Available online at www.sciencedirect.com
`
`Journal of Microbiological Methods 72 (2008) 124 – 132
`
`www.elsevier.com/locate/jmicmeth
`
`Fecal collection, ambient preservation, and DNA extraction for PCR
`amplification of bacterial and human markers from human feces
`Jordan M. Nechvatal a, Jeffrey L. Ram a,⁎, Marc D. Basson b,c,d, Phanramphoei Namprachan a,
`Stephanie R. Niec a, Kawsar Z. Badsha e, Larry H. Matherly d,f,
`Adhip P.N. Majumdar h, Ikuko Kato d,g
`
`a Department of Physiology, Wayne State University, Detroit, MI, USA
`b Surgical Service, John D. Dingell VA Medical Center, Detroit, MI, USA
`c Department of Surgery, Wayne State University, Detroit, MI, USA
`d Karmanos Cancer Institute, Detroit, MI, USA
`e Department of Nutrition and Food Science, Wayne State University, Detroit, MI, USA
`f Department of Pharmacology, Wayne State University, Detroit, MI, USA
`g Department of Pathology, Wayne State University, Detroit, MI, USA
`h Department of Internal Medicine, Wayne State University, Detroit, MI, USA
`
`Received 14 August 2007; received in revised form 26 October 2007; accepted 13 November 2007
`Available online 21 November 2007
`
`Abstract
`
`Feces contain intestinal bacteria and exfoliated epithelial cells that may provide useful information concerning gastrointestinal tract health.
`Intestinal bacteria that synthesize or metabolize potential carcinogens and produce anti-tumorigenic products may have relevance to colorectal
`cancer, the second most common cause of cancer deaths in the USA. To facilitate epidemiological studies relating bacterial and epithelial cell
`DNA and RNA markers, preservative/extraction methods suitable for self-collection and shipping of fecal samples at room temperature were
`tested. Purification and PCR amplification of fecal DNA were compared after preservation of stool samples in RNAlater (R) or Paxgene (P), or
`after drying over silica gel (S) or on Whatman FTA cards (W). Comparisons were made to samples frozen in liquid nitrogen (N2). DNA
`purification methods included Whatman (accompanying FTA cards), Mo-Bio Fecal (MB), Qiagen Stool (QS), and others. Extraction methods
`were compared for amount of DNA extracted, DNA amplifiable in a real-time SYBR-Green quantitative PCR format, and the presence of PCR
`inhibitors. DNA can be extracted after room temperature storage for five days from W, R, S and P, and from N2 frozen samples. High amounts of
`total DNA and PCR-amplifiable Bacteroides spp. DNA (34% ± 9% of total DNA) with relatively little PCR inhibition were especially obtained
`with QS extraction applied to R preserved samples (method QS-R). DNA for human reduced folate carrier (SLC19A1) genomic sequence was also
`detected in 90% of the QS-R extracts. Thus, fecal DNA is well preserved by methods suitable for self-collection that may be useful in future
`molecular epidemiological studies of intestinal bacteria and human cancer markers.
`© 2007 Elsevier B.V. All rights reserved.
`
`Keywords: Bacteroides; DNA extraction; DNA preservation; Enteric bacteria; Feces; Stool
`
`1. Introduction
`
`Feces contain intestinal bacteria and exfoliated epithelial
`cells that may provide useful information concerning gastro-
`
`⁎ Corresponding author. Department of Physiology, Wayne State University,
`540 E. Canfield Avenue, Detroit, MI 48201 USA. Tel.: +1 313 577 1558; fax: +1
`313 577 5494.
`E-mail address: jeffram@med.wayne.edu (J.L. Ram).
`
`0167-7012/$ - see front matter © 2007 Elsevier B.V. All rights reserved.
`doi:10.1016/j.mimet.2007.11.007
`
`intestinal tract health. For example, bacteria activate or me-
`tabolize potential carcinogens (Blaut et al., 2006; Knasmuller
`et al., 2001; Vanhaecke et al., 2006) or can have anti-tumor
`effects (Fukui et al., 2001) that may have relevance to colorectal
`cancer, the second most common cause of cancer deaths in the
`USA. With the gastrointestinal tract being the largest area of the
`body that is constantly exposed to ingested/digested food and
`microorganisms, it is conceivable that luminal exposure may
`play a significant role in the development of colorectal cancer.
`
`Geneoscopy Exhibit 1024, Page 1
`
`

`

`J.M. Nechvatal et al. / Journal of Microbiological Methods 72 (2008) 124–132
`
`125
`
`Epithelial cells in feces represent a potential source of early
`biomarkers of gastrointestinal tract cancers. Although a variety
`of biomarkers have been utilized in epidemiological studies on
`colorectal cancer, most previous markers have been blood-
`based. However, markers analyzed from intestinal samples may
`be more relevant to the onset and detection of colon cancer.
