| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
What's this? |
Phytochemistry |
2Department of Plant Sciences, University of Arizona, Tucson, Arizona 85721 USA; and 3 Department of Medicine, Sarver Heart Center, University of Arizona, Tucson, Arizona 85721 USA
Received for publication April 18, 2005. Accepted for publication January 31, 2006.
ABSTRACT
Pharmacologically active ingredients in plants can cause significant morbidity through their increasingly common use in herbal alternative medicines and dietary supplements. Monitoring consumer products for the presence of toxic plants is encumbered by the lack of rapid and specific assays. To create a sensitive, reliable, fast, and broad-spectrum assay for medicinal or toxic plant species, we tested multiplexed ligation-dependent probe amplification (MLPA), which requires partial genomic DNA sequences from species of plants that are not well represented in currently available genetic databases. Genomic DNA was obtained from 21 species of medicinal and/or toxic plants. The PCR products were amplified from these plants and cloned for sequencing. The MLPA method was successful with DNA samples from many different species. The use of a microarray to facilitate screening of potentially thousands of plants in a single assay also was successful. The combination of the specificity of the MLPA assay with the broad-scale capabilities of microarray technology should make this an especially useful tool in screening in foods and commercial herbal preparations to identify the plant compounds actually present. Other applications could potentially extend to the identification of any plant species in samples for academic botanical studies and for biodefense and forensics applications.
Key Words: CODEHOP • cytochrome P450 • dietary supplement • medicinal plants • microarray • multiplexed ligation-dependent probe amplification (MLPA) • toxic plants
Plants are unique in their ability to produce an extraordinary array of different secondary metabolites. Many of these metabolites have medicinal and/or toxic attributes. The medicinal qualities of plants have been valued by man for centuries, and this tradition continues today in the increasing use of alternative medicines and dietary supplements in our society.
With increased consumption of herbal preparations comes an increased risk of toxic side effects, because of the intended and sometimes unintended plant species that may be found in these preparations. Many herbal preparations are derived from plants collected under uncontrolled conditions, and the collectors may have limited skill in the identification of plant species. Accidental poisonings occur because of mistakes and because of the unwitting use of related herbs that have different properties. The identification of herbs can be confused by the multiple and overlapping names used to describe them (Chan, 2003
). For example, toxicity was reported for infants accidentally exposed to Japanese star anise, which was confused with Chinese star anise, which is not toxic (Ize-Ludlow et al., 2004
).
A diverse range of plants may be included in herbal preparations, and many of them can have serious toxic side effects. Ephedra use has documented effects on cardiovascular health, including death, which has resulted in its sale being prohibited in the United States (Naik and Freudenberger, 2004
; Soni et al., 2004
). Some herbal preparations cause serious drug interactions, especially by altering expression of liver enzymes (Izzo et al., 2002
; Karliova et al., 2000
; Thabrew et al., 2004
). Alternative medicines and supplements have been linked also with hepatotoxicity, nephrotoxicity, neurotoxicity, carcinogenicity, and allergic reactions (Dasgupta, 2003
; Stedman, 2002
).
The use and study of herbal preparations would benefit from a comprehensive method for identifying the species of plants contained in them. To develop such a method, we chose the multiplexed ligation-dependent probe amplification (MLPA) method (Schouten et al., 2002
) in conjunction with microarray technology. The MLPA method uses a ligation reaction for strong specificity, coupled with a PCR amplification step for sensitivity. The ligation reaction joins two oligonucleotide probes that match one continuous sequence of plant genomic DNA. The PCR step uses a single pair of primers that match primer-binding sites added to the outer ends of the ligation probes. The MLPA has some capacity for multiplexing by combining multiple pairs of ligation probes in a single tube reaction. The addition of a microarray step simplifies the expansion of the number of plant species detected with the assay. The microarray consists of oligonucleotides that match each of the possible ligation products, so that these products may be identified upon hybridization. Because DNA microarrays can have tens of thousands of DNA oligonucleotides printed on them, they provide unmatched breadth for parallel testing.
