Monitoring Microbial CommunityStructure and Dynamics during insitu U(VI) Bioremediation with aField-Portable Microarray AnalysisSystemD A R R E L L P . C H A N D L E R , * , †

A L E X A N D E R K U K H T I N , †

R E B E C C A M O K H I B E R , †

C H R I S T O P H E R K N I C K E R B O C K E R , †

D O R A O G L E S , ‡ G E O R G E R U D Y , †

J U L I A G O L O V A , † P H I L L O N G , § A N DA A R O N P E A C O C K |

Akonni Biosystems, Inc., 400 Sagner Avenue, Suite 300,Frederick, Maryland 21701, Microbial Insights, Inc., 2340Stock Creek Blvd., Rockford, Tennessee 37853, PacificNorthwest National Laboratory, Mail Stop K9-33, Richland,Washington 99354, and Haley & Aldrich, 103 NewhavenRoad, Oak Ridge, Tennessee 37830

Received March 1, 2010. Revised manuscript received May26, 2010. Accepted June 10, 2010.

The objective of this study was to develop and validate asimple, field-portable, microarray system for monitoring microbialcommunity structure and dynamics in groundwater andsubsurface environments, using samples representing sitestatus before acetate injection, during Fe-reduction, in thetransitionfromFe- toSO4

2--reduction,and into theSO42--reduction

phase. Limits of detection for the array are approximately102-103 cell equivalents of DNA per reaction. Sample-to-answer results for the field deployment were obtained in 4 h.Retrospective analysis of 50 samples showed the expectedprogression of microbial signatures from Fe- to SO4

2- -reducerswith changes in acetate amendment and in situ field conditions.The microarray response for Geobacter was highly correlatedwith qPCR for the same target gene (R2 ) 0.84). Microarrayresults were in concordance with quantitative PCR data,aqueous chemistry, site lithology, and the expected microbialcommunity response, indicating that the field-portable microarrayis an accurate indicator of microbial presence and responseto in situ remediation of a uranium-contaminated site.

Introduction

A universal theme that has emerged from two decades offundamental microbiology research at contaminated De-partment of Energy (DOE) sites is the importance of metal-and sulfate-reducing bacteria in subsurface bioremediation(e.g., 1-12). A typical strategy for radionuclide bioremediationinvolves injecting electron donors to stimulate metal reduc-tion by microbial communities native to contaminated

aquifers (1, 13), a strategy that has been tested at the DOEIntegrated Field Research Center located in Rifle, CO (1, 10).If the fundamental science from DOE field sites is going tobe translated into practical cleanup solutions for sitereclamation or stewardship, then it becomes important forsite engineers to understand which microbial groups arereducing (or reoxidizing) heavy metal contaminants in situ;which electron donors are most effective in stimulating theirrespective activities; and whether or not underlying changesin microbial community structure through time and spaceare diagnostic indicators of subsurface biogeochemicalprocesses.

Unfortunately, there are still very few methods forassessing in situ microbial community structure, activity, orremediation potential within a time frame or at a price pointthat impacts on-site treatment, remediation, or long-termstewardship decisions. Culture, phospholipids fatty acidanalysis, and various forms of polymerase chain amplificationdominate the remediation service industry, each of whichprovide useful information but suffer from limited coverageof the microbial community and/or ability to provide a moredirect indication of remediation potential, activity, or efficacy.At the other end of the technology spectrum are microarrays,which provide unparalleled opportunities for multiplexeddetection of nucleic acids and microorganisms. Severalgroups have now developed microarrays for the analysis offunctional genes (14, 15) and mRNA (16) in the environment,16S rRNA genes (17), and the direct detection of 16S rRNA(18-20). There is also an Affymetrix array containing probestargeting the entire Ribosomal Database Project (21), a systemthat has been applied to the analysis of microbial com-munities undergoing U reduction and reoxidation (22).Unfortunately, most of these microarray systems are tooexpensive, complex, labor intensive, and/or time-consuming(from sample acquisition and sample preparation throughto data reporting) to find practical use in the industry, letalone in the field (see, e.g., 23).

Rather than using microarrays as a fundamental scienceand discovery tool, an alternative approach is to take theaccumulated biological knowledge of subsurface remediationprocesses, and use low-density arrays as indicators ordiagnostics of microbial community structure and activity.In so doing, there is a corollary opportunity to reduce thefinancial and logistical burden of molecular monitoring, movearray technology closer to the field, and provide informationto site engineers in real time. The technical challenge posedby deploying microarrays in the field has been describedelsewhere (23). The objectives of this study were thereforeto take the fundamental knowledge from prior environmentalscience efforts, develop and validate a simple-to-use, field-portable, microarray-based system for monitoring microbialcommunity structure and dynamics, and monitor microbialcommunity succession in response to in situ U bioreme-diation in a field experiment.

