Community-Level Physiological Profiling of Microbial Communities in Constructed Wetlands: Effects of Sample Preparation Mark Button, Kela Weber, Jaime Nivala, Thomas Aubron & Roland Arno Müller
Applied Biochemistry and Biotechnology Part A: Enzyme Engineering and Biotechnology ISSN 0273-2289 Appl Biochem Biotechnol DOI 10.1007/s12010-015-1921-7
1 23
Your article is protected by copyright and all rights are held exclusively by Springer Science +Business Media New York. This e-offprint is for personal use only and shall not be selfarchived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.
1 23
Author's personal copy Appl Biochem Biotechnol DOI 10.1007/s12010-015-1921-7
Community-Level Physiological Profiling of Microbial Communities in Constructed Wetlands: Effects of Sample Preparation Mark Button 1 & Kela Weber 1 & Jaime Nivala 2 & Thomas Aubron 2 & Roland Arno Müller 2
Received: 5 May 2015 / Accepted: 5 November 2015 # Springer Science+Business Media New York 2015
Abstract Community-level physiological profiling (CLPP) using BIOLOG® EcoPlates™ has become a popular method for characterizing and comparing the functional diversity, functional potential, and metabolic activity of heterotrophic microbial communities. The method was originally developed for profiling soil communities; however, its usage has expanded into the fields of ecotoxicology, agronomy, and the monitoring and profiling of microbial communities in various wastewater treatment systems, including constructed wetlands for water pollution control. When performing CLPP on aqueous samples from constructed wetlands, a wide variety of sample characteristics can be encountered and challenges may arise due to excessive solids, color, or turbidity. The aim of this study was to investigate the impacts of different sample preparation methods on CLPP performed on a variety of aqueous samples covering a broad range of physical and chemical characteristics. The results show that using filter paper, centrifugation, or settling helped clarify samples for subsequent CLPP analysis, however did not do so as effectively as dilution for the darkest samples. Dilution was able to provide suitable clarity for the darkest samples; however, 100-fold dilution significantly affected the carbon source utilization patterns (CSUPs), particularly with samples that were already partially or fully clear. Ten-fold dilution also had some effect on the CSUPs of samples which were originally clear; however, the effect was minimal. Based on these findings, for this specific set of samples, a 10-fold dilution provided a good balance between ease of use, sufficient clarity (for dark samples), and limited effect on CSUPs. The process and findings
Electronic supplementary material The online version of this article (doi:10.1007/s12010-015-1921-7) contains supplementary material, which is available to authorized users.
* Mark Button
[email protected] 1
Environmental Sciences Group, Department of Chemistry and Chemical Engineering, Royal Military College of Canada, Kingston, Ontario K7K 7B4, Canada
2
Helmholtz Center for Environmental Research (UFZ), Environmental and Biotechnology Center (UBZ), Permoserstrasse 15, Leipzig 04318, Germany
Author's personal copy Appl Biochem Biotechnol
outlined here can hopefully serve future studies looking to utilize CLPP for functional analysis of microbial communities and also assist in comparing data from studies where different sample preparation methods were utilized. Keywords CLPP . Domestic wastewater . Horizontal flow . Microbial activity . Treatment wetland
Introduction Community-level physiological profiling (CLPP) predominantly refers to a fast and userfriendly method for the functional characterization of microbial communities via the inoculation of mixed microbial community samples onto BIOLOG® microplates (1). Several types of BIOLOG® microplates are available for different purposes including the characterization of gram-positive bacteria, gram-negative bacteria, or environmental microbial communities. The BIOLOG® EcoPlate was originally designed in response to a request from microbial ecologists for a microplate that contained a set of environmentally relevant carbon sources that helped differentiate the function of soil microbial communities. The method has since expanded into the fields of ecotoxicology, agronomy, and land use (2–4) and is now increasingly employed to monitor microbial activity within various wastewater treatment systems including constructed wetlands (5, 6). Various types of constructed wetlands are currently used for water pollution control. A number of different design variations are employed which can be broadly categorized into two groups: surface flow, where water is seen flowing on the surface of the water treatment system, and subsurface flow, where water flows through a porous media such as sand or gravel. Direction of flow can be categorized as vertical or horizontal, and several operational regimes or operational modifications are possible as well (for example, tidal flow regimes or added aeration). All of the different designs can be planted with various types of vegetation. Any single constructed wetland system can contain many different internal environmental regimes. Therefore, when sampling a constructed wetland system, it is expected that both chemical and microbial community parameters are spatially variable within a single system (5, 7). Microbial communities play a dominant role in the treatment of organic-based pollutants in constructed wetlands, and therefore, functional analysis of microbial communities in this regard is an important analysis parameter (8). Within subsurface flow-constructed wetland systems, microbial communities can be found attached to the bed media within a biofilm, suspended in the interstitial