For this method, species-specific genomic DNA sequences are needed for each plant to be identified. Many plant species have few gene sequences in the public databases or none at all. Some gene sequences are so well conserved between species that they are not useful for this application. Thus, gene samples are needed from many different species of plants. In our studies, PCR was used to amplify gene sequences from the various plant species to be included ultimately in a comprehensive assay. We chose to target the genes for cytochrome P450 enzymes, because of several advantageous properties. They are ubiquitous and heavily represented in plants. They are conserved within some regions of the molecule and widely divergent in others. Some of the cytochrome P450 enzymes are involved specifically in the synthesis of the very unique and diverse secondary metabolites that are of toxic or medicinal interest in some plants. We selected the CYP71 family of enzymes to target, because many of these, and their relatives within the CYP71 clan, are involved in the synthesis of defensive compounds and other secondary metabolites (Morant et al., 2003
; Nelson et al., 2004
).
To amplify cytochrome P450 gene fragments from plants, we used the consensus-degenerate hybrid oligonucleotide primers (CODEHOP) program to design the PCR primers (Rose et al. 2003
; Rose et al. 1998
). This program (http://blocks.fhcrc.org/blocks/codehop.html) uses blocks of conserved protein sequence to design primers and has been employed successfully to amplify plant cytochrome P450 sequences (Morant et al., 2002
). We applied the CODEHOP method to design PCR primers, amplified samples of genomic DNA, and cloned and sequenced the resulting PCR products. We used the sequences to design probes for the MLPA assay and successfully tested these probes. This approach is an effective way to rapidly accumulate sequence information from the long list of plants that will be included in a multiplexed assay.
Ultimately, CODEHOP may help rapidly accumulate hundreds or thousands of genomic DNA sequence samples from diverse plant species. The sequence samples will allow for designing MLPA probes for detecting the plants, and a custom microarray will facilitate parallel testing of potentially thousands of species in a single tube by providing a streamlined procedure for designing the probes and for reading the results.
MATERIALS AND METHODS
Plant species
Plants used traditionally as alternative medicines or as dietary supplements were included, as well as species that one would wish to exclude from these products: Aconitum napellus Herb., (Ranunculaceae); Arabidopsis thaliana L., (Brassicaceae); Artemesia absinthium L., (Asteraceae); Artemisia vulgaris L. (Asteraceae); Atropa belladonna L. (Solanaceae); Capsicum annuum var. glabriusculum L. (Dunal) Heiser & Pickersgill, (Solanaceae); Caulophyllum thalictroides L. Michx. (Berberidaceae); Citrus aurantium L. (Rutaceae); Datura metel L. (Solanaceae); Digitalis lanata Ehrh. (Scrophulariaceae); Echinacea angustifolia, DC., (Asteraceae); Ephedra viridis Coville, (Ephedraceae); Glycyrrhiza uralensis Fisch. ex DC. (Fabaceae); Hypericum perforatum L. (Clusiaceae); Lawsonia inermis L. (Lythraceae); Lobelia inflata L. (Campanulaceae); Mentha pulegium L. (Lamiaceae); Symphytum officinale L. (Boraginaceae); Tanacetum vulgare L. (Asteraceae); Teucrium canadense L. (Lamiaceae); Teucrium chamaedrys L. (Lamiaceae); Tussilago farfara L. (Asteraceae).
Cultivation of plants
Plant specimens and seeds were obtained from Companion Plants, Athens, Ohio, USA; Crimson Sage Nursery, Colton, Oregon, USA; and Richters Herbs, Goodwood, Ontario, Canada. Plants were grown in peat-based potting soil in artificial light. Watering schedules and light intensity were varied to provide conditions appropriate to each species.
