Biotechnology Letters 24: 383–389, 2002. © 2002 Kluwer Academic Publishers. Printed in the Netherlands.
383
Enhanced resistance to the poplar fungal pathogen, Septoria musiva, in hybrid poplar clones transformed with genes encoding antimicrobial peptides Haiying Liang, Catharine M. Catranis, Charles A. Maynard & William A. Powell∗ College of Environmental Science and Forestry, State University of New York, Syracuse, NY 13210, USA ∗ Author for correspondence (Fax: 315-470-6934; E-mail:
[email protected]) Received 16 November 2001; Revisions requested 14 December 2001; Revisions received 4 January 2002; Accepted 7 January 2002
Key words: antimicrobial peptides, biomass species, pathogen resistance, Septoria leaf spot, transgenic trees
Abstract Plasmids, pCA1 and pCWEA1, carrying antimicrobial peptide gene(s), Ac-AMP1.2 and ESF12, were used to transform hybrid poplar clones Ogy and NM6. Peptide Ac-AMP1.2 is an analog of Ac-AMP1 which is one of the smallest chitin-binding proteins. Synthetic peptide ESF12 mimics the amphipathic α-helix found in magainins. Transgene mRNA was detected in the transformed plants. When evaluated for resistance to hybrid poplar pathogen Septoria musiva with an in vitro leaf disk assay, the transformed Ogy plants showed significantly increased pathogen resistance as compared to the untransformed Ogy.
Introduction The hybrid poplar is an important woody biomass and potential bioenergy tree species. This tree species has economic value as a good source for fuel, pulpwood, and solid wood products (Ostry & McNabb 1985). However, utilization of hybrid poplar is limited in the eastern United States due to a fungal pathogen, Septoria musiva. This fungus causes necrotic lesions in the leaves and canker formation on the stems and branches (Moore & Wilson 1983). Leaf spot can reduce the photosynthetic area and cause premature defoliation thereby decreasing annual growth. The cankers on the main stem can reduce growth, predispose the tree to colonization by secondary organisms, and cause severe girdling and breakage of the main stem. In addition, the wood of surviving trees is unsuitable for pulp production (Ostry & McNabb 1983). Biomass losses due to this pathogen vary among clones but losses as high as 63% in a plantation had been reported (McNabb et al. 1982). Although chemical and cultural control of Septoria leaf spot and canker has been tried, it was found to be problematic and only
partially effective (Ostry & McNabb 1983, Yang et al. 1994). Enhancing disease resistance in plants by introduction of small antimicrobial peptide genes has been tried in some annual crops, like tobacco (Huang et al. 1997, Cary et al. 2000, Li et al. 2001), with mixed results regarding pathogen resistance. However, there are no data available on woody plants to date. In this study, we report for the first time on the transformation of hybrid poplar with antimicrobial peptide genes (AcAMP1.2 and ESF12) and the potential resistance of the transgenic plants to fungal disease caused by the hybrid poplar pathogen, S. musiva. Chitin is an important structural component of the cell wall of fungi but is not found in higher plants (Raikhel & Lee 1993). Ac-AMP1 (29 amino acid residues) (Broekaert et al. 1992) is one of the smallest chitin-binding proteins and was isolated from the seed coat of amaranth (Amaranthus caudatus). This peptide, Ac-AMP1, inhibits the growth of several different plant-pathogenic fungi at much lower concentrations than other known antifungal chitin-binding proteins. In addition, this peptide does not agglutinate erythrocytes, indicating low toxicity to mammalian
384 cells (Broekaert et al. 1992). Ac-AMP1.2 is an analog of Ac-AMP1 with the amino terminal valine replaced with methionine to facilitate expression (Powell et al. 1997). ESF12 (Powell et al. 1995) is a synthetic peptide with 18 amino acids and mimics the amphipathic α-helix found in magainins (Zasloff 1987). The growth inhibitory activity of ESF12 to S. musiva and other pathogenic fungi has been demonstrated in vitro (Powell et al. 1995). The hypotheses of this study were that hybrid poplar could be transformed with AcAMP1.2 and ESF12/Ac-AMP1.2 transgenes and that hybrid poplar transformed with these genes would be more resistant to S. musiva than the untransformed controls.
