Eur. J. Biochem. 72, 525-535 (1977)

The Euglena gracilis Chloroplast Genome : Analysis by Restriction Enzymes Helena KOPECKA, Edwin J. CROUSE, and Erhard STUTZ Laboratoire de Gknnttique MolecuIaire, Institut de Recherche en BioIogie Molkulaire, Paris, and Laboratoire de Biochimie, Universite de Neuchgtel (Received July 22/November 15, 1976)

1. Chloroplast DNA was isolated from autotrophically and mixotrophically grown Euglena gracilis cells. 2. Aliquots of chloroplast DNA were mechanically degraded to an average molecular weight of 4-7 x lo6 and G+C-rich DNA fragments (density 1.701 g/cm3) were separated from the bulk DNA (density 1.685 g/cm3) using preparative CsCl density gradients. 3. Total chloroplast DNA and its DNA subfractions, which first were characterized with respect to average G C content and hybridization capacity for chloroplast rRNA, were hydrolysed with restriction endonucleases (endo R . EcoRI, endo R . HindII, endo R . HindIII, endo R . Hind11+III, endo R . HpaI, endo R . HpaII and endo R . HaeIII). The fragments were separated on gels under a variety of electrophoretic conditions. 4. With each enzyme tested, a rather large number of bands was obtained. In all cases, different banding patterns were obtained for total DNA, and the DNA subfractions. 5. Chloroplast DNA from autotrophically and mixotrophically grown cells gave identical banding patterns. 6. Digestion of total DNA with the endo R . HaeIII yielded 51 -52 fragments separated in the gels in a total of 36 bands of which 11 - 12 bands were composed of 2-3 fragments as estimated by densitometry. The molecular weights of all fragments combined was 87 x lo6 or 95% of the genome (92 x lo6). 7. Chloroplast RNA hybridized to 5.1% with total chloroplast DNA, equal to three RNA cistrons per genome ( M , 92 x lo6). These cistrons are located on seven different types of endo R . HaeIII fragments. The hybridising fragments are preferentially found in the G C-rich subfraction and in bands which are composed of 2-3 fragments.

+

+

The unicellular alga Euglenagracilis can grow heterotrophically in the dark, or mixotrophically and autotrophically in the light. Depending on growth conditions, the cell contains either small and undifferentiated proplastids or chloroplasts of variable size and functional capacity. The plastids contain, at all developmental stages, their own double-stranded DNA. Circular chloroplast DNA with a circumference of 44 pm has been isolated [l]. This length corresponds lo7 [l] to an approximate molecular weight of 9 . 2 ~ and matches within limits the chloroplast genome Abbreviations. NaCI/Cit, 0.15 M NaCl and 0.015 sodium citrate; A260 unit, the quantity of material contained in 1 ml of a solution which has an absorbance of 1 at 260 nm when measured in a I-cm

path-length cell. Restriction endonucleases are abbreviated as suggested by Smith and Nathans [22]. Enzyme. Trypsin (EC 3.4.21.4.) ; restriction endonucleases (EC 3.1.4:).

complexity as calculated from DNA renaturation kinetic measurements [2, 31. Moreover, the chloroplast genome is represented in multiple copies per cell [4]. Euglena chloroplast DNA has a very low G + C content (25%) and a considerable compositional heterogeneity as reflected by a multiphasic melting profile and multimodal profiles in alkaline and neutral CsCl density gradients or on hydroxyapatite chromatograms [5-91. According to the latest results [lo], this circular DNA is composed of five segments differing in their average G + C content. Nothing is known about the arrangement of these segments. At an average molecular weight of 5 x lo6, total chloroplast DNA can be subdivided, either in a neutral CsCl density gradient or on a hydroxyapatite chromatogram, into a fraction containing mainly the G + C-rich fragments having a pronounced component with an average buoyant density of 1.701 g/cm3

Euglena Chloroplast DNA

526

(G +C-rich DNA) and a fraction containing mainly the G + C-poor fragments with an average buoyant density of 1.685 g/cm3 (bulk DNA). The G+C-rich DNA carries the majority of cistrons coding for chloroplast rRNA [6, 8, 111. The number of genes per 9.2 x 107'daltonsof total chloroplast DNA is at least two [6, 121. With the introduction of DNA restriction endonucleases, it became possible to fragment the genome in a specific way. In the following, data are presented concerning the degradation of chloroplast DNA with several restriction endonucleases. Subsequent separation and analysis of these fragments are made by gel electrophoresis under a variety of experimental conditions. In particular, the endo R . HaeIII digest fragments are characterized with respect to molecular weight, stoichiometry and capacity to hybridize with chloroplast rRNA. A correlation is sought between the fragment patterns obtained with total chloroplast DNA and its G + C-rich and G + C-poor subfractions. Some of these results were presented at the preFEBS Symposium on Structures, Synthesis and Functions of Nucleic Acids in Organelles, Paris 1975.

MATERIALS AND METHODS Cultivation of Autotrophic Cells Euglena gracilis (Z-strain ; Culture Collection of Algae, Indiana University, No. 753) was grown in the dark in a heterotrophic medium (Difco Euglena broth) to approximately 3 x lo6 cells/ml. About 2500 ml of the etiolated cell suspension was transferred into 13.5 1 of autotrophic medium [13] and kept a t 24-25 "C under 2800 lx of light with continuous aeration for 48 h (ribosome isolation) or 72-96 h (DNA isolation) as previously described [14]. Under these conditions, proplastids develop over a period of approximately 72 - 80 h to fully developed and functional chloroplasts without significant cell division [15]. The cells were harvested and stored at - 20 "C.

