Enhancer and Gene Trap Transposon Mutagenesis in Arabidopsis
[adapted from "Enhancer and Gene trap transposon mutagenesis in Arabidopsis", R. Martienssen and P. Springer, 1998 in "Insertional Mutagenesis: a practical approach" Oxford University Press (G. Coupland, ed.)]
Enhancer traps and gene traps are reporter gene constructs that can respond to cis-acting transcriptional signals when integrated into chromosomal DNA (reviewed in . Insertional mutagenesis using these "traps" involves generating a large number of individuals that have the reporter gene integrated at different sites throughout the genome. Their progeny are collected and examined for expression of the reporter gene and mutant phenotypes caused by insertion. In lines in which the reporter gene is inserted within or near a chromosomal gene, reporter gene expression mimics that of the chromosomal gene. In the last few years enhancer traps and gene traps have been extensively exploited in Drosophila and in mouse developmental genetics , and several modifications have been made to the basic systems. In plants, gene traps have taken a number of different forms, depending on the reporter gene construct and the vector used for insertional mutagenesis. T-DNA and transposons have both been used to introduce promoter and enhancer traps into Arabidopsis . In this summary, we will describe the enhancer trap and gene trap transposon system that we and our colleagues have developed in Arabidopsis .
Enhancer and gene trap transposons
Enhancer trap reporter genes have a minimal promoter that is only expressed when inserted near cis-acting chromosomal enhancers. Gene trap reporter genes (promoter traps and exon traps) have no promoter, so that reporter gene expression can occur only when the reporter gene inserts within a transcribed chromosomal gene, creating a transcriptional fusion. Our system uses both enhancer trap and gene trap reporters, and is based on the maize Ac and Ds transposable elements (Figure 1). The Ds elements carry the b -glucuronidase (GUS) gene as a reporter and the Neomycin phosphotransferase (NPTII) gene (conferring resistance to kanamycin) as a selectable marker. There are two classes of Ds elements: (1) an enhancer trap element which has a minimal promoter fused to the GUS gene, and (2) a gene trap element, which has a multiple splice acceptor fused to the GUS gene. Random insertions of the Ds element throughout the genome allow us to detect chromosomal gene expression through the activation of the GUS gene.
Figure 1. Schematic Transposon constructs.Schematic diagrams of the Ac and Ds elements are shown incorporated into T-DNA (19). IAAH, indole acetamide hydrolase gene (confers sensitivity to naphthalene acetamide, NAM); NPTII, neomycin phosphotransferase gene (confers resistance to kanamycin); GUS, b-glucuronidase reporter gene; 3SA, triple splice acceptor; LB, RB left and right T-DNA borders; 1í, 2í T-DNA promoters; ocs 3í, nos 3í T-DNA transcription terminators; D35S, minimal promoter (-46 truncation of viral 35S promoter). DsG is the Gene trap transposon, DsE is the enhancer trap transposon.
Ds elements are transactivated by crossing to transgenic plants that provide a source of transposase, namely an immobilized Ac element (Figure 1). The Indole acetamide hydrolase (IAAH) gene has been incorporated in the T-DNAs carrying both the Ds elements and the Ac element. Selection against the IAAH gene using the herbicide analog napthalene acetamide (NAM), and for the Ds element using kanamycin allows the recovery of transposition events that have lost (by recombination) the donor locus, and thereby enriches for unlinked transposition events. Since we also select against the Ac element, the insertion is immediately stabilized.
Mutagenesis is initiated by crossing plants homozygous for one of the Ds elements to plants containing the Ac transposase gene (Figure 2). The resulting F1 seed are planted, and the plants are allowed to self-pollinate. The F2 seed are harvested from each individual F1 plant, and plated on media containing kanamycin (selects for the Ds element) and NAM (selects against both T-DNAs). The double resistant F2 seedlings (called transposants) contain a transposed Ds element. The NamR KanR seedlings are transplanted to soil, and allowed to self-fertilize. F3 seed are collected, and grown, and the resulting plants are stained for GUS activity and examined for mutant phenotypes.
Figure 2. Mobilization scheme for unlinked transpositions.Mutagenesis crosses and the expected frequency of mutagenized progeny are indicated in the scheme (Protocols 1 and 2). NAM, naphthalene acetamide; IAAH indole acetamide hydrolase; kan, kanamycin.
