You are on page 1of 9
Seediscussions,stats,andauthorprofilesforthispublicationat: Rapid DNA















Journal of Applied Bacteriology 1993, 74, 70-05

Rapid DNA extraction protocol from soil for polymerase chain reaction-mediated amplification

K. Smalla, N. Cresswell', L.C.

Institute for Biochemistry and Plant

of Biological Sciences, University of Warwick, Coventry, UK,

de Janeiro, llha do Fundao, Rio de Janeiro, Brazil and The Netherlands

Mendonca-Hagler', A. Wolters3 and J.D. van EIsas3

Virology,Biologische Bundesanstalt,Braunschweig, Germany, ' Department

Institute of Microbiology, Federal University of Rio

Institute for Soil Fertility Research, Wageningen Branch,

4197/04/92: accepted 24 July 1992


simple and rapid method of DNA extraction from soil was developed and DNA was made suitable for subsequent efficient amplification by the polymerase chain reaction (PCR). Key features of the extraction and purification were cold lysozyme- and SDS-assisted lysis with either freezing-thawing or bead beating, cold phenol extraction of the resulting soil suspension, CsCl and KAc precipitation and, finally, spermine-HC1 or glass milk purification of DNA. Crude DNA preparations contained 4-20 pg DNA per g of soil extracted, and at least 50% of this was recovered in the final purified DNA preparations. The resulting DNA was pure enough to be restricted by various enzymes, and was amplifiable at concentrations of up to 20 ng of soil-derived DNA per 50 p1 reaction mix. Amplification of a 683 bp target sequence, put, was performed with different Taq DNA polymerases. Application of the protocol enabled us to detect target DNA derived from roughly lo3 introduced Pseudomonaspuorescens (RP4 : : put) cfu per g of soil. The fate of an introduced population in the soil could be followed to this limit with PCR-assisted detection of target DNA. In addition, target DNA was detected in soil 5 months after release, when the introduced organism was no longer detectable on selective agar plates. The extraction and purification protocol applied to various different soil types resulted in DNA of sufficient purity to permit amplification by PCR.


There is a need to develop sensitive methods for the detec- tion of micro-organisms in the environment, e.g. soil, given the putative biohazards of released genetically modified micro-organisms (GMOs). Microbial populations and their dynamics have traditionally been described with either cultivation-based techniques, e.g. by selective plate counts, or specific direct counts, e.g. vra immunofluorescence. Both methods suffer from drawbacks. Plate counts monitor only culturable organisms, whereas immunofluorescence is often plagued by lack of specificity and possibly also enumerates dead cells. The application in microbial ecology of detection tech- niques based on DNA-DNA hybridization has increased recently (Sayler et al. 1985; Ford and Olson 1988; Holben el a/. 1988; Trevors and van Elsas 1989). Such

Correspondence to

Virology, Biologisrhe Bundesanstalt, Messeweg 11/12, Braunsrhweig,



Smalla, Institute for Biochemistry and Plant

hybridization-based detection techniques, which use either previous culturing of cells (e.g. colony hybridization) or direct extraction of DNA (Ogram et al. 1987; Holben et a/. 1988), offer obvious advantages, such as the possibility of specific detection of micro-organisms, e.g. of living, non- culturable, ultramicro- (dwarf) or dead cells, and of genes, by selecting an adequate probe. The polymerase chain reaction (PCR), developed by Saiki et al. (1988) for the detection of target sequences in clinical settings, should lower the limit of detection of genes in the environment by enhancing the target copy number prior to analysis. The application of molecular techniques directly to soil, however, has been hampered by problems caused by the presence of impurities in soil-derived DNA samples, even though successes in detection have been reported (Steffan and Atlas 1988; Cresswell et al. 1991; Pillai et al. 1991; Selenska and Klingmuller 1991a, b; van Elsas ef ul. 199la; Tsai and Olson 1992).

Tracking the fate of introduced GMOs and their genetic material in soil has been the interest of the different labor- atories involved in the development of the simple and rapid DNA extraction/PCR protocol described here (Cresswell et al. 1991; Smalla et al. 1991; van Elsas et at. 1991a, b;

Waalwijk et al. 1991). Pseudomonas fluorescens R2f, carrying

the RP4 : : pat element, was used as a model as PCR and tracking systems were available (Smit et al. 1991; van Elsas et al. 1991a).


Organisms and plasmid

PseudomonasPuorescens R2f, a grassland isolate (van Elsas et

al. 1988) containing plasmid RP4 : : pat,

by inserting the 683-bp potato

'patatin' cDNA sequence, pat, from pPATB2, into the unique Hind111 site of RP4 (van Elsas and Trevors 1990; van Elsas et al. 1991a).

