Most Hydrocarbonoclastic Bacteria in the Total Environment are Diazotrophic, which Highlights Their Value in the Bioremediation of Hydrocarbon Contaminants

Eighty-two out of the 100 hydrocarbonoclastic bacterial species that have been already isolated from oil-contaminated Kuwaiti sites, characterized by 16S rRNA nucleotide sequencing, and preserved in our private culture collection, grew successfully in a mineral medium free of any nitrogenous compounds with oil vapor as the sole carbon source. Fifteen out of these 82 species were selected for further study based on the predominance of most of the isolates in their specific sites. All of these species tested positive for nitrogenase using the acetylene reduction reaction. They belonged to the genera Agrobacterium, Sphingomonas, and Pseudomonas from oily desert soil and Nesiotobacter, Nitratireductor, Acinetobacter, Alcanivorax, Arthrobacter, Marinobacter, Pseudoalteromonas, Vibrio, Diatzia, Mycobacterium, and Microbacterium from the Arabian/Persian Gulf water body. A PCR-DGGE-based sequencing analysis of nifH genes revealed the common occurrence of the corresponding genes among all the strains tested. The tested species also grew well and consumed crude oil effectively in NaNO3 -containing medium with and without nitrogen gas in the top space. On the other hand, these bacteria only grew and consumed crude oil in the NaNO3 -free medium when the top space gas contained nitrogen. We concluded that most hydrocarbonoclastic bacteria are diazotrophic, which allows for their wide distribution in the total environment. Therefore, these bacteria are useful for the cost-effective, environmentally friendly bioremediation of hydrocarbon contaminants.

Hydrocarbonoclastic microorganisms are involved in the mineralization of hydrocarbon pollutants in the total environment. The microbial utilization of such carbon-rich, but nitrogen-poor substrates necessitates the availability of nitrogenous compounds for the synthesis of cell materials such as proteins and nucleic acids. Nitrogen fertilizers have been shown to limit bioremediation in oily environments (5,7,8,10,11,21,22,27). A previous study estimated that 60 mg N was needed for the microbiological consumption of 1 g hydrocarbons (9). Some environments, e.g. vegetation-poor, arid desert areas and seawater bodies, are naturally poor in nitrogenous compounds, as is the case in Kuwait and the whole Gulf area. However, oil spills enrich these environments with hydrocarbonoclastic bacteria (18,19), which implies that indigenous hydrocarbonoclastic bacteria are nutritionally independent of exogenous nitrogenous sources due to their ability to fix atmospheric nitrogen. In 20 years of research on the microorganisms in oil-polluted areas of the Gulf, we identified numerous bacterial species with the coupled potential for hydrocarbon mineralization and nitrogen fixation in many habitats including desert soil, seawater, hypersaline areas, rhizospheres, phyllospheres, and air dust, (1,2,4,6,16,17,25). However, the relative frequencies of these bacteria have not yet been examined in the total environment.
The criteria for diazotrophy of the already studied isolates were growth on nitrogen-free media, the ability to reduce acetylene to ethylene, and, albeit very rarely, the possession of nifH-coding genes in bacterial genomes. These criteria have been used to confirm diazotrophic potential rather than actual in situ nitrogen fixation during growth on hydrocarbon substrates as sources of carbon and energy.
In the present study, we demonstrated that hydrocarbonoclastic bacteria isolated from all over Kuwait exhibited diazotrophic potential, as well as actual diazotrophic activity. Kuwait was suitable for this study because it is a small (approximately18,000 km 2 ), but major oil-producing country. Furthermore, since it is a part of the semiarid zone of the globe, its environmental conditions are harsh; the maximum temperature during the long, dry summer may exceed 70°C while the minimum temperature in winter may reach 0°C. The vegetation-poor desert soil and Arabian Gulf water body are also poor in nitrogenous compounds.

Bacterial isolates
One hundred different species of oil-utilizing bacteria that have been deposited in our private culture collection were used in this study. The individual isolates were previously characterized by comparing the nucleotide sequences of their 16S rRNA genes with those of the closest relatives in the GenBank database. These species had been isolated from various pristine and oil-polluted Gulf habitats, including desert soil, the rhizospheres and phyllospheres of wild and cultivated plants, seawater, hypersaline areas, and air dust, on a mineral medium (24) with oil vapor as the sole source of carbon and energy. This mineral medium had the following composition (g L −1 ): 0.85 NaNO 3  at pH 7.0. Regarding isolations from sea water and hypersaline areas, 0.5 M and 2 M NaCl, respectively, were added to the medium. Details of the culture sources and methods for bacterial isolation and characterization have been described in our previous study (1).