was used for pCA1 construct, WAF1/WAR1 (Figure 1) for pCWEA1, and GUSF/GUSR for pBI121. PCR reactions (25 µl) were performed according to manufacturer’s instructions (Perkin Elmer Cetus, Norwalk, CT), using 10 ng template DNA. The annealing temperatures for CAF1/CAR1, WAF1/WAR1, and GUSF/GUSR were 58 ◦ C, 62 ◦ C and 63 ◦ C, respectively. All amplified DNA was subjected to electrophoresis on agarose gels and visualized by staining with ethidium bromide. Reverse transcription (RT)-PCR analysis
Plasmids pCA1 and pCWEA1 (Figure 1) were modified from the vector pCEA1 (Powell & Maynard 1997). A gene construct was designed using plant preferred codons (Campbell & Gowri 1990) to encode the ESF12 and Ac-AMP1.2 peptides in tandem, each having its own start and stop codons. This construct was subcloned into pBI121 (Clontech, Palo Alto, CA) downstream of the constitutive CaMV35S promoter replacing the GUS gene to form pCEA1. A poplar wound-inducible promoter (win3.12) received from Dr Milton P. Gordon at the University of Washington, USA (Hollick & Gordon 1993) was cloned upstream of the transgene construct ESF12/Ac-AMP1.2 in the pCWEA1. To construct pCA1, the ESF12 gene was deleted by digesting pCEA1 with NcoI and then religating the resulting plasmid.
Total RNA was extracted from young leaves of transformed (unwounded for pCA1 and wounded for pCWEA1) and untransformed plants using the hot borate method (Wan & Wilkins 1994). Pipette tips (10 µl) were employed to wound the pCWEA1transformants by punching holes in the leaf blades. Wounded tissue was then harvested 6 h after wounding. Approximately 0.6 µg total RNA was used for reverse transcription with oligo d(T), using the GeneAmp RNA PCR kit (Perkin Elmer Cetus, Norwalk, CT). Primer pairs HY/ CAR1 and RTEF/RTAR (Figure 1), designed from the coding regions of the transgenes, were used to amplify the transcripts for pCA1 and pCWEA1 constructs, respectively. Primer pairs CAF1/CAR1 and WAF1/WAR1 were used as controls because they would amplify only DNA as a contaminant. After reverse transcription, one half of the cDNA products were used for transcript amplification and the other half for PCR testing for DNA contamination. The annealing temperatures for primer pairs CAF1/CAR1, WAF1/WAR1, HY/CAR1 and RTEF/RTAR were 58 ◦ C, 62 ◦ C, 54 ◦ C and 60 ◦ C, respectively.
Plant tissue culture and transformation
Nuclease protection assay (NPA)
Agrobacterium tumefaciens strain LBA4404, carrying one of the plasmids pCA1, pCWEA1 or pBI121, was employed to transformed hybrid poplar clones Ogy (Populus × euramericana) and NM6 (Populus nigra × P. maximowizii). Plant transformation was done according to Liang et al. (2001).
To detect transgene mRNA, NPA was performed on pCA1 (unwounded) and pCWEA1 (wounded and unwounded, see above) transformants, according to Ausubel et al. (1995) except that S1 was replaced with DNase I. For total RNA, 20 µg, and 75 ng of biotinylated probes, were employed. Probes were made by PCR using biotinylated dNTP mix, with primer pair HY/CAR1 for pCA1 and RTEF/RTAR for pCWEA1. The NPA products were subjected to electrophoresis in a 1.5% (w/v) agarose gel and then transferred onto a Zeta-Probe blotting membrane (Bio-Rad, Hercules, CA). Detection was performed on the membranes ac-
Materials and methods Plasmid binary vectors
PCR analysis To amplify the transgene from putative transformed plants, PCR was performed on the cellular DNA isolated from leaves. Primer pair CAF1/CAR1 (Figure 1)
385
Fig. 1. Vector maps of pCA1 and pCWEA1. The amino acid and DNA sequences of Ac-AMP1.2 and ESF12 peptides as well as the DNA sequences of oligos CAF1, CAR1, HY, WAF1, WAR1, RTEF, and RTAR are indicated. RB: right border of the T-DNA; LB: left border of the T-DNA. The maps are drawn according to scale.
cording to manufacturer’s instructions (NewEngland Biolabs, Beverly, MA).
was repeated three times and totally 45 leaf disks were tested for each transformed line.