RP4 : : pat was constructed

was used. Plasmid

Pseudomonas Juorescens R2f (RP4 : : pat) was cultured in

LB broth supplemented with 50 pg ml-' tetracycline at 28"C, overnight, in a gyratory shaker. For introduction into soil cells were washed twice in sterile demineralized water.

Soil and soil inoculation

Ede loamy sand (van Elsas et al. 1986; Heijnen et al. 1992), obtained from a field microplot at Ital, Wageningen, was inoculated with Ps. jluorescens R2f (RP4 : : pat) in accord- ance with van Elsas et al. (1991a). Population densities applied in the different experiments varied from 10 to lo7 colony-forming units (cfu) per g of dry soil. Flevo silt loam (described by van Elsas et al. 1986) and different soil samples of undefined texture obtained from Schonebeck, Germany, were also used in some DNA extraction and PCR experiments.

Enumeration of inoculant cells from soil

Pseudomonaspuorescens R2f (RP4 : : pat) cfu numbers were

determined by selective plating as described by van Elsas et al. (1991a).


(RP4 : . pat), immunofluorescence of soil samples was applied ; specific FITC-labelled anti-R2f antibodies, as

described by Trevors et al. (1990), were used.








Ps. Juorescens

Extraction of DNA from soil

In analogy to Ogram et al. (1987) we developed a direct DNA extraction protocol, given its high efficiency in recov- ering bacterial DNA from soil (Steffan et al. 1988). The



basic protocol used 5 g soil samples. Samples were sus-

pended in 5 ml of 0.12 mol I-'

supplemented with freshly added lysozyme at a final con- centration of 5 mg ml-'. The resulting suspensions were shaken for 5 rnin at room temperature and incubated at 4°C (incidentally 37°C) for 30 min. After incubation 600 p1 of 20% sodium dodecyl sulphate (SDS) was added and sus- pensions were kept on ice. To complete cell lysis, 2.5 g of acid-washed glass beads (0.17418 mm diameter) were added and the suspensions were shaken in a Braun's homogenizer three times for 90 s at the highest speed. The resulting suspensions were mixed well with equal volumes of cold phenol equilibrated with 0.5 mol I-' Tris (final pH 7.8; Sambrook et al. 1989), after which they were centri-

phosphate buffer (pH 8.0)

fuged for 15 min at 12000 g at room temperature. After transferring the aqueous phases to new tubes the pellets were treated with 5 ml of 0.12 mol 1-' phosphate buffer (pH 8.0) and re-extracted with Tris-equilibrated phenol.

The resulting aqueous phases were combined with the first extracts, and these mixtures were extracted twice with

equal volumes of chloroform-isoamylalcohol

(24 : 1, v/v).

The aqueous phases recovered from these extractions were ethanol-precipitated and the DNA pellets were washed with 70% ethanol. Dried pellets were resuspended in Tris-

EDTA (TE; Sambrook et al. 1989), giving a final volume of 1000 pl.

Purification steps

The crude DNA preparations were subjected to several sequential purification steps. Caesium chloride (CsCI) and potassium acetate (KAc) precipitation steps were first used to selectively remove proteins, RNA and humic substances. Before and after each purification step, DNA samples were used in PCR and restriction assays. In the first purification step, CsCl (1 g) was added to 1000 pl of crude DNA extract. The resulting suspensions were mixed well, kept for 3 h at room temperature, and centrifuged for 20 rnin at 14000 g at room temperature. After recovering the supernatant fluids (about 1 ml), 4 ml of demineralized water and 3 ml of isopropanol were added. The mixtures were kept at room temperature for 15 min and then centrifuged for 15 min at 14000 g. The resulting pellets were dissolved in 1000 p1 of TE buffer. In the next step, the CsC1-treated DNA preparations (1000 pl) were mixed with 200 pl of 8 mol I-' KAc solu- tion and kept at room temperature for 15 min. Suspensions were centrifuged for 15 rnin at 14000 g and the supernatant fluids recovered were treated with 600 pl of isopropanol. After mixing well and keeping at room temperature for 5 min, the suspensions were centrifuged (14000 g, 15 min) and the resulting DNA pellets were washed extensively with 70% ethanol and dissolved in 1000 pl of TE buffer.



lysozyme treatment of 5 g soil


addition of SDS (1.25%) and bead-beating


phenol extraction (2x)


chloroform extraction (2 x)


ethanol precipitation


resolved DNA pellet -crude


Purification of the DNA crude extract

CsCl precipitation (1 g mi-')

