Relative Abundances of Proteobacterial Membrane-Bound and Periplasmic Nitrate Reductases in Selected Environments

Dissimilatory nitrate reduction is catalyzed by a membrane-boundand a periplasmic nitrate reductase. We set up a real-time PCRassay to quantify these two enzymes, using the narG and napAgenes, encoding the catalytic subunits of the two types of nitratereductases, as molecular markers. The narG and napA gene copynumbers in DNA extracted from 18 different environments showedhigh variations, with most numbers ranging from 2 x 10² to 6.8x 10⁴ copies per ng of DNA. This study provides evidence that,in soil samples, the number of proteobacteria carrying the napAgene is often as high as that of proteobacteria carrying thenarG gene. The high correlation observed between narG and napAgene copy numbers in soils suggests that the ecological rolesof the corresponding enzymes might be linked.

INTRODUCTION

Top

INTRODUCTION

Nitrate in the environment can be either assimilated by plantsand microorganisms or reduced to nitrite by one of two microbialdissimilatory processes: denitrification or dissimilatory reductionof nitrate to ammonium. Nitrate reduction by denitrificationis of great importance since the produced nitrite is then reducedto N₂O or N₂ gases, which can lead to considerable nitrogenlosses in agriculture and emissions of greenhouse gases (6,13, 28). The reduction of nitrate present in contaminated watercaused by the nitrate-reducing bacteria living in the humandigestive tract is a potential health problem. As nitrite entersthe bloodstream, it reacts with hemoglobin to form methemoglobin,blocking oxygen transport and causing a disease commonly called"blue baby syndrome" (26). Two types of dissimilatory nitratereductase, differing in their locations, were characterized:a membrane-bound (Nar) and a periplasmic (Nap) nitrate reductase(2, 16, 29). The membrane-bound nitrate reductase is presentin proteobacteria, firmicutes, actinobacteria, and even archaea,whereas the periplasmic nitrate reductase is present only inproteobacteria (21, 24). Nitrate-reducing proteobacteria canharbor Nar or Nap or both (18, 29). In contrast to that of Nar,the physiological role of Nap is still unclear and seems todiffer between bacteria (11, 24). Thus, one proposed role forNap is to support anaerobic metabolism as an alternative toNar (1, 8). It has also been proposed that Nap facilitates theswitch from aerobic respiration to denitrification (27) or scavengesnitrate in some pathogenic bacteria (23). The importance anddiversity of the bacteria containing Nar have been extensivelystudied using both cultivation-based and direct molecular approaches(3, 5, 7, 17, 19, 22). However, only a few studies have focusedon bacteria containing Nap (4, 9, 25). In this study, we investigatedthe relative abundances of the two types of nitrate reductasesin various environments, using a real-time PCR-based assay.

Primer design, standard curves, and real-time PCR procedures.

Top

Primer design, standard curves,...

In order to quantify the two types of nitrate reductases, anew real-time PCR assay was set up, using the narG and napAgenes, encoding the catalytic subunits of the membrane-boundand periplasmic nitrate reductases, respectively, as molecularmarkers. All available sequences were aligned, and narG andnapA primer sets specific to the proteobacteria were designed.The degenerated narG-f (5'-TCGCCSATYCCGGCSATGTC-3'), narG-r(5'-GAGTTGTACCAGTCRGCSGAYTCSG-3'), V17m (5'-TGGACVATGGGYTTYAAYC-3';modified after reference 9), and napA4r (5'-ACYTCRCGHGCVGTRCCRCA-3')primers were used to amplify fragments of 173 (for narG) and152 (for napA) bp. Serial dilutions of linearized plasmids containingthe narG and napA genes from Pseudomonas aeruginosa PAO1 wereused to generate standard curves. The real-time PCR assays werecarried out with a 20-µl reaction volume containing SYBRgreen PCR master mix (ABsoluteTM QPCR SYBR ROX Mix; Abgene,France), 2 µM of each primer, 100 ng of T4 gp32 (QBiogene,France), and 1.25 µl of template DNA (2 to 12.5 ng). Thermocyclingconditions for narG were as follows: 15 min at 95°C; 6 cyclesconsisting of 30 s at 95°C and 30 s at 63°C, with atouchdown of –1°C by cycle; and 40 cycles consistingof 30 s at 95°C, 30 s at 58°C, 30 s at 72°C, and30 s at 80°C. Conditions for napA were similar except thatthe annealing temperature was set at 61°C. All real-timePCRs were performed with an ABI Prism 7900 sequence detectionsystem (Applied Biosystems). Quantification of the 16S rRNAgene was performed as described previously (15). DNAs extractedfrom triplicate samples from 18 different environments, suchas soils (agricultural, industrial, or glacier), river sediments,waters, or biofilms, were used as templates.

