Terminal redundancy and circular permutation of ... - CiteSeerX

Journal of General Virology (1990), 71, 2157-2162.

2157

Printed in Great Britain

Terminal redundancy and circular permutation of mycoplasma virus L3 DNA W a l t e r J u s t I and Giinther Klotz 2 . l Abteilung Klinische Genetik and 2Abteilung Virologie, Institut J~r Mikrobiologie, Universitiit Ulm, Postfach 4066, D-7900 Ulm, F.R.G.

This communication reports the physical map of mycoplasma virus L3 (MV-L3) DNA derived from restriction patterns obtained by digestion with seven different restriction endonucleases. The length of the restriction map is 36 200 bp in contrast to the contour length of native MV-L3 DNA molecules which is 39400 bp as determined by electron microscopy. The difference in length of 3200 bp (corresponding to 8.1% of the native viral DNA contour length) is explained by

terminal redundancy. It was possible to clone all fragments from particular restriction patterns into Escherichia coli vector pAT153, an indication of circular permutation within a population of MV-L3 DNA. However clear evidence has been obtained from the molar ratios of fragments and from hybridization experiments. We suppose that viral DNA is packaged from a concatemeric precursor molecule starting at a specific site called pac.

Mycoplasma virus L3 (MV-L3) is a member of the Podoviridae family (Maniloff et al., 1982), which has been isolated from spontaneous virus plaques on lawns of Acholeplasma laidlawff (Gourlay & Wyld, 1973). The viral D N A is double-stranded, linear with an Mr of 26 × 106, which is equivalent to approximately 39.4 kbp. Host DNA synthesis is shut down within 3 h after infection (Haberer et al., 1979). Virus production starts at 90 min post-infection, reaches a maximum after 10 to 15 h and proceeds for several further hours. The infection is cytocidal rather than lytic. MV-L3 is quite an interesting virus in view of its unusual host, a wall-less, membranebound organism. Earlier investigations using the polyethylene glycolmediated transfection method resulted in failure to transfect A. laidlawii with purified MV-L3 D N A (Sladek & Maniloff, 1985), and it has been discussed that molecular size and linearity might prevent DNA uptake (Razin, 1985). However, Lorenz et al. (1988) obtained MV-L3 virus particles upon electroporation-mediated transfection with MV-L3 viral DNA. Although the efficiency of this procedure was very limited, a surprising observation was made, namely the release of two different viruses from transfected cells. By plaque hybridization techniques MV-L3 particles, derived from transfecting MV-L3 DNA, and MV-L1 particles, produced after induction of an MV-LI provirus, were identified (Just et al., 1989). In early experiments, MV-L3 was grown either on A.

laidlawii JA1 (Liss & Maniloff, 1973) or on strain 1305/K2 (Haberer et al., 1979). Due to the release of induced MV-L1 virus during MV-L3 growth in these strains, further experiments were performed only with strain U1 (a derivative of 1305/K2) which is resistant to MV-L1. Electrophoresis of uncleaved viral DNA molecules did not show any unusual pattern or distribution in the gels. Electron microscopic length measurements showed unit length among observed molecules of native DNA. Incubations with restriction endonucleases were done following the instructions of the enzyme manufacturers. Digestions with two enzymes were performed according to published procedures (Maniatis et al., 1982). Agarose gels used were 0.5% to 1.5% horizontal gels in Trisacetate-EDTA electrophoresis buffer. PAGE was done in 5% to 12% acrylamide vertical gels in Tris-boratebuffered solutions. Endonuclease PstI (fig. 1, lane 6) produced 12 fragments, named A to L, which show equal molarity as determined by scanning the negative of a Polaroid photograph of the separation gel. An additional faint band, however, is located near B, thus introducing the designations B1 (major band) and B2 (minor band). Fragments K and L were detectable only in hybridizations and/or' polyacrylamide gels (6.5% acrylamide concentration). Table 1 summarizes the results for the determination of fragment sizes (mean values from 25 gels). It is noteworthy that the sum of fragment sizes is less than the value from contour length measurements.