`While approximately 55% of dry fecal weight is attributed to
`bacteria, Nair and co-workers report
`that approximately
`1.5 million colonic epithelial cells can also be isolated per
`gram of stool (Desilets et al., 1999; Iyengar et al., 1991). Thus,
`exfoliated gastrointestinal
`tract cells in feces may be an
`alternative for evaluating colon cancer biomarkers.
`Stool sample analysis offers a non-invasive opportunity to
`evaluate both luminal exposure to different types of bacteria as
`well as exfoliated epithelial cell markers for colorectal cancer
`risk. However, one of the major obstacles to introducing fecal
`markers in population studies has been the difficulty in col-
`lecting adequate samples for assays from a large number of
`subjects. This difficulty is exacerbated by the fact that standard
`fecal collection procedures require fresh or frozen samples,
`which limits its application in a community-based setting. As
`a result, epidemiological studies utilizing fecal specimens
`have often been limited in the number of study subjects and
`in controlling potential confounders. Fecal self-collection kits
`have recently been used in large-scale epidemiological studies
`involving the diagnosis of food-borne illnesses, but these kits
`lacked any DNA/RNA preservation method, potentially limit-
`ing their full usefulness (Jones et al., 2004). Since new tech-
`nologies have become available to preserve tissue DNA and
`RNA for a period of time at room temperature, application of
`such technologies to fecal samples may have great potential for
`epidemiological studies.
`In the present feasibility study, multiple methods for fecal
`preservation and DNA extraction were tested. Since a major
`problem with complex samples such as feces is the presence of
`PCR inhibitors, analytical methods were designed to detect,
`quantify, and identify conditions under which PCR inhibition
`was minimal. While this paper focuses on DNA preservation,
`extraction, and quality, the methods studied were also chosen
`for their likely suitability for preserving RNA as well. Alto-
`gether, several ambient temperature preservation and extraction
`combinations were capable of yielding usable DNA; however,
`one combination of ambient preservation and extraction
`methods gave the most consistent yield of relatively inhibitor-
`free DNA.
`
`2. Materials and methods
`
`2.1. Stool samples
`
`Fifteen fresh stool samples, obtained from patients being
`evaluated at the vascular clinic of the John D. Dingell VA
`Medical Center (Detroit, MI), were collected in plastic con-
`tainers that were immediately put on ice. The vascular clinic
`was used for recruitment as it would not be expected that such
`patients would be more likely than the general population to
`have colonic abnormalities, as might be the case for a general
`
`surgery clinic. This research protocol was approved by the
`Wayne State University and VA Medical Center Human In-
`vestigation Committees and written informed consent was
`obtained from each study participant. Samples were further pro-
`cessed or transferred to preservative (see below) within 1 h.
`Although only ten stool samples were needed, fifteen were
`collected since five samples were inadequate for further pro-
`cessing due to poor consistency (i.e., too watery) or inadequate
`quantity and were not used in the study. In addition to the above
`samples collected at the VA Medical Center (referred to, col-
`lectively, in this paper as “VA Samples”), preliminary tests of
`various methods (prior to the above 15 samples) were con-
`ducted with anonymously provided stool samples collected by
`the Ram laboratory, by methods approved by the Wayne State
`University Human Investigation Committee.
`
`2.2. Sample preparation, preservation, and storage
`
`For each VA sample, 0.2 g aliquots (at least five for each
`preservative method) were removed by taking cores of the stool
`sample with a cut-off 1 ml syringe, where 0.2 ml is ≈0.2 g.
`Each 0.2 g core received one of the preservative treatments,
`which included spreading and drying on a Whatman FTA card
`(W; Whatman, Florham Park, NJ.), drying over silica gel beads
`(S), submersion in 1.0 ml RNAlater™ (R; Ambion, Austin,
`immersion in 1.0 ml Paxgene™ (P; PreAnalytiX,
`TX.),
`Hombrechtikon, Switzerland), and refrigerator storage (F).
`Except as noted for pilot tests, the W, S, R, and P preservation
`methods incorporated a five-day “hold” period at ambient
`temperature to mimic the likely delay between self-collection of
`a sample and receipt by an analytical laboratory, for comparison
`to alternative storage procedures utilizing 24 h refrigeration or
`immediate freezing in liquid nitrogen.