In vitro pathogen resistance test Results and discussion To determine the potential S. musiva resistance of the pCA1 and pCWEA1 transformants, an in vitro leaf disk assay was performed as previously described in detail (Liang et al. 2001). Briefly, leaf disks (15 mm in diameter) from each transformed line and untransformed plants were inoculated with 20 µl of 106 conidia S. musiva (strain 92-49C) ml−1 . One week after necrosis became visible on the untransformed leaf disks, necrotic areas of the leaf disks were then measured an using an NIH Image 1.61 system. Resistance of the transformed leaves was scored by comparing the average necrotic area formed in transgenic leaf disks to the average in untransformed controls and represented as a percentage. Data were subjected to ANOVA using the general linear model. This pathogen resistance test
After two rounds of regeneration and selection with kanamycin, two independent pCA1-transgenic lines and one pCWEA1-transgenic line were achieved for Ogy. For NM6, one transgenic line was recovered for pCWEA1 and one for pBI121. PCR with primer pair CAF1/CAR1 produced a 224-bp band in the pCA1-transgenic Ogy lines and the pCA1 plasmid control (Figure 2A). A 325-bp band was detected in the pCWEA1-transgenic Ogy and NM6 lines as well as in the pCWEA1 plasmid control with primer pair WAF1/WAR1 (Figure 2B). With GUS gene-specific primers GUSF and GUSR, a single band of 1.1 kb in length was detected in both pBI121-transgenic NM6 and plasmid DNA (Figure 2C). In contrast, no bands were detected in the untransformed Ogy and NM6
386
Fig. 2. PCR analysis of the pCA1-, pCWEA1-, and pBI121-transformed lines. P: plasmid DNA of pCA1 (A), pCWEA1 (B), pBI121 (C); H: H2 O control; U: untransformed plants; T: transformed plants; o: clone Ogy; n: clone NM6; M1: 100 bp DNA marker; M2: 1 kb DNA marker. The primer pair used for each construct is indicated in parentheses.
Fig. 3. RT-PCR and nuclease protection assay (NPA) of the pCA1transformed Ogy lines. A: RT-PCR. 1: pCA1 plasmid DNA; 2: untransformed Ogy; 3: transformed line No. 1; 4: transformed line No. 2; M: 100 bp DNA marker. B: NPA. 1: transformed line No. 1; 2: untransformed Ogy; 3: digested probes; 4: transformed line No. 2; 5: undigested probes; 6: biotinylated DNA marker (Gibco-BRL, Rockville, MD).
controls. A protocol, including two rounds of antibiotic selection (Liang et al. 2001), was used for plant transformation in which carbenicillin and cefotaxime were not included in the media used for the second round of selection. Since carbenicillin and cefotaxime inhibit the growth of Agrobacterium, this second round of selection also served to test whether the transformants were free of Agrobacterium. As no Agrobacterium contamination was observed visually in this step, this indicated that the transformed plants recovered were free of Agrobacterium. The transgenic plants were propagated by cutting and planted in the field in the spring of 2000. PCR with primer pair WAF1/WAR1 was performed in the second growing season, and all the 25 pCWEA1-transformed Ogy plants being tested still came out positive for the ESF12/Ac-AMP1.2 construct, indicating stable integration of transgene in the plant genome. The transgene mRNA were detected in transgenic plants by RT-PCR (Figures 3A and 4A) and nuclease protection assay (Figures 3B and 4B), using coding region-specific primers (HY/CAR1 for AcAMP1.2 and RTEF/RTAR for ESF12/Ac-AMP1.2, Figure 1). ESF12/Ac-AMP1.2 was regulated by a wound-inducible promoter (win3.12) in the pCWEA1
Fig. 4. RT-PCR and nuclease protection assay (NPA) of the pCWEA1-transformed lines. A: RT-PCR. 1: pCWEA1 plasmid DNA; 2: untransformed NM6; 3: untransformed Ogy; 4: unwounded transformed NM6; 5: wounded transformed NM6; 6: unwounded transformed Ogy; 7: wounded transformed Ogy; M: 100 bp DNA marker. B: NPA. 1: untransformed NM6; 2: untransformed Ogy; 3: unwounded transformed NM6; 4: wounded transformed NM6; 5: unwounded transformed Ogy; 6: wounded transformed Ogy.