0000-9461 © 1990 SGM

2158 1

Short communication 2

3

4

5

6

6

T a b l e 1. F r a g m e n t sizes f r o m difJerent M V - L 3 restriction patterns Enzyme XbaI

~,,.A , ~ BI/z

mA '.'~B

"" A B1 tomB2

~

C

.~ C Cl/2 uD ~D ~E

"~F ""G .,,,. H -,.~,j --I --K ~L

.,~E "" F

"~D

mE

~G H "--I -'-J

~F --'G

".-K ,,--.L "~M

A B C D E F G H I J K L M M2 N

7440* 6558 4883 3230 2621 2527 1732 1665 1444 1274 885 743 700 approx. 690? 483

Enzyme PstI A B1 Bz C D E F G H I J K L

12757 8251 approx. 81007 5734 2772 2285 1280 1066 817 769 679 234 104

Enzyme HinclI A B1/2 C~t z D E F G H I J K L M

6636 5459? 4140 2805 2619 2223 1873 1555 1450 1352 [026 895 60

* Fragment size (bp). These are mean values from 25 gels at least. The standard deviation equals ___0.92 to 2.6~ (the latter value for small fragments). ~ Fragment M2 from the XbaI pattern, B2 from the HincIl pattern, and B2 from the PstI pattern are minor bands in terms of their concentration in the gel.

~H --I

"~N

Fig. 1. Restriction patterns of MV-L3-DNA with different restriction enzymes. Double digestions with these enzymes are included. Separation of fragments was performed on a horizontal 0.7~ agarose gel. Digestions were: HincII/PstI, lane 1 ; HincII, lanes 2; HincII/XbaI, lane 3; XbaI, lanes 4; PstI/XbaI, lane 5; PstI, Lanes 6. The scheme shows the nomenclature of bands within the restriction patterns of MVL3* HincII, MV-L3* XbaI and MV-L3* PstI. Fragments B2 from the HincII pattern, B2 from the PstI pattern, and M2 from the XbaI pattern are minor fragments in terms of their concentrations relative to all other fragments.

Fragment B 2 has not been included in these calculations, and will be discussed later. Using nuclease XbaI, we detected 14 bands, named A to N (Fig. 1, lane 4). In acrylamide gel electrophoresis and hybridization studies no additional bands were visible. On first examination a DNA band comparable to MV-L3* PstI B 2 (MV-L3 fragment B2 from PstI digestion) was not detected in restriction patterns. However, in hybridization studies and in concentrated agarose gels (1-5~ agarose), an additional very faint band in the gel was detected, which had been covered by MV-L3* XbaI M, leading to the designation M2 for the minor band. Without regard to this fragment, the sum of fragment sizes from all fragments did not equal the value

for the size of the native molecule as determined by electron microscopy. Restriction fragment sizes are summarized in Table 1. The HincII restriction pattern is defined by 14 bands, named A to M (Fig. 1, lane 2). In PAGE analysis no additional small bands could be detected. Fragment C is composed of two comigrating bands, which could be resolved in double digestions with two different restriction enzymes. These fragments have been named C1 and C2. They do not share homologies which could be demonstrated by hybridization studies. An additional faint band comparable to the MV-L3* PstI B 2 o r MV-L3* XbaI M 2 w a s not visible at first. Only when a double digestion with two restriction nucleases (HincII and PstI; Fig. 1, lane 1) was applied, a new faint band could be detected at the former position of MV-L3* HincII B. Thus the new band was named MV-L3* HincII B2. The existence of this minor band was confirmed in hybridization studies indicating that it did not originate from partially cleaved material. Summing up fragment sizes, shown in Table 1, resulted in a value for the length of the physical map less than that of the native DNA molecule. Digestion with nuclease KpnI resulted in a restriction pattern characterized by one major band and by at least three minor bands plus several additional very faint bands.,The fragment size of the major band equals approximately 8800 bp, more than 25 ~ of the length of a MV-L3 DNA molecule. The sum of all fragment sizes, including all minor bands, exceeds the genome length by at least twofold.