water, attached to plant roots or root hairs, as well as within a sediment layer (if the system design allows for sediment accumulation). It is also understood that microbial communities within the rhizosphere in the upper parts of a subsurface-constructed wetland will generally exhibit higher levels of activity than communities from other areas within a system (7). Therefore, spatial analysis of microbial community function is an important area of study. Figure 1 depicts a simplified horizontal subsurface flow-constructed wetland highlighting the different areas, expected microbial community sampling methods, and the associated sample characteristics. The BIOLOG® EcoPlate contains 32 wells in triplicate (31 carbon sources plus a blank) in order to allow for sample replication on a single 96-well plate (9). The 31 individual carbon sources can be grouped according to their chemical structure as carbohydrates, polymers, carboxylic acids, amino acids, and amines/amides (10), thus providing a diverse range of
Author's personal copy Appl Biochem Biotechnol
Fig. 1 Simplified schematic of a subsurface horizontal flow-constructed wetland used for water pollution control. Key spatial areas of interest for microbial community assessment are identified with their associated sampling techniques and common sample characteristics
substrates for the assessment of microbial functional diversity. In addition to the carbon sources, each well contains a redox dye indicator, tetrazolium violet. When a mixed microbial community sample is inoculated into each of the wells, the production of NADH via cell respiration reduces the tetrazolium dye to formazan, resulting in the development of a purple color, which can be quantified photometrically. The CLPP approach can provide a sensitive and ecologically relevant measure of heterotrophic microbial community function and functional adaptations over time and space (11, 12). In the field of constructed wetlands, CLPP has been used to study spatial and temporal microbial community dynamics (13, 14), the effects of plant diversity (15, 16), as well as the influence of emerging contaminants (17). The number of studies utilizing microbial assessment to help understand treatment processes in constructed wetlands has increased significantly in the last decade (8). The CLPP approach has been identified to give more information than just microbial community differentiation data (5, 16). Given the diverse set of carbon sources on the plate, these carbon sources can be said to be a relatively broad representation of the types of organic molecules found in a complex wastewater. The differentiation of carbohydrates, polymers, carboxylic acids, amino acids, and amines/amides and quantification of the relative activity levels of the sampled microbial community with regards to each grouping can give useful insight into the water treatment capabilities of the microbial communities being assessed and by extension the constructed wetland from which the microbial community samples originated. Although CLPP is considered a useful method, several important factors must be carefully considered in the application and interpretation of the resulting data. Several reviews have been written on this increasingly diverse subject (1, 11, 18–22). No work to date has compared different aqueous sample preparation methods for application of the CLPP method to microbial communities from water samples of different constructed wetland designs. The physical and chemical characteristics of an aqueous sample to be inoculated onto a BIOLOG® EcoPlate can influence the information gathered relating to the microbial community function. CLPP of aquatic samples offers an advantage over soil or sediment samples in
Author's personal copy Appl Biochem Biotechnol
that the matrix can be inoculated directly into the wells without the need for detachment of biofilm, sample dilution, or major disruption of the microbial community (18). However, when biofilm communities are of interest, turbid samples are common. The data obtained from CLPP is related to the cell density of the inoculum. Differences in inoculum density will alter the extent and pattern of color development on the plate. Using a minimum inoculation, density of 105 cells/mL is the best way to reduce lag times although smaller inoculation densities can be used (11). True cell densities can be difficult to determine; therefore, an alternative inoculation approach is to dilute the sample to an optical density of 0.2 absorbance units (AUs) (11). Irrespective of cell density, turbid or opaque samples should be prepared in a way that ensures sufficient clarity and thus transmission of light during photometric detection of color development over time. Previous work has used an absorbance of 0.25 AU as a cutoff value to determine whether or not a specific carbon source in a well has been utilized [e.g., (16, 23)]. An absorbance of 2.0 AU represents the top end of the linear range for BIOLOG® EcoPlate absorbance values. Absorbance values above 2.0 AU are within the quenching region of the concentration-absorbance curve. Where an aqueous mixed microbial community sample is too dark or turbid, the sample should be prepared in a way that decreases turbidity while minimizing changes to the microbial diversity (24). Despite the common use and strong understanding of the fundamental properties of BIOLOG® EcoPlates, to date there has been no comprehensive investigation comparing different preparation methods for highly turbid samples. The aim of the present study was to compare and contrast different methods for improving the clarity of aqueous samples and explore their effects on the CLPP data from a selection of constructed wetland mixed microbial community samples with differing physical and chemical properties.