Use of the CODEHOP program
Degenerate PCR primers were designed with the aid of the CODEHOP program. The primers were derived from 24 different protein sequences for CYP71, selected from diverse plant species. The program employs a codon usage table to reverse translate the amino acid sequences to DNA sequences. Six different codon usage tables were used to represent six different families of plants: Brassicaceae (Arabidopsis), Asteraceae, Rutaceae (citrus), Fabaceae, Scrophulariaciae, and Solanaceae. The codon usage table for each group (or a representative for the group) was used to generate 2–3 different primers each, for a total of 17 different primers (Table 1). The primers were generated from three different sections of the CYP71 sequence, all near the heme-binding domain of cytochrome P450. For each plant family represented, one primer was chosen to be a sense strand primer (those including the letter C), and two others to be antisense strand primers (D and E, Fig. 1). Each primer was chosen from a group of primer candidates provided by the CODEHOP program. The final sequence of each primer was determined after examining its suitability for PCR, with the aid of Vector NTI (Invitrogen, Carlsbad, California, USA). The primer pairs employed in these studies are given in Table 2.
|
|
|
Cloning and sequencing
The PCR products were cloned with the Zero Blunt Topo kit (Invitrogen, Carlsbad, California, USA). Bacterial colonies were screened by PCR using sequencing primer sites on the plasmid. The positive colonies were grown in CircleGrow medium (Qbiogene, Carlsbad, California, USA) and purified with the Montage Plasmid Miniprep96 kit (Millipore, Billerica, Massachusetts, USA). Bidirectional sequencing was performed by the Genomic Analysis and Technology Core of the University of Arizona. Raw sequences were assembled into contigs and trimmed with the help of ContigExpress/VectorNTI (Invitrogen, Carlsbad, California, USA). The resulting sequences were analyzed by Blastx analysis on the National Center for Biotechnology Information website (http://www.ncbi.nlm.nih.gov/BLAST). The sequences were compared with the AlignX program of VectorNTI (Invitrogen, Carlsbad, California, USA).
MLPA assay
The MLPA assay was performed as described by Schouten et al. (2002)
. Briefly, the MLPA probes were chosen to match the sequences obtained by sequencing. They were designed to have a melting temperature (Tmelt) of approximately 65°C, and PCR primer sequences (the same ones used by Schouten et al. [2002
]) were added to the 5' end of one MLPA probe, and the 3' end of the other. The genomic DNA sample (e.g., 50–200 ng) and the MLPA probes (3.75 fmol) were incubated together in 0.28 mol/L KCl, 0.047 mol/L Tris-HCl pH 8.5, and 0.19 EDTA in 8 µL at 98°C for 5 min, then for 16 h at 65°C. For the ligation step, the following MLPA kit components (MRC-Holland, Amsterdam, Netherlands) were combined with 25 µL water: SALSA PCR buffer 3 µL, NAD buffer 3 µL, and 1 µL Ligase-65. This mixture was added to the annealed probe–genomic DNA mixture and incubated for 15 min at 54°C, followed by 98°C for 5 min. A 10-µL sample of the ligation mixture was transferred to 40 µL of a PCR reaction mixture. The PCR mixture contained Surestart Taq DNA polymerase (Stratagene, La Jolla, California, USA), the appropriate buffer components, and 2.5 µL each of 10 µmol/L stocks of the PCR primers. The PCR protocol consisted of 95°C for 9 min, 34 cycles of (94°C 45 s, 60°C 1 min, 72°C 1 min), followed by 72°C for 7 min. The PCR products were separated on 3% NuSieve (Cambrex East Rutherford, New Jersey, USA), 1% agarose gels.
Microarray printing
A small microarray was created by printing 70-mer oligonucleotides in 3x sodium chloride/sodium citrate buffer (SSC) on aminosilane-coated glass slides (Telechem International, Sunnyvale, California, USA) with an Omnigrid 100 (Genomic Solutions, Ann Arbor, Michigan, USA). The array oligos consisted of two specific oligos, and seven randomly chosen Arabidopsis genome oligos from an Array-Ready Oligo set (Operon Biotechnologies, Huntsville, Alabama, USA). The specific oligos represented the specific portions of the plus strand of the MLPA products for Datura metel and Lobelia inflata. The oligo sequences for Arabidopsis thaliana are proprietary, but they were designed for the following genes: at3g01700, at4g00810, at1g44150, at5g57480, at3g22310, at5g57480, and at1g29430.