per g of packed cells. All subsequent steps, prior to hydroxyapatite chromatography and centrifugation in CsCl density gradients, were performed at 4 "C. The resulting spheroplasts were washed twice to remove the trypsin by diluting the cells with buffer A followed by centrifugation at 3000 rev./min, for 5 rnin in a Sorvall SS-34 rotor. The spheroplast pellet was resuspended in 5 ml of buffer B per initial g of cells, and ruptured using a Waring blendor at full speed for 20-60 s until n o or very few whole cells remained. The suspension was diluted to a total volume of 210 ml with buffer B, filtered through 10 layers of gauze, and layered in 25-ml aliquots over 10 ml of buffer C in each of eight Pyrex centrifuge tubes. Centrifugation was done in two steps, first at 2000 rev./min for 5 min and then at 3000 rev./min for 10 min with a Sorvall GLC-1 rotor. Chloroplasts sedimented to the interface of buffers B and C. The upper suspension, containing mainly mitochondria, was discarded before removing the chloroplast band with a hypodermic syringe fitted with a 16-gauge needle. The resulting chloroplast bands were diluted to 210 ml with buffer C and divided into six nitrocellulose tubes. The chloroplast suspension was then over-layered with buffer A filling each tube. Chloroplasts were then floated to the interface of buffers A and C by centrifugation at 5000 rev./min for 5 min followed immediately by centrifugation at 24000 rev./min for 15 rnin with a Beckman SW27 rotor. The floated chloroplast band was removed with a hypodermic syringe and washed twice by diluting the chloroplasts with buffer D followed by centrifugation at 10000 rev./min for 5 rnin with a Sorvall SS-34 rotor. The resulting chloroplast pellet was used immediately to extract the DNA. Chloroplasts used for rRNA isolation were also prepared by the sedimentation-flotation method described above; however, it was necessary to use buffers E, F, G and H instead of buffer A, B, C and D, respectively. Chloroplast DNA Extraction

Cultivation of Mixotrophic Cells Euglena gracilis was grown under mixotrophic conditions using continuous aeration and 2800 Ix of light in 8 1 of heterotrophic medium [16] with a modified trace element combination [13]. Cultures were harvested in log phase. Cells were stored at - 20 "C. Isolation of Chloroplasts Euglena spheroplasts [14]were prepared from either autotrophic or mixotrophic cells. Between 20 and 30 g of frozen cells were slowly thawed and then incubated for 12 - 14 h at 4 "C in 2 ml of buffer A (see below) and 1 mg trypsin (Sigma, bovine pancreatic type 111)

The chloroplast pellet was lysed in 10 ml of buffer E and deproteinized with 5 ml of phenollm-cresol/ 8-hydroxyquinoline (9O/lO/O.l, v/v/w) by gentle 'rocking' in a graduated cylinder for 5 rnin (redistilled phenol was saturated with buffer D prior to the addition of m-cresol and 8-hydroxyquinoline). Then, 5 ml of chloroform/isoamyl alcohol (24/1, v/v) were added followed by additional gentle 'rocking' for another 5 min. The aqueous phase was either (a) passed through a Sephadex G-50, coarse grade column (1.5 x 55 cm) equilibrated with 0.01 M sodium phosphate buffer pH 6.8 before further purification on a hydroxyapatite column, or (b) diluted with water (1 : 4), adjusted to 0.01 M sodium phosphate buffer pH 6.8 using a 1 M stock solution and

H. Kopecka, E. J. Crouse, and E. Stutz

chromatographed directly on an hydroxyapatite column. The degree of nuclear DNA contamination varied up to 25%. The amount of chloroplast DNA obtained was between 5 and 10 pg per g of wet-packed cells. Chromatography on Hydroxyapatite

Hydroxyapatite chromatography was performed as previously described [7].

527

England). The labelled rRNA was passed through Sephadex G-50, coarse grade column (1.5 x 30 cm) equilibrated in 0.1 x NaCl/Cit to remove low-molecular-weight labelled components and then filtered three times through nitrocellulose filters (Schleicher and Schuell, 0.45 pm, 25 mm). Specific activities up to lo6 counts min-' pg RNA-' were obtained. Radioactivity was measured in a Nuclear Chicago Isocap 300 liquid scintillation counter.

Analytical and Preparative Ultracentrifugation

Hybridization of rRNA to Chloroplast DNA Sumples

Analytical cesium chloride density gradient experiments and sedimentation velocity experiments, were done as described elsewhere [17, 181. Preparative cesium chloride density gradient experiments were performed as previously described [19].

The filter hybridization technique described by Gillespie and Spiegelman [21] was used with some modifications. The hybridization reaction was carried out in glass scintillation vials under the following conditions: 5 ml of 2 x NaCl/Cit, 50pg of lz5Ilabelled rRNA, 10 filters, 5 pg DNA filter, 64 "C, 15-16 h. Under these conditions, all ribosomal cistrons were saturated with rRNA. All hybridization data were corrected for unspecific rRNA binding to a blank filter which were randomly located in the vials. N o correction was made for loss of DNA from the filters during incubation. Such loss would not exceed 10% of the DNA loaded as estimated from control experiments using labelled DNA.