Short communication

Repeating the experiments with MV-L3* KpnI with longer incubation times and with higher enzyme concentration did not affect the restriction pattern. Using the enzyme to cleave cloned MV-L3* XbaI fragments or D N A from phage 2 resulted in regular patterns, defined by equal fragment concentrations. Thus the unusual MVL3* KpnI pattern reflects the structure of the viral D N A and is not influenced by the incubation procedure. Double digestions with either HinclI, PstI or XbaI with KpnI revealed that KpnI recognizes only two cleavage sites in MV-L3 DNA. Restriction enzyme assays with AluI, HpaI and EcoRI showed at least two bands. In combination with enzymes such as HinclI, XbaI or PstI, we could show that the first group of enzymes has only a single cleavage site in MV-L3 DNA. To facilitate the construction of the physical map, we started to clone restriction fragments of individual restriction digests. Fragments were excised from the gel and electroeluted with a Biotrap device (Schleicher & Schfill). The purification of the probe by D E A E Sephacel (LKB-Pharmacia) was as described by Maniatis et al. (1982). The MV-L3* PstI fragments were cloned via the shotgun technique into vector pAT153 (Twigg & Sherratt, 1980). The eluted MV-L3* XbaI fragments were cloned into an XbaI site, which had been introduced into the same vector by replacement of the PstI site by an XbaI linker (pATX153). The host for recombinant D N A was Escherichia coli strain C600 (Appleyard, 1954) grown in LB-broth (Miller, 1972). The plasmids were purified by NACS37 column chromatography (Gibco-BRL) or by caesium chloride-ethidium bromide density gradients. The alkaline lysis procedure prior to column chromatography was adapted from the accompanying manual for the NACS37 matrix (comparable to the published procedure of Birnboim & Doly, 1979). Column chromatography was according to the manufacturer's manual except for the linear salt gradient ranging from 0.5 M-NaC1 to 1.0 M-NaC1 in T r i s - E D T A 10/1 pH 7.2. All restriction fragments from MV-L3* XbaI have been cloned into the modified vector pATX153 (Fig. 2). Only fragment M2, which shares homology with fragment I, has not been cloned, because of its extremely low concentration. All clones have been proved to carry the particular fragment (in colony hybridization as well as in Southern blot analysis). At first, attempts to clone the MV-L3* XbaI fragment I failed completely. Only some clones were obtained which carried the I fragment connected to another fragment of the XbaI restriction pattern. One of these clones, harbouring I and J, was used for further subcloning. After incomplete digestions of the recombinant plasmid and subsequent self-ligation, 5000 clones were tested. Only one single clone carried fragment I as shown in Fig.

XbaI L3 A B C D E

F G H

I

2159

J K L M N

Fig. 2. Panel of clones after introduction of individual MV-L3 XbaI fragments into the vector pATXI53. The MV-L3 XbaI restriction pattern is shown in comparison to the cloned fragments, which were cleavedout of the vector with XbaI. Letters represent the names of the fragments, ranging from A to N. The position of the linearized vector (pATX153)is indicated by the arrowhead. Clone I was an artefact. The insert hybridized to I and showed the same size as I. The nucleotide sequence, however,is differentfrom the originallyisolated component of the MV-L3* XbaI I + J fragment. For more details see text.

2, lane I. However sequence studies showed that one terminal region of this clonable fragment I was no longer homologous to that of the 'master' fragment I, from which it was derived during subcloning. In Southern blots, however, this new fragment hybridized to fragment I. The sizes of the two fragments in question were indistinguishable on a gel. Due to the uncertain state of the clone discussed above, only the XbaI fragment I which orginally had been integrated into the vector together with fragment J was used for further studies. The I band was always cut out of the vector and separated from the other fragments by gel electrophoresis and subsequent electroelution. Filter hybridization studies were used to improve the data from the restriction analyses. Southern blots were performed on nitrocellulose filters using standard capillary soaking procedures. Probes for hybridization were labelled with