Microarray hybridization
An MLPA was performed with oligos for the following plants: Artemesia absinthium, Atropa belladonna, Caulophyllum thalictroides, Datura metel, Digitalis lanata, Echinacea angustifolia, Glycyrrhiza uralensis, Hypericum perforatum, Lobelia inflata, Mentha pulegium, Symphytum officinale, Tanacetum vulgare, Teucrium canadense, T. chamaedrys, and Tussilago farfara, and the genomic DNA for Datura metel alone. The PCR step of the MLPA was modified to include a reverse primer that was Cy5-labeled. The PCR product was purified and used for hybridization with the small microarray described. Before hybridization, the array was rehydrated over a 55°C water bath four times for ~4–5 s each. They were then cross-linked to the surface with UV treatment three times with 120 mJ each time. The array was washed with 0.1% SDS for 10 min, then rinsed in nuclease-free water, and dried with a nitrogen jet. The hybridization reaction was performed with the PCR product in 90 µL of 2x SSC, 6% Liquid Block (Amersham Biosciences, Piscataway, New Jersey, USA), and 0.1% SDS. The mixture, after boiling for 2 min, was applied by capillary action under a 24 x 60 mm Lifterslip (Erie Scientific, Portsmouth, New Hampshire, USA) positioned on a microarray. After incubating in the dark at 58°C for 5 h, the microarray was washed with 2x SSC, 0.5% SDS at 58°C for 5 min. Two subsequent washes were at room temperature with 0.5x SSC. After drying, the slide was scanned at 635 nm in a Genepix 4200AL scanner (Molecular Devices, Sunnyvale, California, USA).
RESULTS
Collection of genomic DNA sequence samples from targeted plants
The PCR was used to amplify portions of the cytochrome P450 genes from target plant species to aid in their identification. In the absence of specific sequence information for the targeted species, the CODEHOP program was employed to design degenerate primers, starting with the protein sequences available for other plant cytochrome P450 genes. Different combinations of primers (Tables 1, 2) and genomic DNA were used to increase the odds of producing a useful product, but emphasis was placed on pairing the genomic DNA with primer sequences produced with the codon usage table for a related plant species. Typically, approximately 25% of the reactions produced products of the appropriate size for the expected cytochrome P450 product. When paired with a C primer, the D and E primers (Fig. 1) produced products of approximately 190 and 420 bp respectively.
After the PCR products were isolated and cloned, the resulting plasmid inserts were sequenced. A summary of the sequenced PCR products is given in Table 3. All of the sequences were trimmed to remove primer sequences. Some of the sequences represented were truncated, because of nonsense or unlikely stop codons in their proposed translations. The 32 PCR product sequences represent 21 different species. Information from submitting the sequences to Blastx analysis indicates that almost all of the sequences are likely derived from plant cytochrome P450 genes. Even for the sequences with the higher expected values (e.g., 0.0004 for the Capsicum annuum sequence), the best match available in the databases is a cytochrome P450 gene. One sequence, obtained from Ephedra nevadensis, had no significant match using Blastx (not shown).
|
), including the CYP98 family in plants (see Fig. 2) and the CYP76, CYP80, and CYP93 families. The alignment in Fig. 2 shows the high level of homology within all five sequences that matched CYP98 family members by Blastx analysis. By comparison, alignments for sequences that matched the CYP71, CYP76, and CYP93 families were significantly more divergent.
|
Design and testing of MLPA probes for targeted plants
The sequence samples obtained from the various plant species were used in the design of probes for the MLPA assay. Primer sequences were taken, as much as possible, from the least conserved sections of each sequence. They were designed to have a Tmelt of approximately 65°C in 100 mmol/L NaCl. Hairpin and dimer formations were avoided. The probes were tested for their suitability for identifying specific target plants in a series of experiments.