Methods Sample Collection Water samples were collected from pilot-scale constructed wetlands at the ecotechnology research facility in Langenreichenbach, Germany. A detailed description of the research facility can be found in Nivala et al. (25). A study on the microbial community metabolic function within the different constructed wetland facility is also available (5). Briefly, the research facility at Langenreichenbach is located adjacent to a community wastewater treatment system. The research facility receives raw wastewater and has a large septic tank that provides common primary treatment for all of the pilot-scale constructed wetlands at the site. The wetland cells each have a footprint between 5 and 6 m2. The influent to all wetlands is the effluent from the septic tank, which for sampling purposes, is collected as a grab sample from the main control building onsite just prior to dosing for a single wetland system. Samples were taken from several different subsurface flow-constructed wetland types (SI Fig. 1). Interstitial water samples were collected in saturated systems from sampling tees located at mid-depth of the water column by using a stainless steel extraction tube connected to a peristaltic pump with polyethylene tubing. The flow rate of the peristaltic pump was approximately 1 L/min, representing a relatively high flow rate allowing for sheering, sloughing, and collection of biofilm within a largely aqueous sample. The tubing was flushed with 500 mL of tap water followed by 500 mL of interstitial water before the biofilm/interstitial water sample was collected into a 500-mL sterilized glass bottle. The samples closer to inlets contained a
Author's personal copy Appl Biochem Biotechnol
significant amount of biofilm and were therefore highly turbid, where the samples close to the outlets were quite clear in comparison. Effluent samples were collected in the main control building, just upstream of the flow measurement device. Five samples representing a range of conditions were selected for this study (SI Fig. 1). All samples were stored in a cooler with ice and promptly transported back the laboratory for preparation and subsequent microbial community analysis.
Sample Preparation All reagents used were analytical grade. All glassware, buffer solutions, and cotton gauze were autoclaved prior to use (121 °C, 30 min). Six different sample preparation methods were investigated with the intention of preparing a suitably clear water sample to allow absorbance measurements to be carried out without interference due to turbidity. The investigated methods were as follows: (1) 10-fold dilution with phosphate-buffered solution (PBS), 10 mM, pH 7.4, 8.5 g/L NaCl), (2) 100-fold dilution with PBS, (3) centrifugation of a 50 mL aliquot (two times 5 min at 3000 rpm), (4) filtration through a sterilized loose-woven cotton gauze, (5) filtration through a sterilized 20-μm filter paper (P8 grade, Fisher), and (6) leaving the sample to settle for a minimum of 10 min. Each of the methods outlined above was applied to five different water samples giving a total of 30 samples for microbial community analysis.
