Optimization of RNA Extraction for PCR Quantification of Aromatic Compound Degradation Genes

Abstract

Seven different bacterial strains and primer sets and a mixed community were used to evaluate the use of reverse transcriptase quantitative PCR (Q-PCR) and Q-PCR of oxygenase genes to assess various approaches for monitoring the bioremediation of polluted sites. Differences in maximum activity were seen when different RNA extraction kits were compared.

Advancements in molecular quantification techniques have facilitated more accurate assessments of pollutant biodegradation and the impact of remediation technologies (2, 4, 11). DNA-based technologies can demonstrate the presence of catabolic genes, but activity can at best only be inferred based on temporal increases in the number of gene copies over time (growth) (2, 11). In contrast, quantification of mRNA is thought to be a more direct approach to assess the current physiological state of microorganisms involved in bioremediation (5), and improvement in mRNA extraction kits has now potentially facilitated the application of mRNA in the evaluation of a community's functional activity in the environment. Our overall goal was to evaluate the practical use of mRNA for assessing the remediation of polluted environments by comparing two RNA extraction kits to determine their effectiveness in monitoring the activity of aromatic hydrocarbon-degrading bacteria over different growth stages. The activity of a simulated bacterial community competing for a single resource was also tested to aid in the interpretation of data from field sites.

Seven previously characterized biodegrading bacterial isolates, seven oxygenase gene-specific primers, and quantitative PCR conditions (3, 8) (see Table S1 in the supplemental material) were used to assess the potential range of variability stemming from different bacteria, primers, and growth stages in the application of reverse transcriptase (RT) (ImProm-II; Promega) quantitative PCR (Q-PCR) (iQ SYBR green Supermix; Bio-Rad) to assessment of a contaminated site. Bacterial strains specific for each primer set were as follows: primer NAH, Pseudomonas putida G7; primer TOD, P. putida F1; primer TOL, P. putida HS1; primer PHE, Pseudomonas sp. strain CF600; primer BPH1, Pandoraea pnomenusa B-356; primer BPH2, P. pseudoalcaligenes KF707; and primer BPH4, Rhodococcus sp. strain RHA1. Clones of each target gene were used as standards to improve quantification accuracy (see Methods and Table S2 in the supplemental material). More total RNA was extracted at all growth phases from all bacteria tested with the MP FastRNA soil kit than with a Qiagen RNA/DNA kit (e.g., for results for P. putida G7, see Fig. 1A) despite the addition of a bead beating step that we previously found improves cell lysis (9).

FIG. 1.

Comparison of temporal differences in total RNA (A) and levels of expression of the upper pathway oxygenase gene based on mRNA copy numbers (B) in Pseudomonas putida G7 using different RNA extraction kits: a modified Qiagen kit and the MP FastRNA kit. Error bars represent standard errors (n = 3). OD600, OD at 600 nm.

The dynamics of oxygenase mRNA quantities measured at the various growth phases differed between the two kits (Fig. 1B). High gene expression was observed from early to mid-exponential growth phase (optical densities [ODs], 0.15 and 0.3) (Fig. 1B) for all the tested strains using RNA extracted using the modified Qiagen RNA/DNA kit, whereas increased expression was not found in samples extracted using the MP FastRNA kit until late exponential phase (OD, 0.7). Despite the greater quantity of total RNA extracted using the MP FastRNA kit, more target mRNA was recovered using the Qiagen RNA/DNA kit; thus, it was chosen for use in subsequent experiments. Using a Qiagen RNA extraction kit, the same gene expression pattern was observed for mmo in Methylosinus trichosporium OB3b (6), rbcL in Phaeodactylum tricornutum CCMP 630 (16), and the Fdh and Hub genes in Dehalococcoides ethenogenes strain 195 (13). With the MP FastRNA kit, substantial activity was not observed until cultures were in late exponential to early stationary phase, when, theoretically, growth was beginning to slow down. The same pattern was observed when the expression of the sulfite reductase gene in Desulfobacterium autotrophicum was quantified using the MP FastRNA kit (12). It appears that less mRNA is recovered below a certain biomass using the MP FastRNA kit despite its greater efficiency at extracting total RNA. It is possible that activity in field samples would be missed if target organisms had low biomass due to slow growth or minimal substrate concentrations.