The MLPA probes (Table 4) were tested in MLPA assays with genomic DNA from different species. Initially, the probes were tested with genomic DNA from the appropriate species and with Arabidopsis thaliana genomic DNA as a negative control. Arabidopsis thaliana DNA was tested also with MLPA probes designed specifically for Arabidopsis thaliana to provide a positive control. The results of two such experiments are given in Table 5. All the ligase-dependent PCR products for these experiments were typically 110–120 bp (Fig. 3), because they represent the ligation product of two MLPA probes, which were from 45 to 60 bp each. The reaction with the Arabidopsis probes and Arabidopsis genomic DNA yielded a product. Otherwise, all but one MLPA probe set produced a product with their cognate plant genomic DNA, but not with Arabidopsis DNA. The exception was for the Symphytum officinale probes, which did not produce a product at all.
|
|
|
|
|
The CODEHOP method (Rose et al., 2003
) for degenerate PCR primer design was developed for amplifying unsequenced genes with homology to known sequences. The method starts with a group of related protein sequences and uses conserved blocks of aligned protein sequences to design PCR primers. To accomplish this, CODEHOP employs a user-selected codon usage table to reverse-translate the protein sequences. The CODEHOP-derived primers have been used previously to amplify successfully cytochrome P450 genes in plants (Morant et al., 2002
). We felt that this would be an ideal method and an ideal family of proteins to use for finding representative sequence tags for species of plants underrepresented in the databases. We chose the CYP71 family of proteins for their ubiquity in plants and because many members of the extended CYP71 clan function in the synthesis of secondary metabolites related to defense, as opposed to the more conserved metabolic functions found in all vascular plants (Nelson et al., 2004
).
We chose as a starting place a well-conserved section of the cytochrome P450 molecule, the heme-binding domain. Indeed, this was the only section of the CYP71 sequence that generated useful primers with the CODEHOP program, without excessively increasing the degeneracy (32–64 or less was preferred) or using a relatively high strictness number (typically 0.4 or less). A higher degeneracy essentially reduces the effective concentration of the PCR primers. A higher strictness number means that the primers are increasingly biased towards the most common base in each position and could limit severely the range of sequences that the primers could amplify. We began by using a limited number of plant codon usage tables to provide a collection of primers to test.
The CODEHOP primers worked very well at producing PCR products. Although not all combinations of primers and genomic DNA samples produced products immediately, 89% of the genomic DNA samples produced products without a need for altering our initial PCR conditions. We were successful at cloning most of these for sequencing.
A Blastx analysis of the PCR product sequences indicated that some likely represented CYP71 family members. Other products were probably from families that are considered part of a greater CYP71 clan, including CYP76, CYP98, and CYP93.
The sequences allowed us to test the MLPA method for identifying the species of origin of samples containing plant genomic DNA. The MLPA method was consistently successful at recognizing genomic DNA from the appropriate species. Preliminary tests for each probe set against DNA samples from Arabidopsis thaliana and two other species of plants demonstrated that 11 of 15 of the MLPA probes were indeed specific to one plant species (within the limitations of our tests). Four of the ligation probe sets tested recognized a DNA sample from a species other than that intended. In three of these cases, another plant in the same genus produced a product. This reveals a predictable problem in trying to construct a reliable assay for any species. Related species may be very difficult to distinguish genetically, unless sequence information is found that helps distinguish one related species from another. The weaker PCR bands that resulted for these false positives indicate that raising the annealing temperature for the MLPA could improve the selectivity of these assays. For the T. vulgare probes, DNA harvested from plants of the same family (Asteraceae) produced false positives, thus indicating that these probes are inadequate for use in plant identification.
A way to improve the specificity of the ligation probes is to use sequence information from less-conserved sections of a gene. In these studies, our probes were all derived from the protein sequences that surround the heme-binding region of CYP71 family members. This is a relatively well-conserved section of the cytochrome P450 molecule, because of its clear importance to enzymatic function. If a section of the molecule closer to the amino terminus were used to produce PCR primers, then longer, less-conserved stretches of sequence could be obtained. To produce effective primers from other regions of the molecule may require the acceptance of a higher level of degeneracy in the PCR primers than tolerated in these experiments. The CODEHOP primers have some ability to compensate for degeneracy of the primers, because all primers include a long stretch of sequence in common on the 5' end, which enables all primers in a degenerate mixture to anneal to amplification products of the other primers, at least at lower annealing temperatures.
