Absence of detectable 5-methylcytosine in DNA of embryos of the brine shrimp, Artemia

DEVELOPMENTAL

BIOLOGY

102,264-267(1084)

Absence of Detectable !I&Methylcytosine in DNA of Embryos of the Brine Shrimp, Artemia A.H. WARNER* AND J.C. BAGSHAW~ * Department of Biology, University of Windsor, Windsor, Ontario N$B SP.& Canudq and fDepartment of Bicchistry, Wayne State University School of Medicine, Detvit, Michigan .&WI Received July 26, 1989;accepted in revised fwm September $0, 1983 DNA from three developmental stages of the brine shrimp, Artemia was found to be undermethylated compared to DNA from calf thymus and salmon sperm. Using high-performance liquid chromatography we found that DNA hydrolysates from both dormant cysts and prefeeding nauplii contain less than 1 residue of 5-methylcytosine per 59 kilobases, placing Artemia DNA in the “insect type” category. The absence of detectable 5methylcytosine in DNA of developing Artemia supports the view that methylation status alone cannot account for regulation of transcription in protostomes, and that DNA methylation may be more common among deuterostomes. INTRODUCTION

The DNA of a wide variety of eukaryotic cells contains 5methylcytosine in amounts ranging from a few percent of total cytosine in mammals to as much as 33% of total cytosine in some plants (Razin and Riggs, 1980; Adams and Burdon, 1982). In animals, 5-methylcytosine (5MeCyt)’ is found almost exclusively in CpG dinucleotide sequences, which occur in the recognition sequences of several restriction endonucleases and differential cleavage with appropriate pairs of enzymes can be used to distinguish methylated from nonmethylated sites (Bird and Southern, 1978). On this basis Bird and Taggart (1980) have categorized animal DNAs as vertebratetype (frequent 5-MeCyt), echinoderm-type (intermediate 5-MeCyt), or insect-type (rare 5-MeCyt). Indeed, UrieliShoval et aL (1982) have recently reported finding no detectable 5-MeCyt in Drosophila DNA using several very sensitive analytical techniques. The function of 5-MeCyt in eukaryotic DNA remains undefined. Clearly, methylation of specific cytosine residues would significantly alter the microenvironment of functional groups in the major groove of the DNA double helix. Attempts to correlate quantitative changes in 5MeCyt content with differentiation in Xenopus and in sea urchins have been inconclusive (Bird et al, 1981; Pollock et aL, 1978). More recent studies of structure and expression of specific genes have led to the general

1Abbreviations used: .5-MeCyt, 5-methylcytosine; ‘I-MeGua, 7methylguanine; 2-MeAde, 2-methyladenine; 6-Me-n-Pur, 6-methylaminopurine; Thy, thymine; Gua, guanine; Cyt, cytosine; Ade, adenine. 6012-x06/&

$9.90

Copyrisht 0 1984 by Academic Press, Inc. All righhts of reproduction in my form maewed.

hypothesis that genes that are transcriptionally active are undermethylated, whereas genes that are transcriptionally inactive are fully methylated (reviewed by Razin and Riggs, 1980 and Weisbrod, 1982). Hypomethylation of specific CpG sequences has also been correlated with structural differences in chromatin in the vicinity of active genes (Weisbrod, 1982). However, hypomethylation alone may not be sufficient to account for transcriptional activity (Adams and Burdon, 1982; Weisbrod, 1982). In contrast to the apparent ubiquity of methylated DNA among vertebrates and echinoderms, 5-MeCyt is largely or entirely absent in DNAs of insects. Restriction endonuclease cleavage has failed to detect methylated CCGG sequences in DNA from several genera and species (Adams et oL, 19’79;Cedar et CL&1979; Rae and Steele, 1979;Eastman et al, 1980). Eastman et al. (1980) reported reaction of antibodies against 5-MeCyt with specific bands of dipteran salivary gland polytene chromosomes, but could find no methylated CCGG sequences by restriction analysis. Urieli-Shoval et al. (1982), using nearest-neighbor base analysis, found no detectable amount of 5-MeCyt (~0.1% of all CpG sequences) in DNA from embryos, larvae, pupae, adults, salivary glands, and cultured cells of D. m&anogaster. It appears that 5-MeCyt is extremely rare or nonexistent in insect DNA. The extent to which this absence may be characteristic of protostome taxa, as compared with the deuterostome taxa typified by echinoderms and vertebrates, remains unknown. We report here the absence of detectable 5-MeCyt in the DNA of embryos and larvae of the brine shrimp, Artmiu, a crustacean which is phylogenetically related to the insects.