Equipment and Reagents
Generation of F1 and F2 seed
Protocol 2:Selecting transposants from F2 families
Equipment and Reagents
Selection of F2 transposants.
a. Wash with 1 ml 95% ethanol for 5 min. The seed will settle and the ethanol can be decanted off.
b. Add 1 ml of bleach solution to the seed, leave 5 min.
c. Fill tube with sterile water to dilute bleach solution. Allow seed to settle, and decant off . Wash once more with sterile water.
Screening transposant lines
Transposant lines can be screened for mutant phenotypes caused by insertion, and for patterns of reporter gene expression. In some cases, it may be beneficial to stain transposants first, and then use the staining pattern to guide phenotypic examination. However, it should be remembered that only a fraction of insertions into genes result in reporter gene expression, so that lines without reporter gene expression can still be very useful. Individual F2 plants can be either homozygous or heterozygous for the transposed Ds element, so any phenotype associated with the insertion could either be present in every F3 plant (in the case of a F2 homozygote) or could segregate 3:1 normal to mutant (if the F2 plant was heterozygous and the mutation is recessive). Segregation ratios substantially more than 3:1 may indicate that the mutation arose after transposition and is not associated with the insertion.
Lethal insertions can be scored most conveniently in F2 transposants, by opening developing siliques and scoring unfertilized ovules and colorless embryos (e.g. ). Lethal mutations in our lines arise at a frequency of about 4%, but should be carefully tested for heritability and association with the Ds insertion. This can be done by plating F3 seed on kanamycin. The Ds insertion carries a kanamycin resistance gene. Thus, if the insertion is responsible for the lethal phenotype, all resistant F3 plants (which are heterozygous for the insertion) should give rise to plants with semisterile siliques or defective embryos. Lethality in the gametophyte or in the embryo should result in poor transmission of the kanamycin resistant trait, with ratios of between 1:1 and 2:1 resistant to sensitive F3 seed.
Screening for reporter gene expression
Screening for reporter gene expression patterns should be done wherever possible in heterozygous plants or in segregating families that include phenotypically normal individuals. This is because homozygosity for the insertion can affect reporter gene expression patterns in many cases. It is wise to examine several individual plants, so that staining in the F3 is recommended. Potassium ferricyanide is included in the staining reaction in order to catalyze dimerization of the indigo monomer, which is the product of the glucuronidase reaction. The colorless monomer is soluble, while the dimer is not, resulting in a blue precipitate at the site of enzyme activity. In the absence of at least 1.5 mM ferricyanide, diffusion of the indigo monomer results in artefactual staining patterns. Unfortunately, ferricyanide ions also inhibit the GUS enzyme at these concentrations. This inhibition can be partially overcome by using long incubation times for weaker staining patterns. Clearing of the tissue following staining is most simply and gently accomplished using 70% ethanol, which we find adequate for all tissues we have examined. However, other more drastic clearing methods can be employed, but need to be carefully controlled with respect to re-dissolving the stain and damaging the tissue . Cleared tissue is mounted in glycerol for optimal Nomarski optics. An example of a line in which the reporter gene is expressed in young trichomes and root hairs is shown in Figure 3.
Figure 3. Enhancer trap reporter gene expression in trichomes and root hairs. In the individual enhancer trap line shown, the reporter gene is expressed specifically in unbranched trichomes (A) and developing trichoblasts (root hair cells, B). Tissue samples were stained and photographed according to Protocol 3.
Protocol 3.Screening F3 seedlings for reporter gene expression
Equipment and Reagents
Plating and Sterilization
Molecular analysis of transposants
Each transposant line that has a staining pattern or phenotype of interest can be analyzed molecularly to determine the location of the transpo sed element in the Arabidopsis genome. Flanking genomic DNA or cDNA corresponding to the gene nearest the insertion site can be rapidly obtained by PCR amplification. The resulting products can be sequenced directly, or used as hybridization probes for further analysis.
Insertions into T-DNA
Although most insertions are randomly distributed around the genome, about 5 percent of our transposants have Ds insertions into the IAAH gene on the T-DNA. These insertions disrupt the IAAH gene resulting in resistance to the negative marker NAM. IAAH insertions typically have strong ubiquitous reporter gene expression in seedling tissues and can often be discarded on the basis of this expression pattern or on the basis of further molecular analysis (see below). However, if this is not possible, flanking DNA can be sequenced as described below.