2160

Short communication

[32p]dCTP or Biotin-11-dUTP (Gibco-BRL). The nonradioactive detection system was as described in the manual for the BluGene detection kit (Gibco-BRL). Probe fragment MV-L3* PstI A hybridized to PstI B2, to several smaller bands, and of course to PstI A (data not shown). Hybridization to the MV-L3* XbaI pattern revealed the existence of the band called M> Fragments MV-L3* PstI B 1 and B2 are so similar in size that they could not be separated from each other. When jointly eluted, the mixture showed as expected hybridization signals with PstI B1 and B 2, but also with A. The names of fragments in Fig. 3 are indicated where needed, particularly in the case of the MV-L3* XbaI pattern. In each lane of the blot shown in Fig. 3 it can be seen that strong and faint hybridization signals are produced. The strong ones derive from hybridization with B1, the other ones from B2. In addition, the faint bands hybridize only to those bands which show homology with MV-L3* PstI A. Therefore B2 and A share homologous sequences. Hybridization with MV-L3* XbaI I led to an unexpected result: fragment I hybridized to itself and to M. Regarding the physical map of MV-L3 DNA in Fig. 4, it is obvious that this specific hybridization signal originates from the homology to M2. No similar results have been obtained with all other cloned probes. Another important result of hybridization studies was the finding that the faint bands MV-L3* HinclI B2 and MV-L3* PstI Bz share homologous sequences with MV-L3* HinclI A and MV-L3* PstI A, respectively. This indicates that faint bands of restriction patterns from HinclI, PstI and XbaI digestions are obviously derived from well defined larger bands of the same restriction pattern, their precursor bands. The computation of MV-L3 DNA restriction fragment sizes obtained with several restriction endonucleases, combined with double digestion analyses and hybridization studies, led to the construction of the physical map shown in Fig. 4. Fragments L and M from the restriction patterns PstI and HinclI, respectively, could be mapped only in comparison with the map position of MV-L3* XbaI B. By cleaving the latter fragment either with HinclI/XbaI or with PstI/XbaI from the vector, it was always found that there was one very small band missing, namely PstI L and HinclI M, respectively. We could show that the restriction enzyme analysis of MV-L3 DNA led to unexpected results. In contrast to other linear DNA, e.g. DNA from bacteriophage 2, there was not a single restriction pattern showing complete homogeneity in terms of the concentrations of fragments. Using restriction endonucleases with only one or two cleavage sites in MV-L3 DNA, fragmentation patterns were obtained which could not be explained easily. On the other hand, digestions with nucleases with

(a) 1

(b) 2

3

4

5

6

7

8

1

2

3

4

5

6

7

Fig. 3. Hybridization of MV-L3 PstI Bt/Bz fragments to D N A of different restriction patterns. The probe has been labelled with BiotinI I-dUTP. Patterns were derived from digestions with: HinclI/XbaI, lanes 1 ; HinclI, lanes 2; Hincll/Pstl, lanes 3; PstI, lanes 4; PstI/XbaI, lanes 5; marker 2HindlI1/EcoRl, lanes 6; XbaI, lanes 7; marker 2HindlIl, lanes 8. Panel (a) shows the ethidium bromide-stained gel whereas (b) shows the results of the Southern blot hybridization. Fragments, which are discussed in detail in the text, are indexed in the figure. The signal deriving from hybridization of MV-L3 PstI fragments B1/B_, to Xbal M_, is rather weak, because M 2 exists in limited amounts only. This is due to circular permutation and the existence of the postulated pac site.

several recognition sites results in a restriction pattern with at least one band of reduced concentration. The mean value for the length of the physical map is 36.2 kbp. This differs by 3.2 kbp (equivalent to 8-1~) from the length of native MV-L3 DNA molecules, that is, 39.4 kbp as determined by electron microscopy. This fact can be explained by either internal duplication or terminal redundancy. Duplication of internal or terminal regions should be recognizable in gels, because certain fragments should than be present in twice the concentration with respect to others. We have never observed such cases. However in the HincII restriction pattern a band with double intensity (C1/C2) has been found but later identified as consisting of two comigrating bands. Terminal redundancy has been found in many viral systems such as Salmonella phage P22 (Tye et al., 1974a, b) and Bacillus subtilis phage SPP 1 (Deichelbohrer et al.,

Short communication

XbaI Hincll Pstl Alul'E'°Rl'ipaI'KpnI

B

K

A~N).

D

/ 1 A

K

2161

L M

H Fig. 4. Physical map of MV-L3 DNA. The map includes data from several restriction patterns. Circles from inside to outside: 1, scale in kbp. The physical map spans 36-2 kbp, whereas a native MV-L3 DNA molecule accounts for 39-4 kbp; 2, restriction map for PstI (fragments' names from A to L) and maps for KpnI, HpaI, EcoRl and Alul (inner part of the circle); 3, restriction map for HinclI (fragments' names from A to M); 4, restriction map for XbaI (fragments' names from A to N). Bold segments of the circle, starting alan arrow, are so-called minor fragments, which could be seen in restriction digests. The arrows point to the postulated pac site where packaging of the viral DNA initiates from a postulated concatemeric precursor.