`For W samples, the 0.2 g of feces was spread over two of the
`four quadrants of the FTA card, allowed to dry approximately
`2 h at room temperature, and then placed in a protective barrier
`pouch with silica gel desiccant packet. For S samples, 0.2 g of
`feces was placed over silica gel beads (∼10 ml) and ∼1 cm of
`glass wool in a 50 ml tightly sealed sterile polypropylene tube.
`R and P samples were stored in 2 ml sterile polypropylene
`tubes. After five days storage at room temperature, W and S
`samples were transferred to −80 °C. Also, after five days, R and
`P samples were centrifuged (2 min at 10,000 × g), the superna-
`tant was removed, and the pellet was stored at −80 °C. For F
`samples, 0.2 g of
`feces was sealed in a sterile 50 ml
`polypropylene tube and placed in a 4 °C refrigerator for 24 h
`and then transferred to −80 °C. On the day of collection,
`remaining portions of each stool sample (designated N2) were
`placed in paper-lined aluminum foil wrappers, flash-frozen in
`liquid N2, and immediately stored at −80 °C. The above
`methods, along with their associated extraction methods (next
`section) are summarized in Table 1.
`
`2.3. Sample extraction
`
`DNA extraction procedures included Mo-Bio Fecal (MB;
`Mo-Bio, Carlsbad, CA.), Qiagen QIAamp DNA Stool Mini
`
`Geneoscopy Exhibit 1024, Page 2
`
`

`

`126
`
`J.M. Nechvatal et al. / Journal of Microbiological Methods 72 (2008) 124–132
`
`Table 1
`Summary of stool sample preservation and DNA extraction methods examined in this study
`Hold time a
`
`Preservation
`method
`
`Method
`abbreviation
`
`Extraction
`method
`Qiagen RNA/DNA Mini c
`5 days
`Paxgene, 1 ml
`Q2N-P
`Qiagen RNA/DNA Minic
`5 days
`RNAlater, 1 ml
`Q2N-R
`Qiagen RNA/DNA Minic
`5 days
`Silica gel beads, 10 ml
`Q2N-S
`Qiagen RNA/DNA Minic
`5 days
`Whatman FTA card
`Q2N-W
`Qiagen RNA/DNA Minic
`1 day
`Refrigeration
`Q2N-F
`Qiagen RNA/DNA Minic
`Immediate
`Liquid nitrogen
`Q2N-N2
`Qiagen QIAamp DNA Stool Mini
`5 days
`Paxgene, 1 ml
`QS-P
`Qiagen QIAamp DNA Stool Mini
`5 days
`RNAlater, 1 ml
`QS-R
`Qiagen QIAamp DNA Stool Mini
`5 days
`Silica gel beads, 10 ml
`QS-S
`Qiagen QIAamp DNA Stool Mini
`5 days
`Whatman FTA card
`QS-W
`Qiagen QIAamp DNA Stool Mini
`1 day
`Refrigeration
`QS-F
`Qiagen QIAamp DNA Stool Mini
`Immediate
`Liquid nitrogen
`QS-N2
`Lysozyme; then Qiagen RNA/DNA Minic
`5 days
`Paxgene, 1 ml
`Q2L-P
`Lysozyme; then Qiagen RNA/DNA Minic
`5 days
`RNAlater, 1 ml
`Q2L-R
`Lysozyme; then Qiagen RNA/DNA Minic
`5 days
`Silica gel beads, 10 ml
`Q2L-S
`Lysozyme; then Qiagen RNA/DNA Minic
`5 days
`Whatman FTA card
`Q2L-W
`Lysozyme; then Qiagen RNA/DNA Minic
`1 day
`Refrigeration
`Q2L-F
`Mo-Bio Fecal
`5 days
`Paxgene, 1 ml
`MB-P
`Mo-Bio Fecal
`5 days
`RNAlater, 1 ml
`MB-R
`Mo-Bio Fecal
`5 days
`Silica gel beads, 10 ml
`MB-S
`Mo-Bio Fecal
`5 days
`Whatman FTA card
`MB-W
`Mo-Bio Fecal
`1 day
`Refrigeration
`MB-F
`a The hold time is the amount of time the sample is held in or with the preservative prior to transfer to the −80 °C freezer.
`b The range of time needed for extractions depends on the number of samples (up to 10) processed simultaneously.
`c The Qiagen RNA/DNA Mini kit also results in the purification of RNA in another step of the two day procedure.
