Yeast Yeast 2001; 18: 261±272.
Research Article
Large circular and linear rDNA plasmids in Candida albicans David H. Huber,{ and Elena Rustchenko* Department of Biochemistry and Biophysics, University of Rochester Medical School, Rochester, NY 14642, USA * Correspondence to: E. Rustchenko, Department of Biochemistry and Biophysics, Box 712, University of Rochester Medical School, Rochester, NY 14642, USA. E-mail:
[email protected] { Current address: Department of Biology, West Virginia State College Institute, WV 25112, USA
Received: 16 August 2000 Accepted: 12 September 2000
Abstract Although plasmids containing rRNA genes (rDNA) are commonly found in fungi, they have not been reported in Candida. We discovered that the yeast opportunistic pathogen Candida albicans contains two types of rDNA plasmids which differ in their structure and number of rDNA repeats. A large circular plasmid of unknown size consists of multiple rDNA repeats, each of which includes an associated autonomously replicating sequence (ARS). In contrast, a linear plasmid, which is represented by a series of molecules with a spread of sizes ranging from 50±150 kbp, carries a limited number of rDNA units and associated ARSs, as well as telomeres. The number of linear plasmids per cell is growth cycle-dependent, accumulating in abundance in actively growing cells. We suggest that the total copy number of rDNA is better controlled when a portion of copies are on a linear extrachromosomal plasmid, thus allowing a rapid shift in the number of corresponding genes and, as a result, better adaptation to the environment. This is the ®rst report of a linear rDNA plasmid in yeast, as well as of the coexistence of circular and linear plasmids. In addition, this is a ®rst report of naturally occurring plasmids in C. albicans. Copyright # 2000 John Wiley & Sons, Ltd. Keywords:
ribosomal RNA; rDNA; plasmids; Candida
Introduction Natural plasmids are commonly found in lower eukaryotes, including yeasts and other fungi (reviewed in Fukuhara et al., 1997; Grif®th, 1995; Meinhardt et al., 1990). Depending on the organism, extrachromosomal DNA can occur in circular or linear form, can have various molecular structures and can be located in the cytoplasm, the mitochondria or the nucleus. A mixed population of different plasmid forms is also found. For example, linear plasmids in yeast are usually found in pairs (Fukuhara, 1995). It is very interesting to note that all previously demonstrated linear plasmids encode either a DNA polymerase, an RNA polymerase or both (Kempken et al., 1992). If a particular organism has more than one linear plasmid, both types of polymerases are usually encoded on different plasmids. An interesting, and perhaps unique, system of three linear plasmids has been described in Debaryomyces hansenii, in which two plasmids can not be maintained in the absence Copyright # 2000 John Wiley & Sons, Ltd.
of the third (Fukuda et al., 1997). In a few instances linear and circular plasmids have been reported together (Feagin et al., 1992). The presence of rRNA genes on extrachromosomal molecules has been reported in organisms throughout the phylogenetic tree, including circular or linear plasmids in lower eukaryotes. However, rDNA linear plasmids in yeast have not been previously reported. The organization of genomic copies of rDNA in C. albicans was studied earlier by us and others (Rustchenko et al., 1993; Iwaguchi et al., 1992; Wickes et al., 1991). It was shown that the rDNA units are clustered as tandem repeats on both homologues of chromosome R, and that the total number of units can vary among different strains by as much as 2.6-fold. The rDNA unit sizes also varied among analysed strains, revealing 11.5, 11.6, 12.2 and 12.5 kbp unit sizes. We found that the number of genomic copies of rDNA in a strain is under complex regulation. On one hand, a physiological condition, the temperature of incubation,
262
de®nes the predominant type of cells having certain number of rDNA units in a population. On the other hand, the number of rDNA units in mutants with spontaneous chromosomal rearrangements seems to be under the control of the same mechanism that controls the chromosomal instability (Rustchenko et al., 1993). Extrachromosomal rRNA genes have not been previously identi®ed in Candida albicans. Here we report two rDNA-containing plasmids in C. albicans, a linear and a circular form. The larger circular plasmid is of unknown size and consists of multiple rDNA units. Its function is totally unknown. In contrast, the linear plasmid, which is represented by molecules with a spread of sizes of 50±150 kbp, carries a limited number of rDNA units. The copy number of this plasmid depends on the growth cycle, with the accumulation of the large amounts of the plasmid in actively dividing cells and a strong decline in copy number during the stationary phase of growth. We suggest that the total copy number of rDNA units in C. albicans is controlled in part by variation in the copy number of the linear extrachromosomal plasmid. We speculate that rapid change in the total number of rDNA copies is important for adaptation and, ultimately, better survival in the mammalian host. This is the ®rst report of an rDNA plasmid in yeast, as well as of the coexistence of circular and linear plasmids. Also, this is the ®rst report of naturally occurring plasmids in C. albicans.
