June 2006 Biol. Pharm. Bull. 29(6) 1207—1211 (2006) 1207

Sequence Analysis of Chloroplast chlB Gene of Medicinal Species and Its Application to Authentication of Ephedra Herb

a a b b b Yahong GUO, Ayako TSURUGA, Shigeharu YAMAGUCHI, Koji OBA, Kasumi IWAI, c,1) ,a Setsuko SEKITA and Hajime MIZUKAMI* a Graduate School of Pharmaceutical Sciences, Nagoya City University; 3–1 Tanabe-dori, Mizuho-ku, Nagoya 467–8603, Japan: b Research and Development Department, Asgen Pharmaceutical Co., Ltd.; 2–28–8 Izumi, Higashi-ku, Nagoya 461–8531, Japan: and c Tsukuba Medicinal Research Station, National Institute of Health Sciences; 1 Hachimandai, Tsukuba, Ibaraki 305–0843, Japan. Received December 26, 2005; accepted February 15, 2006

Chloroplast chlB gene encoding subunit B of light-independent protochlorophyllide reductase was amplified from herbarium and crude drug specimens of Ephedra sinica, E. intermedia, E. equisetina, and E. przewalskii. Se- quence comparison of the chlB gene indicated that all the E. sinica specimens have the same sequence type (Type S) distinctive from other species, while there are two sequence types (Type E1 and Type E2) in E. equisetina. E. intermedia and E. prezewalskii revealed an identical sequence type (Type IP). E. sinica was also identified by di- gesting the chlB fragment with Bcl I. A novel method for DNA authentication of Ephedra Herb based on the se- quences of the chloroplast chlB gene and internal transcribed spacer of nuclear rRNA genes was developed and successfully applied for identification of the crude drugs obtained in the Chinese market. Key words chloroplast chlB; DNA authentication; Ephedra Herb; polymerase chain reaction-restriction fragment length poly- morphism

Ephedra Herb is an important crude drug which has been nucleotide deletions were present in the trnL/trnF spacer of used in Chinese and Japanese traditional (Kampo) medi- E. sinica.5) Thus, neither ITS of rDNA nor the trnL–trnF re- cines. Although about 35 species belonging to the genus gion is suitable as a DNA marker to identify medicinal Ephedra are distributed in Eurasia, North America, and Ephedra species by itself. South America, only a few species contain pharmacologi- The chlB gene encodes a subunit B of light-independent cally active ephedrine alkaloids and are used for medicinal photochlorophyllide reductase that catalyzes reduction of purposes. The Japanese Pharmacopoeia defines Ephedra protochlorophyllide to chlorophyllide in chlorophyll biosyn- Herb as dried aerial parts of Ephedra sinica, E. intermedia thesis, and is located in the chloroplast genome of gym- and E. equisetina.2) Morphological and anatomical character- nosperms, algae and photosynthetic bacteria but not in an- istics have long been investigated to identify and/or discrimi- giosperms.7) The chlB has a higher evolutionary rate than the nate these species since Konoshima described the morphol- rbcL (large subunit of ribulose bisphosphate carboxylase/ ogy of E. equisetina collected in .3) However, because oxygenase) gene which has been widely used as a molecular of their relatively simple organization it is sometimes diffi- marker in plant phylogenetic analyses,8) and might be suit- cult to identify original species of Ephedra Herb, especially able as a DNA marker for identification and/or discrimina- when they do not bear flowers or fruits. Chemical identifica- tion of Ephedra species and the crude drugs derived there- tion of Ephedra mainly based on their alkaloid compo- from. sitions or HPLC fingerprints has also been attempted,4) but In the present investigation we analyzed the nucleotide se- such chemical traits are inevitably affected by environmental quences of the chloroplast chlB gene amplified from the and intra-specific variations. herbarium and crude drug specimens of Ephedra plants and For correct identification of biological materials, DNA established a novel protocol for DNA authentication of profiling (DNA-based polymorphic assay) has several advan- Ephedra Herb based on their chlB and rDNA ITS 1 se- tages over morphological and chemical analyses because quences. Furthermore, we showed that the protocol could be genotypes rather than phenotypes are directly assayed, and so successfully applied to identify the original species of that the results are not affected by environmental factors. Ephedra Herb obtained in the Chinese market. Long et al. compared nucleotide sequences of intergenic transcribed spacer (ITS) 1 and ITS2 of nuclear ribosomal MATERIALS AND METHODS DNA (rDNA) and of the chloroplast trnL–trnF region includ- ing trnL intron, 3-exon of trnL, and trnL/trnF spacer from Herbarium and Crude Drug Specimens Ten herbar- eight Ephedra species.5) They found that the ITS sequence of ium specimens and 22 crude drug specimens were selected E. sinica was identical with that of E. intermedia, whereas from the collection of Central Research Laboratories, there were several polymorphic sites between these species Tsumura & Co. (Table 1). Their morphological features were and E. przewalskii. However, most of these polymorphic sites re-evaluated based on the anatomical traits such as presence were later found to be subjected to intra-specific variation or absence of pith and of fiber bundles in the cortex and within either E. intermedia or E. przewalskii.6) As for the abundance of cuticles in the epidermis to confirm their chloroplast genome there were no polymorphic nucleotide species. The pharmacognostical evaluation of the samples sites in the trnL intron among E. sinica, E. intermedia and E. was performed by Drs. H. Yamaji, K. Kondo and S. Tera- przewalskii. The only difference in these species was that two bayashi of Central Research Laboratories, Tsumura & Co.