(0.6~01%)of DNA in the supernatant fluid KAc precipitation (1.3 mol I-') +centrifugation + isopropanol precipitation (0.6 vol'X) of DNA in the supernatant fluid



isopropanol precipitation




glass milk







Fig. 1 Outline of extraction and purification protocol. SDS, Sodium dodecyl sulphate; PCR, polymerase chain reaction

When necessary, final purification was performed by two alternative methods. The first was based on the capacity of spermine-HC1 to precipitate selectively DNA in a low-salt environment (Hopwood et al. 1985); 25 pl of spermine-HC1 (pH 84L8.5) at 100 mmol 1-' was added to 500 pI KAc- treated DNA preparations. The mixtures were kept at room temperature for 5 min, after which the DNA was precipi- tated by centrifugation (3 min, room temperature, 14000 g).The resulting pellets were washed twice with 1 ml of


1 ~ sodium acetate (NaAc) and 10 pl of 1 mol 1- ' MgCl, , followed by dissolution in 500 p1 of TE buffer. As an alter- native method we used DNA purification with glass milk

(Geneclean 11; Bio 101 Inc). The procedure was performed according to the manufacturer's instructions. For additional purity, the spermine- or glass milk-purified DNA was again treated with spermine or glass milk. An outline of the pro- tocol is shown in Fig. 1.

70Y0 ethanol supplemented with 100 p1 of 3 mol

Determination of DNA recovery rate

Initially, DNA recovery was estimated from the intensity of ethidium bromide-stained DNA bands in agarose gels and comparison with controls of known concentration. In later work Polaroid pictures of DNA bands in ethidium bromide-stained agarose gels were scanned with a Bio- Image scanner (Millipore). For quantification it was neces- sary to have an internal standard of known DNA content on the gel. For calculation, the integrated optical densities were used. The Gibco-BRL 1-kb ladder (No. 52M615SA; molecular size range 0.3-12.2 kb) was used to determine sizes of fragments on agarose gels (Figs 2 and 3).

Slot blot hybridization

DNA preparations (100-200 pl) were applied, both directly


(Hybond, Amersham










RPN 303N, or



to a nylon membrane RPN 303N, or Gene-Screen 12345678 Fig. 2 DNA (10 pl sample per

Fig. 2 DNA (10 pl sample per well) obtained from two unseeded soils using the modified procedure. Lanes: 1-3, DNA extracted from Flevo silt loam ; 4-6, DNA extracted from Ede loamy sand; 1 and 4, crude preparations; 2 and 5, after KAc precipitation; 3 and 6, after glass milk purification; 7, pUC19 DNA; 8, I-kb ladder (see Materials and Methods, arrow indicates I-kb hand). Note the 'cloud' caused by humic substances in the crude DNA from the Ede loamy sand soil














I 2 3 4 5 6 7 8 9 10 II 12 13 Fig. 3 Restriction

Fig. 3 Restriction by HindIII of soil DNA with and without

added pUC19 DNA at different purification steps. Lanes: 1 and 2, crude; 3 and 4, after KAc precipitation; 5 and 6, after glass milk purification; 7 and 8, after spermine-HCI purification;9 and 10, after spermine purification followed by glass milk; 11, pUC19 digestion product (2.7 kb); 12, pUC19 undigested; 13, I-kb ladder (see Materials and Methods, arrow indicates 1-kb band). Uneven numbers, no pUC19 added; even numbers, pUC19 added

held in a Minifold-1 apparatus (Schleicher and Schuell, Dassel). The hybridization procedure was according to standard protocols (Sambrook et al. 1989), with the 683-bp pat fragment as a probe.

Amplification of the pat sequence in soil DNA extracts

Polymerase chain reaction was used to amplify the pat sequence in different soil extracts. For each enzyme used, i.e. AmpliTaq, the Stoffel fragment (both from Perkin- Elmer/Cetus) and SuperTaq (Sphaero-Q), the reaction mixes (50-100 p1) were prepared according to the manufac- turer's instructions. Primers used were as described by van Elsas et al. (1991a). One pl of each DNA preparation was added to the reaction mixture following denaturation. The hot start, as suggested by d'Aquila et al. (1991), was used in the final protocol. Routinely, 35 cycles (94"C, 1 min; 45 or 50°C, 1 min; 72"C, 1-2 min; final extension at 72°C for 9 min) were run. Polymerase chain reaction products, expected to be about 0.7 and, incidentally, 1.2 kb in size (van Elsas et al. 1991a), were analysed on gels after agarose gel electrophoresis (0.8%) and ethidium bromide staining. In addition, detection was performed on Southern blots (Sambrook et al. 1989) to nylon membranes (see above) with the pat fragment as a probe.