`
`Stool mass
`extracted
`
`Time needed
`for extraction b
`
`0.2 g
`0.2 g
`0.2 g
`∼0.01 g
`0.2 g
`0.2 g
`0.2 g
`0.2 g
`0.2 g
`∼0.01 g
`0.2 g
`0.2 g
`0.2 g
`0.2 g
`0.2 g
`∼0.01 g
`0.2 g
`0.2 g
`0.2 g
`0.2 g
`∼0.01 g
`0.2 g
`
`Two 8 h days
`Two 8 h days
`Two 8 h days
`Two 8 h days
`Two 8 h days
`Two 8 h days
`3–5 h
`3–5 h
`3–5 h
`3–5 h
`3–5 h
`3–5 h
`Two 8 h days
`Two 8 h days
`Two 8 h days
`Two 8 h days
`Two 8 h days
`2–3 h
`2–3 h
`2–3 h
`2–3 h
`2–3 h
`
`(QS; Qiagen, catalogue number 51504, Hilden, Germany), and
`modified 2-day Qiagen RNA/DNA Mini (Q2L/N, where 2
`stands for “two-day method” and L/N stands for Lysozyme/No
`lysozyme treatment; Qiagen, catalogue number 14123). In pilot
`tests, a DNA extraction method accompanying Whatman FTA
`cards failed to extract DNA effectively from our sample types.
`This study therefore evaluated MB, QS, Q2N, and Q2L pro-
`cedures as alternatives for extracting DNA from the Whatman
`FTA cards. For samples preserved by R, P, S, and F, full aliquots
`originally weighing 0.2 g were extracted by each method. N2
`samples were extracted only by QS and Q2N procedures. For W
`samples, 20 FTA card-punches (using the Whatman 2.0 mm
`card punch and giving a total of ∼0.01 g of the original fecal
`sample) were extracted by each method. Accordingly, this study
`analyzed a total of 220 DNA extracts: 4 extraction methods per
`each of 5 preservative methods and 2 extraction methods for the
`N2 method, for each of the 10 VA samples).
`All extraction procedures followed original manufacturers'
`standard procedures for fecal DNA extraction except for the
`modified Qiagen 2-day procedure and the previously noted
`alternative to Whatman's procedure. Modifications to the
`Qiagen RNA/DNA Mini kit included the addition of (or lack
`of) lysozyme (5 mg/μl, Sigma L-7651) in 200 μl TE buffer
`(pH 8.0) for an initial room temperature incubation period of
`10 min (Q2L method). Samples that were not treated with
`lysozyme (Q2N = no lysozyme method) were incubated on ice
`for 10 min with 200 μl TE added to them. Following the
`incubations, 0.2 g of sterile DNase-free sand and 1 ml of GITC
`buffer (4 M guanidium thiocyanate, 10 mM Tris HCl [pH 7.0],
`
`and 1 mM EDTA [pH 7.0], 0.5% 2-mercaptoethanol) were added
`to both lysozyme and non-lysozyme samples, and samples ho-
`mogenized for 20 min at maximum speed on a vortex, using a
`horizontal tube adaptor. Q2N/L samples were then centrifuged at
`10,000 ×g for 20 min and supernatant transferred to new tubes.
`Following centrifugation, 0.5 ml of Qiagen solution QRL-1 buffer
`was added to each sample and the new solutions passed through
`an 18 G needle/syringe 10 times. Next, 0.5 ml of Qiagen solution
`QRV-1 was added to the samples, mixed well, and samples
`centrifuged (10,000 ×g) at 4 °C for 20 min. The supernatant was
`then transferred to a new tube, 0.8 volumes of ice-cold iso-
`propanol added, and tubes placed at −80 °C overnight. Day 2 of
`the 2-day procedure began with step #6 of the manufacturer's
`instructions, under the animal cell protocol.
`MB extraction resulted in 50 μl of DNA solution, while QS
`and Q2L/N extractions each resulted in 200 μl DNA (the Q2N/L
`methods also resulted in the subsequent extraction of RNA).
`The above extraction methods varied considerably in time to
`complete, as summarized in Table 1, and this factor may also be
`a consideration in choosing which method to use. Resultant
`DNA samples were stored at −80 °C until quantitation and
`characterization could be performed.
`
`2.4. Picogreen assay and DNA quantitation
`
`DNA was measured by a fluorometric Quant-iT™ Picogreen®
`(Molecular Probes, Eugene, OR) assay using the Bio-Rad MyiQ®
`real-time single-color PCR detection system (Bio-Rad, Hercules,
`CA.) as the fluorometer, comparing relative fluorescence units
`
`Geneoscopy Exhibit 1024, Page 3
`
`

`

`J.M. Nechvatal et al. / Journal of Microbiological Methods 72 (2008) 124–132
`
`127
`
`(RFU) of DNA standard and fecal DNA samples. Phage λ DNA
`was used as the calibration standard in a dilution series ranging
`from 0 ng/μl to 200 ng/μl. Fecal-extracted DNA was measured in
`2.0 μl of duplicate undiluted (designated 1:1), 1:10, and 1:100
`dilutions. Nanodrop® (Nanodrop Technologies, Wilmington, DE)
`spectrophotometer (A260/280) measurements of DNA were also
`performed on most samples, but often indicated variably higher
`levels of absorbance than the fluorometric method would have
`predicted, possibly due to non-DNA contaminants (data not
`shown), some of which may be PCR inhibitors. Picogreen, with its
`high affinity and specificity for dsDNA, provided a more reliable
`measure of DNA.
