Index | Search | Home | Table of Contents

link to pdf file

Bedir, E., I. Khan, R.M. Moraes. 2002. Bioprospecting for podophyllotoxin. p. 545–549. In: J. Janick and A. Whipkey (eds.), Trends in new crops and new uses. ASHS Press, Alexandria, VA.

Bioprospecting for Podophyllotoxin*

Ebru Bedir, Ikhlas Khan, Rita M. Moraes

*This work was supported in part by the United States Department of Agriculture grants No. 97-35501-4886, 99-35504-8261, and Agricultural Research Service Specific Cooperative Agreement No. 58-6408-7-012. We thank USDA/ARS/ NGRP for providing authenticated seeds and Victor Maddox for his expertise and advise on plant collection. We also thank Dhammika Nanayakkara at National Center for Natural Products Research for the lignan standards and Belma Konuklugil, Ankara University, Turkey, for a sample of 5-methoxypodophyllotoxin.

Podophyllotoxin is a natural lignan that is currently being used as a precursor to semi-synthetic anti-cancer drugs etoposide, teniposide, and etopophos. These compounds have been used for the treatment of lung and testicular cancers as well as certain leukemias (Stahelin and Wartburg 1991; Imbert 1998). The patent for the use of etoposide in such therapies expired in 1995 and since then, etoposide has been tested in 167 clinical trials for the use as new investigative cancer treatments or as positive control (Ekstrom et al. 1998; Holm et al. 1998; Ajani et al. 1999). In addition, podophyllotoxin is also the precursor to a new derivative CPH 82 that is being tested for rheumatoid arthritis in Europe and other derivatives for the treatment of psoriasis and malaria (Leander and Rosen 1988; Lerndal and Svensson 2000). Podophyllotoxin preparations are also on the market for dermatological use to treat genital warts (Beutner 1996) and recently, immune-stimulatory activities of podophyllotoxin have been reported (Pugh et al. 2001). Total synthesis of podophyllotoxin is an expensive process and availability of the compound from natural resources is an important issue for pharmaceutical companies that manufacture these drugs (Canel et al. 2000b).

Currently, the commercial source of podophyllotoxin is the rhizomes and roots of Podophyllum emodi Wall., Berberidaceae (syn. P. hexandrum Royle), an endangered species from the Himalayas (Foster 1993). Recent findings concluded that the leaf blades of the North American mayapple (P. peltatum L.) may serve as an alternative source of podophyllotoxin production, since leaves are renewable organs that store lignans as glucopyranosides (Canel et al. 2001; Moraes et al. 2001). Using the new buffer extraction procedure reported by Canel et al. (2000a), podophyllotoxin can be obtained by conversion of podophyllotoxin 4-O-b-D-glucopyranoside into the aglycone. This extraction procedure of P. peltatum leaves yields podophyllotoxin (40.0 mg·g-1 on a dry weight basis) in amounts similar to the ethanol extraction of P. emodi rhizomes and roots (Jackson and Dewick 1984; Canel et al. 2001). After the publication of these findings, there has been an increasing interest in domestication and cultivation of P. peltatum for podophyllotoxin production; however there remains a need for identification and confirmation of the most economical isolated yielding source. Our objective was to quantify podophyllotoxin in other genera and to evaluate their potential production based on yields and abundance. These other genera were Linum, Juniperus, Hyptis, Teucrium, Nepeta, Dysosma, Jeffersonia, Thymus, and Thuja (Kupchan et al. 1965; San Feliciano et al. 1989a,b; Broomhead and Dewick, 1990a,b; Yu et al. 1991; Kuhnt et al. 1994; Konuklugil 1996a,b; Muranaka et al. 1998). In addition, extractions with buffer and other solvents were compared using two abundant US species as sources.


Chemicals and Standards

D. Nanayakkara at the National Center for Natural Products Research, School of Pharmacy, The University of Mississippi, provided all standard compounds. Identity and purity were confirmed by chromatographic (TLC, HPLC) and spectral (IR, 1D- and 2D-NMR, HRESIMS) methods (Bastos et al. 1995). Solvents (water, acetonitrile, methanol, and trifluoroacetic acid) were of HPLC quality and purchased from Fisher Scientific (Fair Lawn, New Jersey).

