A02: Quantitative and qualitative differences in microbial DNA extracted from California soils using three common DNA extraction methods

1
A02 Quantitative and qualitative differences in
microbial DNA extracted from California soils
using three common DNA extraction methods
E. BENT1, R. FISCHER2, J.O. BECKER2 and J.
BORNEMAN1 1 Department of Plant Pathology and
2 Department of Nematology, University of
California Riverside, Riverside CA USA 92521,
bente_at_ucr.edu
Different extraction methods may generate
slightly different community profiles The PCR
template generated by each extraction method may
contain similarly rich, diverse and even
bacterial OTUs, however, it appears that there
may be influences of DNA extraction method on
eubacterial community composition. For example,
a majority of eubacterial rDNA sequences falling
in a section of the phylogenetic tree that seems
dominated by pseudomonads were obtained via the
Zhou et al. (1996) manual extraction method (Fig.
2).
Accurate assessments of microbial soil
communities via molecular biology methods
requires the recovery of DNA from all microbial
taxa in the sample. This work outlines the
differences in yield, in the total quantity of
amplifiable bacterial 16s and fungal 18s rDNA
template and in bacterial species richness,
evenness and diversity observed from DNA samples
from California soils that were extracted using
different methods.

• Amplified fungal 18s small subunit rDNA
• when averaged over all twenty soils, the amount
  of amplified fungal DNA was
• greatest for FP template (2.90 ng/ uL reaction)
• middling for BB template (2.04 ng/ uL)
• least for V template DNA (1.22 ng/ uL)
• - suggests again that the FastPrep kit method
  produces DNA extracts that either contain more
  amplifiable bacterial sequences, or fewer
  PCR-inhibiting contaminants than extracts
  generated with the UltraClean kit
• bead beating may be important for the lysis of
  structures containing fungal DNA in the majority
  of soils sampled.

Table 2 Average total yield of DNA extracted per
sample from 20 soils using FP, BB and V methods.
Minimum significant differences between soils
extracted with the same method (p lt 0.05) are 920
ng for FP, 1200 ng for BB and 170 ng for V.
Determination of bacterial richness, evenness and
diversity via OFRG DNA extracts were diluted
to a concentration of 0.5 ng/uL prior to
amplification by PCR with universal eubacterial
primers. Each purified, amplified product was
used to generate a separate clone library. Clones
were grown on agar plates and a QPix
gridding/picking robot was used to pick 768
clones from each library into 384 well culture
plates containing 50 uL per well of LB medium
containing ampicillin and 8 (v/v) glycerol.
These plates were grown overnight and then frozen
at -80 C until use. A control clone library
consisting of 96 clones of known sequence
replicated four times on a single 384 well plate
was constructed for bacterial rDNA sequences,
using the same vector. Frozen bacterial clone
libraries were thawed and used to provide
template for PCR (1 uL of cell suspension per
reaction, from duplicate plates that had been
freshly grown). 25 384-well PCR plates with 15
uL reaction mix per well were inoculated with
cells, heat sealed with foil, then subjected to
thermal cycling. Primers were designed to amplify
vector inserts. A QPix picking/gridding robot
was used to spot PCR reactions onto nylon
membranes. 48 identical membranes, each
containing a macroarray of 9,600 PCR reactions,
were produced. Each oligonucleotide probe was
labelled with P-33 then allowed to hybridize with
one membrane, after which membranes were washed,
allowed to dry and exposed to a phosphoimaging
screen. Twenty-nine differential probes were
used. Quantified image data was analyzed using
software developed at UCR. Values of 0 (no
probe binding), 1 (probe binding) and n
(indeterminate) were assigned to each clone for
each probe, generating a binary fingerprint 29
digits long. Once fingerprints were assigned,
the clones were grouped according to fingerprint,
with each fingerprint representing an OTU. The
fingerprinting process has previously been shown
to group eubacterial sequences with pairwise
identities of 97 together (Valinsky et al.
2002).
Soil FP yield (ng) BB yield (ng) V yield (ng)
5 3100 2100 290
9 1600 2100 23
11 2200 830 200
12 2200 1000 110
15 1500 790 73
18 2000 510 41
20 1500 670 91
22 2300 990 320
25 630 not detectable not detectable
27 1800 1200 260
30 1600 340 140
32 4900 1600 460
34 1500 120 140
35 1800 200 600
40 2100 1100 230
43 2900 880 370
46 3000 1000 72
48 1800 450 1100
Vanoni 2300 not detectable not detectable
Powell 2000 770 700
Methods

