Plant Cell Reports

Plant Cell Reports (1996) 15:815-818

9 Springer-Verlag 1996

DNA methylation and Dc8-GUS transgene expression in carrot (Daucus carota L.) Yuanxiang Zhou 1, J . M . M a g i l l 2, C . W . Magill 2, and R . J . Newton 1 1 Department of Forest Science, Texas A&M University, College Station, TX 77843, USA 2 Department of Plant Pathology and Microbiology, Texas A&M University, College Station, TX 77843, USA Received 29 August 1995/Revised version received 30 January 1996 - Communicated by G. C. Phillips

Smnmary. DNA methylation has been associated with gone activity in differentiating and developing plant tissues. The objective of this study was to determine the involvement of methylation in the expression of a gone transferred into carrot (Daucus carets L.) tissues by particle bombardment. Expression of the Dc8-GUS gone construct in response to treatments with 5-azacytidine (5-azaC) and to /n vitro methylation by methylases was investigated by hlstochemical assay of GUS activity. The 5-azaC treatment increased the frequency of Dc8-driven GUS expression in both calli and somatic embryos. The increase occurred with treatment either to E. cell containing the plasmid insert or to the carrot tissues before bombardment. GUS expression, increased by the 5-azaC treatment, was enhanced by ABA treatment of both calli and somatic embryos and was more prominent in the latter. Increased digestion of the 5-azaC-treated plasmid DNA with Ec~RII suggested that demethylation had occurred. In vitro methylation of Dc8-GUS by methylases generally resulted in a lower frequency of GUS expression. 3ssI methylase completely inhibited GUS expression. The level of GUS expression was correlated with the extent of methylation of the plasmid. Abbreviations: ABA, Abscisic Acid -- 5-azaC, 5-azacytidine -- GUS, 13-glucuronidase -- DcS, carrot promoter

Introduction DNA methylation oecm~ widely in plants with cytosine being the most commonly methylated base. Cytosine methylation occurs in both the CpG dinucleotido and the CpNpG (N is any base) trinucleotide. The average percentage of total cytosines that are methylated generally ranges from 20-30% but can be as high as 40% (Wagner and Caposius 1981; Vanyushin et al. 1968). DNA methylation has boon shown to have a negative effect Correspondence to: R. J. Newton

on gone expression. Studies of genes carried on T-DNA in transformed plant cells (IOaas et al. 1989; van Slogteren 1984), transposon activity in maiTe (Brettell and Dennis 1991; Kunzo et al. 1988) and genes that code for photosynthetic proteins (Watson et al. 1987) have all suggested an inverse correlation between methylation and gene activity. In this paper, we report the transient expression of a Dc8GUS chimeric gone with different methylation patterns following biolistic introduction into carrot cells. The De8 promoter was obtained from an ABA-induc~le LEA (late embryogenesis abundant) gene cloned from carrot (Daucus carota L.) (Goupil et al. 1992). The objectives were to determine: 1) the relationship between methylation in DcSGUS and its expression, 2) if 5-azaC prevents Dc8-GUS from being silenced after transformation, and, 3) ff ABA influences the expression of Dc8-GUS.

Materials and methods Tissue Culture. The W001 carrot cell line which was provided by Dr. Z.R. Sung (University of California, Berkeley) was maintained as callus on MS 0VIursshige and Skoog 1962) solid medium plus 1 pM 2,4dichioro-phenoxyaeetic acid (2,4-D). Somatic embryos were induced by culturing caili on MS solid medium with 1 v M l-nsphthaleneaeefic acid (NAA) and 1/zM abscisic acid (ABA) under darkness. Calli and somatic embryos near or at the globular stage with no visible cotyledon primordia were used for particle bombardment. Plasmid Construction. The De8 promoter o f carrot (Hatzopoulus et al. 1990) which was also provided by Dr. Z.R. Sung was inserted into PRT99-GUS (Topfer et al. 1988) between the EcoRI and BamHI sites, the site where the CaMV 35S promoter had been removed. The resulting PRT99-Dc8-GUS plasmid was then transformed into E. co/i strain XL1Blue, and was further propagated in LB medium with or without (control) 50 pM 5-azaC at 37~ for 20 h. The cells were collected by centrifugation and the plasmid was extracted from the pellet. 5-Azacyadine and ABA Treamwnts o f Carrot Cells. Five days before bombardm 9 calli and embryos were transferred from MS solid medium into 250 mi flasks containing 100 ml MS liquid medium with lp.M 2,4-D and lptM N A A + Ip.M ABA, respectively,and were cultured