Plant Material

Leaf blades of Podophyllum peltatum L., Linum perenne L., and Linum rigidum Purseh. were collected from plants cultivated at the Medicinal Plant Garden, National Center for Natural Products Research, University of Mississippi. Juniperus virginiana L. was collected in Oktibbeha county, Mississippi (T19N R15E S16) and Hyptis alata in Wayne County, Mississippi (T7N R8W S1). Dysosma versipellis (Hance) M. Cheng, Thuja occidentalis L., Juniperus squamata Lamb., Juniperus sabina L., and Jeffersonia diphylla (L.) Pers. were purchased at Plant Delights Nursery, Inc. Raleigh, North Carolina.

Seeds of Linum austriacum L. (PI 383683), Linum flavum L. (PI 522296 and PI 545704), Linum hirsutum L. (PI 502406), Linum lewisii Purseh. (PI 263911), Linum usitatissimum L. (PI 522333), Nepeta cataria L. (W6 17265), Teucrium polium L. (PI 2440664), and Thymus spp. (PI 2381087 and PI 381045) were supplied by USDA, ARS, NCRPIS, Iowa State University, Ames IA and USDA, ARS, WRPIS Washington State University, Pullman. Seeds were planted in plastic containers (140 × 86 mm × 116 mm diameter) filled with peat moss, perlite, and potting soil. The pots were placed under fluorescent lights ~52 mmol s-1 m-2 for 60 days with 16 hr photoperiod. The plants were watered twice a week as needed.

Extraction and Quantitation of Lignans

Chloroform and dichloromethane extraction methods were scaled down to utilize 50 mg and 500 mg of powdered dried plants respectively (Cairnes et al. 1980; Muranaka et al. 1998). Powdered plant material (500 mg) was extracted with buffer, ethanol, hot ethanol, and dichloromethane (Cairnes et al. 1980; Jackson and Dewick 1984; Broomhead and Dewick 1990b; Konuklugil 1996a; Muranaka et al. 1998). The lignans were separated and quantified by HPLC (Canel et al. 2001). Statistical analysis of data was carried out using SAS, 1999 software.


The buffer extraction yielded significantly more podophyllotoxin in both species than all other solvents, confirming our previous results (Canel et al. 2001) (Table 1). The only exception was the chloroform extraction of J. virginiana leaves short scales in close overlapping pairs (needles), which yielded similar amounts of podophyllotoxin as the buffer extraction. These findings suggest that podophyllotoxin is not stored as the glucopyranoside in the needles. In fact, the HPLC chromatograms of the extractions of J. virginiana needles (Fig. 1A and B) revealed that a compound with the 1.7 min retention time was present in larger amounts in the chloroform extract but drastically reduced in the buffer. This is presumably the evidence of in situ b-glucosidase conversion in the buffer procedure hydrolyzing glycosides other than the podophyllotoxin 4-O-b-D gluco-pyranoside.

Table 1. Evaluation of differential podophyllotoxin extraction from abundant sources in the US.

Extraction Podophyllotoxin content (mg/g dry weight)z
P. peltatum (leaves) J. virginiana (needles) J. virginiana (twigs)
Buffer 21.3 4.7 1.2
Ethanol 2.1 1.7 1.0
Hot ethanol 1.4 3.7 0.9
Chloroform 1.0 4.8 0.8
Dichloromethane 8.8 1.5 0.5
Extraction ***    
Source ***    
Ext Source ***    

zData represent the mean of two extractions of sample and triplicate runs in HPLC.
***Significant at P = 0.001.

Fig 1. HPLC separation of lignans extracted from J. virginiana, (A) lignan fraction of needles extracted by Canel et al. 2001 and (B) by chloroform.

Leaf blades of P. peltatum contained the largest amount of podophyllotoxin on a dry weight basis (21.3 mg·g-1), which is five times more than for J. virginiana. However, needles may be more readily available as a by-product of the lumber industry and another advantage is that podophyllotoxin was the major lignan in the extract (97% of the total lignans) making it amenable to a simple purification process (Table 2).