264

265

BRIEF NOTES

MATERIALS

AND METHODS

DNA Isolution Dormant encysted embryos of the brine shrimp, Artmiu, were purchased from Metaframe Corp. (Newark, Calif.), and sterilized by antiformin treatment as described by Nakanishi et al. (1962). Nauplius larvae were collected 40 or 48 hr after immersion of encysted embryos in sterile artificial seawater (Warner et al, 1979). DNA was prepared by a modification of the method of Vaughn and Petropoulos (1979). Nuclei isolated from embryos or larvae as previously described (Birndorf et a& 1975) were suspended in 10 mlMTris-HCl, pH 8.0,l mM EDTA (TE buffer). To this suspension was added sodium dodecyl sulfate to a final concentration of 1% (w/v) and proteinase K (Sigma) to 50 pg/ml. The resulting viscous lysate was incubated 4-6 hr at 60°C with occasional gentle shaking, then extracted overnight at room temperature with phenol/chloroform (4/l; v/v) with constant gentle shaking, Following centrifugation at 10,OOOg for 10 min to separate the phases, the aqueous phase was reextracted with phenol/chloroform for an additional 6-8 hr. The aqueous phase was then dialyzed against two changes of TE buffer followed by one change of 0.1 1Msodium acetate, pH 4.5. RNase Tz (Sigma) was added to 1 unit/ml, and the sample was incubated 4 hr at 37’C. Following addition of Tris base to 0.1 iIf, cyamylase (Sigma) to 25 pg/ml, and RNase A (Sigma) to 50 pg/ml, the sample was incubated an additional 2 hr at 37”C, then treated with proteinase K as above, extracted twice with phenol/chloroform as above and dialyzed against two changes (24 hr each) of TE buffer at 5’C. Salmon sperm and calf thymus DNAs were purchased from Sigma. Purified DNA (lo-20 Azsounits) was precipitated from TE using 2 vol of ethanol and the precipitate was collected by centrifugation and air dried for 30 min. The precipitate was suspended in trifluoroacetic acid and heated in sealed vials in an oven at 100°C for 4 hr. At the end of this period the trifluoroacetic acid was removed by flash evaporation at 40°C followed by further evaporation of the residue from water. The residue was suspended in 0.5 to 1.0 ml of 0.10 M ammonium phosphate buffer, pH 3.25, and saved for base analysis by high-performance liquid chromatography (HPLC). Analysis of DNA Hydrolysates Using HighPerfomLance Liquid Chromatography Quantities of DNA hydrolysates ranging from 0.3 to 4.3 Azsounits (22 to 350 nmole of bases) in the starting buffer were applied to a 0.4 X 25-cm steel column of Ultrasil CX (Altex) previously equilibrated with 0.10 M

ammonium phosphate buffer, pH 3.25, and the column was eluted isocratically with the same buffer at a flow rate of 1.0 ml/min using a microprocessor (Altex/Beckman, Model 421) controlled high-pressure pump (Altex/ Beckman, Model 1lOA). The column effluent was monitored at 254 and 230 nm using a 20-@IAow cell-detector system (Altex/Beckman, Model 153) and the signal was recorded on 20-cm paper at sensitivities from 0.005 to

100 *

A. STANDARD BASES (0.16AUFS) Thy Cyt 7-MeGua Z-MeAde 6-Me-n-Pur AiJe Is-rfleCyt 1

80. 60.

