Polyphasic characterization of microbial communities under the stressful conditions of nitrate, heav

1
Polyphasic characterization of microbial
communities under the stressful conditions of
nitrate, heavy metals, radionuclides, and acidic
pH in contaminated groundwater M.W. Fields1, T.
Yan2, X. Liu2, C.E. Bagwell3, S.L. Carroll2, P.M.
Jaridne2, C.S.Criddle4, T.C. Hazen5, and J.
Zhou1 Miami University1, Oak Ridge National
Laboratory2, Savannah River Technology Center3,
Stanford University4, and Lawarence Berkeley
National Laboratory5
Abstract Clonal libraries of multiple genes (SSU
rRNA gene, nirK, nirS, amoA, pmoA, and dsrAB)
were constructed from groundwater samples (n6)
that varied in degrees of contamination. The
sites were in proximity to a concentration plume
and the following geochemical parameters were
measured pH, nitrate, uranium, nickel, sulfate,
and TOC (total organic carbon). Principle
components analysis (PCA) was used to compare the
relationships between the sites with respect to
the unique operational taxonomic unit (OTU)
distributions within the different clonal
libraries. When geochemical characteristics were
analyzed, the data suggested that the samples
varied significantly with respect to nitrate and
sulfate. The sites were grouped differently when
compared upon the basis of the SSU rDNA genes or
each of the functional genes. The diversity of
amoA genes was much lower compared to the other
functional genes and the one sample with high
nitrate and circum neutral pH appeared to be
different from the other sites. When all gene
OTUs were used in the analyses, the sites were
more similar than in any other comparison, and
the background site was grouped with the acidic,
contaminated sites. The results indicated that a
combination of the contaminants (i.e., nitrate,
uranium, nickel) and pH levels have impacted the
bacterial communities at the respective sites in
different and dynamic ways. These data suggested
that even though the background site was
phylogenetically distinct from the acidic sites,
the extreme conditions of the acidic samples
might be more analogous to the limiting nutrient
conditions of the background site.
http//www.esd.ornl.gov/nabirfrc/
The pH, nitrate, uranium, and nickel levels in
the groundwater samples from four FRC groundwater
samples. FW-300 represents the background area.
Values were determined by the FRC management team
at Oak Ridge National Laboratory. Well
pH nitratea uraniumb nickelc aluminumc (
mM) (µM) (µM) (mM) FW-300
6.1 0.02 ND 0.85 0.01 FW-005
3.9 6.27 27.0 84.3 1.74 FW-010
3.5 713 0.71 322 41.5 FW-015
3.4 173 32.4 147 22.9 TPB-16 6.3 0.48 4
.62 ND 0.01 FW-003 6.0 17.1 0.04 0.26 0.02
a nitrate was determined via ion chromatography b
uranium was determined via ICP-mass
spectroscopy C nickel and aluminum were
determined via ICP
Number of Screened and Unique Clones for each
Gene Set
a Fields et al. (in review) b Yan et al., 2003
c Yan et al., (in preparation) d Bagwell et
al., (in review)
Principal components analyses. The clonal
sequences for the SSU rRNA genes, nirK, nirS,
amoA/pmoA, and dsr were compared via RFLP and/or
direct sequence determination, and the occurrence
and distribution of phylogenetically distinct
OTUs was determined. The data was reduced with
factor analysis using principal components.
Geochemistry. When the geochemical parameters
of pH, nitrate, TOC, uranium, nickel, sulfate and
aluminum were used, 85 of the variance could be
explained by two components. The background site
and TPB-16 were distinct from the other sites
(least contaminated), and FW-010 and FW-015 were
very similar. The background and TPB-16 had
circum neutral pH and the lowest nitrate, nickel,
and aluminum levels. SSU rRNA sequences. When
the SSU rRNA gene sequences were compared the
background site was grouped with FW-003, TPB-16
was distinct from all other sites, and the three
acidic sites were grouped. The two components
accounted for only 58 of the variance in SSU
rRNA sequences between the sites. amoA/pmoA.
The FW-003 site was distinct from the others when
the amoA and pmoA sequences were compared and
this site had circum neutral pH and increased
nitrate levels. nirK and nirS. When the nirS
sequence were compared, the two contaminated
sites with circum neutral pH were distinct from
the other sites as well as from one another. The
background site and acidic sites were grouped.
The distribution of nirK sequences were even more
similar, however FW-015 was distinct from the
other sites that were all closely
grouped. dsrAB. When the distribution of unique
dsrAB sequences were compared, site TPB-16 and
FW-015 were distinct from the other sites which
were closely grouped. The background site,
FW-003, FW-010, and FW-005 were affected by the
loading factors in a similar fashion. All gene
sequences. When the distribution of all unique
gene sequences were compared, overall the sites
were more closely grouped. Two components could
explain 85 of the variance of gene distribution
among the sites, and this was similar to the
explained variance by the geochemical
characteristics. However, the background site
did group more closely with the acidic sites, and
the circum neutral sites, FW-003 and TPB-16 were
distinct. The PC1 from the geochemical
characteristics explained 65 of the observed
variance. When the geochemical PC1 value was
compared to the diversity index (1/D) estimated
from the SSU rRNA gene sequences, a power
function correlation was observed (r0.85). When
the all gene PC1 value (74) was compared to the
1/D estimation, a significant correlation was not
observed.
The estimated species number and diversity based
upon cloned SSU rRNA genes. The number of
clones screened and the estimated species number
based on less than or equal to 98 nucleotide
identity for the same species (unique OTU). The
Shannon-Weiner and reciprocal Simpsons index are
diversity indices, and a higher number represents
more diversity. The evenness index estimates the
homogeneity of the clone distribution, and a
community has an increased number of different
clones with a similar distribution as E
approaches 1. The site TPB-16 appeared to be
more diverse than background, and the acidic
wells, FW-005, -010, and 015 had significantly
decreased diversity. FW-003 also had decreased
diversity, and this site also had significantly
increased nitrate levels compared to background.
The high nitrate sites also displayed a more
dominated community structure compared to FW-300
and TPB-16.
Conclusions. The measured groundwater chemistry
differed markedly between sampling wells within
the contaminant plume and the background site.
PCA analyses was sufficient for simple separation
of sites by groundwater chemistry, but none of
the measured chemical analytes appeared primarily
or solely responsible for the overall geochemical
variability. High variability in geochemistry
between sampling locations was evident.
Therefore, different combinations of chemical
variables may be necessary to explain spatial
differences in the microbiology within and along
the contaminant plume. In addition, correlations
between groundwater chemistry and the recovery
and diversity of different functional gene
sequences gave very different results. Sites
with drastically different geochemistry could
have similar OTU distributions and sites with
different sequences had similar geochemical
parameters. Clear linear relationships between
the chemistry and microbiology seems unlikely at
the system level. The data reduction with PCA
did consider phylogenetically distinct OTUs, but
the analysis did not consider the relatedness
between the distinct OTUs. Future work is
underway to compare distance matrices of all
sequences between sites in order to capture the
degree of similarity within the genomics
sequences of the sampled community.