Materials and MethodsSite Description and Sample Collection. The Rifle IFC sitelayout and geochemistry are described in detail elsewhere(1, 10), with the location of monitoring wells from the 2008“Big Rusty” gallery illustrated in Suppl. Figure 1 (SupportingInformation). Injection wells were installed to a depth of6.1 m (20 ft) and screened from 1.5 to 6.1 m (∼5 to 20 ft) toencompass the entire saturated interval of the aquifer. Acetateinjections during 2008 included 14 days of 50 mM acetate(Fe-reduction phase), a 7-day water flush (to slow the onsetof sulfate reduction), 10 days of 50 mM acetate (transition

* Corresponding author tel: 734-428-0713; fax: 301-698-0202; eFax:301-542-0120; e-mail: dchandler@akonni.com.

† Akonni Biosystems, Inc.‡ Microbial Insights, Inc.§ Pacific Northwest National Laboratory.| Haley & Aldrich.

Environ. Sci. Technol. XXXX, xxx, 000–000

10.1021/es1006498 XXXX American Chemical Society VOL. xxx, NO. xx, XXXX / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 A

from Fe- to SO42--reduction), and 59 days of 150 mM acetate

(SO42--reduction phase). One- to two-liter water samples

were filtered through a Sterivex 0.2-µm filter at each phaseand depth interval. Phases of the field experiment wereconfirmed by geochemical analysis of Fe(II), SO4

2-, andsulfide levels at all sampling depths (not shown).

Microarray Design and Manufacture. The microarrayused here (TruArray BER; Akonni Biosystems, Frederick, MD)is based on 16S rRNA-targeted arrays described in detailelsewhere (19, 24, 25), with the following modifications.Capture probe sequences were expanded to include fer-mentors, dechlorinators, dissimilatory metal-, sulfate-, andnitrate-reducer 16S rRNA sequences deposited in Genbank,utilizing full-length sequences for which an isolate is availablein a public culture collection, and for which the link betweenphylogeny and function has been established via culture-based methods. The array can be used for both direct rRNAdetection (25) and DNA detection (this study). In total, 150thermodynamically balanced microarray probes serve asindicators for 37 genera, as described in Suppl. Table 1(Supporting Information).

Microarrays were manufactured by a single-step copo-lymerization technique and custom polymer at AkonniBiosystems. Probes were printed in quadruplicate at 0.125mM concentration. Three complete arrays were printed persubstrate and individually fitted with a Grace Biolabs frameseal gasket. Fabricated arrays were stored dry and in thedark under room temperature/ambient conditions until use.

Positive Control Isolates. Nucleic acids were purified fromGeobacter metallireducens, G. sulfurreducens, Clostridiumacetobutylicum, C. glycolicum, Desulfovibrio vulgaris, andThaurea aromatica according to standard techniques, andquantified by fluorometric analysis on a NanoDrop ND-3300(Thermo Scientific, Wilmington, DE). DNA was prepared in10-fold serial dilutions from 1 fg to 100 pg µL-1 in ultrapurewater for protocol development and analytical performanceevaluations.

Nucleic Acid Extraction from Environmental Samples.For the 2008 deployment of the microarray system to theRifle IFC field trailer, we utilized an Akonni SPT TruTip DNA/RNA Extraction Kit (Frederick, MD) and MoBio PowerSoilDNA Isolation Kit (Carlsbad, CA) as per the manufacturer’sinstructions. For subsequent microarray validation and tocorrelate microarray signals with quantitative PCR data, allsamples were extracted with the MoBio kit. The TruTip hasa binding matrix embedded in a 2-mL aerosol barrier pipettip, such that the only equipment required for nucleic acidpurification is a pipettor.

Filter samples (representing 1-2 L of groundwater) wereexcised from their respective cartridges, cut into thin stripswith sterile scissors and added in their entirety to a MoBioPowerBead tube containing 60 µL of MoBio lysis solution.After vortexing, the supernatant was divided evenly betweenthe MoBio and TruTip procedures. The elution volume forboth methods was 100 µL, representing 10-20 mL equivalentof groundwater per microliter purified extract. Purifiednucleic acids were either used immediately (in the fielddeployment) or stored at -20 °C until use (validation study).