Isolation of Chloroplast Ribosomes and Purijkation ofthe vRNA

Chloroplasts from either autotrophic or mixotrophic cells were isolated by the sedimentationflotation method which eliminates 87-S cytoplasmic ribosomes and 70-S mitochondria1 ribosomes. The chloroplast ribosomes and rRNA were prepared as described earlier [14] with some modifications. The ribosome pellet in buffer J was adjusted to 2% sodium dodecylsulphate and shaken for 5 min with a half volume of phenol/m-cresol/8-hydroxyquinoline (saturated with buffer J). A volume of chloroform/ isoamyl alcohol, equal to the phenol/m-cresol/8hydroxyquinoline volume, was added and shaking was continued for 5 min more. After a further extraction with chloroform/isoamyl alcohol only, the RNA was precipitated with 2 vol. of ethanol ( - 20 "C), resuspended in buffer J and reprecipitated. The chloroplast rRNA was further purified in preparative 16-ml convex sucrose gradients in buffer J at 24000 rev./min for 36 h with a Beckman SW-27.1 rotor. The absorption profiles at 260 nm showed a ratio of 16-S rRNA to 23-S rRNA of approximately 1 : 2. The peak fractions corresponding to 16-S and 23-S RNA were combined, dialysed against 0.1 x NaCl/Cit and concentrated in vacuo. The rRNA precipitated by the addition of 2 vol. of ethanol at -20 "C was resuspended in 1 ml of 0.01 M sodium acetate buffer pH 5 and reprecipitated. The rRNA, dissolved in 0.01 M sodium acetate buffer pH 5, was used for iodination in vitro. The yield was 20-40 pg of rRNA per g of wet-packed cells. Iodination of the Chloroplast rRNA

'''I

Chloroplast rRNA was labelled in vitro 1201 with (The Radiochemical Center Ltd, Amersham,

Restriction Enzyme Digestion of Chloroplast DNA

The enzyme nomenclature of Smith and Nathans [22] was used. Isolation and purification of the restriction enzymes from Haemophilus strains ( H . influenme, H. aegyptius and H. parainfluenzae), as well as the specific digestion conditions, were previously described [23]. The digestion conditions for endo R . EcoRI were described by Yoshimori (Ph. D. Thesis, University of San Francisco 1971). Chloroplast DNA (10 -20 pg) was completely digested by an excess of enzyme at 37 "C for 3 h using endo R . EcoRI or for 16 h using Haemophilus endonucleases. Gel Electrophoresis

The digested DNA solution was evaporated under vacuum to 30p1, adjusted to 1% sodium dodecylsulphate, incubated for 30 min at 37 "C. Then 6 p1 of 60% sucrose containing 0.02% bromophenol blue was added. Electrophoresis in vertical slabs (16 x 40 x 0.4 cm) was carried out at 20 "C or at 4 "C using a constant current intensity of 30 mA for 16-30 h depending on the pore size of the gel, until the bromophenol blue reached the bottom of the gel. Pure agarose gels, pure polyacrylamide gels or composite gels containing various concentrations of agarose and polyacrylamide were used. The concentrations

Euglena Chloroplast DNA

528

of gels used in a specific experiment are given in the legends of the respective figures. The details of gel electrophoresis, staining with ethidium bromide, photography of gels, subsequent scanning of the negatives with a Joyce Loebl microdensitometer (Gateshead-on-Tyne, England) have been described elsewhere 1241. The molecular weight of the endo R . HaeIII and endo R . EcoRI fragments of different chloroplast DNA samples were determined using as a secondary standard the endo R . Hue111 fragments of the total chloroplast DNA ;this standard was calibrated against primary standards formed by restriction fragments of phage I DNA from endo R . EcoRI digestion [25], Simian virus 40 (SV40) DNA fragments from endo R . Hind11 I11 digestion and SV40 DNA fragments from endo R . Hue111 digestion [26,27]. RF values of the same or of a different digest were normalized as described by Prune11 et al. [24].

+

Hybridization of the rRNA to Restriction D N A Fragments Transfer of the fragments of DNA from agarose gels to cellulose nitrate filter strips (Millipore, HAWP 304 Fo) and the hybridization with the labelled RNA was done by the method described by Southern [28]. Buffers Buffer A : 0.34 M sucrose, 0.05 M Tris-HC1, pH 7.9, 0.01 M EDTA. Buffer B: 0.70 M sucrose, 0.05 M Tris-HC1, pH 7.9, 0.01 M EDTA. Buffer C: 1.4 M sucrose, 0.05 M Tris-HC1, pH 7.9, 0.01 M EDTA. Buffer D: 0.2 M NaCl, 0.05 M Tris-HC1, pH 7.9, 0.01 M EDTA. Buffer E : 0.34 M sucrose, 0.05 M Tris-HC1, pH 7.9. Buffer F: 0.70 M sucrose, 0.05 M Tris-HC1, pH 7.9. Buffer G : 1.4 M sucrose, 0.05 M Tris-HC1, pH 7.9. Buffer H: 0.05 M Tris-HC1, pH 7.9, 0.1 M KCl, 0.015 M MgCl,, 0.005 M 2-mercaptoethanol. Buffer I : 4 M urea, 2.5% sodium dodecylsulphate, 0.2 M NaC1, 0.05 M Tris-HC1, p H 7.9, 0.01 M EDTA. Buffer J : 0.01 M Tris-HC1, pH 7.9, 0.1 M NaC1, 0.001 M MgC12. RESULTS Preparation of the Various Chloroplast D N A Samples Total chloroplast DNA was subfractionated in a preparative CsCl density gradients (Fig. 1A) after RNA was removed by hydroxyapatite chromatography. From such gradients, three samples were prepared : total chloroplast DNA (equivalent to fractions 30 - 59), sample I (fractions 30 -40) and sample I1 (fractions 41 - 59). Aliquots of the sample I1 were mechanically degraded and reequilibrated in a second preparative CsCl density gradient (Fig. 1B)