Experimental procedures Strains The C. albicans unrelated laboratory strains 3153A, CAF4-2 and WO-1 used in these studies were previously characterized for their electrophoretic karyotypes (Rustchenko et al., 1993; RustchenkoBulgac et al., 1990; Rustchenko-Bulgac and Howard 1993). Strains WO-1 white and WO-1 opaque, which constitute two temperaturedependent interconvertible colonial phenotypes, were obtained in this work from strain WO-1.
Molecular size markers The size of the linear extrachromosomal DNA molecules was determined using the Lambda Copyright # 2000 John Wiley & Sons, Ltd.
D. H. Huber and E. Rustchenko
Ladder PFG Marker (New England BioLabs, Beverly, MA). The chromosomal sizes of S. cerevisiae strain 867, as described in Rustchenko et al. (1993), were used as size markers for C. albicans chromosomes.
Media and growth conditions Yeast extract/peptone/dextrose (YPD) medium, solid or liquid, was previously described (Sherman et al., 1982). Strains were stored and maintained by using procedures that prevent phenotypic changes resulting from chromosome instability (Rustchenko-Bulgac, 1991; Perepnikhatka et al., 1999). All strains were grown at 37uC except the opaque cells of WO-1, which were grown at room temperature, 23±28uC. Cells were either used immediately after growth or were stored at x70uC.
Growth cycle experiments with strains 3153A and WO-1, white and opaque Strains 3153A and WO-1, white and opaque, were grown in 200 ml YPD medium with vigurous shaking. The density of cells was monitored by counting cells with a microscope and a cell counting chamber. Chromosomes were prepared from cells collected at three different phases of growth: logarithmic, early stationary and late stationary. In addition, late stationary cells were re-inoculated into fresh YPD medium for preparation of chromosomes from logarithmic phase cells.
Growth of C. albicans 3153A and CAF4-2 prior to chromosomal preparation Cells stored at ±70uC were streaked on YPD plates and allowed to grow overnight. A mass of cells was then collected and approximately 5000 colony forming units (cfu) were spread and incubated on each of several YPD plates. After growth of relatively young colonies containing approximately 106±107 cells, the colonies were washed from the plates and stored at ±70uC or used immediately for chromosomal preparation.
Growth of C. albicans strains WO-1 white and WO-1 opaque prior to chromosomal preparation Cells of C. albicans WO-1 stored at x70uC were streaked on YPD plates and incubated at room temperature (see above). The colonies initially Yeast 2001; 18: 261±272.
Large circular and linear rDNA plasmids in Candida albicans
appeared on YPD plates as the white phenotype, but after 5 days of incubation microscopic opaque phenotype sectors appeared on the margins of the white colonies. After 6±7 days, when opaque sectors had grown suf®ciently, cells from the white and opaque portions of the same colony were collected, plated for approximately 5000 cfu/plate, incubated at room temperature for 2 days and harvested as above. After development, the colonies were visually screened in order to ensure that the populations of cells spread on the plates were of a similar phenotype.