`
`2.5. Real-time PCR assay
`
`2.5.1. PCR primers and thermocycle conditions
`Preliminary PCR experiments involved testing primers for
`multiple groups of bacterial species, using cycle conditions
`described in each reference (see list of primers, Table 2.).
`Bacteroides DNA was chosen as the primary target in the VA
`samples due to its high abundance and consistent presence.
`Real-time SYBR®-Green (Molecular Probes, Eugene, OR) PCR
`of the VA DNA samples, was accomplished using a 16S rDNA
`Bacteroides target (Bac32F/708R) and a “touch-down” protocol
`(Don et al., 1991). PCR supermix was made using 12.5 μl
`SYBR-Green II master mix (containing Taq polymerase,
`dNTP's, MgCl2, SYBR-Green fluorescent dye, flourescein
`(for signal normalization), and Tris buffer), 11.0 μl water,
`0.25 μl each of 20 pmol/μl Bac32F (5′-AACG CTAG CTAC
`AGGC TT-3′) and 708R (5′-CAAT CGGA GTTC TTCG TG-
`3′) primers, which yields a 676 bp amplicon as initially de-
`scribed by Bernhard and Field (2000), and 1.0 μl of the template
`DNA. The touch-down Bacteroides PCR was performed in
`duplicate on undiluted DNA (1:1) and on dilutions of 1:10,
`1:100, and 1:1000. Bacteroides fragilis (ATCC 25285) DNA, at
`a concentration of 20 ng/μl, served as a positive control. The
`PCR protocol began with an initial denaturation step of 94 °C
`for 2 min, followed by 35 cycles of 94 °C denaturation for 20 s,
`62 °C primer annealing for 20 s (decreasing in decrements of
`0.3 °C per cycle), and 72 °C extension for 45 s; and a final 72 °C
`elongation step for 10 min. PCR products were verified via
`agarose gel(s) and/or melt-curve analysis.
`
`Table 2
`Primers used for various bacterial groups
`
`Bacterial group/species
`
`Primer set
`
`Primer reference
`
`Bacteroides
`Clostridium
`Desulfovibrio
`Lactobacillus
`Escherichia coli
`Enterococcus
`Fusobacterium
`
`Bifidobacterium
`All Eubacteria
`
`Bernhard and Field (2000)
`Bac32F/Bac708R
`Matsuki et al. (1996)
`Ccoc477/Ccoc916R
`Matsuki et al. (1996)
`Dsv691F/Dsv826R
`Lacto157F/Lacto379R Byun et al. (2004)
`16E1F/16E2R/16E3R
`Tsen, Lin and Chi (1998)
`Efs130F/Efs490R
`Matsuki et al. (1996)
`FPR-1/FPR-2
`Wang, Cao and Cerniglia
`(1996)
`Matsuki et al. (1996)
`Matsuki et al. (1996)
`
`Bif164F/Bif601R
`Uni331F/Uni797R
`
`2.5.2. Assessment of PCR inhibition
`Since the presence of PCR inhibitors in DNA extracts could
`affect the accuracy of real-time PCR measurements of DNA
`concentration, the amount of inhibition, if any, was estimated by
`two methods: In the first method, the change in the average
`Ct (Ct is the cycle at which the baseline or threshold RFU value
`is exceeded,) for a 10-fold DNA dilution series ranging from 1:1
`to 1:1000 was determined. In the absence of PCR inhibition, the
`expected result is that higher amounts of starting DNA will
`result in a lower value of Ct. At 100% PCR efficiency (i.e., a
`doubling of the amplicon concentration each cycle), each 10-
`fold dilution would be expected to produce a change of Ct
`(ΔCt) of Log(10)/Log(2) = ∼ 3.32 cycles. By comparing
`average shifts in Ct with this theoretical performance in the
`absence of inhibition, the influence of significant concentrations
`of PCR inhibitors could be estimated.