Table 2. The effect of developmental stage on podophyllotoxin content in needles of Juniperus virginiana.

Stage of development Plant diameter
(mg/g dry wt.)
Juvenile 1 4.5 0.50
Juvenile 2 5.8 1.60
Juvenile 3 (transition stage) 10.2 1.00
Female 41.0 2.90
Male 28.0 2.00
Male 41.0 2.80
Female 45.2 3.40
Male 29.5 4.70

Needles of juvenile J. virginiana yielded less podophyllotoxin than mature plants (Table 2). There appears to be no difference in podophyllotoxin content between female and male plants, however a broader sampling should be considered for further evaluation.

The production of podophyllotoxin from J. virginiana shows good potential, however only few factors were considered in this study, including abundance, podophyllotoxin content in the wild crafted biomass, species growth and development (Tables 1 and 3). However many other factors relevant to podophyllotoxin production such as the actual isolated yield and cost of handling and processing lower yielding biomass still need to be considered.

HPLC analyses reveal that the leaves of Dysosma pleiantha (also known as the Podophyllum from China) and Teucrium polium have some potential for podophyllotoxin production (Table 3). Linum flavum is being recognized (Konuklugil 1996a) as a potential alternative source, however the accessions PI 422296 and PI 502406 grown in our laboratory were low yielding as far as podophyllotoxin content is concerned. The major lignan in the HPLC chromatogram was identified as 5-methoxypodophyllotoxin. Since the precursor for the anticancer drugs is podophyllotoxin, not much attention was given to this analog.

Table 3. Bioprospecting for podophyllotoxin using buffer extraction.

Species Accession Plant organ Lignans (mg/g dry wt.) Total
a-PLTz Pdtox b-PLT
D. pleiantha   leaves 1.24 3.17 0.25 4.67 67.90
J. diphylla   roots - + - + +
J. sabina   needles 0.05 0.07 0.10 0.22  
J .sabina   stem 0.09 0.14 - 0.23  
J. sabina   root - 0.06 - 0.06  
J. squanata   root - 0.09 - 0.09  
J. squanata   stem - + - -  
J. squanata   needles + + - -  
J. virginiana Plant A needles - 3.40 0.10 3.50 97.00
J. virginiana Plant B needles - 4.70 - 4.70 100.00
L. austraium PI 383683 leaves 0.06 - 0.37 0.43  
L. flavum PI 522296 leaves 1.70 0.16 0.41 2.27  
L. flavum PI 545704 leaves 3.50 0.11 0.42 4.03  
L. hirsutum PI 502406 leaves 0.06 0.15 0.19 0.40  
L. lewisii PI 263911 leaves 0.03 + 0.08 0.11  
L. perenne   roots 0.06 - 0.14 0.20  
L. perenne   leaves 0.08 - 0.24 0.32  
L. rigidum   leaves - + 0.11 0.11  
L. rigidum   roots - - 0.41 0.41  
L. usitatissimum PI 522333 leaves 0.05 0.05 + +  
N. cataria   leaves + + - - +
T. polium PI 2440664 leaves 0.21 2.60 - 2.81 92.50
T. occidentalis   stem 0.18 0.01 - 0.19  
T. occidentalis   roots 0.17 0.09 - 0.26  
T. occidentalis   needles 0.11 0.08 - 0.18  
Thymus sp. PI 2381087 leaves 0.08 - - -  
Thymus sp. PI 381045 leaves 0.12 + - -  

These results in this bioprospecting study indicate that the leaves of P. peltatum and needles of J. virginiana may serve as alternative sources of podophyllotoxin. At this stage, however, it is not clear which species is potentially better. The greater podophyllotoxin content of P. peltatum may justify the time and the expenses of commercial field production opposing to needles of J. virginiana in great availability as a by-product of the lumber industry. A closer look at podophyllotoxin recovery from each species as well as a feasibility study comparing the cost of cultivation and price of the land for establishing P. peltatum plantings are needed for any further development.