transformants. At 6 h after wounding, the production of ESF12/Ac-AMP1.2 transcripts in both pCWEA1 lines was found to be enhanced when compared to the unwounded backgrounds (Figure 4B). This win3.12 promoter has been used to regulate the expression of a GUS gene in tobacco and a similar inducible expression pattern was detected (Hollick & Gordon 1993). Such a wound-inducible expression provides an energetic advantage to the transgenic plants over constitutively expressed genes since there is only a low level of expression of the transgene in the absence of wounding or invasion. However, upon wounding or invasion, rapid and systemic activation of the transgenes will occur. One primary concern about RT-PCR is genomic DNA contamination. In order to rule out the possibility of DNA contamination in our RT-PCR results, the total RNA extracted from plant leaves was treated with DNase. At the same time, a different set of primers was designed and used as a negative control in RTPCR (CAF1/CAR1 for pCA1 and WAF1/WAR1 for pCWEA1, Figure 1). These primer sets would amplify only DNA as a contaminant since the forward primers (CAF1 and WAF1) anneal to the DNA se-
387
Fig. 5. In vitro pathogen resistance test. The pictures are representative of three trials. The error bars represent the standard errors derived from 45 conidia-inoculated replicates. In the same graph, the bars with the same letter are not significantly different according to the Tukey’s Studentized Ranged Test. In each plate, the central leaf disk was treated with sterile ddH2 O.
quence upstream of the transcription start sites. PCR with these negative control primer pairs failed to amplify the transgenes from the reverse transcription products (Figures 3A and 4A), excluding the DNA contamination concern in the RT-PCR results. The in vitro leaf disk bioassay used in this study has been used to successfully screen hybrid poplar for resistance to leaf spot caused by S. musiva (Ostry et al. 1988, Yang et al. 1994). Leaf disks of clones inoculated with S. musiva conidia displayed disease resistance similar to that found in field trials (Ostry et al. 1988). In our study, necroses formed in the pCA1- and pCWEA1-transformed Ogy lines were delayed by approximately 2 days. The least necrotic area in the pCWEA1-transformed Ogy leaf disks was 33 ± 3 (SE)% of those formed in the untransformed control. The necrotic areas formed in the pCA1-transformed Ogy line No. 1 and line No. 2 were 60 ± 4 (SE)% and 68 ± 4 (SE)% of the untransformed control, respectively (Figure 5A). There was no significant difference of necrotic area between the two pCA1-
transformed Ogy lines (P = 0.133), even though they represented separate transformation events. However, the pCWEA1-transformed, pCA1-transformed, and untransformed Ogy plants were significantly different from each other (P < 0.001). No necrosis was observed on the water-treated Ogy untransformed controls. Compared to pCA1, the pCWEA1 construct provided higher activity against S. musiva in transgenic Ogy, which may be because of different expression levels of the transgene or to the fact that pCWEA1 contained two antimicrobial peptide genes (ESF12 and Ac-AMP1.2). The accumulation of both peptides in pCWEA1-transformed plants remains to be demonstrated. However, in carrot protoplasts (Putterill & Gardner 1989) and tobacco plastids (Staub & Maliga 1995) such an internal open reading frame can be efficiently translated. The antimicrobial actions of ESF12 and Ac-AMP1 are different because Ac-AMP1 is a chitin-binding, defensin-like peptide (Broekaert et al. 1992), while the predicted structure of ESF12
388 contains an amphipathic α-helix thought to form channels in biological membranes (Powell et al. 1995). Expression of both these antimicrobial peptides may decrease the probability of a pathogen overcoming the multiple resistance mechanisms of two antimicrobial agents (Powell & Maynard 1997), thus providing more durable resistance. Unlike the Ogy clone, necrosis became visible in both the untransformed and transformed NM6 leaf disks at the same time (approximately one week after inoculation) and there was no significant difference among the untransformed, pCWEA1- transformed, and pBI121-transformed NM6 leaf disks (Figure 5B) (P = 0.240), when mean necrotic area was compared. It could be that ESF12/Ac-AMP1.2 expression