Community-Level Physiological Profiling One BIOLOG® EcoPlate was used per water sample and preparation method. There were five mixed microbial community samples and six preparation methods, yielding 30 plates total. Each plate contained 31 carbon sources and a blank well in triplicate (96 wells in total) providing three replicated carbon source utilization patterns (CSUPs) per sample. Each plate was inoculated within 3 h of sample collection. Save the specific sample preparation method, the BIOLOG® EcoPlates were handled according to Weber and Legge (11). Briefly, all laboratory work was performed in an aseptic environment. All equipment involved in the procedure was sterilized using 70 % ethanol. Approximately 20 mL of the prepared sample was poured into a 9-cm petri dish, and then, 100 μL was added to each well of the BIOLOG® EcoPlates (BIOLOG Inc., Hayward CA., USA) using a multichannel pipette. The plates were incubated in the dark at 20 °C on an orbital shaker (Edmund Buehler, Germany) at 100 rpm. Optical density in the form of AUs of the inoculated plates was read using a microplate absorbance reader (Wallac Viktor2TM, Oy, Finland) at an absorbance of 590 nm at various time intervals from immediately after inoculation to 158 h.
Data Processing Analysis of the CLPP data was performed as previously described by Weber et al. (21) and Weber and Legge (11). Two different types of datasets were collated from absorbance data obtained from the microplates: (1) temporal data representing color development on each plate (sample) over the entire incubation period (0–158 h) and (2) data from a single time point, selected based on a combination of greatest variance between well responses and least number of absorbance values above 2.0 (values above 2.0 are above the linear absorbance range). Based on these two factors, the time point of 84 h was chosen for calculation of the average well color development (AWCD) and for multivariate analysis of CSUPs.
Author's personal copy Appl Biochem Biotechnol
The AWCD represents the average metabolic activity over all wells and is calculated as 1 X ðAi −A0 Þ 31 i¼1 31
AWCD ¼
ð1Þ
where Ai represents the absorbance reading of well i and A0 is the absorbance reading of the blank well (inoculated, but without a carbon source). Where there is very little response in a well, small negative values of standardized absorbance may occur, in which case they are coded as zeroes for subsequent data analysis. Principal component analysis (PCA) was performed using the covariance (n−1) matrix of CSUP data to further assess for differences between water samples. If necessary, datasets were subjected to natural logarithm (LN) transformation based on assessment of normality, homoscedasticity, and linear correlations following the recommendations of Weber et al. (21). Statistical significance of differences between samples was assessed using one-way analysis of variance (ANOVA) with Tukey’s post hoc test. All statistical analyses were completed using XLSTAT 2013 (Addinsoft New York, NY).
Water Quality Analysis Field measurements were conducted in the onsite lab at Langenreichenbach, Germany. Water temperature, electrical conductivity, redox potential, and dissolved oxygen were measured using a WTW Multi 350i multimeter. pH was measured using a WTW pH96 meter. Analysis of the remaining water quality parameters were carried out at the UBZ laboratory in Leipzig, Germany, on the same day as the samples were taken. Total organic carbon (TOC) was analyzed according to DIN EN 1484 using a Shimadzu TOC-VCSN device. Total nitrogen (TN) was analyzed according to DIN EN 12660, using a Shimadzu TNM-1 device. Turbidity was measured according to the European standard DIN ISO EN 27027 using a Hach 2100AN Turbidimeter.
Results and Discussion Sample Characterization The five aqueous samples collected from various stages of the subsurface flow-constructed wetlands covered a wide range of physical and chemical characteristics (SI Fig. 1). Most of the variation occurred with TOC concentrations ranging from 15 to over 400 mg/L and turbidity from 3.6 to greater than 170 NTU. Of the chemical properties, conductivity and pH showed less variation between samples than dissolved oxygen (DO) and total nitrogen (TN). Two samples had initial optical densities above the previously recommended minimum AU of 0.2: sample #1 collected from inside of a 50-cm deep horizontal subsurface flow wetland, near the inlet, and sample #2 the influent (septic tank effluent). Sample #1 and sample #2 had optical densities of 0.74 and 0.32, respectively. For sample #1, only 100-fold dilution was sufficient to reduce the sample absorbance (at 590 nm) to below 0.2 AU, while for sample #2, both 10-fold and 100-fold dilutions and filter paper were sufficient (Fig. 2). Samples #1 and #2 showed the largest amount of variation in absorbance between the six preparation methods. Sample #1
Author's personal copy Appl Biochem Biotechnol
showed significant (p
