| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
,1
* Istituto Sperimentale Italiano "Lazzaro Spallanzani," Rivolta d’Adda (CR) 26027, Italy
Istituto di Biologia e Biotecnologia Agraria (IBBA), Consiglio Nazionale delle Ricerche (CNR), Milano 20133, Italy
1 Corresponding author: terenzio{at}ibba.cnr.it
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Key Words: deoxyribonucleic acid detection • feed • milk • traceability
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
To further substantiate these findings and to prove that only chloroplast-based DNA fragments can be successfully used for traceability of the feed components, we analyzed milk samples from 4 distinct experimental groups of goats, each provided with a specific monophytic diet. In this paper, we demonstrate that in milk it is possible, by either DNA sequencing or a PCR-based distinctiveness assay, to identify the plant species in the diet supplied to animals.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Cows.
Stock farms were located in the Mantova and Aosta provinces of Italy. Raw milk samples (50 mL) were collected into sterile tubes and quickly frozen, as reported for goats. Cows were fed with a known polyphytic diet, except for an integrating element whose composition was not detailed on the analytical tag. For the 2 brands of raw milk samples bought on the northern Italian market, compositions of the fodder fed to cows were unknown. Deoxyribonucleic acid was extracted from 4 replicates of each sample on the day of purchase. Once extracted, the DNA solution was stored at +4°C.
Extraction and Purification of Genomic DNA
To avoid any possible laboratory-derived plant DNA contamination that could have occurred during sample handling, PCR analyses done on DNA extracted from vegetative material were kept separate from those performed on milk-derived DNA samples. Laboratory sites, timing, reagents, and pipettes were different for the 2 procedures.
From Plant.
Three grams of each diet was finely ground in liquid nitrogen with a mortar and a pestle. Extraction of DNA with CTAB buffer [1.4 M NaCl, 2% (wt/vol) cetyltrimethylammonium bromide, 100 mM Tris, 15 mM EDTA, pH 8.0] was carried out according to the procedure described by Doyle and Doyle (1987).
From Milk.
Total DNA from both goat and cow milk samples was extracted according to the protocol mentioned above. Goat milk samples were extracted in triplicate. Twenty milliliters of each milk sample was directly mixed with 60 mL of CTAB buffer and 240 µL of proteinase K (20 mg/mL; Sigma-Aldrich, St. Louis, MO). The solution was kept overnight at 65°C under constant agitation. A 75-mL (vol/vol) quantity of phenol:chloroform:isoamyl alcohol (25:24:1) was then added to the samples and mixed gently. The aqueous phase was recovered after centrifugation for 15 min at 8,000 x g. The extraction-centrifugation step was repeated twice. Deoxyribonucleic acid was precipitated overnight at –20°C with 1 vol of 2-propanol and 500 µL of glycogen (20 mg/mL). Samples were then centrifuged at 8,000 x g for 20 min at 4°C. The pellet was washed twice with 75% (vol/vol) ethanol and dissolved in 300 µL of double-distilled sterile water. Samples were finally incubated for 30 min at 50°C with 25 µL of RNase A (100 mg/mL; Sigma-Aldrich).
Purity and quantity of the extracted DNA were determined with a spectrophotometer by measuring the absorbance (A) at A260 and the A260/A280 ratio (DU 730 Beckman spectrophotometer, Fullerton, CA). Only milk samples with a satisfactory DNA purity (A260/A280 = 1.7 to 1.9) were used. Aliquots of each DNA sample were also loaded on a 0.8% (wt/vol) agarose gel, and the DNA quantity was further verified by comparison with known amounts of standard DNA (high molecular weight phage lambda DNA, M-Medical, Cornaredo, Italy).
PCR Methodology and Primer Design
Each PCR-based DNA amplification was carried out in a final volume of 30 µL, in 1x Taq Advanced buffer (Eppendorf HQ, Hamburg, Germany), 2 mM magnesium solution, 200 µM each deoxynucleotide 5'-triphosphate, 1 µM each primer, and 1 U of Taq DNA polymerase (Eppendorf HQ). The amount per reaction of DNA was between 200 and 1,200 ng if extracted from milk (depending on the specific set of primers used) or 0.5 ng when extracted from plant material. The cycling conditions for each primer pair are described below. Polymerase chain reaction amplification was performed in an Eppendorf Mastercycler Gradient thermal cycler (Eppendorf HQ). All PCR were "hot started" by placing the PCR tubes on the preheated (99°C) PCR thermocycler machine.