Amplification of flanking genomic DNA
Chromosomal sequences flanking both gene trap and enhancer trap insertions can be amplified using I (inverse) PCR or TAIL (thermal asymmetric InterLaced) PCR using standard protocols and primers from the Ds element. These PCR products can be sequenced directly after purification on spin columns or gels by cycle sequencing using dye terminators. Each of these methods depends on the fortuitous location of a primer sequence or restriction site close to the insertion, and consequently is only successful about 50% of the time for a given primer/enzyme combination. It is wise, therefore, to use several primer combinations or several approaches when attempting to amplify a given insertion site.
Amplification of cDNA from gene trap fusions
Gene trap insertions result in transcriptional fusions between the reporter gene and the chromosomal gene into which it is inserted. Consequently, flanking sequences can be amplified by 5íRACE PCR using RNA isolated from the gene trap line. This is useful when the transposon is inserted into a large intron, or when multiple introns make chromosomal sequence hard to interpret. RACE PCR products can also be sequenced directly, except that alternate splicing will result in mixed sequence reads in many cases. In these cases, subcloning will be required.
In either case, it is wise to confirm the assignment of a given sequence to a given line by hybridization of the products to genomic Southern blots, or by PCR using specific primers derived from the sequence. The genomic DNA purification procedure described below allows restriction digestion and Southern blotting as well as PCR.
Protocol 5.Preparation of genomic DNA
Flowers and inflorescences have much smaller cells than leaves or roots, and consequently, they have a lot more DNA. 5 µl DNA (0.5-1.0 µg) in a 20 µl digest is sufficient for a Southern blot with single copy probes. Spermidine (pH 7, 2.5 mM final concentration) greatly aids digestion.
Equipment and Reagents
Resuspend pellet in 360 µl TE, add 40 µl 3 M sodium acetate and 800 µl 100% ethanol (cold). Mix on ice and spin down again. Aspirate ethanol, briefly air-dry and resuspend pellet in 50 µl TE.
Protocol 6.Amplification of flanking DNA
Genomic DNA flanking the insertion site can be readily amplified by a number of PCR procedures. The most widely used methods are IPCR and TAIL PCR.
Inverse PCR is performed as described by Long et al. . Genomic DNA is digested with appropriate restriction enzymes and ligated under dilute conditions in order to generate circular fragments. PCR is then used to amplify DNA flanking the insertion site.
Equipment and Reagents
Thermal Asymmetric InterLaced PCR is performed as described by Liu et al and Tsugeki et al. . Three successive rounds of amplification are performed using three semi-nested primers from the Ds element and an arbitrary degenerate (AD) primer. This procedure results in products that can be sequenced directly using dye-terminator chemistry and the last semi-nested primer.
Equipment and Reagents
Repeat the entire procedure using primers Ds5-1 to 5-4 in place of Ds3-1 to Ds3-4. If no products are recovered, use another arbitrary degenerate primer (such as AD5) in place of AD2, or by using Ds3-2 in place of Ds3-1, and Ds3-3 in place of Ds3-2.
Protocol 7.Preparation of RNA from plant tissue
Equipment and Reagents
Make up 5 M NaCl, 3 M NaAc and 1 M magnesium chloride in double-distilled water. Treat with DEPC and autoclave as above. Make up 1 M Tris pH 8, 0.5 M EDTA pH 8 and 10% SDS in DEPC water (already autoclaved) and filter sterilize. Store in plastic tubes.
Protocol 8.Amplification of gene trap cDNA by 5í RACE PCR
Equipment and Reagents
cDNA Synthesis and Homopolymer Tailing
This protocol is described in great detail by Frohman . Alternative protocols (i.e. ) may also be used.
Concentrate in SpeedVac to 10 m l. Add 4 m l 5X TdT buffer (supplied), 4 m l 1 mM dATP and 1 m l TdT. Incubate 5 minutes at 37ƒ C followed by 5 minutes at 65ƒ C to stop the reaction.
2nd Strand Synthesis and Amplification
Products are purified using QiaQuick PCR purification columns (Qiagen). They can be cloned into an appropriate vector by digesting with BamHI (site present in GUS2) and HindIII (present in QI).