to enhance the separation of the G+C-rich DNA from the bulk DNA. From this gradient, samples I11 (fractions 24 - 33) and IV (fractions 34 -46) were formed. The starting material and each of the four subfractions (samples I to IV) were characterized by their buoyant density profile (Fig. l), their average molecular weight as determined from sedimentation coefficients and their hybridization capacity towards chloroplast rRNA (Table 1). The total chloroplast DNA, which was contaminated with approximately 10% of nuclear DNA (1.708 g/cm3), had an initial average molecular weight close to 1 5 x lo6 corresponding to one sixth of a 44-pm-long DNA molecule. Sample I represented a mixture of nuclear DNA (1.708 g/cm3) and G +C-rich chloroplast DNA (1.701 g/cm3) with about 20% of bulk chloroplast DNA (1.685 g/cm3). Its average molecular weight of 10 x lo6 was definity below the average molecular weight of the initial sample. Sample I1 represented pure chloroplast DNA with the bulk DNA as the major component. Samples I11 and IV were generated by shearing with a hypodermic syringe fitted with a 27-gauge needle to an average molecular weight of 4-7 x lo6. The two samples showed very distinct density profiles. Sample I11 was composed of G Crich DNA (60%) and bulk DNA (40%) while sample IV represented mainly bulk DNA fragments. Each of these samples was hybridized with 1251labelled rRNA isolated from chloroplasts (Table I). The hybridization results are expressed either in percentage (w/w) hybridization (RNA/DNA) or as number of rRNA genes per 9.2 x lo7 daltons. Where necessary the values were corrected for contaminating nuclear DNA which is known to bind under these conditions only negligible amounts (below 0.15%) of chloroplast rRNA. The total chloroplast DNA hybridized to a level of 5.1%, a value corresponding to three rRNA genes per 9 . 2 lo7 ~ daltons. Preparations I to IV hybridized to variable amounts; however, the number of genes per 9 . 2 ~ lo7 daltons for these four samples merely reflects a fortuitous enrichment or loss of rRNA cistrons and does not relate to the number of genes per genome. Prevention of any selective loss of DNA segments due to compositional heterogeneity was important for further analysis. With respect to the rRNA cistrons, no significant loss occurred. This can be shown by calculations using the amount of DNA per sample and its corresponding hybridization capacity to rRNA (Table 1). The combined values for samples I and I1 (corrected for nuclear DNA contamination) yield 5.1% hybridization, a value equal to the 5.1% found for total chloroplast D N A ; while the combined values for samples 111and IV yield 4.4% hybridization, a value close to the 4.2% found for sample 11. Since one might argue (see Discussion) that chloroplast DNA could be differently composed depending

+

H. Kopecka, E. J. Crouse, and E. Stutz

1.5

529

I .

1

IT

1 .o

05

E 8

0

z 2 5

35

55

65

35 45 Fraction number

55

45

% m

22 Q

1.5

R

1 .o

0 .E

C

25

Fig. 1. Preparative CsCl densitygrudient ofEuglena chloroplast D N A . (A) 54 AZ6* units of total chloroplast DNA were equilibrated in a 10-ml neutral CsCl density gradient. 6-drop fractions were collected and combined as indicated by the circles. Sample 1, fractions 30-40; sample 11, fractions 41 -59; total DNA, fractions equivalent to 30- 59 (B) 20 A260 units of sample I1 were mechanically sheared and equilibrated in a 10-ml CsCl density gradient. 4-drop fractions were collected and combined as indicated by the circles. Sample 111, fractions 24- 33; sample IV, fractions 34-46. In both cases, the initial density of the gradient was 1.689 g/cm3. Centrifugation was at 33000 rev./min for 64 h at 20 "C in a Beckman type-40 rotor. The tracings of analytical CsCl density gradients from total chloroplast DNA and samples I to IV are given on the right margin. DNA samples were centrifuged at 44000 rev./min for 20 h at 25 "C. DNA from Pseudomonas aeruginosa (density 1.726 g/cm3) was used as density marker Table 1. Some characteristics of the different chloroplast D N A samples Hybridization was calculated as 1OOx pg RNA/pg DNA; average values for three filters are given. The molecular weight of 16-Sf23-S rRNA was taken as 1.6 x lo6 Samples

M , x 10-6

DNA pooled

Hybridization with rRNA uncorrected

A260 unit Total, autotroph

I I1 I11 IV To.

1,

mixotroph

14.6 9.5 17.6 4.4 6.6 9.7

9

I 24 3 6 25

corrected

Number of rRNA genes per 9.2 x lo7 daltons

5.1" 12.2b 4.2 9.1 2.1 5.4

2.9 7.0 2.4 5.2 1.2 3.1

% 4.6 6.1 4.2 9.1 2.1 5.4

* Corrected value accounting for 10% nuclear DNA contamination. Corrected value accounting for 50% nuclear DNA contamination.

on the developmental stage of the organelle, for example, by amplification of selected DNA segments like ribosomal DNA, chloroplast DNA from mixotrophic cells with small and only partly differentiated chloroplasts was introduced in this study. According

to the hybridization values in Table 1 , total chloroplast DNA from either autotrophic or mixotrophic cells hybridized to approximately the same level (5.1% versus 5.4%) indicating that both DNA samples contain the same number of rRNA cistrons per

Euglena Chloroplast DNA

530

9 . 2 lo7 ~ daltons. The DNA density profile (not shown) from mixotrophic cells did not reveal measurable nuclear DNA contamination and, therefore, correction of the respective hybridization valuc~\\ as not necessary.