Two controls for negative amplification (no DNA) were systematically added to every set of PCR. The DNA fragments amplified by PCR were loaded onto 1.5 to 2% (wt/vol) agarose gels with a DNA molecular marker (pUC8 Mix Marker or O’RangeRuler, both from Fermentas, Glen Burnie, MD) that would allow for the rapid identification of amplicon size.
Diverse sets of primers resulting in amplicons of different sizes were used for detection of the chloroplast sequences of rubisco (rbcL), matK, and the intron region of the transfer RNA (tRNA) leucine (trnL) gene. The complete list and description of the primers used in this study is given in Table 1, with the exception of primer pairs β-cas E7F/cas2R and RUB-F2/RUB-R2, which have been reported by Klotz and Einspanier (2001) and Phipps et al. (2003), respectively. Three different DNA extractions and several PCR amplification reactions were performed for each pool of milk samples. The RUB-F2/RUB-R2 set of primers was first tested on the total DNA extracted from the different plant species present in the animal diets, as well as from some cow and goat milk samples. Two newly designed sets of primers (Arub fw/Arub rv and Lrub fw/Lrub rv), both matching the rbcL gene sequence in proximity of the ATG translational initiation codon, were then used for plant DNA detection. Primer specificity was initially evaluated on total DNA separately extracted from each of the 4 different monophytic diets (oats, ryegrass, hay, and silage X-triticosecale).
|
When using the RUB-F2/RUB-R2, the Lrub fw/Lrub rv, and the Arub fw/Arub rv combinations of primers, the thermal cycle conditions were as follows: 5 min at 95°C for the initial denaturation, 45 cycles of amplification (30 s at 95°C, 40 s at 60°C, and 30 s at 72°C), and 5 min at 72°C for the final extension. Except for an annealing step performed for 90 s at 60°C, the same amplification conditions were used with the rbcL-ex fw and rbcL-ex rv and the casE7F/cas2R sets of primers to allow for amplification of the whole rbcL sequence or the β-CN fragments, respectively.
The temperature profile of PCR amplifications carried out with the matK F/matK R combination of primers included an initial denaturation step at 95°C for 5 min, followed by 5 cycles at 94°C for 30 s and 62°C for 30 s. Forty additional cycles at 92°C for 20 s and 62°C for 20 s were then performed. All reactions were brought to completion with a final extension step of 3 min at 72°C.
The PCR conditions for the amplification of the trnL target intron, carried out with the use of the Plant2-forward/trnL-rv set of primers, were as follows: 5 min at 95°C for the initial denaturation, followed by 45 cycles at 94°C for 30 s and 66°C for 30 s, concluding with a final extension performed at 72°C for 3 min.
Sequence Analyses
To confirm the nucleotide sequence composition of the amplicons, DNA bands resulting from PCR were separated by agarose gel electrophoresis, excised, and purified using a QIAquick Gel Extraction Kit (Qiagen, Hamburg, Germany) according to the instructions of the manufacturer. Direct sequencing of the purified PCR products was performed at PRIMM srl (Biomedical Science Park San Raffaele, Milan, Italy). Deoxyribonucleic acid sequencing on both strands of the amplified fragments was repeated at least 3 times from 3 independent PCR amplification experiments.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Preliminary results were obtained according to Phipps et al. (2003), using a set of primers (RUB-F2 and RUB-R2) specific for the large subunit of the rubisco single-copy gene (rbcL). This set of primers was capable of amplifying the 351-bp rbcL fragment in both goat and cow milk samples (data not shown). Two new sets of primers, both matching the rbcL gene sequence in proximity of the ATG initiation codon, were then used (Arub fw/Arub rv and Lrub fw/Lrub rv; see the Materials and Methods section). Representative results of these experiments are shown in Figure 1A, in which a single band of the expected size, that is, 348 bp for the Arub fw/Arub rv combination and 423 bp for the Lrub fw/Lrub rv set of primers, was amplified from the DNA of all the plant samples. The lower intensity of the bands amplified by both sets of primers from the X-triticosecale silage samples is likely to be dependent on the ensiling process, which has been reported to cause a progressively pronounced degradation of DNA (see Discussion section). The lower level of amplification of the 423-bp band in oats is instead occasional, limited to this specific reaction and gel run.
|
The rbcL DNA sequences were also detected from several available cow milk samples. The Arub fw/Arub rv set of primers successfully detected a 348-bp fragment in 2 stock farm and 2 commercially distributed raw milk samples. The identity of the amplified rbcL fragments was verified by DNA sequencing (see below).