Another way to improve the specificity of ligation probes would be to avoid using sequences that likely derive from well-conserved types of cytochrome P450. For example, five different PCR products derived from different plants were very close matches by Blastx analysis to members of the CYP98 family, which are coumaroyl ester 3-hydroxylases (Morant et al., 2002
; Gang et al., 2002
). This group of enzymes is important to the synthesis of lignin monomers and therefore has a conserved, primary function for the plant. Over time, as more sequences are entered in the public databases, we may be able to improve our arsenal of CODEHOP primers, so that they are less likely to amplify sequences from such conserved genes. The list of sequences that may be used to generate primers with CODEHOP increases and may provide for more precise primer design to avoid the more-conserved members of the broad, cytochrome P450 family. Similarly, we may learn to avoid primers that are likely to produce less-useful sequences. In our tests, the P12 primer pair was used to amplify all five of the CYP98 family members in Fig. 2. This primer pair was also by far the most productive one for these experiments, having produced almost half of the cloned sequences. Minor modifications of the P12 primer pair may help avoid the amplification of CYP98 sequences. Ultimately these improvements could help speed the scale-up of this method to a higher throughput.
The MLPA method itself is a good choice for this application, because it uses for selectivity long probes at relatively high annealing temperatures, combined with a strict requirement for a perfect match between probes and target DNA at the point where the two probes meet. Ideally, the point in the genomic DNA sequence chosen for the joining of the probes should be poorly conserved and thus more likely unique for each species.
Applications for MLPA assays typically employ multiple probe sets, and the ligation probes are designed to include a spacer between the PCR primer-binding site and the gene-specific probe sequence (Schouten et al., 2002
). In multiplexed assays, the resulting final products are different sizes so that they may be distinguished by agarose electrophoresis. The discrimination of different size bands by electrophoresis is potentially a strong limitation of the multiplexing of MLPA assays. In our studies, we did not include these spacer sequences and plan to use instead a microarray step to identify any or all amplified sequences in a multiplexed assay. This modified method would use a custom microarray with oligo DNA probes that match perfectly the gene-specific portions of the ligated probes. This approach should make it much easier to design and produce MLPA probes and may enable us to use a higher degree of multiplexing in the ultimate plant diagnostic assay planned.
We tested the concept of using a microarray with the MLPA assay in a simple experiment that included 15 sets of probes in a single tube and genomic DNA from one species. The resulting fluorescent scan clearly indicated that the MLPA assay can be used to test for multiple species simultaneously and that a microarray can distinguish which species is being recognized by the assay. The test study also demonstrated that this method could identify an alternative medicine or other consumer herbal preparation containing potentially toxic Datura plant leaves rather than, or in addition to, the intended plant materials. In the future, this assay could simultaneously monitor for hundreds or thousands of other toxic plant species.
We plan to continue expanding the number of plant species that can be tested in this assay. As more unique sequences are obtained for diverse plant species, more oligos can be added to the custom microarray used for the final readout. We expect that with development, this method will provide a critical tool for monitoring the content of herbal or alternative medicines and dietary supplements. This tool should help protect the consumer and facilitate more reliable research into the efficacy of these products. The combined specificity and scalability of this method means that it could be applied also to identification of plant materials for general botanical studies or studies in archaeology, paleontology, for biodefense, or for forensic studies.
Conclusions
Degenerate, consensus, PCR primers were used to amplify partial cytochrome P450 sequences from 20 different species of plants used in preparation of alternative medicines or considered potential contaminants of these preparations. The resulting PCR products were cloned and sequenced. The sequences were used to design probes for use in MLPA assays for identification of plant specimens that contain genomic DNA. Tests of the MLPA probes indicate that they can be used to identify plant specimens, although care must be taken to consider that some probes may not be specific to one unique species, but may recognize closely related species. When many MLPA probe sets are used in a single tube assay, a microarray can be used to determine the genetic source of genomic DNA in a sample. This method can be applied ultimately to the identification of the source of any sample that contains plant genomic DNA.
FOOTNOTES
1 The authors thank the Center for Disease Control for financial support of this research (Grant no. H75/CCH923339-01) and Payam Morgan for his technical support. 