M

1

2

3

L

5

Restriction Enzyme Analyses of the Various Chloroplast DNA Samples In a first series of experiments, total chloroplast DNA was digested with several restriction enzymes (endo R . HindII, endo R . HindIII, endo R . Hind I1 +III, endo R . Hpal, endo R . HpaII, endo R . Sma, endo R . HaeIII and endo R . EcoR1). Except for the endo R . Sma which did not further degrade the total chloroplast DNA (average molecular weight of 15 x lo6), all other endonucleases tested split the sample into a large number of fragments yielding, upon gel electrophoresis, a characteristic banding pattern for each enzyme. In particular, certain bands were intensified relative to other bands in the same gel region (gel profiles not shown). For a detailed comparative analysis of the various DNA samples (Table l), endo R . HaeIII was used. From the density profiles of total chloroplast DNA and the four subfractions (samples I to IV) shown in Fig. 1, it was obvious that the subfractions represented different populations of the total chloroplast DNA, and thus were expected to yield different gel banding patterns after digestion with endo R . HaeIII. Fig. 2 shows the banding patterns of the various DNA samples analyzed in a 2.5% polyacrylamide/0.5% agarose gel and stained with ethidium bromide. Differences in the overall number and intensity of bands can be seen for total chloroplast DNA and the four subfractions. Total DNA and sample I1 DNA yielded almost identical patterns, the digest of the total DNA having a slightly higher background due to nuclear DNA contamination. The pattern of sample 11, however, was markedly different from the pattern of sample I. The larger fragments of sample I are either lost or very faint, while four bands are relatively intensified. Some bands in the lower part of the gel also changed in relative intensity. This same observation can be better made by comparing sample I11 with sample IV. Sample I and sample 111 represent chloroplast DNA enriched in the G+C-rich DNA, while samples I1 and IV mainly represent the bulk chloroplast DNA. In Fig. 3, the fragment pattern from chloroplast DNA, isolated from either autotrophically or mixotrophically grown cells, is shown. The molecular weight of each preparation was adjusted to approximately l o x lo6. Total chloroplast DNA from autotrophically grown cells was contaminated by nuclear DNA (lo%), while that from mixotrophically grown

Fig. 2. Gel electrophoresis of endo R . HaeIII-generated frugments ,from various chloroplast D N A samples. Electrophoresis patterns of total chloroplast DNA (I), sample I1 (2), sample I (3), sample 111 (4) and sample IV (5) each digested with endo R . HaeIII and resolved by electrophoresis in 2.5% polyacrylamide/0.5'x agarose gel, 4 "C, 30 mA, 30 h. Molecular weight marker (M): phage 1 DNA digested with endo R . EcoRI and SV40 DNA digested with endo R . Hue111

cells was not. No differences in fragment number and band intensity could be seen under these experimental conditions (endo R . HaeIII, 2.5% polyacrylamide/ 0.5% agarose gels) indicating that (a) the nuclear DNA did not yield any discernible bands, but only added to the background and (b) the two types of chloroplast DNA were identical. Genome Size and Compositional Heterogeneity The total DNA sample may be considered to represent the entire chloroplast genome. The genome size can be calculated provided all fragments from

H . Kopecka, E. J. Crouse, and E. Stutz 1

2

Fig. 3. Gel electrophoresis of endo R . HaeIII-generated fragments ,from chloroplast DNA of mixotrophically or autotrophically grown cells. Total chloroplast DNA purified from Euglena cells grown either mixotrophically (1) or autotrophically (2) and digested with endo R . HaeIII. Electrophoretic conditions: see Fig.2

the total DNA can be resolved and possible multiplicities of the fragments can be determined. In order to achieve this, total chloroplast DNA was digested with endo R . Hue111 and the fragments were separated by electrophoresis using different types of gels. The banding patterns of DNA digests as obtained on the different gels are schematically represented in Fig. 4. The six banding patterns are ordered according to the concentration (%) of polyacrylamide in the respective gels: the top left gel was composed in such a

531

way as to maximize resolution of the DNA fragments of highest molecular weight, while the bottom right gel was such as to allow resolution of low-molecularweight fragments. The other four gel types were bridging the resolving power of the two extreme ones. Total DNA, samples 11, 111 and IV were analyzed in all gels; sample I was not included in the first two gels. Heavy solid lines symbolize bands which appeared more strongly than the average band (light lines) ; dashed lines indicate faint bands. Heavy solid lines with an asterisk indicate strong bands which could be resolved into further bands in any of the gels used. The color intensities of the bands were estimated by visual inspection and densitometry (expanded densitograms) of the negative pictures of the ethidiumbromide-stained gels. The complex relationship between band profile on the densitogram and the amount of DNA present and the decrease in fragment yield with increasing fragment size when degraded DNA is used rendered a quantitative evaluation difficult. This problem was partially overcome by tracing the ratio between the surface of the bands and their molecular weights as a function of the molecular weight (semilog plot, not shown). A straight line with some scattering of the points was obtained for a given zone of molecular weights for bands representing simple bands. The multiple bands, however, fitted the straight line only after dividing the surface value by the numbers given in Table 2 (fragment repetitions). We therefore define multiple bands as such bands which, under our experimental conditions, would always stain more strongly than simple bands. In our case multiplicity would be 2 or 3. These figures represent rounded values. (A detailed description of this evaluation procedure will be given elsewhere [24]). As shown in Table 2, 36 bands could be resolved of which 11-12 bands were multiple ones. The molecular weights of the DNA fragments ranged from 5.7 x lo6 to 0.15 x lo6. The values given in the last column represent the sum of the molecular weights of the fragments per band (see Discussion). In Table 3, a summary of the results obtained for total chloroplast DNA and samples I11 and IV DNA is given. The results of samples I and I1 (not shown) were very similar to results of samples I11 and IV, respectively. The added molecular weights of all fragments was 87 x lo6 (upper value) for total chloroplast DNA and 36 x lo6 and 56 x lo6 for samples I11 and IV, respectively. The value for total chloroplast DNA was slightly below the 9.2 x lo7 estimated from the circumference of a 44-pm-long circular DNA. According to these results, about 41% of the total DNA sequences were found, qualitatively, within the heavy DNA fraction and 65% within the light DNA fraction. Between 51 - 52 fragments were obtained from total chloroplast DNA while 38 -40 were from samples I11 and IV.