∗ To whom correspondence should be addressed. e-mail: [email protected] © 2006 Pharmaceutical Society of Japan 1208 Vol. 29, No. 6

Table1. Herbarium and Crude Drug Specimens Used in the Present Investigation

Sample Plant Voucher Place of Date of Herbarium specimen No. name ID collection collection or crude druga)

1 Ephedra sinica THS 41925 Ibaraki, Japan (cultivation) 1999.6 HS 2 THS 42471 1999.8 HS 3 THS 43342 2000.8 HS 4 THS 42443 Liaoning 1984.9 HS 5 THS 42440 Liaoning 1985.6 HS 6 THS 42532 Liaoning 1999.9 HS 7 TMS 10069 1984.12 CD 8 TMS 11094 1987.12 CD 9 TMS 11624 Jilin 1989.6 CD 10 TMS 13963 Liaoning 1994.1 CD 11 TMS 14139 Inner Mongolia 1995.6 CD 12 TMS 14697 Inner Mongolia 1996.4 CD 13 TMS 16929 Shanxi 1999.8 CD 14 TMS 8739 Inner Mongolia 1983.6 CD 15 E. intermedia THS 32595 1988.9 HS 16 TMS 14695 Qinghai 1996.4 CD 17 TMS 16404 Gansu 1998.4 CD 18 TMS 8585 Gansu 1982.12 CD 19 TMS 8586 Gansu 1982.12 CD 20 TMS 8587 Gansu 1982.12 CD 21 TMS 8686 Gansu 1983.6 CD 22 TMS 8687 1983.6 CD 23 TMS 17483 Heilongjiang 2000.1 CD 24 E. equisetina THS 40637 Xinjiang 1997.8 HS 25 TMS 10135 Russia 1985.1 CD 26 TMS 16416 Russia 1983.1 CD 27 TMS 6874 1982.2 CD 28 TMS 8378 Russia 1982.1 CD 29 E. przewalskii THS 32593 Gansu 1988.9 HS 30 THS 40638 Xinjiang 1997.8 HS 31 TMS 6469 Xinjiang 1989.6 CD 32 TMS 8314 Gansu 1984.12 CD