4 Author for correspondence (rogerab{at}ag.arizona.edu
)
LITERATURE CITED
Chan K.. 2003. Some aspects of toxic contaminants in herbal medicines. Chemosphere 52: 1361-1371.[Medline]
Dasgupta A.. 2003. Review of abnormal laboratory test results and toxic effects due to use of herbal medicines. American Journal of Clinical Pathology 120: 127-137.[CrossRef][Web of Science][Medline]
Gang D. Beuerle T. Ullmann P. Werck-Reichhart D. Pichersky E.. 2002. Differential production of meta hydroxylated phenylpropanoids in sweet basil peltate glandular trichomes and leaves is controlled by the activities of specific acyltransferases and hydroxylases. Plant Physiology 130: 1536-1544.
Ize-Ludlow D. Ragone S. Bruck I. Bernstein J. Duchowny M. Garcia-Peña B.. 2004. Neurotoxicity in infants seen with consumption of star anise tea. Pediatrics 114: 653-656.[CrossRef]
Izzo A. Borrelli F. Capasso R.. 2002. Herbal medicine: the dangers of drug interaction. Trends in Pharmacological Sciences 23: 358-359.[CrossRef][Medline]
Karliova M. Treichel U. Maolago M. Frilling A. Gerken G. Broelsch C.. 2000. Interaction of Hypericum perforatum (St. John's wort) with cyclosporin A metabolism in a patient after liver transplantation. Journal of Hepatology 33: 853-855.[CrossRef][Web of Science][Medline]
Morant M. Bak S. Moller B. Werck-Reichhart D.. 2003. Plant cytochromes P450: tools for pharmacology, plant protection and phytoremediation. Current Opinion in Biotechnology 14: 151-162.[CrossRef][Web of Science][Medline]
Morant M. Hehn A. Werck-Reichhart D.. 2002. Conservation and diversity of gene families explored using the CODEHOP strategy in higher plants. Plant Biology 2: 7-12.[CrossRef][Medline]
Naik R. Freudenberger S.. 2004. Ephedra-associated cardiomyopathy. Annals of Pharmacotherapy 38: 400-403.
Nelson D. Schuler M. Paquette S. Werck-Reichhart D. Bak S.. 2004. Comparative genomics of rice and Arabidopsis: Analysis of 727 cytochrome P450 genes and pseudogenes from a monocot and a dicot. Plant Physiology 135: 756-772.
Rose T. Schultz E. Henikoff J. Pietrokovski S. McCallum C. Henikoff S.. 1998. Consensus-degenerate hybrid oligonucleotide primers for amplification of distantly related sequences. Nucleic Acids Research 26: 1628-1635.
Rose T. Henikoff J. Henikoff S.. 2003. CODEHOP (Consensus-degenerate hybrid oligonucleotide primer) PCR primer design. Nucleic Acids Research 31: 3763-3766.
Schouten J. McElgunn C. Waaijer R. Zwijnenburg D. Diepvens F. Pals G.. 2002. Relative quantification of 40 nucleic acid sequences by multiplex ligand-dependent probe amplification. Nucleic Acids Research 30: e57.
Soni M. Ioana G. Carabin I. Griffiths J. Burdock G.. 2004. Safety of ephedra: lessons learned. Toxicology Letters 150: 97-110.[CrossRef][Web of Science][Medline]
Stedman C.. 2002. Herbal hepatotoxicity. Seminars in Liver Disease 22: 195-206.[CrossRef][Web of Science][Medline]
Thabrew I. Munasinghe J. Chackrewarthi S. Senarath S.. 2004. The effects of Cassia auriculata and Cardiospermum halicacabum teas on the steady state blood level and toxicity of carbamazepine. Journal of Ethnopharmacology 90: 145-150.[CrossRef][Web of Science][Medline]
Umemoto N. Kobayashi O. Ishizaki-Nishizawa O. Toguri T.. 1993. cDNAs sequences encoding cytochrome P450 (CYP71 family) from eggplant seedlings. Federation of European Biochemical Societies Letters 330: 169-173.[CrossRef][Web of Science][Medline]
CiteULike Complore Connotea Del.icio.us Digg Facebook Reddit Technorati Twitter What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