Pulsed-®eld gel electrophoresis The isolation of intact C. albicans chromosomes and use of agarose plugs was previously described (Carle and Olson, 1984; Perepnikhatka et al., 1999). We used two versions of pulsed-®eld gel electrophoresis (PFGE): an orthogonal ®eld, alternating gel electrophoresis, or OFAGE, and a contourclamped homogeneous electric ®eld system, or CHEF, with the CHEF Mapper version (Bio-Rad Laboratories, Hercules, CA). OFAGE running conditions, designed to optimize chromosomal separation of differently sized groups of chromosomes, short or bottom (B), medium-sized (M) and long or top (T), were applied singly or in combinations, as described previously (Rustchenko-Bulgac and Howard, 1993). Care was taken to select a running time that would not allow the fastmigrating linear extrachromosomal band to run off the gel. CHEF was used for the optimal separation of the long chromosomes (see Rustchenko-Bulgac, 1991). In order to determine the size of the extrachromosomal band identi®ed on the bottom of the gel, and also to investigate whether it migrates as a linear molecule, the following additional conditions were used; OFAGE with constant voltage of 9 V/cm, a 50 s pulse interval and 1% (w/v) of agarose, leaving the other parameters the same as described in Rustchenko-Bulgac et al. (1990), as well as CHEF Mapper automated condition for 1% (w/v) of agarose in gel. The optimum condition for visualization of the upper extrachromosomal band using OFAGE was as follows: constant voltage of 7.5 V/cm; a 250 s pulse interval; Copyright # 2000 John Wiley & Sons, Ltd.
263
1% (w/v) agarose; a 0.5rcooling buffer at approximately 13uC, and a 30±40 h running time.
Preparation of linear extrachromosomal DNA from PFGE The linear extrachromosomal DNA was prepared from C. albicans WO-1 opaque by excising corresponding ethidium bromide stained bands of 50±150 kbp in size from several 1% agarose OFAGE gels. These agarose blocks were then embedded in a conventional 3.5% NuSieve GTG low-melting agarose gel (FMC BioProducts, Rockland, ME) in an orientation parallel to the electrical ®eld, and run at low voltage of 1.5 V/cm for 3±5 days. The concentrated DNA was identi®ed by ethidium bromide staining, and low-melting agarose blocks were excised and digested with b-agarase I according to the manufacturer's protocol (FMC BioProducts, Rockland, ME), followed by standard ethanol precipitation of DNA.
Hybridization probes A rDNA probe was created from a 270 bp HindIII±XhoI fragment derived from plasmid WOL25. This fragment hybridizes to the middle band of approximately 3.7 kbp of three EcoRI bands that hybridize with the Ca5 rDNA probe published by Sadhu et al. (1991) (C. Sadhu and M. J. McEachern, unpublished data). The telomeric probe was composed of a 2.5 kbp BamHI fragment from the pBR322 subclone of the original Ca7 telomeric fragment (Sadhu et al., 1991). The probe contained a total of 376 bp of the 16 telomeric repeats and 8 bp of subtelomeric sequence (McEachern and Hicks, 1993). The mitochondrial DNA (mtDNA) probe consisted of the EcoRI 4.7 kb mtDNA fragment E5 (Wills et al., 1984), which was previously inserted in pBR322. All three fragments carrying the rDNA, as well as mitochondrial and telomeric segments, were gel-puri®ed prior to labelling. The probe for the autonomously replicating sequence (ARS) was composed of a 489 bp fragment derived from plasmid pRC2312 (Cannon et al., 1990). Because this ARS was originally cloned on the same fragment with 5S rDNA, care was taken to separate the ARS from the 5S rDNA in the following way. The Sau3AI± AvaII fragment of 670 bp containing an ARS was excised from the plasmid and gel-puri®ed. This Yeast 2001; 18: 261±272.
264
D. H. Huber and E. Rustchenko
fragment was then used as a template, together with primers ATGAGGTAGTGCAAGTTATAC and CGGTTGTTAGATTGAAGTTTG, to construct the probe by incorporating digoxigenin-dUTP during PCR, according to the manufacturer's protocol (Boehringer-Mannheim, Indianapolis, IN). A hybridization probe was created from the linear 50±150 kbp plasmid described below. Plasmid fragments under 10 kbp in size, required for ef®cient labelling, were produced by digestion with HindIII after elution of plasmid DNA from an OFAGE gel (see below). In order to remove all traces of agarose from the solution, the digested DNA was additionally puri®ed with a Qiagen gel extraction kit (Valencia, CA).