a) HSherbarium specimen; CDcrude drug. and voucher samples of these specimens were deposited in Central Research Laboratories, Tsumura & Co. For DNA au- thentication, 21 crude drug samples were obtained in the Chinese market during the period between 1993 and 2003. The voucher samples of the crude drugs were stored in the Research and Development Department, Asgen Pharmaceuti- cal Co., Ltd. Preparation of DNA Total DNA was prepared from the powdered samples (about 40 mg) using a DNeasy Plant Mini Kit (Qiagen). DNA content in the preparation was estimated by a DyNA Quant 200 Fluorometer (Amersham) using calf thymus DNA (Sigma) as a standard. Amplification of chlB Gene Polymerase chain reaction Fig. 1. Sequence Strategy of the Chloroplast chlB Genes from Ephedra (PCR) primers (1F and 4R) were designed based on the chlB Species gene sequence of E. altissima retrieved from the DDBJ/ Primers 1F to 4F and 1R to 4R indicate the annealing position of the forward and re- EMBL/Genebank Nucleotide Sequence Database (accession verse, respectively, primers for PCR amplification of the chlB gene and its fragments. no. U21315) so that the whole open reading frame (ORF) of the chlB gene can be amplified. In addition, the 1.5 kb chlB GTTTTCCCAGTCACGACACTGTTGTTTTTGGTGA- ORF was also divided into four partially overlapped frag- TGC-3; 1R5-ATTTAGGTGACACTATAGAATACCC- ments (Fragment 1 to 4) of about 0.5 kb and each fragment GATGATATTTACAGAAGG-3;2R5-ATTTAGGTGA- was individually amplified using internal primers (2F, 3F, 4F, CACTATAGAATACTGAAAACCAAACAGCTTGGG-3; 1R, 2R, and 3R) designed based on the chlB sequence deter- 3R5-ATTTAGGTGACACTATAGAATACTCGTACCC- mined. The nucleotide sequences of the PCR primers were as CATAAAAGGACG-3; 4R5-ATTTAGGTGACACTATA- follows: 1F5-GTTTTCCCAGTCACGACATGAAATT- GAATACCAATTGTAATCTTAACTACACCC-3. The un- AGCTTATTGGAT-3; 2F5-GTTTTCCCAGTCACGAC- derlined sequence in the forward and reverse primers shows ATTCAAGCAGCAGATAGGAC-3; 3F5-GTTTTCCC- the M13 universal forward primer and SP6 primer sequence, AGTCACGACTCGGTGAATCCTTTTGCTTC-3; 4F5- respectively, for direct sequencing of the PCR products. The June 2006 1209 approximate positions where the primers anneal to the tem- and the filtrate was subjected to HPLC column (Develosil plate DNA are shown in Fig. 1. ODS-5, 4.6150 mm, Nomura Chemicals) with isocratic The PCR mixture contained 10 mM Tris–HCl (pH 8.8), elution by 27 mM sodium lauryl sulfate/acetonitrile/phos- 25 mM potassium chloride, 5 mM ammonium sulfate, 2 mM phoric acid (640 : 360 : 1). Flow rate was 1.3 ml/min and the magnesium sulfate, 0.2 mM each dNTP, 0.4mM of each elution was monitored at 210 nm. primer, 0.4 ng/ml template DNA and 0.02 unit/ml Taq DNA Polymerase (Roche). Amplification was carried out under the RESULTS following conditions: precycling at 94 °C for 3 min, 30 cycles of denaturation at 94 °C for 30 s, annealing at 55 °C for 30 s Sequence Analysis of the chlB Fragments Amplified and elongation at 72 °C for 1 min, and then final elongation from Medicinal Ephedra Species Ten herbarium speci- at 72 °C for 10 min. PCR products were detected by ethidium mens and 22 crude drug specimens of Ephedra sinica, E. in- bromide staining following 2% agarose gel electrophoresis termedia, E. equisetina, and E. przewalskii whose identities using 5 ml of the PCR mixture. were confirmed based on their morphological characteristics Sequencing of the PCR Products Usually a 10 ml were sampled from the collection of Central Research Labo- aliquot of the PCR mixture was incubated with 5 ml of Ex- ratories, Tsumura & Co. oSAP-IT solution (Amersham) at 37 °C for 15 min and then PCR products containing the whole ORF of the chlB gene at 80 °C for 15 min. A 7 ml aliquot of the resulting mixture were amplified from some of the samples using a PCR was used for sequencing reaction using Thermo Sequenase primer set (1F and 4R) designed based on the chlB sequence Primer Cycle Sequencing Kit (Amersham) and the nu- of E. altissima, and their nucleotide sequences were deter- cleotide sequence was determined for both strands with mined. The sequences thus obtained were used for designing DSQ-2000L (Shimadzu). internal PCR primers. In some samples, DNA was highly de- PCR-Restriction Fragment Length Polymorphism graded into small fragments and did not give a PCR product (RFLP) Analysis Usually a 5 ml aliquot of the PCR mix- when DNA preparations from these samples were used as ture was digested with 10 unit of Bcl I at 50 °C for 1 h in a templates. In such cases the chlB gene was divided into four final volume of 10 ml. The digested fragments were separated fragments (Fragment 1 to 4, as shown in Fig. 1) of about by agarose gel electrophoresis followed by ethidium bromide 500 bp, which were individually amplified using the internal staining. primer sets and sequenced. The nucleotide sequence of the Amplification and Sequence Determination of ITS1 of chlB was determined by aligning the sequences of the four rDNA For PCR amplification of the 3-end of the ITS1 re- fragments. gion from Ephedra Herb obtained in the market, we used the Comparison of the nucleotide sequences indicated that following primers: ITS-F5-GTTTTCCCAGTCACGACC- there are four genotypes in the chlB gene amplified from four GGCGGACGCGGTT-3; ITS-RATTTAGGTGACACTA- Ephedra species as shown in Table 2. All of the 14 samples TAGAATACCCCCGCGGCCGACGA-3. The underlined se- identified as E. sinica exhibited Type S. The samples identi- quence in the forward and reverse primer shows the M13 fied as E. intermedia (nine specimens) or E. przewalskii (four universal forward primer and SP6 primer sequence, respec- specimens) had the identical genotype, Type IP. In contrast, tively, for direct sequencing of the PCR products. The ITS-F four of five E. equisetina specimens exhibited Type E1 and and ITS-R primers were designed to anneal at nucleotides one revealed Type E2. E. equisetina specimens with Type E1 835 to 848 and 926 to 912, respectively, of the ITS1 region genotype were from either Xinjiang or Russia and that with of E. intermedia (accession no. AY394070) and E. przewal- Type E2 genotype was from Shaanxi. The chlB sequences of skii (accession no. AY394072) and expected size of the PCR six other Ephedra species: E. campulopoda, E. gerardiana, products is about 100 bp. PCR amplification and sequencing E. minuta, E. pachyclada, E. procea, E. regerianaand were of the PCR products were carried out as described above for also determined to find that their chlB sequences were differ- the chlB gene. ent from any of the above-mentioned genotypes, although HPLC Analysis of Ephedrine Alkaloids Quantitative identity of these species was not confirmed (data not shown). analysis of ephedrine alkaloids was performed according to Identification of E. sinica by PCR-RFLP Fragment 1 The Japanese Pharmacopoeia.2) Briefly, the powdered crude of the chlB comprises about 500 bp from the 5-end of the drug (ca. 0.5 g) was extracted with 20 ml of 50% methanol, gene (Fig. 1). A Bcl I restriction site is present at nt. 243—