1982). However, it is obvious that terminal redundancy alone cannot explain the restriction data obtained. Terminal redundancy without any additional structural feature can be excluded by means of another experiment: it was shown that all fragments from the MV-L3* XbaI restriction pattern could be cloned into the E. coli vector pATI53. It is not very likely that the termini of a linear d s D N A would carry a structure identical to the one produced by cutting with XbaI. That all fragments from a particular restriction enzyme pattern can be cloned, however, is a strong indication for circular permutation of the D N A in question. This structural feature is also well known in other viral systems (Schnitzler et al., 1987; Soltau et al., 1987). Compared to the D N A of bacteriophage ~b149 (Chowdhury et al., 1987), having a circular permutation which is restricted to a certain segment of the genome, we assume that for MV-L3 D N A the permutation is complete. This means that all parts of the genome may be located at terminal positions. Otherwise some fragments might appear in major, others in minor fragment concentrations in the gel. Scanning the photographic negatives of

an ethidium bromide-stained gel did not show any irregular fragment concen~trations other than those discussed above. The minor fragment B2 from the PstI restriction pattern was recovered. Additionally, the minor fragment B2 from the HinclI restriction pattern could also be identified after double digestion with HinclI/PstI. In this case the dominant fragment B 1 was cleaved by PstI whereas the uncleaved HinclI B2 fragments remained at the original position. However, the minor fragment from the XbaI pattern, M2, could not be detected by this method. One of the best known bacteriophages having a circularly permuted and terminally redundant genome is Salmonella phage P22. For its D N A replication cycle the term "pac' for packaging site has been introduced (Jackson et al., 1978a, b). With respect to the P22 model which has been supported by a variety of other bacteriophage systems (B. subtilis phage SPP 1 (Deichelbohrer et al., 1982), E. coli phage P1 (Sternberg & Coulby, 1987a, b) we apply this term for our viral D N A system too. As already mentioned, restriction nucleases having several cleavage sites in MV-L3 D N A (e.g. PstI, HinclI and XbaI) produce regularly concentrated bands but also one defined minor band. We assume that the regularly concentrated bands are produced by cuts with a restriction endonuclease whereas the single less concentrated fragment is due to the headful packaging cleavage at pac and one restriction site. Pac site-specific cleavage is assumed to occur in a concatemeric precursor D N A molecule during the processing of replicative intermediates. Concatemeric molecules have been found in several bacteriophage systems, particularly in those with circularly permuted and terminally redundant genomes. Haberer & Maniloff (1980) have reported fast sedimenting material isolated from A. laidlawii cells infected with MV-L3. These findings will be further studied in our laboratory. From the relative positions of the restriction sites and from the discussion of the model, a circular representation of the physical m a p was derived (Fig. 4). The inner circle is a length scale. Physical maps for PstI, HinclI and XbaI are shown as concentric circles. The inner part of the PstI m a p also carries the restriction sites for EcoRI, AluI, HpaI and KpnI. Bold lined sections represent the minor fragments of each restriction pattern, showing that all of them are located in the same area. The arrow points to the start site of packaging, the pac site, which is common to all replicative intermediate molecules to be processed and determines the left-hand side of the so-called minor fragments of the first unit molecule. This work was supported by the Deutsche Forschungsgemeinschaft grant KL404/5-1 to G.K.