Sequencing of the Different rbcL Fragments
To confirm that the amplified DNA fragments from the milk samples were actually derived from the diets fed to the animals, the rbcL PCR products amplified from the 4 goat milk samples were sequenced on both strands. Similarly, the rbcL PCR fragments amplified from each monophytic diet were sequenced to allow for a direct and unambiguous comparison with the fragments amplified from the corresponding milk samples. Incidentally, this procedure permitted the identification of an entirely novel nucleotide sequence, the rbcL fragment of X-triticosecale, registered in GenBank with accession number FM202540.
Pairwise alignment between the rbcL DNA sequences amplified from plant material composing the diet fed to the animals and the sequence of the corresponding milk sample showed a 100% match of nucleotide identity. This is shown in Figure 2 for a representative portion of the multiple sequence alignment of the rbcL gene amplified from each of the 3 different species provided as monophytic diets and that of the corresponding milk samples. The chromatogram-based information obtained from each rbcL DNA sequence present in milk was readily comparable with the sequence of the corresponding diet. As such, the same few nucleotide changes that distinguish the rbcL sequence of oats, ryegrass, and X-triticosecale were detected in the DNA amplified from the related milk samples. This correlation was maintained for X-triticosecale whether the type of fodder that was provided to the goats was hay or silage. Table 2 presents the number and the related percentage of SNP substitutions resulting from the comparison of the rbcL and matK gene sequences of the 3 plant species used as diets. Table 3 shows the comparison between the number of unique polymorphic sites identified in the 348-bp milk-derived rbcL amplicons and the number of polymorphic sites present in the 1,383-bp rbcL fragments amplified from plant material of each specific diet. The rbcL DNA sequence detected in milk samples of goats fed ryegrass contained an ecotype-specific nucleotide mismatch when compared with the L. perenne rbcL sequence registered in GenBank (see Discussion section). In fact, this is the only ecotype-specific nucleotide change, of the 3 present in the rbcL 1,383-bp fragment amplified from ryegrass (Table 3, numbers in parentheses), which falls within the 348-bp rbcL fragment amplified from the total DNA derived from milk.
|
|
|
As is readily observable in Figure 2, the chromatogram profiles of the DNA sequences we obtained from the single diets and from the corresponding milk samples were neat, with no background and double peaks. This is different from the profiles observed in the plant DNA amplified from milk samples produced by cows fed a polyphytic diet. Figure 3 shows a portion of the nucleotide DNA sequence of the rbcL fragment that was amplified from cow milk samples obtained from a stock farm located in the province of Mantova. As was predictable from the polyphytic diet fed to the cows, several double peaks were evident in the chromatogram sequence, the most relevant of which are indicated by arrows. Consistently, these double peaks largely matched the SNP that were expected to be present because of the polyphytic fodder composition, known in the case of the Mantova stock farm samples, where cows were provided daily with the following ration composition: 64% corn, 20% alfalfa, 16% commercial feed (soybean and other unspecified plant components that provide a high protein value). Accordingly, single peaks were noted in the chromatogram region where nucleotide identity was 100% among the 3 different plant species that constituted the main components of the diet (Figure 3, from nucleotide +224 to +244). We could then observe double peaks, corresponding to those nucleotides that, according to the available sequence data, were different and could therefore identify 2 or 3 of the different plant species present in the diet. The thick arrow in Figure 3 points to nucleotide position +263, where the nucleotide polymorphism could not be assigned to any of the 3 plant species but is a likely contribution of one of the unknown components of the commercial mix fed to the cows. Similar data were obtained by sequencing the rbcL fragment amplified from samples of raw cow milk sold on the northern Italian market.
|
We attempted to define an assay for distinctiveness depending on a simple PCR run, with primer combinations that could possibly differentiate the 3 monophytic diets, based on the minimal differences found when comparing the 3 rbcL DNA sequences. Although we met with occasional success, our efforts to establish a simple rbcL PCR-based assay for distinctiveness were hampered by the high degree of nucleotide identity (more than 97%; National Center for Biotechnology Information (NCBI)/Nucleotide DataBase and Vector NTi tool (http://www.invitrogen.com/site/us/en/home/Products-and-Services/Applications/Cloning.html); see also Table 2) shared by the 3 DNA sequences. Polymerase chain reactions were successful but not discriminatory.