Table2. Sequence Comparison of the Chloroplast chlB Genes from Four Ephedra Speciesa)

Nucleotide positionb)

11111 124901334 Sequence Ephedra species 7468441183 type 8276879597

Type S C A C G C A A T T C E. sinica Type IP C A T G C A A T T C E. intermedia or E. przewalskii Type E1 G C T C T A CCCC E. equisetina Type E2 G C T C T C A C C T E. equisetina

a) The nucleotide sequences reported here will appear in the DDBJ/EMBL/GenBank Nucleotide Seuences Database under the accession numbers AB244096 (Type S), AB244097 (Type IP), AB244098 (Type E1), and AB244099 (Type E2). b) Nucleotide positions are counted from the 5-terminal of the chlB open reading frame. 1210 Vol. 29, No. 6

248 (from the 5-end of the fragment) in this fragment from products were incubated with Bcl I. As shown in Fig. 2B, all E. sinica. This site is Bcl I resistant in seven other Ephedra the PCR products of Fragment 1 from E. sinica specimens species so far examined (E. campulopoda, E. gerardiana, E. were digested to two fragments of about 250 bp while the minuta, E. pachyclada, E. procea and E. regeriana) or de- products from the other Ephedra species remained undi- posited in the DNA database (E. altissima) due to transition gested. This protocol may provide us with a convenient from cytosine to thymine at nt. 267 (Fig. 2A). We amplified method for identification of E. sinica without sequencing. Fragment 1 from various Ephedra specimens and the PCR Authentication of Ephedra Herb Obtained in the Chi- nese Market It was shown that E. sinica and E. equisetina can be identified based on the chlB sequence while the nu- cleotide sequence of the chlB gene of E. intermedia is identi- cal with that of E. przewalskii. However, a nucleotide substi- tution was reported to be present at nt. 884 in ITS1 of the ri- bosomal DNA between these two species.6) Therefore, we ex- amined DNA authentication of 21 Ephedra Herb samples ob- tained in the Chinese market by analyzing the nucleotide se- quences of both chlB and ITS1. Because E. przewalskii was practically devoid of ephedrine alkaloids, we also analyzed ephedrine alkaloid contents in these crude drug samples (Table 3). Of the 21 samples nine were identified as E. sinica and three as E. equisetina based on their chlB sequence type. The rest of the samples were identified as either E. interme- dia or E. przewalskii because their chlB sequence exhibited Type IP. The nucleotide sequences of the ITS1 region were analyzed for these samples, and seven were identified as E. intermedia and two as E. przewalskii based on the nucleotide at nt. 884 (adenine for E. intermedia and cytosine for E. przewalskii). The identification of these two samples as E. przewalskii was also confirmed based on their very low con- Fig. 2. (A) Nucleotide Sequences and Positions of the Restriction Sites of tent of the ephedrine alkaloids. the Fragment of the chlB Gene Sequences recognized with Bcl I are underlined. An asterisk indicates the nucleotide DISCUSSION the same as that in Type S (Ephedra sinica). (B) PCR-RFLP Analysis of the Fragment 1 of chlB Amplified from Nine Sequence comparison of the chloroplast chlB sequence Specimens of Four Ephedra Species: Sin., Ephedra sinica; Int., E. interme- among the herbarium and crude drug samples belonging to dia; Equ., E. equisetina; Prz., E. przewalskii The fragment 1 amplified from the samples was digested with Bcl I and separated by four Ephedra species revealed that the chlB sequence type 4% agarose gel electrophoresis. For the samples, see Table 1. correlated well with the original species. Ten nucleotide sub-