Genetic analysis of transposants
In many cases, insertion of a transposon will be associated with a mutant phenotype. However, spontaneous mutants arise at a surprisingly high frequency in transposon lines, and it is important to determine whether the transposon is responsible for any observed phenotype.
The most important application of the second approach is in the analysis of revertants. The restoration of a wild-type phenotype when the transposon excises provides strong evidence that the transposon was responsible for the mutation. A second application is in the disruption of nearby genes. These procedures are described below.
Mapping of transposed elements can be most readily accomplished molecularly, by amplifying flanking DNA (see above) and using the resulting PCR products as probes to hybridize to anchored libraries, or to Southern blots of DNA from mapping populations such as recombinant inbred lines. Most powerfully, the sequence of the Arabidopsis genome is accumulating rapidly as the Arabidopsis Genome Initiative nears conclusion. Currently, this means that 25-30% of all flanking sequences can be mapped immediately by matching PCR products with genomic sequence.
Phenotypically, the DsG and DsE transposons each carry a kanamycin-resistance gene. This means that each insertion can be mapped with respect to any associated mutant phenotype, as well as to previously mapped phenotypic and molecular markers. Plants heterozygous for the insertion are outcrossed to wild-type plants, and F1 progeny are allowed to self-pollinate. F2 families are sown once to screen for any mutant phenotype, and again to screen for kanamycin resistance. Mutations that are caused by the insertion will only be found in F2 families with kanamycin resistant progeny. If some kanamycin resistant families have no mutant progeny, this might indicate poor penetrance of the mutation, or the presence of a second insertion elsewhere in the genome. If the insertion causes a lethal mutation, the ratio of kanamycin resistant to kanamycin sensitive seedlings should be less than 3:1 on self-pollination of a heterozygous F1 plant. The ratio will be 2:1 for an embryo lethal, and 1:1 for a gametophyte lethal (18).
By using a wild-type parent from a different ecotype (Columbia), F2 seed can be used to map the insertion. This is accomplished by preparing DNA from pooled F2 seedlings from kanamycin resistant and kanamycin sensitive families, and screening the DNA samples with PCR-based polymorphic markers.
Insertion of a Ds transposon results in the duplication of 8 bp of target sequence immediately flanking the insertion site. When Ds excises, the target duplication is partially removed, resulting in small insertions and deletions at the original locus. If the Ds is inserted into the coding region of the gene, only those events that restore the reading frame and result in a functional protein will revert the mutant phenotype back to wild-type. In contrast, almost all reversions from non-essential sequences such as introns will result in reversion of the mutant phenotype.
Reversion is accomplished by crossing mutant plants to transgenic plants that carry the transposase gene. The resulting F1 plants are then planted and allowed to self-pollinate. The F2 progeny will now include mutants that carry the transposase gene. The transposon responsible for the mutant phenotype will excise in these plants resulting in somatic sectors of tissue that have lost the transposon. If these plants are mosaic for the mutant phenotype, this is good evidence that the phenotype can be reverted by transposase. More importantly, a proportion of the F3 progeny of these plants should be wild-type, in contrast to the progeny of mutants that do not carry the transposase gene, which should be true-breeding mutant. Revertant alleles can be amplified from wild-type progeny using primers that flank the insertion site. These products can then be sequenced to determine the nature of the reversion event. In the special case of lethal mutations, reversion can be observed in the F1 plants themselves. Most of the siliques on these plants should be semisterile, or carry dead seed, depending on whether the mutation is lethal at the gametophytic or the embryonic stage. However, reversion early in development will result in normal, fully fertile siliques provided reversion occurs early enough to detect revertant branches.
There are many circumstances when reinsertion of a transposon by short-range transposition is advantageous. For example, enhancer trap transposon insertion frequently leads to reporter gene expression even if the insertion does not disrupt the gene (a "near-miss"). This is because many enhancers can act at some considerable linear distance from target (reporter) promoters. Even gene trap insertions may not disrupt gene function, as insertions within introns may be spliced from the RNA transcript without phenotypic effect. In these cases, it can be useful to obtain a secondary insertion into the nearest exon by inducing a short-range transposition. A high proportion of transpositions of Ds are to closely linked sites when these transpositions are not counter-selected (see introduction). A protocol for remobilization is given below.