Molecular biology techniques applied to detection from soil

The manual of Sambrook et, al. (1989) was used for all molecular work.



Restriction analysis was performed on soil DNA extracts of different degrees of purification, with HindIII (Boehringer 656321), EcoRl (Boehringer 703737), BamHl (Boehringer 567604), PvuII (Boehringer 642690) and Bgfl (Boehringer 404101). The reaction mixtures (20 pl) con- tained the recommended buffer, 5 p1 of soil-derived DNA and 5-10 units of the enzyme, and were incubated at 37°C for at least 3 h. The same mixtures were also incubated with an additional 200 ng of the cloning vector pUC19 to follow more readily the inhibition of different restriction enzymes by the soil DNA extract. After incubation the digests were precipitated with ethanol. The washed pellets were resuspended in 10 p1 of TE and analysed after agarose gel electrophoresis and ethidium bromide staining. The pat probe was prepared according to van Elsas et al. (1991a). Nick translation was used to label probes for initial hybridizations, following the protocol of Sambrook et al. (1989). As a non-radioactive alternative for nick translation, labelling was performed with the digoxigenin (Dig) label- ling kit (Boehringer 1175033). Prehybridization and hybrid- ization were done at 68°C according to the Boehringer protocol. After washing at high stringency (68"C, 0.1 x SSC), chemiluminescent detection was performed with the Dig luminescent detection kit (Boehringer 136514) as described by the manufacturer.



Development of rapid DNA extraction-PCR protocol

In a previous paper (van Elsas et al. 1991a), the pat frag- ment was amplified from Ps.Jluorescens (RP4 : : pat)-inocu- lated Ede loamy sand after purification of crude DNA extracts from soil via agarose gel electrophoresis and electro-elution of fragments of desired size. In addition, background pat or other DNA amplifiable with pat-specific primers, were shown to be absent from this soil. A dis- advantage of this method might be the inherent limits to DNA recovery in gel electrophoresis and the bias in electro-elution towards smaller DNA fragments. We there- fore decided to base a modified protocol on more quantitat- ive extraction and purification, omitting gel electrophoresis as well as time-consuming CsCl gradients and polyvinyl poly-pyrrolidone adsorption of humics.

Cell lysis and DNA extraction and recovery

The experiments performed in the development of the pro- tocol used soil samples seeded with roughly lo6 or lo7 cfu of Ps.Jluorescens per g, which contained up to 10-fold less cfu at the time of sampling. Cell lysis was originally based on freezing/thawing of a soil suspension in the presence of lysozyme and SDS, followed by rapid extraction of the soil lysate with cold phenol. In later work lysis was improved


Purification step








After potassium acetate precipitation

After spcrmine-HCI

After glass milk

caesium chloride precipitation

Digestion by













Amplification by

















+ +







Table 1 Effect of purification of

soil-derived DNA on restrictabilityand

polymerase chain reaction (PCR) amplification*

* Soil crude DNA extracts were derived from soil containing about lo6 cfu g- ', using the modified protocol described with Braun's lysis. pUC19 was added to DNA preparations to serve as a marker of the activity of restriction enzymes. t E/R, EcoRI/Bgfl double digestion. -, No digestion or amplification; +/ -, partial digestion; +, complete digestion or posi- tive amplification; ND, not determined.

by using a Braun's homogenizer. Crude DNA preparations obtained with cold phenol were consistently less brown than those obtained after a hot phenol extraction step (60°C, 30 min), as in the Ogram protocol, indicating that

these contained less humic substances. However, crude DNA prepared in this way and pure pUC19 DNA added to the preparations were insensitive to restriction by HindIII,


partial cuts in the DNA (Table 1 and Fig. 3). In addition, crude DNA was not amplifiable. This suggested the pres- ence of substances hindering both digestion by restriction enzymes and amplification of the DNA. Figure 2 shows the appearance of DNA and humic substances following gel electrophoresis and ethidium bromide staining. Several attempts to improve purity at this stage, i.e. the

use of mercapto-ethanol/spermidine during lysis, the addi-

tion to soil, prior to lysis, of salmon sperm DNA to complex humic substances and the use of RNase and pronase digestion following extraction, did not lead to improved purity as evidenced via agarose gel electro- phoresis and 0.D.260/0.D.280ratios (1.2).