`A second measure of the presence of PCR inhibitors com-
`pared the relative fluorescence (RFU) produced by the final
`PCR product of the undiluted DNA sample to the final RFU for
`diluted, potentially less inhibited samples. The RFU of the final
`PCR product
`is a measure of the total amount of DNA
`produced, possibly modified by quenching or autofluorescence.
`A lower final RFU for the undiluted DNA sample, compared to
`that obtained at 1:10 or 1:100 would indicate the presence of
`PCR inhibition.
`
`2.5.3. Calculation of DNA concentration
`The amount of Bacteroides DNA was calculated based on
`the relative Ct values, using the formula [Cal]⁎2^(Ctcal-
`Ctu)⁎dil, where [Cal] is the concentration of a known reference
`DNA measured in a PCR reaction run at the same time, Ctcal is
`the Ct obtained for the reference DNA sample, Ctu is the Ct
`obtained for the sample with unknown concentration of the
`target DNA, and dil
`is the dilution factor of the sample
`compared to the solution for which the concentration is being
`calculated. This calculation assumes a doubling of the amount
`of DNA for each additional cycle of Ct, an assumption that
`is justified if no PCR inhibition is occurring. In the present
`experiments, this calculation was applied to extracts that had
`been diluted 1:100 (i.e., dil = 100), for which data will be
`presented showing no inhibition.
`
`2.6. Amplification of human genomic DNA
`
`Aliquots of DNA were also analyzed for a specific human
`target DNA, human reduced folate carrier (SLC19A1) genomic
`sequence (Genbank accession number U19720), using a nested
`PCR procedure capable of detecting small amounts of human
`DNA. In the primary PCR, reactions contained 5 μl GeneAmp®
`10× PCR Buffer II (Applied Biosystems, N8080130), 4 μl
`dNTPs (Applied Biosystems, N8080007), 3 μl 25 mM MgCl2
`(Applied Biosystems N8080130), 2.5 μl dimethylsulfoxide,
`1.0 μl of each primer (10 pmol/μl each of hRFC2308R (5′-
`AAGA GCAC CAAG GATG ACCA GCAA TGTC-3′) and
`hRFC1525F (5′-AGGA GAAG GCAG CACA GGCA CTAG)-
`3′, 0.5 μl 5 U/μl Taq DNA polymerase (Promega, PR-M8291),
`0.2 μl–4 μl template DNA solution, and 28–32 μl water to fill
`
`Geneoscopy Exhibit 1024, Page 4
`
`

`

`128
`
`J.M. Nechvatal et al. / Journal of Microbiological Methods 72 (2008) 124–132
`
`the final reaction volume to 50 μl. Second round PCR mixture
`was the same, but utilized 2.0 μl of first round PCR product for
`the DNA template and used as primers hRFC1857R (5′-GCGC
`CCGA GAAT CACT TGGT TTCA CATT-3′) and hRFC1643F
`(5′-GGAG CAGA GACA GAGC GACC CATA CCTG-3′).
`The primary PCR thermocycle consisted of 94 °C for 3 min
`initial denaturation, 35 cycles of amplification (30 s 94 °C, 45 s
`64 °C primer annealing, 1 min 72 °C elongation), and 7 min
`final 72 °C elongation. Second round PCR was identical except
`only 32 cycles were used and the annealing step was at 62 °C.
`PCR products were separated on 2% agarose gels stained with
`ethidium bromide.
`
`pilot tests of the various DNA extraction techniques on 5–10
`stool samples each, the relative amounts of PCR products, as
`judged by lower Ct values, for the various bacterial groups was
`Bacteroides N Clostridium ∼ Desulfovibrio ∼ Fusobacterium N
`Lactobacillus NBifidobacterium NN Escherichia coli and Entero-
`coccus. The qualitative results identified Bacteroides spp. as being
`reliably present and at a generally higher level than other targeted
`bacterial groups. Accordingly, subsequent quantitative studies
`on the VA samples focused on Bacteroides spp. Before present-
`ing the quantitative results, however, we consider the presence
`of PCR inhibitors, which can affect PCR-based detection and
`quantitation.
`
`3. Results
`
`3.1. DNA yield
`
`Total amounts of DNA extracted with different preservative
`and extraction combinations varied considerably (Fig. 1), with
`some combinations being significantly different from others
`(One Way ANOVA, p b 0.001). The highest yields tended to be
`obtained for DNA preserved with either R or P; viz., the top four
`average yields were for QS-P, QS-R, Q2L-R, and Q2N-R, with
`yields of 12–25 μg total DNA from the 0.2 g (wet weight) fecal
`starting material. The QS extraction method accounted for 4 of
`the top 6 average DNA yields. The MB method consistently
`gave lower yields than the other methods.