8. 6ALMON DNA (0.32AUFS)

E

t

8 60. (u

\ 16

s40.

20

24

.

1

\ C. ARTEMIA

-

*

DNA (0.32 AUF.9

I fO.005 AUFS

I1

80. 60. 16

40.

AAL 4

8

L , 12 “DL”%

20

24

4 20

24

28

(YL)

FIG. 1. High-performance liquid chromatography of DNA bases. DNA was hydrolyzed to the free bases as described under Materials and Methods. Twenty microliters of the hydrolysate, corresponding to about 22-35 nmole of bases, was chromatographed on Ultrasil CX, as described under Materials and Methods. The effluent was monitored at 254 nm (solid lines) and 230 nm (broken lines). The broken lines are shown in the inserts only. In subsequent runs the sensitivity of the detector was increased to 0.010 and 0.005 absorbance units full scale (AUFS) as shown in the inserts in B and C, respectively.

266

DEVELOPMENTALBIOLOGY

1.32 absorbance unit full scale (AUFS). Peaks that eluted from the column were identified by their retention time and 254/280 ratio compared to standard bases run separately under identical conditions. The standard bases run were thymine, guanine, cytosine, adenine, 7-methylguanine, 5-methylcytosine, 2-methyladenine, and 6methylaminopurine and all were purchased from Sigma. 5-Methylcytosine eluted from the column after adenine (and ‘I-MeGua if present) as observed by Singer et aL (1977) using slightly different conditions. The area under each peak was determined using the formula A = hull2 and each peak was converted into nanomoles using the following absorption data at 254 nm (or 280 nm) and 0.08 AUFS sensitivity: Thy, 165.6; Gua, 278.2; Cyt, 104.2; Ade, 287.6, and 5-MeCyt, 189.6 (280 nm) mm’/nmole. RESULTS AND DISCUSSION

The elution pattern of Artemia cyst DNA acid hydrolysate from a high-performance cation column is shown in Fig. 1, compared to the elution patterns of standard bases and salmon sperm DNA hydrolysate. The order of elution of the bases is thymine, guanine, cytosine, and adenine. Other bases such as ‘I-methylguanine, 5-methylcytosine, N-2-methyladenine, and 6methylaminopurine, if present, elute sequentially after adenine (see Fig. 1A). However, the only modified base detected in any DNA hydrolysate analyzed in this study was 5-methylcytosine. The elution composition of Artemia embryo DNA compared to salmon sperm and calf thymus DNAs are shown in Table 1. These results have been calculated from data similar to that shown in Fig. 1. These data show that A&da DNA is undermethylated at all developmental stages analyzed regardless of whether the DNA is from metabolically inactive cysts or from nauplii actively engaged in DNA synthesis. In contrast, salmon sperm and calf thymus DNAs contain 1.6 to 1.7% 5-

VOLUME102,1984

methylcytosine or 7% of the total cytosine in these DNAs. Given the fact that the limit of detection of 5-methylcytosine with our system is about 5 pmole at the highest level of sensitivity of the UV detector (0.005 AUFS), and the fact that none could be detected in large samples containing 350 nmole total bases, we have concluded that 5-methylcytosine, if present in Artemiu embryos, is less than 0.01% of the cytosine content or less than 1 residue per 59 kilobases. This low level of methylation places Artemia in the “insect-type” category of Bird and Taggart (1980). On the other hand salmon sperm DNA and calf thymus DNA with ‘7% of the cytosine methylated are consistent with their position in the “vertebrate-type” category. We found that the method of DNA isolation, and particularly the steps to rid the preparation of RNA contamination, were critica for subsequent base analysis. In particular, the addition of a RNase Tz digestion step was necessary to assure complete removal of RNA. In some Artemia embryo DNA samples treated only with RNase A, we found trace amounts of 5-methylcytosine (~0.1% of total cytoaine) as well as other modified bases. These samples also showed aberrant A/T and G/C ratios, especially elevated levels of adenine. RNase T2 is not generally used in preparation of DNA, but our results indicate that, at least in Artemia, RNase A alone is not sufficient to prevent contamination of the preparation with small amounts of RNA. This might explain the findings of Antonov et al (1962) who reported that Artern&x DNA contains 0.1% 5-MeCyt, or the occasional detection of 5-methylcytosine in hydrolysates of insect DNA (Adams et aL, 1979) or in fixed polytene chromosomes (Eastman et aL, 1980). It is plausible that the DNA of Artemia, a crustacean, and of insects is in fact devoid of 5-methylcytosine. Current evidence indicates that the Uniramia, the phylum