+ Bgll, BamHl and Pvu1; only Sau3A produced

The amount of DNA


in the crude prep-

arations, estimated from agarose gels, was initially about 1-5 pg of DNA per g of dry soil, vs up to fourfold less for parallel preparations obtained with the protocol of Ogram et al. (1987). DNA obtained by freeze/thaw lysis appeared in a band of about 1&25 kb size, whereas DNA obtained following lysis in a Braun's homogenizer was somewhat more sheared (Fig. 2). However, the degree of shearing was less than that described by Ogram et al. (1987). Cell lysis induced in the Braun's homogenizer resulted in a consider- able improvement of DNA recovery, i.e. between 4 and 20 pg of DNA was routinely obtained in the crude prep- arations. In one representative experiment, recovery, as determined by scanning, was 11 pg DNA per g of soil after mechanical lysis and 7 pg DNA following freeze/thaw lysis. The 0.D.260/0.D.280values of all crude preparations were

between 1.2 and 1.4, indicating the presence of large amounts of contaminating matter. The approximate DNA amount was determined from agarose gels, because O.D.260 values as well as the values obtained by fluorimetric DNA measurements indicated that the latter two methods were

unsuitable. The DNA recovery values (4-20 pg g- ' of soil) obtained by us were comparable to those reported by Steffan et al. (1988), Porteous and Armstrong (1991) and Selenska and


counts) commonly


assuming a cellular DNA content of 5-8 fg (McCoy and Olson 1985), theoretically between 5 and 80 pg of bacterial DNA per g of soil should be extractable from a 'common' soil. Assuming the likely interference by some fungal DNA, and neglecting DNA from non-microbial sources, this still

suggests that

those of the other















workers, represent sizeable parts of the bacterial popu- lations in soil.

Purification steps, restrictionand amplification

The objective of the purification procedure was to obtain DNA clean enough to be amplifiable, and at the same time to attain quantitative recovery of DNA. As criteria for judging the effectiveness of each purification step, we used restrictability with various restriction enzymes (Steffan et al. 1988), and direct amplifiability in the PCR. Most of the work was carried out with DNA samples derived from soil microcosms containing about lo6 cfu of Ps.jluorescens R2f (RP4 : : pat) per g at the time of sampling. Initial purification via CsCl precipitation of impurities (protein, RNA, humic substances) resulted in a substantial loss of brown colour and a drastically improved efficiency of the following KAc precipitation step. Little to no loss of DNA was recorded. However, DNA after CsCl clean-up was not restrictable with most enzymes tested (except for partial digestion with Sau3A) and, in addition, was not



123456789 Further development of PCR applled to soil samples

9 Further development of PCR applled to soil samples Fig. 4 Southern blot of polymerase chain

Fig. 4 Southern blot of polymerase chain reaction (PCR) products amplified with the Stoffel fragment using DNA from Ede loamy sand. DNA preparationswere purified twice with glass milk. Lanes contain PCR product from soil containing per g: 1, about lo6 cfu of Pseudomonasjuorescens R2f (RP4 : : pat); 2, as in 1, plus 20 ng of free RP4 : : put DNA; 3, as in 1, plus 0.2 ng of

free RP4 : : put DNA; 4, 3 x lo3 cfu of Ps.j?uorescens R2f (RP4 : : put); 5, as in 4, plus 20 ng of free RP4 : : pat; 6; 20 ng of

free RP4 : : pat DNA ; 7; 0.2 ng of

positive control (Ps.juorescens R2f (RP4 : : put) added); 9, negative control. Amplification product (indicated by arrow) was 0.7 kb as expected (van Elsas el al. 1991a)

free RP4 : : par DNA ; 8,

amplifiable (Table 1). Potassium acetate precipitation of contaminating matter was then applied to CsC1-treated DNA preparations. This, again, removed humic substances and other impurities at negligible loss of DNA. DNA prep- arations were only partially restrictable with Sau3A and BamH 1, and only incidentally amplifiable after dilution to 0.5-1 ng per PCR reaction mixture. As final purification steps, two strategies were followed. The first was based on DNA precipitation by spermine-HCI, and the second used precipitation of DNA on glass milk. Both methods resulted in virtually complete removal of the brown colour from the DNA preparations at recovery values of 50-100% for spermine-HC1 and 75-100% for glass milk. In addition, the DNA was now restrictable with HindIII, EcoRI/Bgll, PvuII, Sau3A and BamHl (partially for spermine-HC1; Table 1). Application of the PCR to spermine- and glass milk-cleaned preparations routinely produced the expected 683-bp amplification product, as evidenced by gel electro- phoresis and Southern blot analysis (Fig. 4). Both DNA from soil containing about lo6 and 3 x lo3cfu g-’, and target DNA freely added to soil (0.2 and 20 ng g - of soil), and mixes of both, were detected via PCR amplification and Southern hybridization (Fig. 4); unseeded soil did not show an amplification product.