• Details of the methods used here can be found at
  http//www.cs.ucr.edu/zliu/rRNA/CSM2004method.doc
  
• Soil Samples Twenty soils of varying
  physicochemical characteristics were sampled from
  farms across Southern California (Table 1).
• DNA extraction for evaluation of yield and
  amplifiable rDNA
• DNA was extracted in duplicate from 0.5 g samples
  of all twenty soils using three different
  methods
• FP - the FastPrep DNA extraction kit for soil
  (Bio 101),
• V - the UltraClean Soil DNA Extraction kit
  (MoBio), following the manufacturers standard
  protocol, in which samples are vortexed
  horizontally for ten minutes
• BB - the UltraClean Soil DNA Extraction kit
  (MoBio), following the manufacturers standard
  protocol except that instead of vortexing,
  samples were bead beaten.
• DNA extraction for evaluation of bacterial
  richness, evenness and diversity
• DNA was extracted in triplicate from the Powell
  soil using the above three methods, as well as
• A manual hot detergent/ proteinase K lysis method
  (Zhou et al., 1996). This method required 5 g
  soil samples.
• Determination of amplifiable fungal and bacterial
  rDNA

Table 4 Average amplified fungal DNA from
reaction mixtures containing DNA template from 20
soils extracted using FP, BB and V methods.
Minimum significant differences between mixtures
containing template from soils extracted with the
same method (p lt 0.05) are 2.65 ng/ uL for FP,
2.19 ng/ uL for BB and 3.23 ng/ uL for V.
Soil FP amplified DNA (ng/ uL) BB amplified DNA (ng/ uL) V amplified DNA (ng/ uL)
5 3.45 2.26 1.08
9 1.95 0.93 1.02
11 4.35 1.88 1.21
12 4.35 1.88 1.21
15 1.44 0.76 1.45
18 1.87 0.41 1.58
20 2.09 2.89 0.39
22 1.94 2.75 1.72
25 2.14 not detectable not detectable
27 2.76 2.97 0.74
30 1.48 0.45 0.29
32 3.80 2.31 0.92
34 5.34 4.58 3.69
35 4.59 2.28 3.29
40 3.91 4.50 1.13
43 3.91 3.91 2.90
46 1.39 1.69 0.30
48 2.67 2.73 1.40
Vanoni 3.00 0.04 not detectable
Powell 1.61 1.66 0.13

• Amplified eubacterial 16s small subunit rDNA
• When averaged over all twenty soils, the amount
  of amplified eubacterial DNA was
• greater for FP template (2.03 ng/ uL reaction)
• than for BB (1.52 ng/ uL reaction)
• or V template DNA (1.41 ng/uL, respectively)
• - suggests the FastPrep kit method produces DNA
  extracts that either contain more amplifiable
  bacterial sequences, or fewer PCR-inhibiting
  contaminants than extracts generated with the
  UltraClean kit.

