polyisoprene can reach up to 97% [5, 6]. Both materials undergo a process of vulcanization, which consists of heating the rubber at high temperatures in the presence of sulfur, forming sulfide bridges between isoprene chains that can contain up to five sulfur atoms, which are inserted with the double bonds contained in the polymer. The vulcanization gives the final product the desired physical properties such as greater heat resistance and elasticity and better ability to maintain its shape [5, 6].

In spite of its importance and diverse industrial applications, the main downside of rubber products is the high contamination potential of soil and water. Due to its cross linked structure and the presence of chemicals added during manufacturing, most rubber materials are not recycled; disposal of rubber products, such as tires, in landfills has been banned in many countries due to the decrease in available sites for this purpose, risk of explosion, and mosquito breeding. Burning in power plants is one method employed to degrade rubber materials. Even though this solves the storage complication, it also causes further environmental pollution [7].

Bioremediation methods employing microorganisms have drawn an increasing interest due to some advantages over physical-chemical techniques. They do not produce toxic compounds and consume less energy. Many microorganisms can degrade rubber. Among the Gram-negatives, Steroidobacter cummioxidans 35Y (former Xanthomonas sp. strain Y35) [8–10] and Rhizobacter gummiphilus NS21 have been well characterized, as well as their enzymatic system (LatA1, RoxA, and RoxB), which are responsible for the first steps in rubber degradation [11]. Several Gram-positive rubber degraders are well known and most of them are Actinobacteria that belong to the Gordonia [12, 13], Nocardia [14–16], Streptomyces [17, 18], and Rhodococcus genera [19]. Rubber degraders can be divided into two groups: the first forms clear zones on agar plates containing NR (R. gummiphilus, S. cummioxidans, and Streptomyces spp.) [5, 20–22], while the second does not show clear zones on agar plates; instead, members of the second group grow adhered to rubber, forming a biofilm (Nocardia, Gordonia, Corynebacterium, and Mycobacterium) [14, 23–26].

Hiessl et al. [27] reported the main steps of rubber degradation in Gordonia polyisoprenivorans VH2. Initially, the product of one or more lcp (Latex Clearing Protein) genes is required. These enzymes cleave the isoprene molecule, adding O₂ in the double bond, forming two molecules, one with an aldehyde group and the other with a ketone group. These molecules will enter several metabolic steps generating Acetyl-CoA and intermediaries used in β-oxidation, which may generate energy used for cellular growth [27]. The Gram-negative bacterium S. cummioxidans 35Y lacks the lcp, but is able to degrade rubber due to the action of the Rubber oxygenase A (RoxA) [10]. Myxococcus fulvus, Haliangium ochraceum, Corallococcus coralloides, all gamma-proteobacteria, and also R. gummiphilus NS21 have the gene roxA [28, 29]. Recent studies demonstrated that RoxA from R. gummiphilus is able to cleave polyisoprene to the C15 oligoisoprenoid [28]. Recently, a novel type of rubber oxygenase was reported in S. cummioxidans 35Y (RoxB) [30] and R. gummiphilus NS21 (LatA) [11, 31]. Comparisons of the amino acid sequence and other properties showed RoxB and LatA are homologues, but RoxA is not a homologue of RoxB/LatA [28–30]. Two species of Gordonia (G. westfalica and G. polyisoprenivorans) have been associated to rubber degradation; both have the Lcp and are able to use this substrate as sole carbon source [12, 13].

In this paper, we describe a novel rubber-degrading Gordonia species, G. paraffinivorans strain MTZ041, isolated from a compost unit at São Paulo Zoo, São Paulo, Brazil, which is able to grow on NR and IR as sole carbon source. The genome of MTZ041 was sequenced, and bioinformatics analysis revealed that it contains a lcp gene. This strain does not form clear zones on agar plates; therefore, it belongs to the second group of rubber degraders; also, it is able to form biofilm on NR and IR surfaces, which was observed by scanning electron microscopy (SEM). This is the first report of a G. paraffinivorans capable of using rubber as sole carbon source.