Eugiena Chloroplast DNA

532

M

T

I

Hae

IE

Hae

M

XIK

I

T

I

I

IE I

I

U

X

? .O

5.0

0.2

2.5

i .o

0.5

0.4

3.5 1 .o

1 .o

0.6 3.25 1.5

0.8 3.5

3.15 1 .c

Y

2

0.2

1 .i 1 .S%

2 OIo

a

b

U

M

I

T I T

._

s

-a L

Hae IE

Hae IE

(L

+,

r 0)

3 0 G,

T E I I I I I K M

X I E M

10

0.6

2 .o

0.5

0 .! 1

2

.o

D .50

0.5

0.4

1 .o 0.3

0.5

1 .I

0.25

1.5

1.0

0.20

0.25 I.

0.25 2 .o

1.5 2 I. 2 .o

0.125

0

3% d

4% e

0.1

2.5

5% f

Fig. 4. Scheme of banding patterns of various chloroplast D N A samples unalysed under various polyacrylumide gel concentrations. Total chloroplast DNA (T), along with samples I, 11, 111, I V (I, 11, 111, IV, respectively) were digested with endo R . HaeIII, and the DNA fragments were resolved in (a) 1.S% polydcryhmide/0.7% agarosegel, (b) 2% polyacrylamide/0.5% agarose gel, (c) 2.5%, polyacrylamide/0.5% agarose gel, (d) 3% polyacrylamide gel, (e) 4% polyacrylamide gel, and (05% polyacrylamide gel, 4 "C, 30 mA, 16 h. Various marker DNA fragments (M) were included: (a) left marker: phage 3, DNAiendo R . EcoRI plus SV40 DNA/endo R - HindII + 111, right marker: phage 3, DNA/endo R . EcoRI plus SV40 DNA/endo R . HaeIII; ( b )left marker: phage iDNAiendo R . € m R I , right marker: phage i DNAiendo R . EcoRI plus SV40 DNAiendo R . HueIII; (c) phage L DNA/endo R . EcoRI plus SV40 DNA/endo R . HaeIII; (d) left marker: phage i DNA/endo R . EcoRI plus SV40 DNA/endo R . Hind11 + 111, right marker: phage 1 DNA/endo R . EcoRI plus SV40 DNAiendo R . HoeIII; (e)SV40 DNA/endo R . Hind11 + I l l ; (f) left marker; SV40 DNA/endo R . HaeIII, right marker: SV40 DNA/endo R . HindII I l l . Bands which were never further resolved using the six different types and concentrations of gel (a- 4, were marked with an asterisk. Depending on the gels, bands in the upper or lower part may be missing

+

Restriction DNA Fragments Currying rRNA Genes From the hybridization results given in Table 1, it became clear that (a) the total chloroplast DNA accomodates three copies of the rRNA gene per 9.2 x lo7 daltons and (b) the sample 111 DNA carried 4.5 times more sequences complementary to rRNA

than sample IV DNA. Furthermore, the results from the gel analysis showed that (a) certain bands are proportionally stronger and (b) sample I11 DNA preferentially contains such bands. To correlate these results, endo R . Hue111 fragments from total DNA, samples 111 and IV DNA were separated on a 1% agarose gel, transferred to millipore filter strips and

H. Kopecka, E. J. Crouse, and E. Stutz

533

hybridized with '251-labelled rRNA. Hybridization was monitored by autoradiography. Under these analytical conditions, 25 bands were resolved as schematically shown in Fig.5. The DNA in seven gel regions as indicated by squares hybridized with

1251-labelled rRNA. These seven zones correspond most likely to the band numbers 6, 7, 11, 19-21, 25, 27-28 and 31. An unequivocal correlation is not possible since this 1% agarose gel did not resolve all the bands listed in Table 2.

Table 2. Molecular weights of endo R . HaeIII DNA fragments from Euglena chloropast DNA The bands resolved in any of the six gels described in Fig.4 are ranged according to molecular weights. The numerical values in the third column indicate the relative staining intensities of the bands as explained in the text; the term 'fragment repetitions' does not express any qualitative statement concerning the relationship of fragments found in a band x M,

Band number

DNA

of

Fragment repetitions

x M , of DNA per band

1 2 3 4 5 6 1 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36

5.7 5.3 4.3 3.9 3.1 3.2 2.4 2.2 2.1 1.9 1.8 1.6 1.4 1.2 1.1 0.88 0.82 0.81 0.80 0.78 0.13 0.12 0.69 0.68 0.64 0.59 0.51 0.50 0.38 0.36 0.29 0.28 0.20 0.19 0.16 0.15

3 2 1 1 2 2 3 1 1 1 1-2 1 1 1 1 1 1 1 3 1 2 1 1 1 2 1 1 1 1 1 3 1 2 2 1 1

17.1 10.6 4.3 3.9 7.4 6.4 7.2 2.2 2.1 1.9 1.8-3.6 1.6 1.4 1.2 1.1 0.88 0.82 0.81 2.40 0.78 1.46 0.72 0.69 0.68 1.28 0.59 0.51 0.50 0.38 0.36 0.87 0.28 0.40 0.38 0.16 0.15

Hae XI

T

IU

I

I

M

5.0

2.5

Fig. 5. Scheme ofbanding patterns of various chloroplast DNA samples in a 1% agarose gel. Total chloroplast DNA (T), sample 111 (111) and sample IV (IV) were digested with endo R . HaeIII and the resulting fragments were resolved in 1% agarose gel, 4 "C, 30 mA, 16 h. Fragments were transferred to cellulose nitrate strips and hybridized with '251-labelledrRNA. Hybridizing zones were located by autoradiography using Kodak medical X-Ray film. Marker (M): phage 1 DNA/endo R . EcoRI plus SV40 DNA/endo R . Hind II+III

Table 3. Summary of results of endo R . Hae III hydrolysis of chloropast DNA Sample