816 in darkness on a rotary shaker at 250 rpm. Experiments were performed in two phases. In phase 1, replications were made by taking carrot cells from different flasks; while in phase 2, they were made by taking samples from the same flask. For both calla and somatic embryos, the treatments were: nontreated (control), 5/aM ABA, 50 ~M 5-azaC, and 5/~M ABA + 50 laM 5-szaC. For phase 2 experimentation, treatment with ABA and 5-azaC together for 1 d was also included. For this treatment, the ABA and 5-azaC were added 4 d after culturing either calla or somatic embryos in liquid medium. Five d after liquid culturing, the cells were passed through a filter with the appropriate mesh size so that the unscattered calla and/or somatic embryos were retained on the filter. Ten ml of the filtered mixture were uniformly distributed onto filter paper (diameter 5.5 cm) which was then placed on solid MS medium in petri dishes with the same treatment composition and maintained for bombardment.

Particle Bombardment and GUS Assay. Plasmld D N A was adhered to pretreated, 1.6/an gold particles (1 #g DNA/1 mg of particles) by CaCl 2 and spermidine precipitation. For each bombardment, 0.5 nag of particles with coated DNA were placed on a macroearrier and bombarded onto either calla or somatic embryos placed on filter paper (see above) under 650 psi helium pressane and 28.5" Hg vacuum, using the Dupont Biolistic particle delivery system (PDS-1000He). Afler bombardment, the petri dishes were incubated at room temperature for 48 h. Then, transient GUS expression was performed. For Phase 1 experimentation and Phase 2 in vitro methylation effects (see below), 0.5 ml GUS assay buffer (Jefferson et al. 1987) was added to the surface of the filter paper, and the dishes were incubated at 37"C for 24 h. For the other Phase 2 experiments, the dishes were maintained at room temperature for 5 days

after adding 0.5 ml GUS assay buffer (Kirehner et al. 1993). The numbers of transformants expressing GUS activity were determined by counting the number of blue spots visible under light microscopy.

Methylation Assay of dam and dcm Sites in Plasmid DNA. The PRT99DcS-GUS plasmid from 5-azaC-treated and nontreated E.coli strain XL1Blue was digested by MboI (sensitive to 6mA in the dam site: GATC) and EcoRlI [sensitive to 5mC in the inner C of the dcm site: CC(A/T)GG]. The resulting fragments were then separated on 0.8% agarose gel and visualized under UV light after ethidium bromide staining. Site-Specific Methylation and Dc8-GU$ Expression. The PRT-99-DeSGUS from the dcm- and dam-deficient E. cola strain (JMII0) was methylated by methylases under the conditions recommended by the vendor (New England Biolabs). For phase 1 experimentation, the enzymes used were: SssI, A/uI, HhaI and MspI methylases; while for phase 2, the enzymes were S,vsI, A/uI, MspI and TaqI mcthylases. The non-methylatedand methylase-treated plasmids were then bombarded into carrot embryos, and transient GUS expression was assessed using the procedures described above. Dry Weight Measurement. The filter papers (diameter 5.5 cm) used for distributing carrot tissue and subsequent GUS assay were dried at 80"C for 24 h, cooled to room temperature in a desiccator and weighed. After GUS assays, they were dried again at 80"C for 24 h with carrot tissue on them, and then weighed. The difference between the two measurements was taken as the dry weight of the carrot tissue.