Table3. Authentication of Ephedra Herb Obtained in the Chinese Market

Alkaloid composition (% of dry weight)a) Sample Sample ChlB ITS1 Place of collection Identification No. code type nt. 884 EPNMTotal

1GS35 Gansu IP A E. intermedia 0.15 1.38 0.01 0.03 1.57 2 HB14 S ndb) E. sinica 0.93 0.21 0.03 0.00 1.17 3 HL18 Heilongjiang S nd E. sinica 0.61 0.34 0.01 0.16 1.12 4 JL24 Jilin S nd E. sinica 0.73 0.36 0.02 0.07 1.18 5 JL53 Jilin S nd E. sinica 0.45 0.41 0.01 0.06 0.93 6 LN15 Liaoning S nd E. sinica 0.80 0.13 0.01 0.10 1.04 7 NM84 Inner Mongolia S nd E. sinica 1.03 0.26 0.02 0.11 1.42 8 QH2 Qinghai IP A E. intermedia 0.04 1.00 0.08 0.03 1.15 9 SH5 Shanxi IP A E. intermedia 0.43 0.19 0.02 0.03 0.67 10 SX5 Shaanxi S nd E. sinica 0.42 0.22 0.02 0.13 0.79 11 TJ16 Tianjin S nd E. sinica 0.40 0.45 0.02 0.05 0.92 12 XJ4 Xinjiang IP C E. przewalskii 0.00 0.00 0.00 0.00 0.00 13 U11 Russia E1 nd E. equisetina 0.46 1.12 0.01 0.03 1.62 14 QH1 Qinghai IP A E. intermedia 0.00 1.19 0.00 0.00 1.19 15 XJ2 Xinjiang E1 nd E. equisetina 1.37 0.20 0.03 0.06 1.66 16 XJ10 Xinjiang IP C E. przewalskii 0.02 0.00 0.00 0.00 0.02 17 GS6 Gansu IP A E. intermedia 0.45 1.72 0.01 0.10 2.28 18 GS11 Gansu IP A E. intermedia 0.02 1.47 0.01 0.00 1.50 19 GS18 Gansu IP A E. intermedia 0.01 1.44 0.01 0.01 1.47 20 SC12 Sichuan E2 nd E. equisetina 0.71 0.30 0.03 0.03 1.07 21 SC13 Sichuan S nd E. sinica 0.33 0.49 0.01 0.05 0.88