Nucleic Acid Amplification and Labeling. For the in-field experiment, each of two replicate extracts per filtersample was amplified in an individual 50 µL reaction; for thevalidation study, three separate amplifications were per-formed from each individual extract. Optimized amplificationconditions were 1X Phusion Hot Start High Fidelity DNAPolymerase PCR buffer (New England Biolabs, Ipswich, MA),0.25 mM of each dNTP, 10 µL of TruArray BER asymmetricprimer mix, 1 µL of Phusion Hot Start High Fidelity DNApolymerase (New England Biolabs), and 3 µL of MoBio-purified nucleic acid (30-60 mL equivalents groundwater).The asymmetric primer mix contains one labeled universal

16S rRNA primer and one unlabeled universal 16S rRNAprimer in unequal ratios, which leads to a nonexponentialamplification reaction and accumulation of a single-stranded,labeled product that can be hybridized directly to themicroarray without additional fragmentation, labeling, orpurification. PCR amplification was conducted in a Pikothermal cycler (Finnzymes, Woburn, MA) at 98 °C for 30 s,25 cycles of [98 °C for 5 s, 59 °C for 5 s, 72 °C for 5 s], 35 cyclesof [98 °C for 5 s, 64 °C for 5 s, 72 °C for 5 s], and a finalextension at 72 °C for 1 min.

Quantitative PCR. Quantitative PCR for specific bacteriaor groups was based on Microbial Insights’ series of CENSUSqPCR assays, utilizing 3 µL of MoBio-purified environmentalnucleic acid and PCR tests for eubacteria, iron- and sulfate-reducing bacteria, Geobacter, and dissimilatory sulfite re-ductase. Results are presented as cell equivalents per mL ofgroundwater.

Microarray Hybridization and Wash. For each replicateamplification, 15 µL of (unpurified) amplified product wasbrought to 30 µL total volume with microarray hybridizationbuffer, 1 µL of BSA (100 mg mL-1), 0.8 nM Cy3-labeledTruArray BER internal positive control, and ultrapure water.The internal positive control is a synthetic 92-mer with noknown homology to any sequence in any database, and servesas a positive control for target hybridization only. Duringassay development and verification, experiments were alsoperformed with chaperone/helper probes (one for each genusrepresented by the array) based on prior work (19, 24, 25)that showed chaperone/helper probes improve (planar- andbead-) array hybridization specificity and sensitivity bydisrupting secondary and tertiary structures in the 16S rRNAtarget. Target nucleic acids were heat denatured for 3 minat 93 °C and 28 µL was applied to the array. The array wassealed with parafilm and then statically incubated for 3 h ateither ambient temperature (20-24 °C, field experiment) orovernight at 37 °C in an MJ Research in situ tower (validationstudy). After hybridization, parafilm covers were removed,arrays were placed in a histology slide holder and washed ina bulk container containing 500 mL of 6X SSPE, 0.05% SDSfor 10 min with intermittent agitation. Thereafter, the arrayswere sequentially transferred through two containers ofultrapure water. Excess liquid in the frame seal chamber wasremoved with a Kimwipe, and arrays were air-dried (∼5 min)before imaging.

Imaging and Data Analysis. Dried arrays were imagedon an Akonni field-portable TruDx 1000 imager for up to10 s. Exposure time was adjusted to avoid pixel saturation.Images were saved as 16-bit raw .tif files and exported toSpotfinder v3.1.1 software. A custom-written PERL scriptautomatically applied an analysis grid to the image, with theoperator verifying that the grid was properly aligned prior todata extraction. Data extraction used an Otsu threshold fordefining foreground from background pixels, and localbackground was subtracted from each gel element pixel.Integrated, background-corrected signal intensities wereaveraged across all replicate probes (including the nonsenseprobes, N), and a signal-to-noise ratio (SNR) was calculatedrelative to the average, background-corrected nonsenseresponse. SNR data for the n ) 2 (field trial) or n ) 3(validation) replicate arrays were then averaged, resulting inan average microarray profile for each sample. For thepurpose of this study, we used an average SNR > 3 fordeclaring probes detectable over all sources of noise.Microarray data and profiles were visualized as “heat map”displays using TreeView v1.6 (26). For quantitative com-parison to qPCR data, SNR values for all probes within afunctional group were first summed and then log10-transformed. Other comparisons and data analysis methodsare described in text.

B 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. xxx, NO. xx, XXXX

Results and DiscussionAnalytical Sensitivity and Specificity. Some of the assayoptimization data are shown in Suppl. Figure 2 (SupportingInformation) for a 3 h static hybridization at 37 °C. Limits ofdetection (LoD) for the microarray method against a G.metallireducens genomic DNA target were 10 fg (approxi-mately 2 cell equivalents) per reaction (Suppl. Figure 2A andB). Chaperone probes in the hybridization mix led to a generaldecrease in microarray LoD (from 10 fg to 1 pg), as shownin Suppl. Figure 2B for all Geobacter-specific probes. Limitsof detection for other positive control isolates were generally100 fg to 1 pg per reaction in the absence of chaperone/helper probes (not shown). We therefore estimate the LoDto be between ∼100 and 1000 genomes per reaction for a 3 hhybridization protocol, a result that is comparable with real-time PCR data, known copy-number requirements forreproducible PCR, and intrinsic biomass of subsurfaceenvironments prior to in situ remediation.