`
`3.2. PCR amplification of bacterial DNA
`
`3.2.1. Qualitative survey of bacterial groups
`In preliminary experiments, fecal samples that had been
`directly frozen in a −80 °C freezer prior to extraction were
`tested with a variety of primer sets (Table 2) that target various
`bacterial groups expected to be present in fecal samples. In
`
`3.2.2. PCR inhibition
`Undiluted DNA extracts sometimes produced less PCR
`product than extracts that had been diluted 10-fold, provid-
`ing clear evidence of the presence of PCR inhibition. The
`amount of inhibition was estimated by two methods in order
`to compare the efficacy of various methods at removing the
`inhibitors and also to determine conditions under which com-
`paratively little inhibition was present. Fig. 2 shows results
`of the first method, in which ΔCt, the shift in Ct for each 10-
`fold dilution of the sample, was compared to 3.32, the the-
`oretical shift in the absence of inhibition. For some samples,
`such as Q2L-R and QS-S, ΔCt is negative, i.e., the average
`Ct for the undiluted sample, 1:1, is higher than the average
`Ct for the 1:10 dilution, clearly indicating the presence of
`PCR inhibitors. By this standard, QS-R, Q2N-W, and Q2N-F
`samples had the least amount of PCR inhibition, comparing
`ΔCt values determined for undiluted (1:1) v. 1:10 samples.
`Also, by this criterion, no PCR inhibition occurred for any
`DNA sample diluted to 1:100, which showed ΔCt values N3
`for all methods (Fig. 2B).
`The second measure to assess the influence of PCR in-
`hibitors compared the relative fluorescence (RFU) produced by
`
`Fig. 1. Amounts of DNA extracted from feces preserved and extracted by various methods. Labels are of the form X–Y, where X is the DNA extraction method and Y
`is the preservative method. Abbreviations for the methods are identified in the text. Starting material in each case is 0.2 g feces, except for W samples, which were
`extracted from 20 punches, with an estimated fecal weight of ∼0.01 g. Averages are mean ± sem of stool samples obtained from 10 VA subjects. [DNA] was measured
`by a fluorometric method, using Picogreen and the Bio-Rad iCycler as the fluorometer.
`
`Geneoscopy Exhibit 1024, Page 5
`
`

`

`J.M. Nechvatal et al. / Journal of Microbiological Methods 72 (2008) 124–132
`
`129
`
`Fig. 2. Analysis of PCR inhibition, estimated by comparing the change in Ct values for 10-fold dilutions of the DNA extract with the theoretical change in Ct expected
`in the absence of inhibition (3.32 for a 10-fold change in template DNA concentration). (A) Bars represent mean ± sem Ct values at dilutions of 1:1 (undiluted, hatched
`bars), 1:10 (open bars), 1:100 (filled bars) and 1:1000 (cross-hatched bars). (B) Data for the same experiments as in (A), plotted as ΔCt, the difference between the Ct
`value obtained at one concentration minus the Ct value for the 10-fold more diluted extract of the same sample. Bars represent Ct1:10–Ct1:1 (open bars), Ct1:100–Ct1:10
`(filled bars), and Ct1:1000–Ct1:100 (cross-hatched bars). The line at 3.32 represents the expected value if no PCR inhibition were present.
`
`the final PCR product. By this criterion, QS-R and Q2N-W
`again had relatively little PCR inhibition, while the Q2N-F RFU
`was reduced by 40% (Fig. 3).
`
`3.2.3. Amounts of Bacteroides DNA
`The amounts of PCR-measured Bacteroides DNA were
`compared to total DNA measured with Picogreen (Fig. 4). For
`calibration, positive control Bacteroides DNA, at a concentra-
`tion of 20 ng/μl, gave an average Ct of 12.2 ± 0.1 (n = 15). The
`amount of Bacteroides DNA in experimental samples was
`determined from the Ct values measured for the 1:100 samples,
`a dilution at which the above experiments indicated that PCR
`inhibition did not occur. The relationship between the amounts
`of Bacteroides DNA and total DNA measured by Picogreen is
`illustrated by the least squares line fitted to the left-hand 11
`points on Fig. 4A. The slope of the line indicates that the
`average percentage of total DNA in the sample that is Bacter-
`
`oides DNA is 38% (r2 = 0.77, p b 0.001) of the total. The two
`points not included in calculation of the linear regression curve
`were for methods Q2L-R and Q2N-R. Although the measure-
`ments of total DNA amounts for these two methods were quite
`high (Fig. 1), they were nevertheless lower than the estimated
`amount of Bacteroides DNA (percentages N 100%, Fig. 4B),
`possibly indicating the occurrence of fluorescence quenching in
`the Picogreen measurements of total DNA for these samples.