Table 1. 5-azatcytidinetreatment of carrot tissues wadE. coli contaiffmg PRT99-DeS-GUS on the transient expression of DeS-GUS after bombardment Control Carrot Tissue Phase

Tissue

Plasmid (n)

Callaa

One

Mean +/- SD

Control

(3)

60.0

5-azaC

(3)

88.7

Control

(2)

483.5

5-azaC

(2)

537.0

% Inorease Over Control Plasmid

47.8

5-azaC-~.ated Carrot Tissue (n)

Mean +/- SD

(2)

148.0

(3)

358.7

% Inorease Over Control Plasmid

142.0

a

Somatic ombr~s

11.1

(2)

1184.0

(2)

1320.5

11.5

a

Calli

Control

(5)

15.4 +/- 4.2

5-azaC

(6)

14.0 +/- 3.4

Control

(6)

106.5+/- 14.3

5-azaC

(6)

120.8+/- 20.3

Control

(6)

0.7 +/- 0.2

5-azaC

(6)

0.7 +/- 0.2

0.0

Control 5-azaC

(6) (6)

4.4 +/- 1.1 4.9 +/- 0.9

9.3

-9.1

(5)

112.8+/- 34.7

(6)

122.7 +/- 41.9

(6)

423.8 +/- 47.8

(6)

483.0 +/- 42.2

(5)

4.1 +/- 1.4

(6)

4.8 +/- 1.8

18.9

(6) (6)

14.8 +/- 3.0 16.9 +/- 1.8

14.6

8.8

a

Somatic embryos Two Calli

b

Somatic emb~,os

b

11.9

13.9

a: Transient expression is expressed as number of blue spots per petri dish b: Transient expression is expressed as number of blue spots per mg dry weight SD: Standard Deviation which is given only for phase 2 experiment Numbers in the parenthesis represent replications Control plasmid was obtained fronaE, coli not lre__atedwith 5.-azaeytidine,5-azaC plasmid was from E. coli treated with 5-azaC P>0.01 for the difference between the 5-azaC-Uealedcarrot and the control

817 Results and Discussion

To determine ff methylation affects Dc8-driven GUS expression in transformed caw and somatic embryos, the effects of 5-azaC treatments prior to bomhaxdment of both the carrot tissue and the E. coli bearing the plasmid to be transformed were investigated. When the demethylating agent was used to treat carrot tissues, the Dc8-GUS activity was significantly increased in both calli and embryos, and the increase tended to be more significant in calli than in embryos. In the case of treating E. cell, there was a general trend for the plasmid from the 5-azaC-treated E. coli to be more active in expressing Dca-GUS regardless of whether the carrot tissues had been treated with 5-azaC or not (Table 1). It is understandable that 5-azaC treatment of replicating E. coli results in a reduced level of methylation to the plasmid in the E. coli cells, hence an increased activity of the gene inserted in the plasmid after transformation. As for the treatment of carrot tissues with 5-azaC, the methylation level of the PRT99-Dc9-GUS construct will not be influenced unless it is integrated into the carrot genome and replicates in the mitotic process. Therefore, the effects of 5-azaC on GUS expression, when used to treat carrot tissues, appear to be indirect. Since there have been reports indicating the inh~itory effects of 5-azaC on DNA methylases in bacteria and animals (Creusot et al. 1982; Santi et al. 1983; Jnttennann et al. 1991), we speculate that one of the roles of 5-azaC, when used to treat carrot tissue, is to prevent the transformed plasmid from being methylated due to the reduced methylase activity.

Digestion of plasmids from both 5-azaC-treated and nontreated E. coli strain XL1-Bhe with MboI, which is sensitive to 6mA in the GATC and EcoRII which is sensitive to 5mC of the inner C of CC(A/T)GG, revealed that plasmids from the non-treated E. coli were completely methylsted at both sites while those from the treated E. coli were partially demethylated at the dcm sites and remained methylated at the dam sites (Fig. 2) Demethylation at the dcm sites of the plasmid from 5-azaC-treated E. coli may account for the increased transient GUS expression. The results from restriction digestion and GUS transient activity 3500 3000