Assessment of diazotrophic potential
These 100 bacterial isolates were examined for "presumptive" nitrogen-fixation potential by testing their growth on the solid mineral medium (24) after removing NaNO 3 , the sole nitrogen source available. An amount of 0.3% (w/v) crude oil (Light Kuwaiti crude), 0.3%, (w/v), was added as the sole source of carbon and energy. In order to confirm nitrogen-fixation potential, the 15 most dominant species in their habitats (oily desert soil and seawater) during initial plate-counting were tested for nitrogenase activity using the acetylene to ethylene reduction test (13) (the names of the bacterial species tested have been provided in the "Results" section). A cell suspension was prepared for each organism by mixing a loopful from a pure 48-h culture in 5 mL of the mineral nitrogen-free medium in a Hungate tube with a tight rubber cap. Using a gas-tight syringe, 10% of the head-space gas was replaced by acetylene. The medium used in this test contained 0.3% crude oil as the carbon and energy source. The tubes were incubated at 30°C for 12 d, and the head-space gas was analyzed for ethylene by gas-liquid chromato graphy (GLC) using a WCOT fused silica CP-SIL-5CB capillary column and temperature program 50-80°C in which the temperature was increased at 10°C min −1 . Nitrogen was the carrier gas; the detector temperature was adjusted to 250°C and the injector temperature to 150°C. The appearance of an ethylene peak in the GLC profile indicated a positive test. Diazotrophy was confirmed by extracting the genomic DNA of the tested organism and amplifying the nifH-coding gene therein using the primers PolF (5'-TGCGAYCCSAARGCBGACTC-3') and PolR (5'-ATSGCCATCATYTCRCCGGA-3') (1,15). The gene amplicons were subsequently separated and detected by gel electrophoresis. A loopful of the fresh culture was homogenized in 100 µL of PrepMan Ultra Sample Preparation Reagent (Applied Biosystems, USA) for genomic DNA extraction. The mixture was incubated in a water bath for 10 min at 100°C, then cooled for 2 min, and finally centrifuged at 14,000×g for 3 min to collect the DNA-containing supernatant. The PCR reaction mixture contained puReTaq Ready-To-Go PCR Beads (Amersham Biosciences, UK), 1 µL of DNAcontaining supernatant, 1 µL each of the nifH primers (20 pmol µL −1 ) and the final volume was increased to 25 µL with molecular water (Sigma, UK). Amplification was achieved using the Veriti Thermal Cycler (Applied Biosystems) starting with initial denaturation for 5 min at 94°C, followed by 30 cycles of denaturation for 30 s at 94°C, annealing for 30 s at 58°C, and primer extension for 30 s at 72°C. The final extension was conducted for 5 min at 72°C. Gel bands carrying nifH DNA were cut out and the nifH fragments were reamplified with the same primers and then purified using the QIA quick PCR purification Kit (Applied Biosystems). Twenty nanograms of the DNA template was added to 2 µL of the BigDye version 3.1 terminator; 2 µL of the BigDye terminator 5× buffer, 1 µL of either PolF or PolR, and the final volume was increased to 10 µL with ultrapure water. The sequencing reaction was completed in the Veriti Thermal Cycler (Applied Biosystems) using a cycle of 96°C for 1 min, followed by 25 cycles of 1 min at 96°C, 5 s at 50°C, and 4 min at 60°C. Sequencing was performed in a 3130xl genetic analyzer (Applied Biosystems) with sequencing analysis software version 5.2 (Applied Biosystems). Sequences were subjected to a basic local alignment search tool analysis with the National Center for Biotechnology Information (NCBI; Bethesda, MD, USA) GenBank database (1). The classic nitrogen fixer Klebsiella pneumonia ATCC13883 was used as a positive control. The GenBank accession numbers are KP192902-KP192907.
Nitrogen fixation by the most frequently detected bacterial species (Table 1) was confirmed from a novel experiment designed for this purpose. Bacterial growth was measured in liquid mineral medium (24) aliquots containing NaNO 3 and aliquots lacking this nitrogen source. Flask cultures were sealed and incubated at 30°C for 10 d. Growth in terms of optical density at 660 nm was measured at time zero and in 2-day intervals. Three parallel replicates were run for each reading and the mean values were calculated. Growth curves were plotted. The head-space gas of the cultures in some experiments was air (containing approximately 80% N 2 ), but was completely substituted with a gas mixture consisting of 80% CO 2 and 20% O 2 (v/v) (i.e. no N 2 gas was available) in other experiments. Tween 80, 0.3% (w/v), was used as the sole carbon and energy source in this experiment. This substrate had the advantage of being completely miscible with the aqueous medium. This was essential for subsequent optical density measurements. Another advantage was that Tween 80 contains long chain alkyl moieties, the carbon skeleton of aliphatic hydrocarbons. In this context, Tween 80 is the substrate utilizable by the unique group of the obligate hydrocarbonoclastic bacteria (OHCB) recently considered responsible for most of the natural removal of hydrocarbon contaminants in the marine ecosystem (28).