Exploration of both the literature and nucleic acid database (Hilu et al., 2003; NCBI/Nucleotide DataBase) suggested the use of matK, another single-copy chloroplast-based molecular marker gene. Multiple alignment of the 3 matK gene sequences of oats, ryegrass, and X-triticosecale allowed us to design a set of primers that could better discriminate ryegrass from the other 2 plant species because of the presence of 3 to 4 nucleotide mismatches in both the forward and reverse primers. The primers were also longer than previously used primers, 26 and 24 nucleotides, respectively, allowing for greater specificity (Figure 4A). A ryegrass-specific combination of matK primers (matK F/matK R; Table 1) was first tested for successful amplification on total DNA extracted from plant material that composed the 4 monophytic diets. Consistent with this, Figure 4B shows that the combination of ryegrass-specific matK primers was capable of exclusively amplifying a 294-bp matK fragment from the total DNA extracted from ryegrass. The same assay for ryegrass distinctiveness was then applied to total DNA extracted from each of the 4 different goat milk samples. As shown in Figure 4C, the 294-bp fragment was amplified only from milk-derived DNA samples obtained from goats fed ryegrass. Two different amounts, 200 and 400 ng, of total DNA per reaction were successfully used in all analyzed samples, with no appreciable difference in selectivity and no evidence of cross-reaction. The absence of amplification in the milk samples obtained from goats fed plant species different from ryegrass was due neither to the poor quality of the DNA nor to the absence of plant DNA because the generic rbcL 348-bp long fragment (Figure 4D) and the 453-bp β-CN fragment (data not shown) were successfully amplified from these samples.
|
A multiple alignment of the trnL plastidial sequences of several plant species, including those of interest for this work, was done preliminarily to design commonly shared PCR-oligonucleotide primers that correspond to conserved intron sequences flanking a region of variable length. This provides an easy assay for intron length polymorphisms (ILP; Supplemental Figure 1: http://jds.fass.org/content/vol92/issue11/; Clustal W program). Different protocols and preliminary experimental attempts eventually led us to rely on a previously described Plant2-forward primer (Einspanier et al., 2001) and a newly designed trnL-rv reverse primer that locates in a very conserved region of the intron, 3' proximal to the trnL exon II.
The use of this set of primers allowed for the amplification of the same noncoding region from the total DNA extracted from the 3 plant species supplied as fodder to our goat groups, yet the amplicon varied in size between each plant species. As shown in Figure 5A, oats, ryegrass, and X-triticosecale could easily be distinguished based on the ILP of their trnL-amplified fragment, which was 153, 367, and 400 bp, respectively. The 400-bp long trnL fragment specific for X-triticosecale was obtained from both DNA samples, whether extracted from hay or silage. When the same analysis was performed on the DNA extracted from the milk samples, we obtained the results shown in Figure 5B. The same ILP of the trnL marker found in the single-species diets were found and confirmed in the corresponding milk samples.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Milk is a complex matrix, particularly when trying to detect DNA fragments of plant origin undoubtedly present in trace amounts. To achieve this goal, several factors should be evaluated carefully, such as the complexity of the matrix, the presence of PCR inhibitors, and the amount and heterogeneity of the DNA fraction that originates from different sources, such as animal or bacterial cells. The use of chloroplast-based markers such as the rubisco (rbcL), the maturase matK, and the leucine tRNA (trnL) genes is certainly one of the factors that greatly facilitates plant DNA detection in milk. It takes advantage first of the specificity of the organelle, because chloroplasts are mainly restricted to plants, and second of the high copy number, ranging from 5,000 to 50,000 copies per cell. In addition, the plastidial genome of many crops has been fully sequenced, providing readily available information for the design of primers and related experimental approaches. For this purpose, the overall number of nucleotide substitutions proves to be sufficiently high, although sequence variation of the chloroplast genome is reported to be significantly lower compared with that of the plant nuclear genome. A high copy number of the target sequence was shown to be important for detection of plant DNA in milk, as was similarly reported by other laboratories that failed to detect either plant nuclear-based single-copy genes or nuclear-delivered transgenic sequences (Phipps et al., 2003, 2005; Rizzi et al., 2008). In accordance with those studies, we failed to amplify even short (250-bp) DNA fragments from the small nuclear plant β-tubulin gene family in the same milk samples from which the rbcL sequences were successfully amplified. In this regard, detection of nuclear-based plant DNA sequences can indeed be considered indicative of airborne contamination, as suggested by Agodi et al. (2006), because the plant material remains substantially intact and makes detection of even long fragments (more than 1 kb) feasible when using a highly sensitive protocol. This was also confirmed previously when using the combinatorial tubulin-based polymorphism (cTBP) method (Breviario et al., 2007) for the identification of trace amounts of single plant species in commercially sold fodder mixtures (D. Breviario, unpublished data).