Materials and methods

Bacterial isolation, strain identification, and growth conditions

Gordonia paraffinivorans MTZ041 was isolated from a compost chamber, designated as ZC4, located at “Unidade de Produção de Compostos Orgânicos” (UPCO), Fundação Parque Zoológico de São Paulo (FPZSP), Brazil. The composition and samples retrieved from this compost chamber (ZC4) have been previously reported [32]. Compost samples were collected as previously described by Bitencourt et al. [33] and microbial isolation was done according to Dutra et al. [34]. Briefly, 1 g of the compost was diluted in 5 mL saline (0.9% NaCl), vigorously agitated, and incubated for 1 h at room temperature to decant. Next, 2.5 mL of the upper phase was inoculated in 500 mL flask with defined medium (250 mL M9 medium, 0.5% dextrose and carboxymethyl cellulose) and maintained on rotary shaker at 150 rpm at 30 °C. Every 72 h, the culture was harvested by centrifugation and the cell pellet was resuspended in fresh medium. After 10 days of cultivation, the cells were platted on nutrient agar for colony isolation. Bacterial identification was done by MALDI-TOF (Autoflex Bruker Daltonics), software FlexControl (Bruker Daltonics). The spectra were generated at 40-HzA laser, mass range 2000 a 20,000 Da. Identification was done with Biotyper Software (Brucker Daltonics) and the 16S rDNA gene sequencing was performed as previously described [34, 35]. Routinely, the MTZ041 isolate was maintained on nutrient agar plates as source of inoculum. Inoculums for growth curves were prepared on LB broth at 30 °C in various volumes according to the experimental design; cells were harvested by centrifugation at 4000 rpm for 15 min, washed, and resupendend in saline (0.9% NaCl). Growth curves were done on Bushnell Haas (BH) mineral salts medium (Difco 257820) containing the sole carbon source under analysis, cis-1,4-polyisoprene (Sigma-Aldrich 182141) or NR (latex gloves) according to the formulation of Linos et al. (2000) [36].

IR solid medium was prepared in Petri dishes containing BH plus 20 g/L agar overlaid with a thin layer of 3% IR solution (described below) and after evaporation of the solvent, the remaining IR film was covered with an additional coat of BH agar, forming an IR sandwich plate. Alternatively, the IR solution was spread as a thin film directly on the mineral agar. NR solid medium was prepared according to Jendrossek et al. [22] and Broker et al. [23]. BH medium added with 1% fructose was overlaid with a thin layer of NR latex concentrate being dispersed into BH agar at a concentration of 0.2% (wt/vol). Incubation of inoculated plates was performed at 30 °C. The negative control plate did not have any carbon sources and the positive control had an addition of 2% fructose.

IR and NR preparation and growth curve

Rubber as carbon source was prepared according to Linos et al. 2000 [36]. Briefly, NR from natural latex glove pieces were cut, weighed (0.25 g/fragment), and extracted with 100 mL acetone during 24 h. The glove fragments were washed, dried, and added to 250-mL Erlenmeyer flasks with BH medium and autoclaved. IR was prepared with 3 g of cis-1,4-polyisoprene, which was extracted in 100 mL of acetone for 24 h and then diluted in 100 mL of chloroform for 48 h (3% solution). Aluminum fragments (1 cm²) were immersed in the rubber solution (6 times), forming six layers of synthetic rubber that covered the aluminum completely. These pieces were sterilized in ethanol (100%) for 30 s and added to a previously autoclaved BH medium. The growth curves were held for 80 days in shaker with 150 rpm rotation and 30 °C in 100 mL BH containing NR or IR. Each treatment was inoculated at an optical density of 0.1 (OD₆₀₀) at the beginning of the experiment. In each time point of measurement, a 1-mL aliquot was collected and OD₆₀₀ was registered. Samples were collected every 24 h until day 5 and then every 5 days until the end of the experiment (80 days). Statistical significant differences among treatments were calculated by one-way ANOVA multiple comparison (p < 0.0001) with GraphPad Prism 7.0 Software.

Schiff’s staining

At the end of the growth curves, the NR fragments were removed and colored according to the protocol described by Linos et al. (2000) [36]. In a sealed bottle, 10 mL of the fuchsin reagent (1% fuchsin, 25% glacial acetic acid, 5% sodium metabisulfite, and 50% hydrochloric acid 0.1 N, diluted in water) were added to each fragment for 10 min at room temperature. Excess reagent was discarded and 10 mL of sulfite solution (5% sodium metabisulfite, 5% hydrochloric acid in water) was added to suppress nonspecific color reaction. The glove fragments were washed with distilled water and dried in the air flow.