Number of bands

Number of fragments

Sum of M , of fragments

36 25 30

51 -52 38 38-40

87 x lo6 36 x lo6 56 x lo6

~~

(% total) Total DNA Sample I11 Sample IV

( 100) ( 41) ( 65)

Euglena Chloroplast DNA

534

DISCUSSION From renaturation data, the genome size was estimated to be in the range of 9.2 x lo7 daltons [2,3] suggesting that the entire chloroplast genome can be accommodated on one circular molecule. However, the renaturation data were not too accurate in view of the very low G + C content and the compositional heterogeneity of the DNA. Therefore, it was important to determine the genome size by another method independent of the compositional heterogeneity. The upper value of 8.7 x lo7 daltons found in this analysis using endo R . HueIII matches reasonably well the 9.2 x lo7 daltons calculated from the circumference of the 44-pm-circumference circle, considering the inaccuracies inherent in the analytical methods used. The amount of DNA per EugZena chloroplast seems to vary depending on the growth conditions [12], and it may be anywhere between 1.2 x lo-’’ to 12 - 16 x lo-’’ g [4, 291 which corresponds to 8 or up to 100 circles of 44-pm circumference, respectively. This quantitative variability might also be due to selective gene amplification [8]. Milenz and Hershberger 1301 argued that certain segments of the chloroplast genome might be selectively amplified under certain conditions of mixotrophic cell growth. The results presented here using endo R . HueIII d o not support this hypothesis since chloroplast DNA isolated from mixotrophically grown cells yielded exactly the same restricted DNA fragment pattern as chloroplast DNA from autotrophically grown cells. Furthermore, no differences in the relative quantity of G + C-rich segments, carrying rRNA genes, seemed to exist between chloroplast DNA from autotrophically and mixotrophically grown cells according t o the results from the hybridization experiments (Table 1). Asymmetric density profiles are characteristic of randomly fragmented Euglena chloroplast DNA samples. Up to three different classes of chloroplast DNA fragments could be discerned having average buoyant densities (and G + C contents) of 1.685 (25%G+C), 1.692 (33%G+C) and 1.701 (41% G + C ) g/cm3 (reviewed in [9]). The amount of 1.692-g/cm3 or 1.701-g/cm3 fragments, relative to the 1.685-g/cm3 fragments, depended on the molecular weight of the DNA sample. A further subdivision of the chloroplast genome into five segment classes was postulated [lo], having G + C contents of 22, 26, 31, 36 and 41%. According to this study, two segment classes with low G + C contents of 22% and 26% made up about two thirds of the genome, while the remainder of the genome was composed of segments with G + C contents of 31-41%. Both of these reports were based on analysis of randomly degraded DNA and therefore remain very approximate values.

A more precise picture concerning the qualitative and quantitative aspects of compositional heterogeneity should be obtained by the analysis of restricted DNA fragments. In this study, the G C-rich segments mainly found in sample I11 yielded fragments (endo R . HueIII) of lower molecular weight (e.g. bands 19, 21, 25) while sample IV, composed mainly of A + T-rich segments, yielded fragments (endo R . HueIII) of higher molecular weight (e.g. bands 1-5). Such a result might have been expected in view of the sequence requirement of the endo R . HueIII enzyme : -G-G-C-C-

+

-++J-+

All fragment patterns in Fig.2 show some of the bands as particularly pronounced (multiple bands) and analysis by densitometry revealed that these bands contained approximately two to three times more DNA than the simple bands, in the same gel region. This proportionately higher DNA content may have one of two reasons: (a) existence of nonidentical (size, sequence) DNA fragments having the same electrophoretic mobility under the given experimental conditions; (b) existence of identical fragments (size, sequence) which are cut out of reiterated sequences of the genome. We cannot exclude that some of the multiple bands are composed of non-identical fragments. Digestion of the DNA with a second restriction enzyme and analysis of its fragment pattern would be required to solve that problem. However, the observation that several of the multiple bands hybridized to rRNA, and therefore contained DNA fragments with ribosomal DNA sequences which are repeated three times per 92 x lo6 daltons, strongly suggests that some of the multiple bands are indeed composed of identical fragments. Working with degraded DNA renders a quantitative evaluation of band intensities difficult, since, for obvious reasons, relative losses or gains of fragments may occur. The values given in Table 2 only indicate that certain bands stained 2 - 3 times more strongly relative to comparable ones (same range of molecular weight) and therefore have a proportionately higher DNA content. We consider this multiplicity not to be fortuitous or artefactual since the banding patterns from different total chloroplast DNA preparations (see Fig. 3) were qualitatively and quantitatively very similar in having identical bands intensified to about the same extent. According to these hybridization results the Euglena chloroplast genome contains three rRNA cistrons. This value is in line with earlier reports 16,291 but disagrees with some of our own reported values [8] and reports from other laboratories [ I l , 121 where only one or two cistrons per 92 x lo6 daltons were found. The discrepancy may be due to a selective loss of ribosomal DNA during the DNA preparation;