The possibility of detecting DNA of plant origin in milk is further increased by the search for short-sized amplicons (less than 500 bp). The DNA of the diet undergoes extensive degradation during its route through the digestive tract of the animals; consequently, smaller sized fragments are more prevalent than larger fragments in both milk and animal fluids (Phipps et al., 2003). Our results corroborated this effect. First, we failed to detect 1,383-bp rbcL fragments in any of the milk samples analyzed. Second, a stronger amplification signal was consistently obtained when using the Arub fw/Arub rv set of primers compared with the Lrub fw/Lrub rv primer combination, even though the amplified fragment of the former is shorter and should, as a consequence, incorporate less ethidium bromide. However, nucleotide sequence specificities may also have contributed to this discrepancy. Nemeth et al. (2004) also reported a 10% decrease in the detection of rbcL sequences in milk when the sizes of the amplified fragments increased from 173 to 500 bp. Degradation of genomic DNA into small fragments is reported to occur progressively because of mechanical treatments and bacterial attacks on feed (Lutz et al., 2006). We compared X-triticosecale in 2 forms: as it was provided to goats as fresh hay, or as a 2-wk-old ensiled forage. However, in contrast to other reports (Einspanier et al., 2001; Phipps et al., 2005), we were able to amplify rbcL DNA fragments in both hay- and silage-derived milk samples, most likely because the time of silage was restricted to 2 wk. Consistently, the genomic DNA samples appeared only moderately degraded on agarose gels. By taking into consideration all the key factors involved and the resultant application of experimental condition adjustments, eventually we were able to obtain a frequency of 100% plant DNA detection in the goat milk samples.
Detection of feed-derived DNA fragments was also assessed in different raw cow milk samples either sold on the market or obtained from stock farms. These samples were checked successfully for the presence of rbcL DNA fragments. Although the percentage of detection (approximately 70% of the total analyzed samples) was lower with respect to goat milk, it was close to that reported by Nemeth et al. (2004). This may be explained either by a potentially higher level of contamination with PCR inhibitors or by the different physiology of the cow digestive tract (Doerfler, 2000). Even so, analysis of the rubisco DNA sequence data obtained from raw cow milk samples has made valuable information available concerning fodder composition. The chromatogram profile was not uniformly clear, as was evident for the goat milk. Few but significant overlapping single base peaks, representative of the different SNP expected to be present because of the multispecies composition of the diet, were detected. Where the nucleotide sequence was identical for all the components of the fodder, the chromatogram was clean and unambiguous, but double peaks emerged, corresponding to predicted species-specific nucleotide changes. In addition, DNA sequencing data allowed us to infer a priori the composition of the diet supplied to those cows that produce a brand-labeled raw milk commonly sold on the northern Italian market. Our diagnosis, based on the rbcL sequence data, was compatible with the information that was later recovered from the producer. This also highlights the possibility of using DNA sequencing for the identification and diagnosis of uncertified components (undisclosed on the fodder tag) possibly present as contaminants in animal feed.
To our knowledge, this is the first time that nucleotide sequencing data, with reference to components of the diet detectable in milk samples, have actually been shown. This potentially provides insight into the differences that can be found when comparing milk samples produced by animals fed different diets.
We have also shown and discussed our DNA sequence data with reference to the different diets consumed by the animal, whether polyphytic or monophytic. With regard to the latter, DNA sequences were found to be 100% identical when comparing the amplified rbcL DNA fragments of the goat milk sample with that of the corresponding diet. This was found for all 3 plant species investigated in the present study. Of additional relevance is the finding that the DNA amplified from milk samples contained the same few nucleotide changes that distinguish the rubisco DNA fragments of oats, ryegrass, and X-triticosecale. For oats, the amplified rbcL sequence was identical to the sequence registered in GenBank. For ryegrass, the rbcL DNA sequence amplified from the milk of goats fed ryegrass contained the same nucleotide change (at nucleotide position +334) found in the corresponding sequence of the ryegrass ecotype used as the diet. This sequence is different from the rbcL DNA sequence of the ryegrass reference variety in GenBank. In fact, this is 1 of 3 mismatches (the other 2 located at positions +678 and +1,144) we found when comparing our ryegrass variety with that registered in the database (NCBI, accession number NC009950) for the whole length of the rbcL 1,383-bp fragment (Tables 1 and 3). For X-triticosecale, no reference sequence was present in the database, as mentioned above.