In brief, the transposon is remobilized by crossing to transposase, and is then stabilized by selecting against the transposase gene in the next generation (Figure 4). The parental transposon is not selected against by this procedure, and so a large proportion (more than half) of the resulting plants will still have the transposon inserted at the original location. Those plants that have new insertions therefore need to be identified molecularly, phenotypically, or by staining for reporter gene expression. In our experience, a collection of 2,000 plants selected in this manner will carry between 500 and 1,000 new transpositions. About 20% of these will be within 100 kb which should be sufficient to saturate the nearby region with new insertions (C. Yordan and R. Martienssen, unpubl. observations).
Figure 4. Remobilization scheme for linked transpositions.Mutagenesis crosses and the expected classes of mutagenenized progeny are shown as described in Prtotocol 10. In this case, the Ds element at the original locus is not flanked by the IAAH gene, so that Ds at the original locus and linked transpositions are not selected against. They will therefore predominate among the F2 progeny. F2 plants carrying the Ac transposase gene are selected against in order to stabilize transposed elements. NAM, naphthalene acetamide; IAAH indole acetamide hydrolase; kan, kanamycin.
Protocol 10.Remobilization of transposed elements
1. Feldmann, K. A. and Marks, M. D. (1987). Mol. Gen. Genet., 208, 1.
2. Bechtold, N., Ellis, J., and Pelletier, G. (1993). Comptes Rendus de l'Academie des Sciences Serie III Sciences de la Vie, 316, 1194.
3. Azpiroz-Leehan, R. and Feldmann, K. A. (1997). Trends Genet., 13, 152.
4. Wilson, C., Bellen, H. J., and Gehring, W. J. (1990). Ann. Rev. Cell Biol., 6, 679.
5. Skarnes, W. C. (1993). Current Opinion in Biotechnology, 4, 684.
6. O'Kane, C. J. and Gehring, W. J. (1987). Proc. Natl. Acad. Sci. USA, 84, 9123.
7. Bellen, H. J., O'Kane, C. J., Wilson, C., Grossniklaus, U., Pearson, R. K., and Gehring, W. J. (1989). Genes Dev., 3, 1288.
8. Bier, E., Vaessin, H., Shepherd, S., Lee, K., McCall, K., Barbel, S., Ackerman, L., Carretto, R., Uemura, T., Grell, E., Jan, L. Y., and Jan, Y. N. (1989). Genes Dev., 3, 1273.
9. Mlodzik, M., Hiromi, Y., Weber, U., Goodman, C.S., and Rubin, G.M. (1990) The Drosophila seven up gene, a member of the steroid receptor gene superfamily, controls photoreceptor cell fates. Cell 60:211-224.
10. Wilson, C., Pearson, R. K., Bellen, H. J., O'Kane, C. J., Grossniklaus, U., and Gehring, W. J. (1989). Genes Dev., 3, 1301.
11. Friedrich, G. and Soriano, P. (1991). Genes Dev., 5, 1513.
12. Wurst, W., Rossant, J., Prideaux, V., Kownacka, M., Joyner, A., Hill, D. P., Guillemot, F., Gasca, S., Cado, D., Auerbach, A., and Ang, S.-L. (1995). Genetics, 139, 889.
13. Takeuchi, T. (1997). Develop. Growth Differ., 39, 127.
14. Townley, D. J., Avery, B. J., Rosen, B., and Skarnes, W. C. (1997). Genome Research, 7, 293.
15. Kertbundit, S., De Greve, H., Deboeck, F., Van Montagu, M., and Hernalsteens, J.-P. (1991). Proc. Natl. Acad. Sci. USA, 88, 5212.
16. Topping, J. F., Wei, W., and Lindsey, K. (1991). Development, 112, 1009.
17. Fedoroff, N. V. and Smith, D. L. (1993). Plant J., 3, 273.
18. Springer, P. S., McCombie, W. R., Sundaresan, V., and Martienssen, R. A. (1995). Science, 268, 877.
19. Sundaresan, V., Springer, P., Volpe, T., Haward, S., Jones, J. D. G., Dean, C., Ma, H., and Martienssen, R. (1995). Genes Dev., 9, 1797.
20. Tsay, Y.-F., Frank, M. J., Page, T., Dean, C., and Crawford, N. (1993). Science, 260, 342.
21. Bancroft, I., Bhatt, A. M., Sjodin, C., Scofield, S., Jones, J. D. G., and Dean, C. (1992). Mol. Gen. Genet., 233, 449.