Scanning electron microscopy

IR and NR fragments described before were removed from liquid cultures at the time of analysis and fixed with a 2.5% glutaraldehyde solution in 0.1 M phosphate buffer pH 7.4 for 24 h. The fragments were washed with sterile water and dehydrated for 10 min in graded ethanol (30, 50, 70, 90% and absolute ethanol). Dehydrated samples were subjected to the critical point drying with liquid CO₂ according to the manufacturer’s protocol (815 Autosamdri, TOUSIMIS). The samples were mounted on aluminum specimen stubs using electrically conducting carbon and gold-sputtered with a Desk V vacuum sputter device (Denton Vacuum). Micrographs were recorded digitally using a scanning electron microscope (JEOL KAL-6610LV) with secondary electrons at 20 kV acceleration voltages under high vacuum conditions. SEM analyses were performed in the Scanning Electron Microscopy Facility at the Universidade Federal de São Paulo (UNIFESP), Diadema, São Paulo, Brazil.

Genome sequencing, assembly, and annotation

The genomic DNA was extracted with Wizard Genomic DNA Purification Kit (Promega) according to the instructions provided by the manufacturer. Nucleic acid quantification was done by Nanodrop 2000c (ThermoFisher). The genome of strain MTZ041 was sequenced using Illumina MiSeq Plataform. Shotgun genomic library was prepared using Illumina Nextera DNA library preparation kit with total DNA input of ~ 35 ng. The resulting DNA fragment library was cleaned up with Agencourt AMPure XP beads (Beckman Coulter) and fragment size within the range of 400–700 bp was verified by running in the 2100 Bioanalyzer (Agilent) using Agilent High Sensitivity DNA chip. Fragment library quantification was performed with KAPA Library Quantification Kit and sequencing run was done with the MiSeq Reagent kit v2 (500-cycle format, paired-end (PE) reads). On average, Illumina PE read1 and read2 presented, respectively, > 80% and > 75% of bases with quality score at least 30 (Q30). Raw reads were assembled with MIRA 4 [37] and ABACAS [38]. The NCBI Prokaryotic Genome Annotation Pipeline (https://www.ncbi.nlm.nih.gov/genome/annotation_prok/) was used to carry out the annotation. The genome has been deposited in GenBank, and is available under the following identifiers: Bioproject PRJNA287716; Biosample SAMN03785434; accession LFLR00000000.

Bioinformatics analysis

The amino acid sequences of the Lcp protein from G. westfalica DSM44215 (WP_074850806.1) and G. polyisoprenivorans VH2 (WP_00636873) were used to search the MTZ041 genome with BLASTp and tBLASTn [39]. Once the coding sequence of G. paraffinivorans Lcp was found, the amino acid sequence was used to search various groups of bacteria. The amino acid sequences found by this search were used to construct a phylogenetic tree of Lcp amino acid sequences using the MegAlign program (DNA Star Lasergene).

RNA extraction and cDNA synthesis

An isolated colony was inoculated in 30 mL of LB and incubated at 30 °C and 150 rpm overnight. The bacterial culture was transferred to a 50-mL plastic tube and centrifuged at 5000 rpm for 15 min. The cell pellet was washed twice with 50 mL of milli-Q sterile water and resuspended in 5 mL of sterile milli-Q water. The cell number was adjusted to 5 × 10⁹ according to McFarland standard prior to the inoculation in 10 mL BH medium containing the carbon sources: 2% fructose, 1% NR, or 1% IR. The cultures were incubated in 50-mL Erlenmeyer flasks at 30 °C and 150 rpm for 2 h or 24 h. After 24 h, cell concentration in all culture conditions was adjusted to 5 × 10⁹ prior to RNA isolation. The minimum of four biological replicates were made for each treatment. RNA extraction was performed with the Midi RNeasy Kit (Qiagen) according to the manufacturer’s recommendations, adding 60 min of incubation in a water bath with 50 mg/mL lysozyme at 37 °C for cell lysis. Residual genomic DNA in the RNA samples was removed using RNase free DNase I (Roche), according to the manufacturer’s instructions. Efficiency of DNase treatment was checked by PCR amplification of 16S rDNA gene using as template the treated RNA samples and genomic DNA as positive control. The amplification products were separated by electrophoresis in 0.8% agarose gel stained with 250 μg/mL of ethidium bromide and visualized in ultraviolet light. cDNA synthesis was made with the RevertAid kit First Strand cDNA (ThermoFisher) using hexamers as initiators, reverse transcriptase and 3 μg of RNA.