TABLE 1 BASE COMPOSITION OF ARTEMIA DNA COMPARED TO OTHER DNAs”

Base Thymine Guanine Cytosine Adenine 5-Methylcytosine

0 hr (n = 3)

40 hr (n = 3)

48 hr (n = 4)

0.326 + 0.022 0.163 + 0.004 0.163 f 0.006 0.347 + 0.014 n.d.b

0.294 + 0.009 0.157 zb0.009 0.178 + 0.005 0.370 k 0.002 n.d.b

0.312 + 0.012 0.181 + 0.003 0.170 + 0.006 0.336 + 0.003 n.d.”

Salmon sperm (n = 4)

Calf thymus (n = 4)

0.271 t0.203 + 0.213 f 0.296 + 0.016 +

0.247 + 0.009 0.214 + 0.008 0.223 + 0.606 0.279 -+ o.oos 0.017 * 0.002

0.012 0.005 0.062 0.002 0.001

aThe values represent the mean + standard deviation of replicate samples (n) assayed from the same hydrolysate. bNone detectable. Limit of detection of 5-methylcytosine in these experiments is about 5 pmole. The maximum amount of total free bases applied to coiumns without detecting 5-methylcytosine was 350 nmoie. Therefore, 5-methylcytosine, if present, is less than 0.01% of the cytoaine content.

BRIEF Noms

that includes modern insects, and the Cmtacea arose independently more than 500 million years ago, probably from ancestral forms that were already diverged from one another (Manton and Anderson, 1979). Thus Arten& is nearly as distant evolutionarily from insects as both are from the deuterostomes. Methylation of specific cytosine residues may be a phenomenon more common to the deuterostomes than to the protostomes. However, the extent of DNA methylation among the protostome and deuterostome taxa will require analysis of DNA from additional species in these groups. Regardless of the phylogenetic distribution, the absence of 5-MeCyt in insects and at least one crustacean has implications for mechanisms of gene regulation. There is a frequent correlation in vertebrates between transcriptional activity and absence of methylation at specific CG sequences (Razin and Riggs, 1980; Weisbrod, 1982). In two cases there is experimental evidence for a necessary, if not sufficient, causative relationship between absence of methylation and transcription of specific genes (Compere and Palmiter, 1981; Christy and Scangos, 1982). If vertebrates, and perhaps all deuterostomes, use selective methylation as a regulatory mechanism, how are the equivalent genes regulated in protostomes such as Artemia and insects? Answering this question will require a careful comparison of regulatory mechanisms in members of both branches of animal phylogeny. This work was supported in part by the Natural Sciences and Engineering Research Council of Canada (Grant A2909) and by NIH Research Grant GM21376. REFERENCES ADAMS, R. L. P., and BURDON,R. H. (1982). DNA methylation karyotes. CRC G-it. Rev. Biochmx 13.349-384.

in eu-

ADAMS,R. L. P., MCKAY, E. L., CRAIG,L. M., and BURDON,R. H. (1979). Methylation of mosquito DNA. Biochem Biophya Acta 563, ‘72-81. ANTONOV,A. S., FAVOROVA,0. O., and BELOZERSKII,A. N. (1962). The nucleotide composition of the desoxyrihonucleic acids of animal and higher plants. Dok Akad Nadc. SSSR 147,1480-1483. BIRD, A. P., and SOUTHERN,E. M. (1978). Use of restriction enzymes to study eukaryotic DNA methylation. I. The methylation patterns in rihosomal DNA from Xenopus la&s. J. MoL Bid 118,27-47. BIRD, A. P., and TAGGART,M. H. (1980). Variable patterns of total DNA and rDNA methylation in animals. Nucleic Acids Rea 8,14851497.