`Among the sample types known to have relatively low levels
`of PCR inhibitors, QS-R extracts had higher amounts of Bac-
`teroides DNA than Q2N-W (compare point I with point B in
`Fig. 4A). Q2N-W extracts contained only 2.0 ± 1.4 ng/μl of
`Bacteroides DNA; whereas, the estimated amount of Bacter-
`oides in the QS-R extracts was 19.8 ± 7.5 ng/μl. The percentage
`of total DNA identified as Bacteroides DNA in extracts made
`by the QS-R method (34% ± 9%; see Fig. 4B) did not differ
`significantly from the QS method applied to fecal samples that
`
`Geneoscopy Exhibit 1024, Page 6
`
`

`

`130
`
`J.M. Nechvatal et al. / Journal of Microbiological Methods 72 (2008) 124–132
`
`Fig. 3. Analysis of PCR inhibition, estimated by comparing the relative florescence at the end of 35 cycles of SYBR-Green real-time PCR. The relative fluorescence (in
`relative fluorescence units, RFU) for the reaction with the 1:1 dilution was calculated as a percentage of the “expected value” if no inhibition were present, where the
`expected value is estimated from the RFU obtained for the 1:10 or 1:100 dilution (whichever had the higher RFU). Values below 100% would indicate that production
`of the PCR was inhibited in the more concentrated sample. Values close to 100% (dotted line in graph) indicate relatively little PCR inhibition. Samples with low RFU
`at 1:10 and 1:100 due to low [DNA], as determined by the presence of a still rising slope of the fluorescence at cycle 35, are not included in the calculations.
`
`had been quick frozen in liquid nitrogen (25% ± 6%; not
`significantly different by paired t-test) and was the closest
`among all preservation methods to the percentage extracted by
`the QS method.
`
`3.3. Amplification of human genomic DNA
`
`PCR using primers for human reduced folate carrier (hRFC)
`demonstrated that human genomic DNA was present in QS-R
`extracts (Fig. 5) and QS-N2 extracts (data not shown). Fig. 5
`shows representative positive results obtained for 6 of the
`extracts tested with 4 μl of undiluted extract per reaction.
`Several of the other extracts did not produce a product when
`tested at this template concentration but when diluted (only 1 μl
`or 0.2 μl of extract were used, equivalent to 4-fold and 20-fold
`dilutions), product was obtained, indicating that PCR inhibition
`may have been present at the higher concentrations. Altogether,
`positive results were obtained at one or the other concentration
`for human hRFC for 9 out of 10 QS-R extracts and 10 out of 10
`QS-N2 extracts.
`
`Fig. 4. Amount of Bacteroides spp. DNA compared to the total DNA extracted
`for samples preserved and extracted by several methods. (A) Extraction methods
`for each point are labeled by letters and listed here in the order illustrated from
`left to right: (A) QS-W, (B) Q2N-W, (C) Q2N-F, (D) Q2N-S, (E) QS-S, (F) Q2N-
`N2, (G) Q2N-P, (H) QS-F, (I) QS-R, (J) QS-N2, (K) QS-P, (L) Q2L-R, and
`(M) Q2N-R. Points represent means ± sem for both variables. the sem for Bac-
`teroides measurements of points L and M is written out due to the great
`variability in the measurements for these two points. The line is a least squares
`regression of the left-hand 11 points. (B) Bars represent mean ± sem of the
`Bacteroides spp. DNA as a percent of the total DNA for each sample, obtained
`for each method. Total DNA was measured by Picogreen, as in Fig. 1; Bacter-
`oides DNA was estimated by quantitative real-time PCR from the Ct value
`relative to standard Bacteroides DNA, assayed at 20 ng/μl, which gave an
`average Ct of 12.2 ± 0.1 (n = 15).
`
`Fig. 5. PCR identification of human DNA in fecal DNA extracts preserved in
`RNAlater and extracted by the QS method. Lane 1, 100 bp DNA ladder (labels
`at left indicate DNA size); lanes 2 and 9, negative controls; lanes 3–8, PCR
`products from the second PCR round of the nested PCR procedure to detect a
`214 bp sequence in the human reduced folate carrier gene, amplified for 6 of the
`10 QS-R fecal DNA extracts.
`
`Geneoscopy Exhibit 1024, Page 7
`
`

`

`J.M. Nechvatal et al. / Journal of Microbiological Methods 72 (2008) 124–132
`
`131
`
`4. Discussion
`
`Both bacterial and human DNA can be extracted from stool
`samples stored at room temperature in RNAlater (R)