level in NM6 was lower than the level in Ogy, since the effectiveness of transgenes in conferring resistance depends on their expression level (Florack et al. 1995, Ko et al. 2000). Variations in expression level have been found among independent transformants, which may be determined by copy number and/or integration site (Pröls & Meyer 1992, Hobbs et al. 1993). An alternative explanation could be the difference in S. musiva virulence on NM6 versus Ogy. Ogy is highly susceptible to S. musiva (Heuchelin et al. 1997). In contrast, the NM6 is a moderately resistant clone (Lo et al. 1995). If a more virulent isolate of S. musiva had been used in the pathogen resistance test, enhanced resistance to S. musiva in pCWEA1 transformed NM6 may have been more obvious (Glen Stanosz, University of Wisconsin, pers. commun.). We have attempted to detect the peptide expression in the transgenic plants using HPLC and LC-MS but so far without success. We believe this may be due to the physical and biochemical characteristics of these small-size peptides. For example, ESF12 is only 18amino acid long (approximately 1.9 kDa) and very basic (Powell et al. 2000). In our preliminary study, the strong alkalinity of the ESF12 peptide seemed to inhibit the ionization in the mass spectrometry. It required very high voltages to ionize the peptide and any ionization is totally suppressed by the presence of Tris or phosphate buffers. We could detect the commercially synthesized ESF12 peptide that was dissolved in water but by including plant protease inhibitors the signal was decreased by as much as 75% and no signal was detected in the untransformed plant samples ‘spiked’ with known amount ESF12 peptide. Difficulty in the detection of transgenically expressed foreign peptides in plants has also been reported in
other research groups (Hancock & Lethrer 1998, Cary et al. 2000, Li et al. 2001). In conclusion, Agrobacterium-mediated transformation of hybrid poplar clones Ogy and NM6 with genes encoding antimicrobial peptides Ac-AMP1.2 and ESF12 was successful in this study. The coding regions were actively transcribed in all transgenic lines and the presence of these transcripts was associated with enhanced resistance in the Ogy transformants to invasion by S. musiva during the in vitro pathogen resistance test. The hypotheses that hybrid poplar could be transformed with ESF12 and Ac-AMP1.2 antimicrobial peptide genes and that the resulting transgenic plants would show enhanced resistant to S. musiva were supported by these data. Acknowledgements We thank Dr Gregory L. Boyer for his kind assistance with the HPLC and LC-MS work. This research was supported in part by grants from the USDA McIntire-Stennis Program and William Heckrodt of the Community Foundation for the Fox Valley Region, Inc. References Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA, Struhl K (1995) Short Protocols in Molecular Biology. Toronto: John Wiley & Sons, Inc. Broekaert WF, Marien W, Terras FRG, Bolle MFC, de Proost P, van Damme J, Dillen L, Claeys M, Rees SB, Vanderleyden J (1992) Antimicrobial peptides from Amaranthus caudatus seeds with sequence homology to the cysteine/glycine-rich domain of chitin-binding proteins. Biochemistry 31: 4308–4314. Campbell WH, Gowri G (1990) Codon usage in higher plants, green algae, and cyanobacteria. Plant Physiol. 92: 1–11. Cary JW, Rajasekaran K, Jaynes JM, Cleveland TE (2000) Transgenic expression of a gene encoding a synthetic antimicrobial peptides results in inhibition of fungal growth in vitro and in planta. Plant Sci. 154: 171–181. Florack D, Allefs S, Bollen R, Bosch D, Visser B, Stiekema W (1995) Expression of giant silkmoth cecropin B genes in tobacco. Transgenic Res. 4: 132–141. Hancock RE, Lehrer R (1998) Cationic peptides: a new source of antibiotics. Trends Biotechnol. 16: 82–88. Heuchelin SA, McNabb Jr HS, Klopfenstein NB (1997) Agrobacterium-mediated transformation of Populus × euramericana Ogy using the chimeric CaMV 35S-pin2 gene fusion. Can. J. For. Res. 27: 1041–1043. Hobbs SLA, Warkentin TD, DeLong CMO (1993) Transgene copy number can be positively or negatively associated with transgene expression. Plant Mol. Biol. 21: 17–26. Hollick JB, Gordon MP (1993) A poplar tree proteinase inhibitorlike gene promoter is responsive to wounding in transgenic tobacco. Plant Mol. Biol. 22: 561–572.