Evaluation of assay specificity and sensitivity.

Top

Evaluation of assay specificity...

A functional gene pipeline interactive tool (http://flyingcloud.cme.msu.edu/fungene/)was used for in silico evaluation of primer specificity. Searchesfor the narG and napA primer sequences among 36 and 40 narGand napA sequences from complete genomes of proteobacteria showedthat 78, 70, 86, and 86% did not exhibit any mismatch with thenarG-f, narG-r, V17m, and napA4r primers, respectively. Primerspecificity was further confirmed experimentally using a collectionof 19 strains (Table 1). Four firmicutes and one actinomycetenitrate-reducing strain were selected as negative controls,and 14 nitrate-reducing strains belonging to the alpha-, beta-,and gammaproteobacteria were selected as positive controls.None of the gram-positive nitrate reducers, which were usedas negative controls, gave an amplicon. For Proteobacteria,an absence of PCR products with both narG and napA primers wasrecorded only with Alcaligenes faecalis ATCC 8750. Sequenceanalysis of 84 and 79 narG and napA real-time PCR products fromagricultural soils (Côte Saint André and Yvetot),glacier soil (Rotmoosfermer), cave biofilm (Padirac), and riverphototrophic biofilm (Garonne) revealed that all sequences wererelated to the narG or napA genes. A high level of diversityamong the sequences of the real-time PCR products was observed,with identities as low as 67% for narG and 69% for napA to thesequences used for the designs of the primers (see Fig. S3 andS4 in the supplemental material). This indicates that our newlydeveloped real-time PCR systems are suitable for general detectionof proteobacterial nitrate reductase genes. However, 3 out of84 narG sequences fell into a cluster containing only narG fromactinobacteria, indicating that the designed narG primers werenot entirely specific to proteobacteria.

View this table:
[in this window]
[in a new window]

TABLE 1. Bacterial strains used in this study to test the specificities of the narG and napA primers^a

The detection limit of our assay was around 10 copies per ngof template DNA, and no signal was detected in the no-templatecontrol. The PCR efficiencies of the narG and napA real-timePCR assays were 86 and 83%, respectively. Genomic DNA from Pseudomonasaeruginosa PAO1, for which the theoretical 16S rRNA, narG, andnapA gene copy numbers per ng of DNA were calculated, was usedas an external control. The presence of PCR inhibitors coextractedwith DNA was tested as described previously (12), and this testdid not reveal any significant inhibition. Therefore, our assaysbased on novel primer sets narG-f-narG-r and V17m-napA4r providean efficient and sensitive method for quantifying either Naror Nap in environmental samples.

Quantification of narG and napA genes.

Top

Quantification of narG and...