267

BIRD, A., TAGGART,M., and MACLEOD,D. (1981). Loss of rDNA methylation accompanies the onset of ribosomal gene activity in early development of X Zoeuia CeU 26,381-390. BIRNDORF,H. C., D’ALESSIO, J., and BAGSHAW,J. C. (1975). DNAdependent RNA-polymerases from Artemia embryo. Charaeterization of polymerases I and II from nauplius larvae. Dev. Bid 45, 34-43. CEDAR,H., SALAGE,A., GLASER,G., and RAZIN, A. (1979). Direct detectioqof methylated cytosine in DNA by use of the restriction enzyme Mspl. Nucleic Acids Res. 6,2125-2132. CHRISTY,B., and SCANGOS,G. (1982). Expression of transferred thymidine kinase genes is controlled by methylation. Proc Natl Acad Sci USA 79,6299-6363.

COMPERE,S. J., and PALMITER,R. D. (1981). DNA methylation controls the inducibility of the mouse metallothionein-1 gene in lymphoid cells. Ce.!.!25, 233-240. EASTMAN, E. M., GOODMAN,R. M., ERLANGER,B. F., and MILLER, 0. J. (1980). 5methylcytosine in the DNA of the polytene chromosomes of the diptera Seinro wprophila, Drosophila melanogaater and D. per&m& Chrcnwxwmu 79,225-239. MANTON, S. M., and ANDERSON,D. T. (1979). In “The Origin of Major Invertebrate Groups” (M. R. House, ed.), pp. 269-321, Academic Press, New York. NAKANISHI, Y. H., IWASAKI, T., OKIGAKI, T., and KATO, H. (1962). salina. I. Embryonic development Cytological studies of Ark&a without cell multiplication after the blastula stage in encysted dry eggs. Annot. Zoo1 Japan 35,223~228. POLLOCK,J. M., SWIHART,M., and TAYLOR,J. H. (1978). Methylation of DNA in early development: 5-methyl cytosine content of DNA in sea urchin sperm and embryos. Nucleic Acids Rea 5.4855-4863. RAE, P. M. M., and STEELE,R. E. (1979).Absence of cytosine methylation at CCGG and GCGC sites in the rDNA coding regions and intervening sequencesof Drosophila and the rDNA of other higher insects. Nucleic Acids Res. 6, 2987-2995.

RAZIN,A., and RIGGS,A. D. (1980).DNA methylation and gene function. Scim 210, 604-610. SINGER,J., STELLWAGEN,R. H., ROBERTS-EMS,J., and RIGGS,A. D. (1977). 5Methylcytosine content of rat hepatoma DNA substituted with bromodeoxyuridine. J. Biol Chem 252,5509-5513. URIELI-SHOVAL,S. GRUENBAUM,Y., SEDAT, J., and RAZIN, A. (1982). The absence of detectable methylated bases in Drosophila melanogaster DNA. FEBS L&t. 146,148-152. VAUGHN, J. C., and PETROPOULOS, C. J. (1979). DNA sequence organization in the genome of the brine shrimp, Artemia salina In “Biochemistry of Artemia Development” (J. C. Bagshaw and A. H. Warner, eds.), pp. 190-268. University Microfilms, Ann Arbor, Mich. WARNER,A. H., MACRAE, T. H., and WAHBA, A. J. (1979). The use of Artemia sdina for developmental studies: Preparation of embryos, tRNA, ribosomes and initiation factor 2. In “Methods in Enzymology” (K. Moldave and L. Grossman, eds.), Vol. 60, Part H, pp. 298-311. Academic Press, New York. WEISBROD,S. (1982). Active chromatin. Nature (London) 297,289~295.