2162

Short communication

References APPLEYARD, R. K. (1954). Segregation of new lysogenic types during growth of a doubly lysogenic strain derived from Escherichia coli KI2. Genetics 39, 440 452. BIRNBOIM, H. C. & DOLY, J. (1979). A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucleic Acids Research 7, 1513 1523. CHOWDHURV, R., RAY, A., RAY, P. & DAS, J. (1987). Replication and packaging of choleraphage phi 149 DNA. Journal of Virology 61, 3999-4006. DEICHELBOHRER, I., MESSER, W. 8¢. TRAUTNER, T. A. (1982). Genome of Bacillus subtilis bacteriophage SPPI: structure and nucleotide sequence of pac, the origin of DNA packaging. Journal of Virology 41, 83 90. GOURLAY, R. N. 8z WYLD, S. G. (1973). Isolation of mycoplasmatales virus-laidlawii 3, a new virus infecting Acholeplasma laidlawii. Journal of General Virology 19, 279 283. HARBERER, K. & MANILOFF, J. (1980). Virus and host cell DNA synthesis during infection ofAcholeplasma laidawii by MVL3, a nonlytic cytocidal mycoplasmarirus. Journal of Virology 33, 671 679. HARBERER, K., KLOTZ, G., MANILOFF, J. & KLEINSCHMIDT, A. K. (1979). Structural and biological properties of mycoplasmavirus L3: an unusual virus-procaryote interaction. Journal of Virology 32, 268-275. JACKSON, E. N., JACKSON, D. A. & DEANS, R. J. (1978a). EcoRI analysis of bacteriophage P22 DNA packaging. Journal of Molecular Biology 118, 365-388. JACKSON, E. N., MILLER, H. I. & ADAMS, M. L. (1978b). EcoRI restriction endonuclease cleavage site map of bacteriophage P22 DNA. Journal of Molecular Biology 118, 347 363. JusT, W., DA SILVA CARDOSO, M., LORENZ, A. & KLOTZ, G. (1989). Release of mycoplasmavirus L1 upon transfection with homologous and heterologous viral DNA. Archives of Virology 107, 1 13. LISS, A. & MANILOFF, J. (1973). Infection of Acholeplasma laidlawii by MVL51 virus. Virology 55, 118-126. LORENZ, A., JUST, W., DA SILVA CARDOSO,M. & KLOTZ, G. (1988). Electroporation mediated transfection of Acholeplasma laidlawii with mycoplasma virus L1 and L3 DNA. Journal of Virology 62, 3050 3052.

MANIATIS, T., FRITSCH, E. F. & SAMBROOK, J. (1982). Molecular Cloning." A Laboratory Manual. New York: Cold Spring Harbor Laboratory. MANILOFF, J., HABERER, K., GOURLAY, R. N., DAS, J. & COLE, R. (1982). Classification and nomenclature of mycoplasma viruses. Intervirology. 18, 177- 189. MILLER, J. H. (1972). Experiments in Molecular Genetics, p. 433. New York: Cold Spring Harbor Laboratory. RAZIN, S. (1985). Molecular biology and genetics of mycoplasmas (Mollicutes). Microbiological Review 49, 419-455. SCHNITZLER, P., SOLTAU,J. B., FISCHER, M., REISNER, H., SCHOLZ, J., DELIUS, H. & DARAI, G. (1987). Molecular cloning and physical mapping of the genome of insect iridescent virus type 6: further evidence for circular permutation of the viral genome. Virology 160, 66 74. SLADEK, T. L. 81. MANILOFF, J. (1985). Transfection of REPmycoplasmas with viral single-stranded DNA. Journal of Virology 53, 25 31. SOLTAU, J. B., FISCHER, M., SCHNITZLER, P., SCHOLZ, J. • DARAI, G. (1987). Characterization of the genome of the insect iridescent virus type 6 by physical mapping. Journal of General Virology 68, 2717-2722. STERNBERG, N. 8~ COULBY,J. (1987 a). Recognition and cleavage of the bacteriophage P1 packaging site (pac). I. Differential processing of the cleaved ends in vivo. Journal of Molecular Biology 194, 453~468. STERNBERG, N. 8,~COULBY, J. (1987b). Recognition and cleavage of the bacteriophage P1 packaging site (pac). II. Functional limits of pac and location of pac cleavage termini. Journal of Molecular Biology 194, 469~479. TWlGG, A. J. & SHERRATr, D. (1980). Trans-complementable copynumber mutants of plasmid ColE1. Nature, London 283, 216 218. TvE, B.-K., CHAN, R. K. & BOTSTEIN, D. (1974a). Packaging of an oversize transducing genome by Salmonella phage P22. Journal of Molecular Biology 85, 485 500. TYE, B.-K. HUBERMANN, J. A. & BOTSTEIN, D. (1974b). Non-random circular permutation of phage P22 DNA. Journal of Molecular Biology 85, 501-532.

(Received 12 January 1990; Accepted 8 May 1990)