To compare with accuracy the numbers of genes in the differentenvironments, results were expressed as gene copy numbers perng of extracted DNA. Quantification of the narG and napA genesrevealed high variations between environments, with most numbersranging from 2 x 10² to 6.8 x 10⁴ copies per ng of DNA (Fig.3). The highest copy numbers for both narG and napA genes wereobserved in the river sediment samples. The napA gene copy numberswere lower than those of narG in most of the freshwater samples,whereas similar narG and napA gene copy numbers were observedin all soil samples except the Yvetot soil and the Rotmoosfermerglacier soil (Fig. 1). The numbers of 16S rRNA genes were between1 and 3 logs higher than those of narG or napA genes. However,up to 12 copies of 16S rRNA may be found in the same bacterialgenome (10), while only 1 napA copy and a maximum of 3 narGcopies have been identified in the same strain (18). Hence,our study provides evidence that in soils, the numbers of proteobacteriacontaining periplasmic nitrate reductase are similar to thenumbers of proteobacteria with the membrane-bound nitrate reductase.This is consistent with the results from Roussel-Delif et al.(25) and Carter et al. (4), which showed that a large proportionof nitrate-reducing, gram-negative isolates contain Nap. Unfortunately,it is not possible to conclude from our study whether the similarnarG and napA copy numbers in many samples resulted from a majorityof nitrate reducers possessing both types of nitrate reductaseor from similar numbers of bacteria possessing either Nar orNap. The narG genes from proteobacteria were mainly targetedin our assay, whereas these are also present in gram-positivebacteria and archaea, in contrast to napA. Therefore, the factthat the narG gene copy numbers were either similar to or higherthan the napA gene copy numbers indicates that Nar is probablypredominant in the environment.

View larger version (45K):
[in this window]
[in a new window]

FIG. 1. Abundance of 16S rRNA, narG, and napA genes expressed as gene copy numbers per ng of extracted DNA, analyzed from three independent replicates per site. Error bars indicate standard deviations. S indicates significant differences (P < 0.05) between the narG and napA gene copy numbers (Student's t test) for each site.

Interestingly, a high correlation coefficient of 0.9 betweennarG and napA gene copy numbers was calculated for the soiland sediment samples, whereas this coefficient was only 0.4for samples from river biofilm and water samples (Fig. 2). Incontrast, the correlation between narG or napA and 16S rRNAgene copy numbers did not exceed 0.5. The physiological roleof Nap is still unclear and probably differs between strains(11, 21, 24). The high correlation observed between the numbersof narG and napA genes but not with 16S rRNA suggests that theecological role of Nap in the majority of soil proteobacteriamight be similar or complementary to that of Nar. Our resultsalso showed that the abundances of the two types of nitratereductases differed between environments, which could be dueto the selection of nitrate reducers in some habitats. Sincedetection of functional genes is only a weak hint of the presenceof the corresponding activity (20), investigation of the relativecontributions of the two types of nitrate reductases to thetotal nitrate reduction activity in the different environmentsis of interest. Unfortunately, the activities of the two typesof nitrate reductase are simultaneously monitored by the nitratereduction assay developed by Kandeler (14), and the activitiesof Nar and Nap can be distinguished only on bacterial isolates(4). In the future, integrated studies are needed to compilefurther information on the physiology, diversity, and distributionof nitrate reducers for a more comprehensive understanding ofnitrate reduction in the environment.

View larger version (15K):
[in this window]
[in a new window]

FIG. 2. Correlation between narG and napA gene copy numbers in the soil and sediment samples (white triangles) and in water and biofilm samples (black circles). Each replicate of the 18 selected environments is plotted.

Nucleotide sequence accession numbers.

Top

Nucleotide sequence accession...

The GenBank accession numbers for the sequences reported hereare EF217059 to EF217221.

ACKNOWLEDGMENTS

We are grateful to A. Brauman, G. Braker, F. Garabetian, S.Hallin, E. Kandeler, S. Nazareth, and B. Stres for providingreference strains or environmental samples. We also thank S.Henry and S. Hallet for their assistance.

This work was supported by the French Ministry of Research (ACIPNBC MUTEN).

This work is dedicated to the memory of FrédéricMarmont and Ivan Mahne.

FOOTNOTES

* Corresponding author. Mailing address: INRA, University of Burgundy, Soil and Environmental Microbiology, CMSE, 17 rue Sully, B.P. 86510, 21065 Dijon Cedex, France. Phone: 33 3 80 69 33 46. Fax: 33 3 80 69 32 24. E-mail: philippo{at}dijon.inra.fr

Published ahead of print on 13 July 2007.

Supplemental material for this article may be found at http://aem.asm.org/.

REFERENCES

Top