The microarray response (the totality of detectable signalsover all probes) was fairly specific to the target organisms(Suppl. Figure 2C and 2D). At 100 pg DNA per reaction andin the absence of chaperone probes, strong cross-hybridiza-tion (SNR > 10) beyond the targeted genus was only observedfor a few probes, as described in the Suppl. Figure 2C legend.No template reactions were negative over all probes or onlyintermittently positive at a SNR > 3, including all testsconducted in the field. As in prior studies with 16S rRNA-targeted arrays (19, 24), chaperone/helper probes improvedthe analytical specificity of the array response, but did notentirely eliminate the observed cross-hybridization (notshown). We therefore chose to exclude chaperone probesfrom the method and analysis of environmental samples,favoring the improvement in LoD over some loss in analyticalspecificity

The limited extent of cross-hybridization is not necessarilya technology flaw or impediment to diagnostic use in anenvironmental setting, as discussed in detail elsewhere (25).For community profiling applications, we advocate thatindividual probes are simply indicators of the cognate geneor organism, that correlated samples are required forinterpreting microarray data, and that it is the relative changein probe A versus probe A across the correlated sample setthat carries biological or ecological information. In the eventthat a specific “name” of an organism or group of organismsis important to the user, then the microarray result shouldbe followed by a targeted qPCR assay. Thus, the analyticaldata are consistent with prior 16S rRNA-targeted arrays, anddisplay analytical sensitivity and specificity on par with morecomplex, lab-based microarray analysis systems (e.g., 15, 21,22, 27).

Field Deployment. The microarray system was deployedto the Rifle IFC field trailer for a one-week trial in July 2008.Approximately 100 microarray analyses were conducted bya single technician over a 4-day period. Some data for theMoBio and Akonni sample preparation comparison areshown in Figure 1 for background well U01 and downstreamwell D01 at t ) 0 and t ) 4 days. Average SNRs for these dataranged from 3.1 to 64.2. No template reactions were eithernegative for all probes or sporadically positive for randomprobes (i.e., no consistently positive probes for any notemplate reaction; not shown), indicating no cross-con-tamination between samples (or from other 16S rRNA-targeted PCR amplification activity occurring simultaneouslyin the field trailer). Relative to the MoBio sample preparationmethod, TruTip-purified nucleic acids showed an increasein the number of reactive probes and the detection of theexpected microbial community shift toward Fe reduction att ) 4 days. We naturally expect that different samplepreparation methods will lead to different lysis, extraction,and purification efficiencies and, ultimately, molecular

signatures (28-30). Considering both techniques utilized thesame (MoBio) lysis method, differences between the twosample preparation methods are most likely related toextraction and purification efficiency. What is more important(and encouraging) about these results is that the vast majorityof microarray signatures detected from the MoBio extractwere also detected via the TruTip method (Sørensen similarityindex)0.81 for TruTip-generated profiles and 0.71 for MoBio-generated profiles at t ) 0), which indicates that a 5-min,in-field TruTip sample preparation procedure is efficient andeffective for filtered groundwater samples.

Logistically, sample acquisition took approximately ∼15min per filter, and the entire assay (sample-to-answer) wastranslated into microbial community profiles within 4 h ofsample receipt. We conclude from these experiments thatthe microarray and field portable system can provideecologically relevant information to site engineers within asingle shift, and in the field. Analytical method complexity,equipment infrastructure, and assay turnaround time aretherefore no longer an impediment to field deployment ofmicroarray technology and real-time monitoring of microbialcommunity response to environmental conditions.

Assay Validation and Microbial Community DynamicsDuring in situ Bioremediation. To validate the microarray,

FIGURE 1. In-field microarray analysis of microbial communityresponse to acetate injection, showing an expected bloom inGeobacter at t ) 4 days. Only average signal-to-noise ratios(SNR) > 3 are plotted in the heat map. Band intensity iscorrelated with the absolute average SNR values for eachsample, but is not a quantitative reflection of SNRs due to thenumber of bins used to generate the heat map.

VOL. xxx, NO. xx, XXXX / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 C

FIG

URE

2.(A

)H

eat

map

disp

lay

ofav

erag

e(n

)3

arra

yspe

rsa

mpl

e)m

icro

arra

yre

spon

sefo

ral

lpr

obes

with

aSN

R>

3.Ea

chse

tof

four

profi

les

isor

gani

zed

bydi

stan

cedo

wn-

grad

ient

from

the

inje

ctio

nga

llery

(Sup

pl.F

igur

e1

inSu

ppor

ting

Info

rmat

ion)

,by

dept

h,an

dby

phas

eof

the

acet

ate

inje

ctio

n.(B

)Ta

qMan

qPCR

data

(n)