389 Huang Y, Nordeen RO, Di M, Owens LD, McBeath JH (1997) Expression of an engineered cecropin gene cassette in transgenic tobacco plants confers disease resistance to Pseudomonas syringae pv tabaci. Phytopathology 87: 494–499. Ko K, Norelli JL, Reynoird JP, Boresjza-Wysocka E, Brown SK, Aldwinckle HS (2000) Effect of untranslated leader sequence of AMV RNA 4 and signal peptide of pathogenesis-related protein 1b on attacin gene expression, and resistance to fire blight in transgenic apple. Biotechnol. Lett. 22: 373–381. Li Q, Lawrence CB, Xing HY, Babbitt RA, Bass WT, Maiti IB, Everett NP (2001) Enhanced disease resistance conferred by expression of an antimicrobial magainin in transgenic tobacco. Planta 212: 635–639. Liang H, Maynard CA, Allen RD, Powell WA (2001) Increased Septoria musiva resistance in transgenic hybrid poplar leaves expressing a wheat oxalate oxidase gene. Plant Mol. Biol. 45: 619–629. Lo MH, Abrahamson LP, White EH, Manion PD (1995) Early measures of basal area and canker disease predict growth potential of some hybrid poplar clones. Can. J. For. Res. 25: 1113–1118. McNabb Jr HJ, Ostry ME, Sonnelitter RS, Gerstenberger PE (1982) The effect and possible integrated management of Septoria musiva in intensive, short-rotation culture of Populus in the north central states. In: Zavitkovski WIJ, Manhattan EH, eds. Proceedings of the North American Poplar Council Meeting, July 20–22, Rhinelander. Wisconsin: Division of Extension, Kansas State University, pp. 51–58. Moore LM, Wilson LP (1983) Recent advances in research of some pest problems of hybrid Populus in Michigan and Wisconsin. Gen. Tech. Rep. North Centr. For. Exp. Station 91: 94–101. Ostry ME, McNabb Jr HS (1983) Diseases of intensively cultured hybrid poplars: a summary of recent research in the north central region. Gen. Tech. Rep. North Centr. For. Exp. Station 91: 102– 109. Ostry ME, McNabb Jr HS (1985) Susceptibility of Populus species and hybrids to disease in the North Central United States. Plant Dis. 69: 755–75.
Ostry ME, McRoberts RE, Ward KT, Resendez R (1988) Screening hybrid poplars in vitro for resistance to leaf spot caused by Septoria musiva. Plant Dis. 72: 497–499. Powell WA, Maynard CA (1997) Designing small antimicrobial peptides and their encoding genes. In: Klopfenstein N, Chun YW, Kim M-S, Ahuja MR, eds. Micropropagation, Genetic Engineering and Molecular Biology of Populus, Fort Collins, CO: Rocky Mountain Forest and Range Experiment Station, pp. 165–172. Powell WA, Catranis CM, Maynard CA (1995) Synthetic antimicrobial peptide design. Mol. Plant-Microb. Interact. 8: 792–794. Powell WA, Catranis CM, Maynard CA (2000) Design of selfprocessing antimicrobial peptides for plant protection. Lett. Appl. Microbiol. 31: 163–168. Pröls F, Meyer P (1992) The methylation patterns of chromosomal integration regions influence gene activity of transferred DNA in Petunia hybrida. Plant J. 2: 465–475. Putterill JJ, Gardner RC (1989) Initiation of translation of the betaglucuronidase reporter gene at internal AUG codons in plant cells. Plant Sci. 62: 199–205. Raikhel NV, Lee H-L (1993) Structure and function of chitin0 binding proteins. Annu. Rev. Plant Physiol. Plant Mol. Biol. 44: 591–615. Staub JM, Maliga P (1995) Expression of a chimeric uidA gene indicates that polycistronic mRNAs are efficiently translated in tobacco plastids. Plant J. 7: 845–848. Wan CY, Wilkins TA (1994) A modified hot borate method significantly enhances the yield of high-quality RNA from cotton (Gossypium hirsutum L.). Anal. Biochem. 223: 7–12. Yang D, Bernier L, Dessureault M (1994) Biological control of Septoria leaf spot of poplar by Phaeotheca dimorphospora. Plant Dis. 78: 821–825. Zasloff M (1987) Magainins, a class of antimicrobial peptides from Xenopus skin: isolation, characterization of two active forms, and partial cDNA sequence of a precursor. Proc. Natl. Acad. Sci. USA 84: 5449–5453.