Using the RNA extracted with the modified Qiagen kit, gene expression was highest during exponential growth phase and decreased significantly around the beginning of stationary phase, as illustrated for strains HS1 and JI104 (Fig. 2). These gene expression dynamics followed the expected pattern for substrate metabolism reported by others (7). Most natural environments are oligotrophic (10, 14) or have nutrient bioavailability constraints (15), with microbes likely spending most of their time in stationary phase. Exponential growth occurs when nutrients (e.g., organic pollutants) become available, similar to the scenario we created in the lab. An RT Q-PCR approach would be useful for determining microbial gene expression commensurate with nutrient input, for example, with oxygenase genes expressed in remediation of sites experiencing fluxes in levels of organic pollutants.

FIG. 2.

Temporal expression based on mRNA copy numbers of the upper pathway oxygenase gene (▴) and the lower pathway dioxygenase (C23DO) gene (⧫) in Pseudomonas putida HS1 (A) and Pseudomonas aeruginosa JI104 (B). Growth over time (○) is measured as the OD₆₀₀. Error bars represent standard errors (n = 3).

Simultaneous extraction of RNA and DNA from a single sample has been shown to be helpful for comparing metabolisms and the presence of specific genes (1, 17, 18). The Qiagen kit had the added benefit of coextracting DNA to estimate cell numbers using the same sample. Competition between different bacteria for the same growth substrate was studied by comparing levels of gene expression and the copy numbers of the different dioxygenase genes that they carry. Experiments were performed using a mixture of Pandoraea pnomenusa B356, P. pseudoalcaligenes KF707, and Rhodococcus sp. RHA1 competing for biphenyl as the sole carbon source. Differences in the levels of fitness of the three strains were evident; P. putida KF707 appeared to be most fit because BPH2 gene expression was the highest, even greater than when it was grown in pure culture (Fig. 3A). Expression of the BPH1 gene in B356 and the BPH4 gene in RHA1 was notably lower when the strains were grown in a mixed community than when they were grown in pure cultures (Fig. 3B and C), with RHA1 being least fit. This was substantiated by the very low number of BPH4 gene copies quantified from mixed-culture DNA. In contrast, gene copy numbers of BPH2 in KF707 and of BPH1 in B356 were significantly higher in a mixed community than when the strains were grown singly. The lack of BPH1 gene expression by B356 but an increase in DNA gene copy number indicated that their growth did not occur from biphenyl utilization but was instead syntrophic with that of KF707.

FIG. 3.

Comparison of upper pathway gene copy numbers (triangles) and the levels of expression of this gene (circles) based on mRNA copy numbers in a single strain (open symbols) versus a mixed community (filled symbols). (A) Pseudomonas putida KF707 using primer BPH2; (B) Pandoraea pnomenusa B356 using primer BPH1; (C) Rhodococcus sp. RHA1 using primer BPH4. Error bars represent standard errors (n = 3).

Despite the increasing use of Q-PCR and, more recently, RT Q-PCR in microbial ecology, there are few papers that include an evaluation of the specific methods being used. These methods are valuable because mRNA quantification can be used as an indicator of current activity, while DNA quantification is the outcome of that activity. Commercial nucleic extraction kits and RT Q-PCR were able to measure oxygenase gene expression in bacteria grown in pure culture or a mixed community. However, we show that results can vary depending on isolate, growth stage, primers, and nucleic acid extraction method chosen. Despite the tremendous value in obtaining quantitative measures of functional activity, there is a need to understand the precision and accuracy of methods being used and to recognize limitations of data interpretation, particularly when applied to field studies.