a) Eephedrine; Ppseudoephedrine; Nnorephedrine; Mmethyl ephedrine. b) nd, not determined. June 2006 1211 stitutions were found among four chlB genotypes, almost all had the identical chlB sequence as E. przewalskii. Sequence of them synonymous with only one non-synonymous substi- analysis of rDNA of E. intermedia and E. przewalskii re- tution at nt. 1119. The nucleotide sequences of the chlB gene vealed that these two species could be discriminated based were completely identical between E. intermedia and E. on the nucleotide at nt. 884 (adenine in E. intermedia and cy- przewalskii and only one synonymous nucleotide substitution tosine in E. przewalskii) of the ITS 1 region.6) Therefore, me- was found between these two species (Type IP genotype) and dicinal Ephedra species can be identified by analyzing nu- E. sinica (Type S). In contrast, there are seven to nine nu- cleotide sequences of both chlB and ITS1 regions. We suc- cleotide substitutions including a non-synonymous substitu- cessfully applied this strategy to Ephedra Herb obtained in tion between Type S or IP genotypes and Type E1 or E2 the Chinese market. genotypes (E. equisetina). The present results indicate that E. In conclusion, we have shown that the DNA profiling is a sinica and E. equisetina can be authenticated based on the powerful tool for identification of the original species of chlB sequence while E. intermedia and E. przewalskii cannot Ephedra Herb. Direct sequencing of the Fragment 1 of the be identified although they are discriminated from E. sinica chlB sequence (or alternatively PCR-RFLP analysis of Frag- or E. equisetina. The chlB sequence types of the additionally ment 1 for E. sinica) will lead to identification of E. sinica examined six Ephedra species and of one species deposited and E. equisetina. Determination of the nucleotide at nt 884 in the DNA database were different from either type S, IP, E1 in the ITS1 of the rDNA will then allow us clear identifica- or E2. Although there are still many other Ephedra species tion of E. intermedia. remaining to be analyzed, it is unlikely that such species ap- pear in the Chinese or Japanese crude drug market as crude Acknowledgements This work was supported, in part, drugs. Therefore, the results shown in this report provide us by a Grant-in-aid for Scientific Research from the Ministry with basic information for application of the chlB gene se- of Health, Labor and Welfare of Japan. We thank Drs. H. Ya- quence for authentication of the crude drug, Ephedra Herb. maji, K. Kondo and S. Terabayashi, Central Research Labo- Comparison of the nucleotide sequences of trnL intron and ratories, Tsumura & Co., for morphological evaluation and trnL–trnF intergenic spacer located in chloroplast genome supply of the herbarium specimens and crude drug speci- among Ephedra plants revealed that there are no nucleotide mens. substitutions among E. sinica, E. intermedia and E. preze- walskii but one or two substitutions between these three REFERENCES species and E. equisetina.5) Yamaji et al.9) also reported that 1) Present address: Faculty of Phamaceutical Sciences at Kagawa Cam- E. intermedia and E. przewarskii had the identical chloro- pus, Tokushima Bunri University, 1314-1 Shido, Sanuki, Kagawa plast rbcL sequence. They also showed that there is one nu- 769–2193, Japan. cleotide substitution between these two species and E. sinica 2) Ministry of Health, Welfare and Labor of Japan, “The Japanese Phar- whereas several substitutions between these three species and macopeia,” 14th ed., 2001, pp. 975—976. E. equisetina. Analysis of ITS 1 and ITS 2 of rDNA indi- 3) Konoshima M., Yakugaku Zasshi, 65, 67—75 (1945). 4) Schaneberg B. T., Crockett S., Bedir E., Khan I. A., Phytochemistry, cated that E. sinica and E. intermedia had the identical se- 62, 911—918 (2003). quence while E. prezewalskii had several nucleotide sites dif- 5) Long C., Kakiuchi N., Takahashi A., Komatsu K., Cai S., Mikage M., ferent from E. sinica and E. intermedia, although there are Planta Med., 70, 1080—1084 (2004). some intra-specific variations in E. intermedia and E. prze- 6) Long C., Kakiuchi N., Zhong G., Mikage M., Biol. Pharm. Bull., 28, walskii.5,6) All of these results as well as our results on the 285—288 (2005). 7) Buchanan B. B., Gruissem W., Jones R. L., “Biochemistry & Molecu- chlB sequences indicated that E. sinica, E. intermedia and E. lar Biology of Plants,” American Society of Plant Physiologists, przewalskii are phylogenetically close to each other while E. Rockville, 2000, pp. 574—578. equisetina is an outgroup of the four Ephedra species. 8) Bolvin R., Martine R., Beauseigle D., Bousquet J., Bellemare G., Mol. The present investigation showed that E. sinica and E. eq- Phylogenet. Evol., 6, 19—26 (1996). uisetina can be identified based on the chlB sequence by di- 9) Yamaji H., Shiba M., Kondo K., Terabayashi S., Amagaya S., Mizukami H., Tsuruga A., Chikada H., Abstract of the 49th Annual rect sequencing. E. sinica might also be able to be distin- Meeting of the Japanese Society of Pharmacognosy, Fukuoka, 2001, p. guished from other species by PCR-RFLP analysis. In con- 148. trast, E. intermedia cannot be identified because this species