qPCR

The cDNA samples were diluted 1:10 in ultrapure sterile water and used in quantitative PCR assays performed in real-time PCR System Thermal Cycler StepOnePlus™ (Applied Biosystems) with Fast Optical MicroAmp plates 96-Well Reaction Plate with Barcode and SyberGreen, according to the manufacturer’s protocol. Due to the high content of CG (cytosine/guanine) present in the samples, 5% dimethyl sulfoxide (DMSO) was added to the reactions. The gene-specific primers used were lcpf GAAGTTCCGCATCACCGTCG and lcpr CACCTACCTCGGCGTCATC, including a primer pair targeting the 16S rDNA gene used for qPCR normalization 16Sf GTATTACCGCGGCTGCTGGC, 16Sr CCAGACTCCTACGGGAGGCAGC [15, 40]. The amplification conditions were adjusted through StepOne V2.3 Software using the default settings for the Cycle Threshold (CT) and assays were performed in triplicate for each sample. Relative quantification was calculated by the 2^-ΔΔCT method [41] based on three biological replicates.

Results

Gordonia paraffinivorans MTZ041 isolation and identification

MTZ041 was isolated from a composting operation unit which has been intensively studied in the last few years [32, 33, 42] and has been a source of various microorganisms with potential biotechnological applications [34, 35, 43, 44]. The strain MTZ041 was subjected to MALDI-TOF and by comparison of the spectrum generated to a spectrum library (Biotyper software) showed similarity to Gordonia rubripertincta with a score of 2099. However, a score such as this enables secure identification only at the genus level. Therefore, we concluded that MTZ041 belongs to the genus Gordonia. For identification of the isolate at the species level, we sequenced a 1392-bp amplicon of the 16S rDNA sequences and used it to search GenBank. The result showed that the MTZ041 16S gene has 99–100% identity to 11 GenBank entries for G. paraffinivorans 16S sequences. This result indicated that MTZ041 is an isolate that belongs to G. paraffinivorans, which was further confirmed by the whole genome sequencing of this strain: Bioproject PRJNA287716; Biosample SAMN03785434; accession LFLR00000000.

The lcp gene from G. paraffinivorans MTZ041

The genome of MTZ041 encodes a latex clearing protein (Lcp) with 99% identity at the amino acid sequence to Lcp from G. paraffinivorans NBRC108238 (accession WP_006898821). The lcp gene is described as required for the assimilation of IR and NR as carbon sources in several actinobacteria species, including Gordonia sp. [36]. The gene is 1224 bp long and encodes a protein with 407 amino acids residues, with a predicted molecular mass of 45 kDa. Using the program MegAlign (DNA Star Lasergene), it was possible to construct a phylogenetic tree which shows that MTZ041 lcp gene is phylogenetically closer to G. westfalica than to G. polyisoprenivorans (Fig. 1).

Fig. 1.

Phylogenetic tree showing Lcp protein similarities between G. paraffinivorans MTZ041 and its closest matches from the BLAST search at the NCBI database. Sequences were aligned using the Clustal W Method. Numbers denote the sequence accession number from the BLAST search database. G. rubripertincta (WP_005200885.1); G. alkanivorans (WP_035754826.1); G. namibiensis (WP_006868302.1); G. westfalica (WP_074850806.1); G. amicalis (WP_024499951.1); G. paraffinivorans (WP_006898821); Gordonia sp. HS-NH1 (WP_055475823.1); Gordonia sp. KTR9 (WP_014925816.1); G. lacunae (WP_086537135.1); G. terrae (WP_101820330.1); G. polyisoprenivorans VH2 (ABV6892); G. polyisoprenivorans (WP_00636873); Gordonia sp. i37 (WP_079931519.1); G. iterans (WP105941566.1); Williamsia muralis (WP_062795088.1); Williamsia sp. 1138 (WP_084839014.1); Nocardia sp. 348MFTsu5.1 (WP_020107582.1); Rhodococcus kunmingensis (WP_068280989.1); Nocardia farcinica NVL3 (API85527.1); Nocardia farcinica E1 (ABC59140.1); Streptomyces sp. CFMR 7 (ALC27020.1); and Streptomyces sp. K30 (AAR25849.1). Bootstrap trails = 1000. Seed = 111