1)fo

rea

chgr

ound

wat

ersa

mpl

e.Th

edo

tted

line

acro

ssal

lgr

aphs

repr

esen

tsth

eap

prox

imat

elim

itof

dete

ctio

nfo

rth

em

icro

arra

y,as

estim

ated

from

Supp

l.Fi

gure

2B.(

C)G

roup

-spe

cific

mic

roar

ray

sign

alin

tens

ityfo

rth

ose

func

tiona

lgr

oups

that

wer

eal

som

easu

red

byqP

CR.S

igna

lto

nois

era

tios

over

all

prob

esw

ithin

the

desc

ribe

dfu

nctio

nal

grou

pw

ere

first

sum

med

,and

then

log 1

0tr

ansf

orm

edto

coin

cide

with

the

qPCR

data

scal

e.Ch

arts

and

indi

vidu

alpo

ints

inpa

nels

Ban

dC

are

alig

ned

with

the

corr

espo

ndin

gw

ell,

dept

h,ph

ase,

and

mic

roar

ray

heat

map

sign

atur

ein

pane

lA

.

D 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. xxx, NO. xx, XXXX

we applied the system to a retrospective analysis of 50groundwater filter samples associated with the U02 to D10transect (Suppl. Figure 1 in Supporting Information). Theoverarching purpose of the 2008 field experiment was (inpart) to understand the microbiological processes contribut-ing to U(VI) removal from groundwater during the in situtransition from iron- to sulfate-reduction and into the sulfate-reduction phase. Changes in microbial community com-position as a function of distance from the injection gallery,depth, and phase of in situ remediation are plotted as heatmaps in Figure 2A. Matching no template controls (n ) 33)were all negative or sporadically positive for one or tworandom probes (as in the field deployment). The expectedbloom in Geobacters is clearly evident in the Fe-reductionphase, with a corresponding reduction in Geobacter signa-tures (number of probes) and intensity as the systemtransitioned to and entered sulfate reduction. Wells and alldepth intervals nearest the injection gallery showed thestrongest Geobacter response, with well D10 showing a lagin community response during all phases of the fieldexperiment, a response that is consistent with site lithologyand groundwater flow through the gallery.

The general pattern of Geobacter probe response is alsoreflected in the qPCR data (Figure 2B) and by plotting thelog10(Σ average SNR) over all Geobacter-specific probes (Figure2C). Importantly, only those samples with >102 Geobactercell equivalents mL-1 were detectable by the microarray,consistent with the estimated lower limit of detection for thearray-based method (Suppl. Figure 2B). Equally importantis that the total Geobacter-specific microarray signal washighly correlated with the Geobacteraceae qPCR results (R2

) 0.84, Suppl. Figure 3).

As measured by qPCR, the total eubacterial communityfluctuated between ∼105 and 107 cell equivalents mL-1

whereas the combined iron- and sulfate-reducer assay variedover 5 orders of magnitude (Figure 2B). The dsr gene wasconsistently detected at <103 copies mL-1 by PCR, but therewas no correlation between dsr gene abundance and theiron- and sulfate-reducer microarray values (R2)0.008), evenduring the sulfate-reducing phase of the field experiment(not shown). Likewise, there was no correlation betweeneither the dsr qPCR and Desulfobacter or total sulfate reducermicroarray signals (R2 < 0.002). These results suggests thatthe dsr TaqMan assay may actually be too specific formonitoring sulfate-reducer community dynamics, function,or activity at the Rifle site. On the other hand, averagedmicroarray SNR values for the eubacteria and (more specif-ically) iron- and sulfate-reducer groupings did track with therise and fall in the corresponding qPCR data (Figure 2C).

The correlation between total microarray SNR and qPCRwas rather weak for the eubacterial and iron/sulfate reducergroupings (R2 ) 0.36 and 0.33 respectively; Suppl. Figure 3).We believe that the weaker correlation between microarrayand TaqMan values at the higher taxonomic groups is dueto the fact that the qPCR and microarray are measuringdifferent components of the total microbial community atthese functional levels. The microarray, for example, hasbroader taxonomic coverage of iron- and sulfate-reducersthan does the qPCR (which does not detect Desulfotobac-terium, for example), whereas the universal eubacterial qPCRis much more encompassing than the low-complexity,targeted microarray.

Other elements of the microarray response during thevalidation study were also consistent with prior knowledge

FIGURE 3. Fold-change of genus-level microarray SNR relative to the correlated U02 background sample. Total SNR for eachfunctional group was summed over each depth interval within the well, and then averaged across the three downstream wells toarrive at a global average SNR for each genus during each phase of the field experiment. Averaged SNR values were then dividedby the corresponding genus-level SNR in up-gradient sample U02 to calculate a fold-change in array response. For those genera thatwere undetectable in the upstream or downstream gradients (SNR < 3), the SNR was set to 1 prior to the calculation to avoid a“divide by zero” error. Thereafter, calculated fold-changes were log2 transformed as an estimate of an underlying doubling effect inthe detected genera. *The Ferribacterium probe also has perfect sequence identity with Dechloromonas aromaticum, and is thereforenot in and of itself an absolute indicator of Ferribacterium presence.