El

9-

2500

[]

b

-~

2000

[]

d

~. 500 D00 5OO 2000 _

[] ~

~

[]

b

[]

d

1000 500 -

J

o

Compared to controls, Dca-driven GUS expression was more frequent in both call/ and somatic embryos in the presence of ABA, and was more prevalent in somatic embryos (Fig. 1). When both 5-azaC and ABA were added to the liquid culturing media, GUS activity was higher than either single treatment (Fig. 1). Even a 1 d treatment siL,nificantly increased GUS activity (P > 0.01). Furthermore, the effects were increasingly apparent when cells were treated from 1 to 5 d. The De8 promoter is known to have an ABA-responsive element (ABRE) (Guiltinan et al. 1990). Therefore, a possible mechanism for the increased GUS expression after ABA treatment of carrot tissue might be that ABA enhanced the binding of ABA-responsive proteins to ABRE in the De8 promoter. Treatment of calli with either 5-azaC alone or 5-azaC and ABA together significantly increased the GUS transient expression (P>0.01) while ABA alone did not (P0.01

Ref~ Fig. 2. Methylation at dcm and dam sites of PRT99-Dcg-GUS from 5azacytidme 1rested aad nontreated E. coli strfm XL1-Blue, as shown by Er andMboI digestions. PRT99-Dcg-GUS plasmids were extracted fromE, eoli cells that were cultured in LB medium with or without 50 pM 5-azacyfidinsfor 20 k The plamaid was first linearized byHindIII digestion followed by pte~ai~on and resuspension, and then was digested by EcoRII andMboL M: Lambda DNA digested byHindIII; lanes I and 2, plasmid from E. coli not treated with 5-azacytidine; lanes 3 and 4, plasmid from E. r whh 5-azacytidine. Lanes 1 and 3,MboI digestion; lanes 2 and 4,EcoRII digestion, lames I, 2 and 3, 300 ng DNA; lane 4, 500 tag DNA.

In conclusion, we have shown that DeS-driven GUS expression ~om a transgene introduced by particle bombardment into carrot tissues was enhanced with treatment by 5-azaC, a dernothylating agent. The enhancement was associated with presmned demethylation of the transgene. Furthermore, the expression enhanced by 5-azaC was even greater when the transformed tissues were incubated with ABA, and was higher than when either of them was used alone. Also, we have shown that in vitro methylation by methylases prior to bombardment resulted in reduced transgene expression.

Brettell ILLS, Dennis ES (1991) Mol Gen Genet 229:365-372 Creusot F, Acs G, Christman JK (1982) J Biol Chem 257:2041-2048 Gonpil P, Hatzopoulos P, Franz G, Hempel FD, You R, Sung ZR (1992) Plant Mol Blol 18:1049-1063 Guillinan MI, Marcotte WR, Quatrano RS (1990) Science 250:267-271 Hatzopoulos P, Fong F, Sung ZR (1990) Plant Physiol 94:690-695 Jefferson RA, Kavanagh TA, Bevan MW (1987) EMBO J 6:3901-3907 Juttermann R, Li E, henisch R (1991) Proc Nail Acad Sci USA 91:11797-11801 Kirohner G, Kinslow CJ, Bloom GC, Taylor DW (1993) Plant Idol Biol Pep 11:320-325 Kisas M, John MC, Crowell DN, Amasino RM (1989) Plant Mnl Biol 12:413-423 Kun~ R, Stadinger P, Schwartz D (1988) Mol Gen Genet 214:325-327 Murashige T, Skoog F (1962) Physiol Plant 15:473-497 Palmgren G, MaUson O, Okkelm FT (1993) Plant Mol Biol 21:429-435 Santi DV, Garrett CE, Barr PJ (1983) Cell 33:9-10 Topfer R, Schell J, Steinbiss HH (1988) Nucleic Acids Res 16:8725 Van Slogteren GMS, Hooykaas PJJ, Schilperoort RA (1984) Plant Mol Blol 3:333-336 Vanyushin BF, Belozersky AN, Kokurlna, NA, Kadirova DX (1968) Nature (London) 218:1066-1067 Wagner I, Capesius I (1981) Bloohim Biophys Acts 654:52-56 Watson JC, Kaufman LS, Thompson WF (1987) J Mol Biol 193:15-26 Weber H, Graossmann A (1989) FEBS I,eU 253:163-166

DNA methylation and Dc8-GUS transgene expression in carrot (Daucus carota L.).

DNA methylation has been associated with gene activity in differentiating and developing plant tissues. The objective of this study was to determine t...
489KB Sizes 0 Downloads 0 Views