The PCR was optimized with pure pPATB2 DNA (van Elsas and Trevors 1990) as a target. Variations in the PCR cycle, the number of cycles, the enzyme, the buffer used and the primers were tested using different concentrations of target DNA. The protocol described (Materials and Methods) summarizes the results of these efforts. With this optimized protocol and either one of the three enzymes mentioned, it was possible to obtain amplification of the highest dilution used, i.e. 0.01 fg of target DNA, corre- sponding to 16 copies of the target.

AmpliTaq was used in initial amplifications from soil, but was easily inhibited by soil compounds. SuperTaq was shown to be relatively insensitive to soil compounds, per- mitting the amplification of up to 2 pl of soil DNA in 50 jd

reaction mix ; PCR products were often smeared,

making interpretation difficult. The cause of this smearing

is unknown, and PCR applied to a different target sequence produced distinct bands of defined size (not shown). The use of the Stoffel fragment avoided this problem with the

pat amplification system, and also allowed for efficient amplification, but a second round of glass milk or spermine was sometimes necessary to obtain consistent amplification. To determine the limit of detection of the PCR the two purification protocols were applied to day-1 soil samples inoculated with 0, 3 x lo2, 3 x lo3, 3 x lo4, 3 x lo’,


3 x lo6 and 3 x lo7 cfu of Ps. fltlorescens R2f (RP4

: : pat)

per g of soil. Colony-forming unit counts at sampling were,

lo’, 3.0 x lo2, 5.4 x lo3, 4.4 x lo4,

respectively, 0, 1.3 x

4.5 x lo5, and 1.0 x lo7 per g dry soil. The application of the PCR resulted in the appearance of 683-bp amplification products only in the soil DNA preparations with counts of 3 x 10’ cfu (inoculum level 3 x lo3) per g of dry soil or

higher. Theoretically, at a level of lo3 inoculant cfu per g


copies of pat would be present in the final volume (1000 pl). Up to 1 p1 of this final volume, corresponding to 20 ng total soil DNA, could serve as a template for amplification. Therefore, assuming 100% recovery of DNA, and the necessity of 10-20 copies of the target for consistent posi- tive amplification, about lo3 inoculant cfu per g of soil would still be detectable, which is about the limit of detec- tion found. Tsai and Olson (1992), targeting a 7-copy 16s ribosomal DNA sequence by PCR, reported a limit of detection of 5 x 10’ Esrherichia coli cells per g of soil, or the equivalent of about three cells (21 copies of the target) in the PCR reaction mixture. Pillai et at. (1991) reported PCR-assisted detection of 1-10 transposon Tn5-labelled Rhizobium leguminosarum cells per g of soil with PCR on pellets obtained from soil-derived suspensions. However, two rounds of PCR were needed which enhances the risks of non-specific amplification. Moreover, signals of low

of soil, and three copies of RP4 : : pat

per cell,

1.5 x


inoculum densities were only positive in dot-blots of PCR reaction mixes. A more precise method for quantification of PCR seems to be possible using temperature gradient gel electro- phoresis (TGGE), as developed by DIAGEN. We are cur- rently looking for possibilities to apply this methodology to soil DNA extracts.

Application of PCR to the description of the fate of introducedPseudomonas fluorescens (RP4 : :pat)

As shown in Table 2, introduced Ps. Puorescens R2f (RP4 : : pat) showed a progressive decline in cfu numbers in the Ede loamy sand soil. The corresponding imrnuno- fluorescence specific cell counts also showed a progressive decline, albeit at a significantly lower rate. This suggested the (temporary) occurrence of non-culturable cells in the system, which might be either viable, non-viable or even dead. Attempts to track the introduced cells by specific detec- tion of put via slot-blot hybridization of crude extracts showed that only at cfu and cell counts around or above lo6 per g of dry soil, a (weak) hybridization signal was detected with the pat probe. In some cases, dilution of the DNA preparations was needed in order to obtain a positive signal. As shown in Table 2, PCR accurately detected the presence of the pat sequence throughout the course of the microcosm experiment, accompanying and complementing the cfu and total specific cell counts. We also applied detection via PCR to a study over a longer period in the same soil, with a starting inoculum of

per g of

about lo5 cfu of Ps. jluore~censR2f (RP4 : : pat)

soil. Total DNA was extracted from the soil 5 d and 5 months after inoculation. Both DNA preparations purified with glass milk could serve as a template for PCR. Whereas Ps.Juorescens (RP4 : : pat) cfu were still detectable after 5 d, no cfu were obtained after incubation for 5 months. These observations suggested the presence of non- culturable cells or liberated target DNA in the latter

Table 2 Survival of PseudomonasJluorescens R2f (RP4 : : pat) in

soil, using selective plating, immunofluorescenceand

molecular-based techniques

Log cfu or cells (IF) per g dry soil on (day):




Cells (IF)








































IF, I mmunofluorescence; ND, not determined ; PCR, polymerase chain reaction.

sample. The possibility of detecting such cells or the target represents the major advantage of the use of PCR detection for monitoring of GMOs following release into soil.