Measurement of oil consumption
Crude oil consumption by individual bacterial species was measured quantitatively in batch cultures under the conditions specified above. One hundred-milliliter aliquots of the mineral medium (24) amended with 0.3% (w/v) crude oil (instead of Tween 80) were dispensed in 250-mL flasks. Each flask was inoculated with 0.25 mL of the bacterial cell suspension (a loopful of 48-h biomass in 5 mL sterile water). The cultures were sealed and incubated on an electric shaker at 120 rpm for 2 weeks. Residual crude oil was extracted with three successive 10-mL portions of pentane. The combined extract was increased to 35 mL using pentane, and 1 µL was analyzed by GLC. Regarding GLC, we used a Chrompack (NJ, USA) CP-9000 instrument equipped with a flame ionization detector and WCOT fused silica CP-SIL-5CB capillary column at a temperature of 10°C min −1 . Nitrogen was the carrier gas; the detector temperature was 250°C and the injector temperature 150°C. The percentage loss of total peak areas in the GLC chromatograms based on the peak areas of the control samples (similarly prepared, but using previously autoclaved bacteria) was calculated as a quantitative measure of oil consumption. Three parallel replicates were run for each reading and the mean and standard deviation values were calculated.

Results and Discussion
Growth on nitrogen-free medium and possession of nifH genes Eighty-two out of the 100 hydrocarbonoclastic bacterial species tested were capable of growing on the NaNO 3free mineral medium with oil as the sole source of carbon and energy. Table 1 shows the names of the 15 species selected for further study, as well as information related to their 16S rDNA sequencing. These species had their origin in Kuwaiti desert soil samples and seawater from the Arabian Gulf; they belonged to the bacterial subdivisions (classes); Alphaproteobacteria, Betaproteobacteria, Gammaproteobacteria, and Actinobacteria. The individual species exhibited 99 to 100% similarities in their 16S rDNA-sequences to their closest GenBank counterparts, with two species showing only 98 and 97% similarities.
These 15 bacterial species tested positive for nitrogenase, as revealed by the GLC-detection of ethylene peaks in their culture top space gases (GLC profiles not shown). The gel electrophoresis profiles in Fig. 1 confirmed that all the 13 species selected for analysis had nifH genes in their genomic