VOL. xxx, NO. xx, XXXX / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 E

of site microbiology, anticipated community shifts due toacetate injection, and the progression of in situ geochemicalconditions from iron- to sulfate-reduction. For example, aconsistent Desulfotomaculum signature was detected in theU01 background and all downstream samples at all depthsand during all phases of the field trial, a signature that wasalso detected via bead array analysis of 16S rRNA (25).However, the fold-change in the Desulfotomaculum signature(when averaged over all downstream samples, and relativeto the corresponding U02 background at each phase of thefield experiment), suggested little to no change in thiscomponent of the community over the course of the fieldexperiment (Figure 3). This conclusion is also supported byfine-scale analysis of the microarray signatures, wherein theDesulfotomaculum response was restricted to the same twoprobes throughout the study. Such an interpretation is slightlydifferent than might otherwise be deduced by looking solelyat the heat map in Figure 2A, where the absolute intensityof the Desulfotomaculum probe response does show adecrease in SNR intensity relative to samples collected beforeacetate treatment and during iron reduction. A similar resultwas obtained for the nitrate reducers; the heat map displayin Figure 2 shows very strong probe responses for Dechlo-romonas and Dechlorosoma that constitute up to 70% of thetotal array SNR signal (depending on the sample), but thefold-change plot indicates that neither of these genera wasvery responsive to the changes in site geochemistry over thecourse of the field experiment. On the other hand, there isa clear Geobacter, Geothrix, and Pelobacter response (relativeto the U02 background) during iron reduction, a relativeresponse that held through sulfate reduction even thoughdata in Figure 2 show a decline in total Geobacter SNR signalbeginning with the iron-sulfate transition phase and coin-cident with distance downstream of the injection gallery.

There was likewise an obvious Desulfobacter and sulfate-reducer response to the acetate treatment, beginning withthe onset of iron reduction and continuing well into thesulfate-reducing phase. From Figure 3 it would seem thatthe global Desulfobacter response was sustained over theentire field trial, but fine-scale microarray analysis showsthat the Desulfotobacter signature was quite sporadic duringthe iron-reduction phase and was then consistently detectedat all depths and all downstream wells beginning with theiron-sulfate transition (Figure 2A). We likewise see in Figure2A that the diversity of sulfate-reducing bacteria increasedas the system was driven into sulfate reduction, withresponses from a more diverse set of Desulfobacter andDesulfitobacterium species-specific probes than is seenduring Fe reduction or the iron to sulfate transition. Thesedata, in combination with the decline in Geobacter and othermetal-reducer signatures during the Fe- to SO4

2--transition,support a hypothesis of Geobacter stress, and represent adiagnostic signal of a fundamental shift in communitycomposition and in situ activity.

Other views of the microarray data are consistent withthe physical and hydraulic properties at the site, and provideinsight into nutrient delivery (hence bioremediation ef-fectiveness) to the downstream gallery. For example, signal-to-noise ratios at the probe, genus and higher taxonomicgroupings were stronger in the 12-15 ft interval than at the20 ft depth through all phases of the field trial (Suppl. Figure 4in the Supporting Information). If the upstream U02 mi-croarray signals are added to the down-gradient well data,then there is an even more pronounced signature at the 12ft depth than the 15 and 20 ft depths (not shown). This resultmay be related to the physical proximity of U02 to theinjection gallery and sample collection, in that the samplingmethod and/or total sample volume may have pulledgroundwater (hence, stimulated biomass) from near theinjection gallery. Regardless, the array recorded depth-related

signatures that are consistent with subsurface microbiology,subsurface lithology, and the delivery of acetate to thoseareas in the injection gallery. We therefore conclude that themicroarray system can be used as a diagnostic indicator ofmicrobial response and U(VI) reduction in the subsurface.

AcknowledgmentsWe are indebted to Dr. Ken Williams, Richard Dayvault, andthe entire Rifle IFC project for field support and sampleacquisition. This work was supported by Phase II SBIR grant200-2006-19011 from the U.S. Department of Energy (DOE),and project 51882 from the Environmental Research SciencesProgram, Office of Science, DOE. Pacific Northwest NationalLaboratory is operated by Battelle for the United StatesDepartment of Energy under contract DE-AC06-76RL01830.

Supporting Information AvailableSupplemental Table 1 and Suppl. Figures 1-4. This materialis available free of charge via the Internet at http://pubs.acs.org.