Environmental sample
Table 3 Average amplified eubacterial DNA from
reaction mixtures containing DNA template from 20
soils extracted using FP, BB and V methods. The
minimum significant difference between mixtures
containing template from soils extracted with FP
(p lt 0.05) is 2.33 ng/ uL no significant
differences were observed between mixtures
containing template extracted with either BB or V
methods
Extract DNA, PCR amplify rDNA
rDNA clone library
Table 1 Physicochemical characteristics of soils used in this study Table 1 Physicochemical characteristics of soils used in this study Table 1 Physicochemical characteristics of soils used in this study Table 1 Physicochemical characteristics of soils used in this study Table 1 Physicochemical characteristics of soils used in this study Table 1 Physicochemical characteristics of soils used in this study Table 1 Physicochemical characteristics of soils used in this study
Soil pH organic matter Est. water content (uL/g soil) sand silt clay
5 8.0 1.24 99 13 49 38
9 7.5 1.16 95 8 42 50
11 8.0 0.57 123 66 18 16
12 8.1 0.51 3 70 15 15
15 7.8 0.94 33 15 50 35
18 7.9 0.95 45 7 54 39
20 7.9 0.80 91 8 51 41
22 7.9 0.50 64 71 20 9
25 7.5 1.65 83 52 30 18
27 7.6 1.68 59 23 52 25
30 7.3 0.72 65 57 30 13
32 7.2 1.97 61 47 36 17
34 6.9 2.65 44 46 36 18
35 7.8 2.69 60 44 33 23
40 7.1 0.61 75 88 5 7
43 7.0 1.21 93 19 42 39
46 5.9 0.74 94 60 25 15
48 7.9 0.07 56 94 2 4
Vanoni 7.7 2.85 75 1 65 34
Powell 5.9 2.13 64 54 30 16
Array rDNA on nylon membranes
Hybridize with labeled small DNA probe
Soil FP amplified DNA (ng/ uL) BB amplified DNA (ng/ uL) V amplified DNA (ng/ uL)
5 2.49 2.75 1.82
9 2.16 0.07 1.47
11 2.55 0.45 1.44
12 2.12 1.60 1.40
15 2.99 2.19 2.15
18 1.32 1.47 0.56
20 1.72 1.12 1.07
22 1.03 1.56 1.71
25 2.02 not detectable not detectable
27 2.26 2.66 1.60
30 2.59 1.22 1.44
32 2.06 2.38 1.20
34 3.24 1.23 2.38
35 1.88 2.06 2.10
40 1.92 2.23 1.52
43 1.82 1.99 1.78
46 0.88 1.39 1.44
48 1.14 1.93 0.82
Vanoni 2.60 not detectable 0.89
Powell 1.82 2.12 1.54
Bacterial richness, evenness and diversity via
OFRG Each fingerprint is considered here to
represent an operational taxonomic unit (OTU)
which may consist of one bacterial species or a
group of related bacterial species. There were
no significant variations in total bacterial OTU
richness, in diversity as estimated via Shannons
index, or in evenness (plt0.05), indicating that,
all four DNA extraction methods produced PCR
template that can be considered equally rich,
diverse and even in eubacterial OTU content, and
therefore probably in eubacterial species content
(Table 5).
Fig. 2 Section of eubacterial 16s rDNA clone
phylogenetic tree. 101 FP method, MBV V
method, MBBB BB method, and Zhou Zhou et al.
(1996) method. A, B and C stand for replicates.
Analyze signals from all probes
Hybridization fingerprints
UPGMA analysis
Summary
Identification of Microorganisms

1. Of the kit methods, FP seems to extract
   amplifiable eubacterial and fungal rDNA template
   more reliably from a variety of soils
2. Bead beating seems to be important in extraction
   of amplifiable fungal rDNA template but less
   important for extraction of eubacterial template
3. All four extraction methods produce eubacterial
   template that seems equally rich, diverse and
   even in OTU composition
4. However, extraction method may influence the
   community profile that is generated in terms of
   which genera or species are represented

Fig. 1 Overview of the OFRG (oligonucleotide
fingerprinting of rRNA genes) process.
Results and Discussion

• DNA Yield The amount of DNA extracted using
  three different techniques from 20 different
  California soils is summarized in Table 2.
• When averaged over all twenty soils, the total
  yield of DNA (per sample) as determined via
  fluorometry was
• 2100 ng for the FastPrep kit (FP),
• 890 ng for the UltraClean kit with bead beating
  (BB)
• 210 ng for the UltraClean kit used with
  vortexing (V)
• significant differences (p lt 0.05) between all
  three methods.
• There were no statistically significant
  correlations between soil pH, organic matter
  content, estimated water content, clay content,
  sand content or silt content and the total ng of
  DNA extracted for each technique (not shown).

Table 5 Average richness, diversity and evenness
of eubacterial DNA template as estimated from
OFRG analysis of eubacterial fingerprints. DNA
was derived from Powell soil via four different
extraction methods.
Extraction method Average richness Average Shannons index of diversity Average evenness
FP 547 6.210 0.99
BB 543 6.201 0.98
V 537 6.196 0.99
Zhou et al. (1996) 565 6.210 0.98
Literature Cited
Valinsky, L., G. Della Vedova, A. J. Scupham, S.
Alvey, A. Figueroa, B. Yin, J. Hartin, M.
Chrobak, D. E. Crowley, T. Jiang, and J.
Borneman. 2002. Analysis of bacterial community
composition by oligonucleotide fingerprinting of
rRNA genes. Applied Environmental Microbiology
683243-3250. Zhou, J., M. A. Bruns, and J. M.
Tiedje. 1996. DNA recovery from soils of diverse
composition. Applied Environmental Microbiology
62316-322.