Concluding remarks

DNA extraction and purification methodology from soil was developed to obtain, via a simple, inexpensive and rapid protocol, DNA preparations suitable for PCR analysis. The protocol described here represents an improvement over a formerly applied method (van Elsas rt al. 1991a), as the gel electrophoresis and electro-elution steps could be replaced by three simple, rapid and quanti- tative precipitation steps. Furthermore, hydroxyapatitc purification, which often resulted in poor (2.5'XI) DNA recovery, and tirne-consuming CsCl gradients could be omitted. Our scheme for extraction and purification permits amplification and restriction enzyme analysis of DNA derived from different soils. For instance, Flevo silt loam as well as different Schonebeck soil samples extracted by the protocol presented also showed amplification by PCR (not shown); the number of purification steps needed after KAc precipitation was dependent on soil type. Therc- fore, the developed protocol is flexible in that it allows for sufficient purity of DNA for PCR and restriction analysis after varying final purification steps, depending on the soil type (Fig. 1). Purification of soil DNA oia different steps was devel- oped for routine testing in GMO field releases as well as for control investigations in areas surrounding contained use of GMOs. A major factor limiting the usefulness of PCR for environmental studies is that it is still qualitative or, at best, semi-quantitative. The possibility of using end-point dilution/most-probable-number quantification or of ampli- fication in the presence of internal standards, e.g. cia TGGE, still has to be explored for soil-derived DNA. A second factor is the interpretation of the appearance of PCR products, i.e. whether these indicate the presence of whole living or possibly dead bacterial cells or even free DNA. However, the analysis presented here suggests the usefulness of PCR for the detection of GMOs (and/or sequences) of interest in a non-culturable state or at reduced levels in soil.


We thank A. Hagler, G. Berben, L. van Overbeek and C. Clegg for contributing to some of the experiments. This work was supported by an EC grant awarded to J.D.v.E. and by OECD short-term fellowship awards to K. Smalla and N. Cresswell. L.C. Mendonca-Hagler was supported by an EC: fellowship.


Cresswell, N., Saunders, V.A. and Wellington, E.M.H. (1991) Detection and quantification of Streptomyces violaceolatus plasmid DNA in soil. Letters in Applied Microbiology 13, 193-


D’Aquila, R.T., Bechtel, L.J., Videler, J.A., Eron, J.J., Gorczca, P. and Kaplan, J.C. (1991) Maximizing sensitivity and speci- ficity of PCR by preamplification heating. Nucleic Acids Research 19, 3749. Ford, S. and Olson, B.H. (1988) Methods for detecting genetically engineered microorganisms in the environment. Advances in Microbial Ecology 10, 45-73. Heijnen, C.E., Hok-A-Hin, C.H. and Van Veen, J.A. (1992) Improvement to the use of bentonite clay as a protective agent, increasing survival levels of bacteria introduced into soil. Soil Biology and Biochemistry (in press). Holben, W.E., Jansson, J.K., Chelm, B.K. and Tiedje, J.M. (1988) DNA probe method for the detection of specific micro- organisms in the soil bacterial community. Applied and Environ- mental Microbiology 54, 703-71 1. Hopwood, D.A., Bibb, M.J., Chater, K.F., Kieser, T., Bruton, C.J., Kieser, H.M., Lydiate, D.J., Smith, C.P., Ward, J.M. and Schrempf, H. (1985) Genetic Manipulation of Streptomyces: A Laboratory Manual. Norwich, UK: The John Innes Founda- tion. McCoy, W.F. and Olson, B.H. (1985) Fluorimetric determination of the DNA concentration in municipal drinking water. Applied and Environmental Microbiology 49, 81 1-817. Ogram, A,, Sayler, G.S. and Barkay, T.J. (1987) DNA extraction and purification from sediments. Journal of Microbiological Methods 7, 57-66. Pillai, S.D., Josephson, K.L., Bailey, R.L., Gerba, C.P. and Pepper, I.L. (1991) Rapid method for processing soil samples for polymerase chain reaction amplification of specific gene sequences. Applied and Environmental Microbiology 57, 2283-


Porteous, LA. and Armstrong, J.L. (1991) Recovery of bulk DNA from soil by a rapid, small-scale extraction method. Current Microbiology 22, 345-348. Saiki, R.K., Gelfand, D.H., Stoffel, S., Scharf, S.J., Higuchi, R., Horn, G.T., Mullis, K.B. and Erlich, H.A. (1988) Primer- directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239, 487494.