Literature Cited(1) Anderson, R. T.; Vrionis, H. A.; Ortiz-Bernad, I.; Resch, C. T.;

Long, P. E.; Dayvault, R.; Karp, K.; Marutzky, S.; Metzler, D. R.;Peacock, A.; White, D. C.; Lowe, M.; Lovley, D. R. Stimulatingthe in situ activity of Geobacter species to remove uranium fromthe groundwater of a uranium-contaminated aquifer. Appl.Environ. Microbiol. 2003, 69 (10), 5884–5891.

(2) Coates, J. D.; Phillips, E. J. P.; Lonergan, D. J.; Jenter, H.; Lovely,D. R. Isolation of Geobacter species from diverse sedimentaryenvironments. Appl. Environ. Microbiol. 1996, 62 (5), 1531–1536.

(3) Cummings, D. E.; Snoeyenbos-West, O. L.; Newby, D. T.;Niggemyer, A. M.; Lovley, D. R.; Achenbach, L. A.; Rosenzweig,R. F. Diversity of Geobacteraceae species inhabiting metal-polluted freshwater lake sediments ascertained by 16S rDNAanalyses. Microb. Ecol. 2003, 46, 257–269.

(4) Lovley, D. R. Dissimilatory Fe(III) and Mn(IV) reduction.Microbiol. Rev. 1991, 55, 259–287.

(5) Lovley, D. R. Fe(III)- and Mn(IV)-reducing prokaryotes In TheProkaryotes; Dworkin, M., Rosenberg, S. F. E., Schleifer, K. H.,Stackebrandt, E., Eds.; Springer-Verlag: New York, 2001.

(6) Nevin, K. P.; Finneran, K. T.; Lovley, D. R. Microorganismsassociated with uranium bioremediation in a high-salinitysubsurface sediment. Appl. Environ. Microbiol. 2003, 69 (6),3672–3675.

(7) Murillo, F. M.; Gugliuzza, T.; Senko, J.; Basu, P.; Stolz, J. F. Aheme-C-containing enzyme complex that exhibits nitrate andnitrite reductase activity from the dissimilatory iron-reducingbacterium Geobacter metallireducens. Arch. Microbiol. 1999,172, 313–320.

(8) Senko, J. M.; Stolz, J. F. Evidence for iron-dependent nitraterespiration in the dissimilatory iron-reducing bacterium Geo-bacter metallireducens. Appl. Environ. Microbiol. 2001, 67 (8),3750–3752.

(9) Treude, N.; Rosencrantz, D.; Liesack, W.; Schnell, S. Strain FAc12,a dissimilatory iron-reducing member of the Anaeromyxobactersubgroup of Myxococcales. FEMS Microb. Ecol. 2003, 44, 261–269.

(10) Vrionis, H. A.; Anderson, R. T.; Ortiz-Bernad, I.; O’Neill, K. R.;Resch, C., T.; Peacock, A. D.; Dayvault, R.; White, D. C.; Long,P. E.; Lovley, D. R. Microbiological and geochemical hetero-geneity in an in situ uranium bioremediation field site. Appl.Environ. Microbiol. 2005, 71 (10), 6308–6318.

(11) Lovley, D. R.; Widman, P. K.; Woodward, J. C.; Phillips, E. J.Reduction of uranium by cytochrome c3 of Desulfovibriovulgaris. Appl. Environ. Microbiol. 1993, 59, 3572–3576.

(12) Tebo, B. M.; Obratsova, A. Y. Sulfate reducing bacterium growswith Cr(VI), U(VI), Mn(IV) and Fe(III) as electron acceptors.FEMS Microbiol. Lett. 1998, 162, 193–198.

(13) Istok, J. D.; Senko, J. M.; Krumholz, L. R.; Watson, D.; Bogle,M. A.; Peacock, A.; Chang, Y. J.; White, D. C. In situ bio-reductionof technetium and uranium in a nitrate-contaminated aquifer.Environ. Sci. Technol. 2004, 38 (2), 468–475.

(14) Palumbo, A. V.; Schryver, J. C.; Fields, M. W.; Bagwell, C. E.;Zhou, J.-Z.; Yan, T.; Liu, X.; Brandt, C. C. Coupling of functionalgene diversity and geochemical data from environmentalsamples. Appl. Environ. Microbiol. 2004, 70 (11), 6525–6534.

F 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. xxx, NO. xx, XXXX

(15) Wu, L.; Liu, X.; Schadt, C. W.; Zhou, J. Microarray-based analysisof subnanogram quantities of microbial community DNAs byusing whole-community genome amplification. Appl. Environ.Microbiol. 2006, 72 (7), 4931–4941.

(16) Gao, H.; Yang, Z. K.; Gentry, T. J.; Wu, L.; Schadt, C. W.; Zhou,J. Microarray-based analysis of microbial community RNAs bywhole-community RNA amplification. Appl. Environ. Microbiol.2007, 73 (2), 563–571.