H. Kopecka, E. J. Crouse, and E. Stutz

the ribosomal segments have a G + C content which is 22% higher than the average G + C content of the bulk DNA. The ribosomal DNA was found spread over seven different kinds of endo R . HaeIII DNA fragments. Some of these fragments could stem from within the repeated cistrons and generate multiple bands. The combined molecular weight of the hybridising fragments amounts to 11.4 x lo6, not taking into account the multiplicity of some fragments. This is sufficient space to accomodate three rRNA cistrons including precursor and spacer sequences [31, 321. Total chloroplast DNA from autotrophic cells and samples I11 and IV DNA were also restricted with endo R . EcoRl (see E. Stutz et al. in Proceedings of the Genetics and Biogenesis of Chloroplasts and Mitochondria, Munich, 1976). Total DNA yielded 22-24 bands resolved in a 1% agarose gel. The molecular weights were in the range of 15 x lo60.48 x lo6. Three bands were recognized as multiple ones. These three bands contained DNA fragments which hybridized to rRNA. The respective molecular weights were 1.58 x lo6, 2.10 x lo6 and 5.25 x lo6. Sample I11 DNA yielded almost exclusively these three bands while in sample IV they were almost absent. While this work was in progress, Slavik and Hershberger [lo] reported that four endo R . EcoR1generated fragments from Euglena chloroplast DNA hybridized with chloroplast rRNA. Their result cannot be correlated with our data since they did not report the size of these fragments. This investigation was supported by Fonds National Suisse de la Recherche Scientifique, grant number 3.0300.73 and Fonds Dr Sauberli to E. S . The assistance of Dr Graf and Messrs Montandon, Jenni and Blaser (UniversitC de Neuchltel) is gratefully acknowledged. Some of the analytical ultracentrifugation analyses were performed by Mr Macaya and Mrs Fonty (Institut de Biologie Moleculaire, Paris). Chemical base determination of the chloroplast rRNA was made by Mrs Fonty. Endo R . EcoRI enzyme was a generous gift of Dr Albert0 Bernardi (Institut de Biologie Moleculaire, Paris). Appreciation is extended to Dr Giorgio Bernardi and to Dr Clark-Walker for their helpful criticism in the preparation of this manuscript, and especially to Dr Giorgio Bernardi for the liberal use of his research facilities.

535

REFERENCES 1. Manning,J.E. & Richards,O.C. (1972) Biochim. Biophys. A m , 259,285 -296. 2. Stutz,E. (1970) FEBS Lett. 8,25-28. 3. Slavik,N.S. & Hershberger,C.L. (1975) FEBS Left. 52, 171 -174. 4. Edelman,M., Cowan,C.A., Epstein, H.T. & Schiff, J. A.( 1964) Proc. Naif Acad. Sci. U.S.A. 52, 1214-1219. 5. Stutz,E. & Rawson,J.R.Y. (1970) Biochim. Biophys. Acta, 209, 16-23. 6. Stutz,E. & Vandrey,J.P. (1971) FEBSLett. 17,277-280. 7. Stutz,E.&Bernardi,G.(1972)Biochimie(Paris)54,1013 - 1021. 8. Vandrey,J.P.&Stutz,E.(1973)FEBSLett.37,174-177. 9. Crouse,E. J., Vandrey,J. P. & Stutz, E. (1974) in Proc. 3rd Int. Congr. on Photosynthesis (Avron,M.,ed.) pp. 1775- 1786, Elsevier, Amsterdam. 10. Slavik,N.S. & Hershberger,C.L. (1976) J . Mol. Biol. 103, 563 - 581. 11. Rawson,J.R.Y. & Haselkorn,R. (1973) J . Mol. Biol. 77, 125-132. 12. Gruol,D., Rawson,J.R.Y. & Haselkorn,R. (1975) Biochim. Biophys. Acta, 414, 20-29. 13. Cramer,M. & Meyers,J. (1952) Arch. Mikrobiol. 17, 384-402. 14. Rawson,J.R.Y. & Stutz,E. (1969) Biochim. Biophys. Acta, 190, 368 - 380. 15. Schiff,J.A. & l3psteiqH.T. (1968) in The Biology of Euglena (Buetow,D.E.,ed.)pp. 285-333, Academic Press, NewYork. 16. Greenblatt,C.L. & Schiff,J.A. (1959) J . Protozool. 6, 23-28. 17. Bernardi,G., Piperno,G. & Fonty,G. (1972) J. ,4401. Biol. 65, 173- 189. 18. Prunel1,A. & Bernardi,G. (1973) J . Biol. Chem. 248,3433- 3440. 19. Filipski,J., Thiery,J.P. & Bernardi,G. (1973) J . Mol. Biol. 80, 177-197. 20. Orosz, J. M. & Wetmur, J. G. (1974) Biochemistry, 13, 5467 5473. 21. Gillespie,D. & Spiegelman,S. (1965) J . Mol. Biol. 12. 829-842. 22. Smith,H. & Nathans,D. (1973) J . Mol. Biol. 81,419-423. 23. Kopecka,H. (1975) Biochim. Biophys. Acta, 391,109-120. 24. Prunell, A., Kopecka, H., Straws, F. & Bernardi, G. (1977) J . Mol. Biol. in the press. 25. Allet,B., Jeppsen,P. G.N.,Katagiri,K. J. & Delius,H. (1973) Nature (Lond.) 241, 120-123. 26. Danna,K. & Nathans,D. (1971) Proc. Natl Acad. Sci. U.S.A. 68,2913-2917. 27. Lebowitz,P., Siegel,W. & Sklar,J. (1974) J . Mol. Biol. 88, 105-123. 28. Southern,E.M. (1975) J. Mol. Biol. 98, 503-517. 29. Scott,N.S. (1973) J. Mol. Biol. 81, 327-336. 30. Mielenz, J. R. & Hershberger,C. L. (1974) Biochem. Biophys. Res. Commun. 58,769 777. 31. Carrit,B. & Eisenstadt,J.M. (1973) FEBS Lett. 36, 116-120. 32. Heizmann,P. (1974) Biochem. Biophys. Res. Commun. 56, 112 - 118. -

-

H. Kopecka, Laboratoire de GtnCtique Moleculaire, Institut de Recherche en Biologie Moleculaire, Universite de Paris VII, 2 Place Jussieu, F-75221 Paris-Cedex-05, France E. J. Crouse and E. Stutz, Laboratoire de Biochimie, Universite de NeuchPtel, Rue de Chantemerle 18, CH-2000 Neuchltel 7, Switzerland