Sambrook, J., Fritsch, E.F. and Maniatis, T. (1989) Molecular Cloning. A Laboratory Manual, 2nd edn. Cold Spring Harbor, New York. Sayler, G.S., Shields, M.S., Tedford, E.T., Breen, A., Hooper, S.W., Sirotkin, K.M. and Davis, J.W. (1985) Application of DNA-DNA colony hybridization to the detection of catabolic genotypes in environmental samples. Applied and Environmental Microbiology 49, 1295-1303. Selenska, S. and Klingmuller, W. (l991a) Direct detection of nf- gene sequences of Enterobacter agglomerans in soil. FEMS Microbiology Letters 80, 243-246.



Selenska, S. and Klingmuller, W. (1991b) DNA recovery and direct detection of Tn5 sequences from soil. Letters in Applied Microbiology 13, 21-24. Smalla, K., Isemann, M., John, G., Weege, K.-H., Backhaus, H. and Wendt, K. (1991) A risk assessment of industrial pro- duction of alpha-amylase with an rDNA production strain. In Biological Monitoring of Genetically Engineered Plants and Microbes ed. MacKenzie, D.R. and Henry, S.C. pp. 205-220. Bethesda, Maryland : Agricultural Research Institute. Smit, E., van Elsas, J.D., van Veen, J.A. and de Vos, W. (1991) Detection of plasmid transfer from Pseudomonas Jluorescens to

indigenous bacteria in soil by using phage @R2f for donor counterselection. Applied and Environmental Microbiology 57,


Steffan, R.J. and Atlas, R.M. (1988) DNA amplification to enhance detection of genetically engineered bacteria in environ- mental samples. Applied and Environmental Microbiology 54, 2185-2 I9 1. Steffan, R.J., Goksoyr, J., Bej, A.K. and Atlas, R.M. (1988) Recovery of DNA from soils and sediments. Applied and Environmental Microbiology 54, 2908-291 5. Trevors, J.T. and van Elsas, J.D. (1989) A review of selected methods in environmental microbial genetics. Canadian Journal of Microbiology 35, 895-902. Trevors, J.T., van Elsas, J.D., van Overbeek, L.S. and Starodub, M.E. (1990) Transport of a genetically engineered Pseudomonas Jluorescens strain through a soil microcosm. Applied and Environmental Microbiology 56, 401408. Tsai, Y.-L. and Olson, B.H. (1992) Detection of low numbers of bacterial cells in soils and sediments by polymerase chain reac- tion. Applied and Environmental Microbiology 58, 754-757. van Elsas, J.D. and Trevors, J.T. (1990) Plasmid transfer to

indigenous bacteria in soil and rhizosphere : problems and


spectives. In Bacterial Genetics in Natural Environments ed. Fry, J. and Day, M.J. pp. 188-199. London: Chapman & Hall. van Elsas, J.D., Govaert, J.M., Dijkstra, A.F. and van Veen, J.A. (1986) Survival of Pseudomonas Jluorescens and Bacillus subtilis introduced into two soils of different texture in field microplots.

FEMS Microbiology Ecology 38, 151-160. van Elsas, J.D., Trevors, J.T. and Starodub, M.E. (1988) Bac- terial conjugation between pseudomonads in the rhizosphere of wheat. FEMS Microbiology Ecology 53, 299-306. van Elsas, J.D., van Overbeek, L.S. and Fouchier, R. (1991a) A specific marker, pat, for studying the fate of introduced bacteria and their DNA in soil using a combination of detection tech- niques. Plant and Soil 138,49-60. van Elsas, J.D., van Overbeek, L.S., Feldmann, A.M., Dull- emans, A.M. and de Leeuw, 0. (1991b) Survival of genetically engineered Pseudomonas Juorescens in soil in competition with the parent strain. FEMS Microbiology Ecologv 85, 53-64. Waalwijk, C., Dullemans, A.’and Maat, C. (1991) Construction of a bioinsecticidal rhizosphere isolate of Pseudomonas Jluorescens. FEMS Microbiology Letters 77, 257-264.