(17) Loy, A.; Lehner, A.; Lee, N.; Adamczyk, J.; Meier, H.; Ernst, J.;Schleifer, K.-H.; Wagner, M. Oligonucleotide microarray for 16SrRNA gene-based detection of all recognized lineages of sulfate-reducing prokaryotes in the environment. Appl. Environ.Microbiol. 2002, 68 (10), 5064–5081.

(18) Small, J. A.; Call, D. R.; Brockman, F. J.; Straub, T. M.; Chandler,D. P. Direct detection of 16S rRNA in soil extracts usingoligonuleotide microarrays. Appl. Environ. Microbiol. 2001, 67(10), 4708–4716.

(19) Chandler, D. P.; Jarrell, A. E. Automated purification andsuspension array detection of 16S rRNA from soil and sedimentextracts using tunable surface microparticles. Appl. Environ.Microbiol. 2004, 70 (5), 2621–2631.

(20) Eyers, L.; Smoot, J. C.; Smoot, L. M.; Bugli, C.; Urakawa, H.;McMurry, Z.; Siripong, S.; Fantroussi, S. E.; Lambert, P.; Agathos,S. N.; Stahl, D. A. Discrimination of shifts in a soil microbialcommunity associated with TNT-contamination using a func-tional ANOVA of 16S rRNA hybridized to oligonucleotidemicroarrays. Environ. Sci. Technol. 2006, 40 (19), 5867–5873.

(21) Wilson, K. H.; Wilson, W. J.; Radosevich, J. L.; DeSantis, T. Z.;Viswanathan, V. S.; Kuczmarski, T. A.; Andersen, G. L. High-density microarray of small-subunit ribosomal DNA probes.Appl. Environ. Microbiol. 2002, 68 (5), 2535–2541.

(22) Brodie, E. L.; DeSantis, T. Z.; Joyner, D. C.; Baek, S. M.; Larsen,J. T.; Andersen, G. L.; Hazen, T. C.; Richardson, P. M.; Herman,D. J.; Tokunaga, T. K.; Wan, J. M.; Firestone, M. K. Application

of a high-density oligonucleotide microarray approach to studybacterial population dynamics during uranium reduction andreoxidation. Appl. Environ. Microbiol. 2006, 72 (9), 6288–6298.

(23) Chandler, D. P.; Jarrell, A. E. Taking arrays from the lab to thefield: trying to make sense of the unknown. BioTechniques 2005,38 (4), 591–600.

(24) Chandler, D. P.; Newton, G. J.; Small, J. A.; Daly, D. S. Sequencevs. structure for the direct detection of 16S rRNA on planaroligonucleotide microarrays. Appl. Environ. Microbiol. 2003,69 (5), 2950–2958.

(25) Chandler, D. P.; Jarrell, A. E.; Roden, E. R.; Golova, J.; Chernov,B.; Peacock, A. D.; Long, P. E. Suspension array analysis of 16SrRNA from Fe- and SO4

2--reducing bacteria in uraniumcontaminated sediments undergoing bioremediation. Appl.Environ. Microbiol. 2006, 72 (7), 4672–4687.

(26) Eisen, M. B.; Spellman, P. T.; Brown, P. O.; Botstein, D. Clusteranalysis and display of genome-wide expression patterns. Proc.Natl. Acad. Sci. U.S.A. 1998, 95 (25), 14863–14868.

(27) Wu, L.; Thompson, D. K.; Liu, X.; Fields, M. W.; Bagwell, C. E.;Tiedje, J. M.; Zhou, J. Development and evaluation of microarray-based whole-genome hybridization for detection of microor-ganisms within the context of environmental applications.Environ. Sci. Technol. 2004, 38, 6775–6782.

(28) Wintzingerode, F.v.; Gobel, U. B.; Stackebrandt, E. Determinationof microbial diversity in environmental samples: pitfalls of PCR-based rRNA analysis. FEMS Microbiol. Rev. 1997, 21, 213–229.

(29) Frostegård, Å.; Courtois, S.; Ramisse, V.; Clerc, S.; Bernillon, D.;Gall, F. L.; Jeannin, P.; Nesme, X.; Simonet, P. Quantification ofbias related to the extraction of DNA directly from soils. Appl.Environ. Microbiol. 1999, 65, 5409–5420.

(30) Martin-Laurent, F.; Philippot, L.; Hallet, S.; Chaussod, R.;Germon, J. C.; Soulas, G.; Catroux, G. DNA extraction from soils:Old bias for new microbial diversity analysis methods. Appl.Environ. Microbiol. 2001, 67 (5), 2354–2359.

ES1006498

VOL. xxx, NO. xx, XXXX / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 G