Also, the whole genome sequencing of MTZ041 provided information about genes for fatty acid degradation and beta-oxidation (Table 1). The comparison of the genes found in MTZ041 with the references strains G. paraffinivorans NBRC 108238 and G. polyisoprenivorans VH2, DSM 44266 suggests that these metabolic pathways are very similar among these strains (Table 1). In the context of rubber degradation, this result indicates that once rubber is assimilated into the cell, G. paraffinivorans MTZ041 would able to use it as carbon source.

Table 1.

Genes for fatty acid degradation and beta-oxidation (italic) present in MTZ041, Gordonia paraffinivorans NBRC 108238, and Gordonia polyisoprenivorans VH2, DSM 44266 genomes

Function ID	Name	MTZ041	Gordonia paraffinivorans NBRC 108238	Gordonia polyisoprenivorans VH2, DSM 44266
EC:4.2.1.17	Enoyl-CoA hydratase.	18	16	14
EC:2.3.1.9	Acetyl-CoA C-acetyltransferase.	15	13	11
EC:1.3.8.7	Medium-chain acyl-CoA dehydrogenase.	10	12	8
EC:1.1.1.1	Alcohol dehydrogenase.	8	10	7
EC:1.2.1.3	Aldehyde dehydrogenase (NAD(+)).	6	3	4
EC:6.2.1.3	Long-chain-fatty-acid--CoA ligase.	5	5	7
EC:1.1.1.284	S-(hydroxymethyl)glutathione dehydrogenase.	3	5	4
EC:1.2.1.8	Betaine-aldehyde dehydrogenase.	3	2	3
EC:2.3.1.16	Acetyl-CoA C-acyltransferase.	2	3	3
EC:1.1.1.35	3-hydroxyacyl-CoA dehydrogenase.	2	1	1
EC:5.1.2.3	3-hydroxybutyryl-CoA epimerase.	2	1	1
EC:1.3.99.-	Oxidoreductases. Acting on the CH-CH group of donors. With other acceptors.	1	1	1
EC:1.14.14.1	Unspecific monooxygenase.	1	1	1
EC:1.18.1.3	Ferredoxin--NAD(+) reductase.	1	3	4
EC:1.3.8.6	Glutaryl-CoA dehydrogenase (ETF).	1	1	2
EC:1.2.1.10	Acetaldehyde dehydrogenase (acetylating).	1	1	1
EC:1.3.3.6	Acyl-CoA oxidase.	1	1	1
EC:1.3.8.1	Short-chain acyl-CoA dehydrogenase.	1	1	1
EC:6.2.1.20	Long-chain-fatty-acid--[acyl-carrier-protein] ligase.	1	1	0
EC:1.3.8.8	Long-chain-acyl-CoA dehydrogenase.	1	1	1
EC:1.14.15.3	Alkane 1-monooxygenase.	0	0	0
EC:1.1.1.211	Long-chain-3-hydroxyacyl-CoA dehydrogenase.	0	0	0
EC:1.18.1.1	Rubredoxin--NAD(+) reductase.	0	0	0
EC:1.2.1.31	L-aminoadipate-semialdehyde dehydrogenase.	0	0	0
EC:1.2.1.47	4-trimethylammoniobutyraldehyde dehydrogenase.	0	0	0
EC:1.3.8.9	Very-long-chain acyl-CoA dehydrogenase.	0	0	0
EC:1.3.99.12	2-methylacyl-CoA dehydrogenase.	0	0	0
EC:1.6.2.4	NADPH--hemoprotein reductase.	0	0	0
EC:2.3.1.21	Carnitine O-palmitoyltransferase.	0	0	0
EC:2.3.1.40	Acyl-[acyl-carrier-protein]--phospholipid O-acyltransferase.	0	0	0
EC:5.3.3.8	Dodecenoyl-CoA isomerase.	0	0	0

Assessment of growth in presence of rubber as carbon source

To address the question whether MTZ041 has the ability to use rubber as sole carbon source and to place it in one of the two groups which classify the rubber-degrading bacteria, we inoculated MTZ041 on agar plates containing IR as sole carbon source. The isolate grew well and formed colonies on either type of plates (IR sandwich plates or rubber covering the surface of the medium) (Fig. 2). The comparison of the growth pattern of the negative (Fig. 2a) and positive controls (Fig. 2b) with the colonies on rubber (Fig. 2c, d) demonstrated that MTZ041 can grow robustly in these substrates. There was a lack of a clear halo around the colonies as seen in Fig. 2e.