DNA extracts.
The results of sequencing of the nifH gene bands in Fig. 1 are presented in Table 2. The bands of the 13 species tested revealed identical nifH gene sequences to those of heterotrophic bacteria in the GenBank database. The 3 species, Sphingomonas dokdonensis, Marinobacter hydrocarbonoclasticus, and Pseudomonas xanthomonas in Table 1 were not subjected to this analysis and, therefore, were absent in Table 2.  Band I and Band II for every species were cut, amplified, and sequenced, and the sequences obtained were then compared with the nearest GenBank matches (see Table 2). in our laboratory for diazotrophy and tested positive. Furthermore, we only selected one of the two Pseudomonas sp. for this analysis. On the other hand, we subjected Pigmentiphaga daeguensis to this analysis (Table 2) even though it was not among the species in Table 1.
Nitrogen gas was needed by hydrocarbonoclastic bacteria for growth and oil consumption in the absence of NaNO 3 The growth curves in Fig. 2 showed that all the hydrocarbonoclastic strains tested grew well in the NaNO 3containing mineral medium with Tween 80 (a hydrocarbonlike substrate) as the sole carbon and energy source. However, in the absence of NaNO 3 , active growth only occurred when the top space of the cultures contained nitrogen gas. Table 3 shows the quantitative data of oil consumption by the individual bacterial species in the presence and absence of NaNO 3 (in the medium) and nitrogen gas (in the head space). The results obtained clearly demonstrated that oil consumption was markedly effective in the NaNO 3 -containing medium irrespective of whether nitrogen gas was present in the top space. However, the effective consumption of hydrocarbons only occurred in the NaNO 3 -free medium when nitrogen gas was available in the top space, and was minimal in its absence. This was also clearly shown in Fig. 3, which presents the typical GLC profiles of residual hydrocarbons recovered from Nesiotobacter exalbescens growing in the Fig. 2. Growth curves of bacterial species in the presence and absence of NaNO 3 and nitrogen gas. Open triangles, NaNO 3 -containing medium with nitrogen gas; closed triangles, NaNO 3 -containing medium without nitrogen gas; open circles, NaNO 3 -free medium with nitrogen gas; closed circles, NaNO 3 -free medium without nitrogen gas. There was only minimal growth, if any, in the NaNO 3 -free medium with the top space free of nitrogen gas. presence and absence of NaNO 3 and nitrogen gas for 2 weeks.
The novelty of this study is not only related to its methodology, but also to its major findings. Previous studies, mainly from our laboratory, described the combined potential of certain bacterial species for hydrocarbon utilization and nitrogen fixation. In addition to the habitats specified in the "Introduction" section, such bacteria have also been recorded in coastal mud-flats (4), epilithic biofilms (20), and hypersaline soils and waters (3).
One important result of the present study is that the nifH genes were detected unequivocally in all the environmentally predominant hydrocarbonoclastic bacterial species tested. Another novel result is that these bacterial species did not grow or exhibit any hydrocarbonoclastic activity in nitrogenfree media unless nitrogen gas was available. This result provides conclusive evidence for the ability of hydrocarbonoclastic bacteria to fix nitrogen in situ.
Microbiological nitrogen fixation is an activity that requires large amounts of energy. Therefore, when nitrogenous compounds and nitrogen gas are simultaneously available, diazotrophs prefer to utilize the former, which explains why nitrogen fertilizers enhance bacterial hydrocarbon bioremediation, as commonly reported, even though most hydrocarbonoclastic bacteria are diazotrophic. Based on their diazotrophy, such bacterial species have the ecological advantage of survival in environments extremely poor in (or free of) nitrogenous compounds.
It is experimentally challenging to unequivocally demonstrate diazotrophy in the 82 different bacterial species that gave positive presumptive tests. Therefore, we examined the most frequently detected species in desert soil and seawater in Kuwait. Most (12 species) of these species belonged to Alphaproteobacteria and Gammaproteobacteria, which was consistent with previous findings on hydrocarbonoclastic bacteria in contaminated sediments and seawater from a refinery harbor in the north of Tunisia (12). The occurrence, in all samples, of two gene bands instead of only one (Fig. 1) confirmed earlier findings that several versions of the nifH gene naturally occurred in the same diazotrophs (14, 15). The mineral medium used was provided with 0.3% crude oil. Incubation was for 2 weeks. Data were the means of 3 readings ± the standard deviation values. The very low consumption values in the right column were attributed to nitrogenous impurities in the medium that supported minimal bacterial growth and oil consumption. Fig. 3. Typical GLC profiles showing the markedly effective consumption of crude oil by Nesiotobacter exalbescens in NaNO 3 -free medium in the presence of nitrogen gas and poor consumption in its absence. Oil consumption was also markedly effective in the NaNO 3 -containing medium with and without nitrogen gas in the head space.
Weaker peaks indicate more effective consumption.
The results of this study are interesting from the view of recent studies in which hydrocarbons were shown to be utilized by conventional diazotrophic bacteria including Rhizobium (17), Azotobacter (26), and Klebsiella (1,23). Therefore, we assumed that, in the course of early evolution, microorganisms that developed the ability to utilize nitrogenpoor substrates, such as oil, simultaneously developed the potential for nitrogen fixation in order to synthesize their nitrogenous cell materials (including proteins and nucleic acids). However, many molds and yeasts (eukaryotes that are incapable of nitrogen fixation) can also utilize hydrocarbons. Eukaryotic organisms appeared later in the course of evolution.