22. Aarts, M. G. M., Dirkse, W. G., Stiekema, W. J., and Pereira, A. (1993). Nature, 363, 715.
23. Long, D., Martin, M., Sundberg, E., Swinburne, J., Puangsomlee, P., and Coupland, G. (1993). Proc. Natl. Acad. Sci. USA, 90, 10370.
24. Aarts, M. G., Corzaan, P., Stiekema, W. J., and Pereira, A. (1995). Mol. Gen. Genet., 247, 555.
25. Honma, M. A., Baker, B. J., and Waddell, C. S. (1993). Proc. Natl. Acad. Sci., USA, 90, 6242.
26. Altmann, T., Felix, G., Jessop, A., Kauschmann, A., Uwer, U., Pena-Cortes, H., and Willmitzer, L. (1995). Mol. Gen. Genet., 247, 646.
27. Bhatt, A. M., Page, T., Lawson, E. J., Lister, C., and Dean, C. (1996). Plant J., 9, 935.
28. Long, D., Goodrich, J., Wilson, K., Sundberg, E., Martin, M., Puangsomlee, P., and Coupland, G. (1997). Plant J., 11, 145.
29. Cardon, G. H., Frey, M., Saedler, H., and Gierl, A. (1993). Plant J., 3, 773.
30. Swinburne, J., Balcells, L., Scofield, S. R., Jones, J. D. G., and Coupland, G. (1992). Plant Cell, 4, 583.
31. Lawson, E. J., Scofield, S. R., Sjodin, C., Jones, J. D., and Dean, C. (1994). Mol. Gen. Genet., 245, 608.
32. Das, L., and Martienssen, R.A. (1995) Plant Cell, 7, 287.
33. Fedoroff, N. V. (1995). Current Topics in Microbiology and Immunology, 197, 143.
34. Chomet, P. S., Wessler, S., and Dellaporta, S. L. (1987). EMBO J., 6, 295.
35. Machida, C., Onouchi, H., Koizumi, J., Hamada, S., Semiarti, E., Torikai, S., and Machida, Y. (1997). Proc. Natl. Acad. Sci. USA, 94, 8675.
36. Jones, J. D. G., Carland, F. M., Maliga, P., and Dooner, H. K. (1989). Science, 244, 204.
37. Long, D., Swinburne, J., Martin, M., Wilson, K., Sundberg, E., Lee, K., and Coupland, G. (1993). Mol. Gen. Genet., 241, 627.
38. Bancroft, I. and Dean, C. (1993). Genetics, 134, 1221.
39. Keller, J., Lim, E., and Dooner, H. K. (1993). Theor. Appl. Genet., 86, 585.
40. Klee, H. J., Horsch, R. B., Hinchee, M. A., Hein, M. B., and Hoffmann, N. L. (1987). Genes Dev., 1, 86.
41. Stougaard, J. (1993). Plant J., 3, 755.
42. O'Keefe, D. P., Tepperman, J. M., Dean, C., Leto, K. J., Erbes, D. L., and Odell, J. T. (1994). Plant Phys., 105, 473.
43. Dotson, S. B., Lanahan, M. B., Smith, A. G., and Kishore, G. M. (1996). Plant J., 10, 383.
44. Jefferson, R. A., Burgess, S. M., and Hirsh, D. (1986). Proc. Natl. Acad. Sci. USA, 83, 8447.
45. Sylvester, A. W. and Ruzin, S. E. (1994). In The Maize Handbook (ed. M. Freeling and V. Walbot), pp. 83-95. Springer-Verlag, New York.
46. Gallagher, S. R. (1992). GUS protocols: Using the GUS gene as a reporter of gene expression. Academic Press, Inc., San Diego.
47. Klimyuk, V. I., Carroll, B. J., Thomas, C. M., and Jones, J. D. (1993). Plant J., 3, 493.
48. Liu, Y.-G., Mitsukawa, N., Oosumi, T., and Whittier, R. F. (1995). Plant J, 8, 457.
49. Tsugeki, R., Kochieva, E. Z., and Fedoroff, N. V. (1996). Plant J, 10, 479.
50. Frohman, M. A. (1993). Methods in Enzymology, 218, 340.
51. Chen, Z. (1996). Trends Genet., 12, 87