Fig. 2.

G. paraffinivorans MTZ041 growth on rubber as substrate. BH medium without carbon source (a), BH supplemented with 2% de fructose (b), synthetic rubber (IR) sandwich plates (BH medium plus 3% IR, covered with an additional layer of BH agar) (c); BH medium with a surface layer of rubber (d), and BH medium with latex overlay medium (e)

Growth curve of MTZ041 in IR and NR

We performed growth curves of MTZ041 in IR and NR separately as sole carbon sources to characterize the growth pattern in these different rubbers. All growth curves were performed in duplicate. IR promotes rapid growth, with a very short exponential phase, entering stationary phase in the first 5 days (Fig. 3a solid triangles). During growth, there was pigment production in the culture, which is presumed to be carotenoids. The non-inoculated control presented no growth (Fig. 3a open triangles). Statistical analysis was performed and there is significant difference between these two treatments (p < 0.0001).

Fig. 3.

Growth of G. paraffinivorans MTZ041 in mineral salts medium (BH) with synthetic (IR) and natural rubber (NR) as carbons sources (a). Non-inoculated synthetic rubber (IR) control (white triangle), non-inoculated natural rubber (NR) control (white square), growth of G. paraffinivorans MTZ041 with 1% synthetic rubber (IR) (black triangle) and 1% natural rubber (NR) (black square) as the sole carbon source at 30 °C (numbers are average of two readings, semi logarithmic plot; double determination of growth). Natural rubber latex gloves stained with Schiff’s reagent to detect accumulation of aldehyde intermediates after incubation with G. paraffinivorans MTZ041

The growth in NR had a lag phase that lasted for 10 days and after this period, the culture entered exponential growth (Fig. 3a solid squares). There is an exponential increase until the end of the curve, and even after 80 days, the stationary phase was not reached. The non-inoculated control for NR (Fig. 3a open squares) showed increasing OD₆₀₀ over time; however, there is a significantly difference between the two treatments (p < 0.0001), as well as there is no growth when aliquots were plated in nutrient agar. In the final weeks of the experiment, it was possible to observe tiny pieces of rubber floating in the culture medium, which is probably the reason why OD₆₀₀ showed a slight increase over time. At the end of the growth curves, the fragments of NR were removed and stained with Schiff reagent to detect the presence of biofilm in the rubber against the negative control. The fragments from the negative control did not present any staining; therefore, we presumed that no biofilm formation occurred. As for the fragments from the inoculated medium, they presented a bright pink color, indicating the presence of exopolysaccharides produced in the biofilm.

G. paraffinivorans MTZ041 forms a biofilm on NR and IR surfaces

Scanning electron microscopy analyses were performed on the fragments of NR and IR during 11 weeks in order to address the question if G. paraffinivorans can form a biofilm. In the treatment with IR, it was possible to observe that, in the negative control, surface remains unchanged throughout the period of the experiment (Fig. 4a). Small holes can be observed in the structure of the material, but they result from chloroform evaporation during the preparation of synthetic rubber fragments. In Fig. 4a, no bacterial cells adhered to the material was observed after 80 days, demonstrating that there was no degradation or change in the structure of the material during the experiment for the non-inoculated controls. As for the treatment of synthetic rubber with G. paraffinivorans MTZ041, it was possible to observe the presence of cells in the first 4 days. The bacilli-shaped cells of G. paraffinivorans MTZ041 were not inside the biofilm, but some cells were starting to adhere to the synthetic rubber. Most bacteria were still suspended in the medium, which may explain why the optical density of the growth curve was higher at this stage (Fig. 4a). From the second week, the biofilm began to establish and some cells were attached to the material (Fig. 4b). As the biofilm evolved, micro colonies were formed and there was a well-developed matrix that served as a substrate, so other cells could adhere to the biofilm. In the sixth week, the IR was almost completely taken over by bacterial biofilms (Fig. 4c). In the last sample (11 weeks), IR was fully covered by biofilm; it was not possible to observe the IR or distinguish individual cells (Fig. 4d). The biofilm matrix was observed as a three-dimensional variable structure, with holes and channels in which the medium can flow through to ensure the survival of the cells.