The rubisco sequencing data demonstrate that DNA fragments of plastidial origin can be recovered equally in cow and goat milk samples and that DNA sequence analysis can be used to assess the composition of monophytic as well as polyphytic diets. Therefore, DNA sequencing can be used as a tool for traceability of plant sequences in milk.
The unambiguous nucleotide correlations found when comparing rbcL fragments amplified from the diet and the corresponding milk sample indicate that contamination did not interfere with our data. The occurrence of contamination or cross-reactions with rubisco sequences of other organisms, such as photosynthetic bacteria, fungi, and algae, was excluded by our data on matK. The latter is a gene that encodes for the only known maturase of higher plant plastids (Vogel et al., 1997), with no corresponding paralogous or orthologous sequences in organisms different from higher plants.
Although the issue regarding the length of feed-derived DNA fragments that can be detected in the animal gastrointestinal track is still ongoing (Chowdhury et al., 2004), and despite the precautions taken when collecting the milk samples, we cannot rule out with absolute confidence (100%) that negligible amounts of feces could have been accidentally included in the milk samples. Such a possibility, although much less probable than that of some DNA release during milking, would have resulted in greater variation in the detection of diet-specific DNA fragments in all the samples analyzed.
The alignment of the 3 rbcL DNA sequences obtained from the different plant species used to feed goats revealed the presence of sporadic SNP occurring within an overall nucleotide identity amounting to 97.4%. Despite this, we tried to exploit these SNP to determine PCR conditions that would permit us to selectively distinguish each of the 3 diets. However, specific detection was not consistently achieved. For this reason, it was necessary to use matK because the ryegrass sequence of this gene contains a higher number of nucleotide mismatches, which could be useful in designing selective primers.
The matK gene, located in the chloroplast genome within an intron of group II, nested between the 5' and the 3' exons of the trnK (tRNA-lysine) gene (Barthet and Hilu, 2008), represents an excellent tool for the selective identification of plant-specific DNA fragments. The reason is that its sequence variability allows for the design of distinctiveness assays based on PCR, a much more convenient and faster means of detection compared with DNA sequencing (see below).
With regard to the use of a matK-based distinctiveness assay, it must also be pointed out that detection is favored by the fact that the amount of DNA used for each PCR reaction can be 10-fold higher than that used for the rbcL marker. The greater number of polymorphisms in matK is likely a consequence of the higher mutational rate that has affected the matK gene sequence compared with rbcL (Soltis and Soltis, 1998), which is attributable to minor selective pressure on the former. Table 2 provides some data regarding the number and frequency of SNP substitutions, which better support this view and explain the success obtained using the matK-based primers for a ryegrass-specific distinctiveness assay. In fact, although the total number of SNP found when comparing the whole rbcL gene sequences of the 3 species amounted to 37, this number was increased to 143 when comparing the matK gene sequences. This discrepancy was also reflected in the frequency of SNP present in the PCR-amplified fragments, which ranged between 2.5 and 3.8% for the differently sized rbcL fragments, and increased to 13.3% for matK, thereby confirming that matK is potentially a better and more promising target for the development of PCR-based distinctiveness assays.
Feed traceability done using matK sequences successfully confirmed the unambiguous correlation between the milk and the diet from which the milk was eventually produced. The higher percentage of SNP present in the plant matK compared with the rbcL nucleotide sequence raises an important point for the future development of distinctiveness assays that aim to increase the number of species that can be detected.
In fact, this possibility was anticipated with the data obtained when using the intron of the trnL chloroplast-located gene as the source for detection of ILP. The trnL gene has already been used successfully for evolutionary studies (Tsai et al., 2006) and as a genetic marker (Taberlet et al., 1991; Gielly and Taberlet, 1994). Here, the different sizes of the trnL-amplified bands were more conveniently used for the identification of a plant species-specific length-based polymorphism in what was a true assay for distinctiveness.
Further improvements along the line of plant species traceability in milk can also be foreseen. This approach for multiple-species detection could effectively replace DNA sequencing, the success of which always depends on the clear output and reading of the multiple peaks found in the chromatogram.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
Received for publication March 25, 2009. Accepted for publication July 23, 2009.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGMENTSREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