Fig. 4.

Scanning electron micrographs of the synthetic rubber (IR) surface after incubation with G. paraffinivorans MTZ041. a The non-inoculated controls; b growth of G. paraffinivorans MTZ041 after 2 weeks; c after 6 weeks; and d after 11 weeks. Bars, 5 μm (a–c) and 50 μm (e, f). a–c At × 5000 magnification; d at × 450 magnification

The fragments of NR latex gloves incubated in the presence or absence of G. paraffinivorans MTZ041 were analyzed after the second week of treatment. Non-inoculated controls of these fragments (Fig. 5a) showed a rougher texture compared with the controls of synthetic rubber, but even with a more irregular surface, it was possible to observe the difference between the negative control (non-inoculated) and the sample inoculated with MTZ041. In 6 weeks of growth, there were small holes on the surface of the NR (Fig. 5b), which showed bacterial colonization. The holes on the surface of the NR increased gradually and at 9 weeks, the formation of biofilm within these spaces was observed (Fig. 5c). Some G. paraffinivorans MTZ041 distinct cells could be observed, but most of the culture was inside the biofilm matrix. At 11 weeks, bigger and relatively deeper holes in fragments of NR were noted when compared to what was seen previously (Fig. 5d).

Fig. 5.

Scanning electron micrographs of the natural rubber (NR) surface after incubation with G. paraffinivorans MTZ041. a The non-inoculated controls; b growth of G. paraffinivorans MTZ041 after 6 weeks; c after 9 weeks; and d after 11 weeks. Bars, 50 μm (a, c, and d) and 5 μm (b). a, c, and d At × 450 magnification; b at × 5000 magnification

G. paraffinivorans MTZ041 lcp gene expression

The fact that MTZ041 has a lcp gene, combined with the findings obtained by the growth curves, as well as, scanning electron microscopy, confirmed the ability of this organism to use rubber as sole carbon source. Next, we asked the question if the lcp gene would be induced by the presence of the substrate in the culture medium. Therefore, the quantification of lcp gene expression was evaluated by RT-qPCR.

After 2-h induction in the presence of IR, the lcp expression was 0.9-fold compared with the reference condition (fructose). In the presence of NR after 2-h induction, lcp also showed a slight decrease (0.76-fold) compared with the reference. After 24-h induction in IR as substrate, lcp was 1.14-fold, whereas at the same time in NR, lcp was 0.76-fold relative to the non-inducing condition. All these values when compared with the reference control (fructose) were not statistically significant (p < 0.001), suggesting lcp is not regulated by the presence of the substrate, and rather, it may be a constitutive gene.

Discussion

In this work, we were able to isolate a Gordonia paraffinivorans (MTZ041) from compost. Its genome was sequenced, assembled, and analyzed, confirming the taxonomic identification by MALDI-TOF and 16S rDNA gene sequencing. The isolation of this strain was not directed to obtain a rubber degrader microorganism; however, the bioinformatics showed that the genome contains an lcp gene, which has been described as a genetic element required for the assimilation of IR and NR as carbon sources [14, 15, 18, 27, 36, 45]. While the presence of the lcp in the genome suggests MTZ041 could be a rubber degrader, to the best of our knowledge, there were no reports in the literature regarding G. paraffinivorans ability to use any form of rubber as a carbon source.

In the assessment of growth in presence of rubber as a sole carbon source, it became clear that MTZ041 is a rubber degrader; however, the lack of a clear halo around the colonies (Fig. 2e) suggests that G. paraffinivorans MTZ041 belongs to the second group of rubber-degrading bacteria, that is, the one that attaches to the substrate and forms biofilm [46]. In addition, it was possible to see an orange pigment on the colonies growing on rubber, which is probably a carotenoid pigment. Other Gordonia species also produce this kind of metabolite throughout growth in different substrates [47, 48]. However, the exact nature of the pigment remains to be shown in this MTZ041 isolate.

Hiessl et al. [27] conducted growth curves with IR and reported that, after 10 days, G. polyisoprenivorans had entered stationary phase. In their experiment, the OD₆₀₀ 1.0 was reached in 5 days; however, this was their first measurement since the beginning of the experiment. Therefore, it is not possible to compare the initial growth profile of this strain with MTZ041 that reached OD₆₀₀ 1.0 in the first 24 h and maintained this optical density throughout the experiment. During the growth curves with NR, we observed tiny pieces floating in the culture medium, which are supposed to be fragments of the glove that began to disintegrate. This is consistent with the data reported by Linos et al. [36], who stained these pieces and proved that they correspond to degraded rubber. Due to the results obtained by the growth curves, it can be concluded that isolate MTZ041 is able to grow on these substrates, but NR supports higher growth than IR. Ibrahim et al. [14] described a growth pattern in N. farcinica E1 that seems similar to the one observed for MTZ041 [14]. The growth curves performed with fragments of IR and NR showed that the microorganism was able to achieve higher growth in the presence of NR over a 50-day period. Berekaa et al. [49] reported that, in the growth curves of Gordonia westfalica Kb2, the colonization of NR gloves increased over time. Despite the fragment being completely covered by a dense biofilm after the first week of cultivation, the number of planktonic cells in culture was low. On the third week of cultivation, there was an increase in optical density of the medium and at the end of 8 weeks, a significant number of planktonic cells in the culture medium were observed.

The formation of biofilm is a microbial strategy for survival in environments with few nutrients as well as for using a solid substrate [50, 51]. Some literature data suggest that the production of mycolic acids and biosurfactants is essential for the formation of biofilms, which allow the direct contact of bacterial cells with synthetic rubber, allowing the degradation of this material by species of Gordonia [52]. G. paraffinivorans, has been described as producing mycolic acids with lengths from 52 to 62 carbon atoms [53]. Based on these data, we can suggest that isolate MTZ041 may also produce essential mycolic acids for the formation of biofilm.

In order to gain more insights on the genetic mechanism that rubber is consumed by this strain, real-time PCR was used to detect the levels of lcp expression in the presence of rubber. No statistical significant induction by the substrate was verified, suggesting that lcp maybe a constitutive gene. Contrasting to this result, Linh et al. [15] studied lcp gene expression in Nocardia sp. NVL3 in the presence of IR and reported an increase of 1596 times in the expression when compared with the control without rubber. Yikmis et al. [54] reported a qualitative increase in lcp gene transcription for Streptomyces sp. K30 in the presence of cis-1.4-polyisoprene relative to the reference condition (glucose), which can be observed by a stronger band on agarose gel [54]. Bröker et al. [23] also reported a qualitative increase in lcp gene expression in G. polyisoprenivorans in the presence of cis-1.4-polyisoprene when compared with the reference control with sodium acetate. So far, the lcp expression in Gordonia species had not been quantified by real-time PCR experiments. The growth curves in synthetic rubber revealed that G. paraffinivorans MTZ041 reached exponential growth at 24 h, at which point the optical density reaches 1.0. This result suggests that the isolate uses rubber as carbon source at this stage and, consequently, the Lcp protein, which allows this ability, must be expressed. The comparison of the levels of expression between the treatments with rubber and fructose (24 h) indicates that the lcp gene is not induced in the presence of the substrate, suggesting that it has constitutive expression in G. paraffinivorans MTZ041; however, the fact that lcp is not induced does not mean that is not expressed. The protein must be sufficiently expressed in order to degrade rubber.

In summary, this paper presents the genomic and phenotypic characterization of MTZ041, a G. paraffinivorans isolate from a compost chamber, which encodes a lcp gene and is capable of rubber assimilation as sole carbon source and biofilm formation. Since rubber is a toxic pollutant and a difficult substrate to degrade, these findings may be of great relevance for biotechnological applications.

Funding information

This work was supported by research grants from Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) to RCP (2016/14542-8) and to JCS (2011/50870-6) and by a FAPESP fellowship (2016/07360-0) to SPB. This work was also supported by CNPq research fellowships to AMDS and JCS and by a CAPES research grant (3385/2013) to JCS.

Compliance with ethical standards

Conflict of interest

The authors declare that there is no conflict of interest.

Ethical approval

This article does not contain any studies with human participants or animals performed by any of the authors.