Other protocols Gels were blotted on Nytran nylon membranes (Schleicher and Schuell, Keene, NH) with a pressure blotter (Stratagene, La Jolla, CA). Both blotting and Southern hybridizations were carried out using regular protocols (Sambrook et al., 1989). The non-radioactive DIG DNA Labelling and Detection kit containing digoxigenin-dUTP (Boehringer-Mannheim, Indianapolis, IN) was used for labelling probes. For regular extractions of DNA from agarose gels, the Qiagen gel extraction kit (Valencia, CA) was used.
Results The presence of large linear and circular extrachromosomal DNAs in different laboratory strains Two extrachromosomal bands were identi®ed while analysing the electrophoretic karyotypes of the unrelated C. albicans laboratory strains, 3153A, CAF4-2, and two forms of WO-1, white and opaque. One band, which varied in size from approximately 50±150 kbp (Figure 1), ran in the same position independent of the running condition or type of apparatus (see Experiment and procedures), as indicated by lambda ladder size marker, a property characteristic for a linear con®guration (Beverley, 1988). This band also had variable amounts of DNA, in contrast to the ®xed amounts of chromosomal DNA present in each chromosome at a stochiometric ratio. This observation strongly suggests that this molecule replicates autonomously. Copyright # 2000 John Wiley & Sons, Ltd.
Figure 1. The visualization of the linear plasmid in strains WO-1 opaque (lane 1) and 3153A (lane 2). The OFAGE gel was run under conditions designed to determine the size distribution of a pool of linear plasmid molecules (see Experimental procedures). These conditions do not resolve C. albicans chromosomes, which are compressed within a single band denoted TMB at the top of the gel. Numbers indicate the range of sizes in kbp. S. cerevisiae (S.c.) chromosomes and lambda ladder (l) were used as size markers. The smallest brightly stained band in the S.c. lane corresponds to the 2m plasmid
Another low-intensity sharp band with an unusual shape having an abnormally large width was observed under optimal running conditions at the top of the gels, immediately below the three longest unresolved chromosomes (also denoted T for the top group of chromosomes), in strains CAF4-2 and WO-1, white or opaque, but not in 3153A (Figure 2A). Sometimes, under a combination of running conditions, which separated the short and medium chromosomes, the upper band was also visualized. However, the relative position of this band compared to the position of the chromosomes was variable, suggesting that its mobility is highly sensitive to minor uncontrolled differences in the preparation of the gel or in the running condition. The band could be found either immediately under the longest chromosomes (T group), or between the medium-sized chromosomes (M group) (data not shown). This band was not revealed under other standard conditions routinely used by us to separate Yeast 2001; 18: 261±272.
Large circular and linear rDNA plasmids in Candida albicans
265
Figure 2. The visualization of linear and circular plasmids. (A) Ethidium bromide stained gels of the electrophoretic karyotypes of C. albicans strains 3153A, CAF4-2 and WO-1 white (W). B (bottom), (M) (middle) and T (top) denote the three groups of the chromosomes, short, medium and long, respectively. The OFAGE conditions were chosen to accentuate the separation of the upper extrachromosomal band and the M group. Group B is well separated in the WO-1 strain, but not in the two other strains. The long chromosomes are compressed into a single wide band above the circular plasmid. For the complete electro-karyotypes of 3153A and CAF4-2, see Janbon et al. (1998). The arrows at the top and bottom of the gels indicate, respectively, the circular and linear plasmids. Note that the circular plasmid is not visible in 3153A. The linear plasmid appears as a tight band because the running condition caused compression of the corresponding area. (B) and (C) A Southern blot of the gel from (A) hybridized with the probes constructed from the telomere and ARS, respectively. For explanation, see (A)
chromosomes, which can be explained by a total loss of mobility, by co-migration with the chromosomes, or its distribution along the length of the gel. The last situation does not allow visualization by ethidium bromide staining due to the lack of brightness. Such lack of migration, as well as variability in position is the characteristic of circular DNA (see e.g. Beverley, 1988). The formation of this circular band could be due to continuous looping-out from chromosome R, containing the rDNA cluster (see below), or due to a slow rate of autonomous replication, which would explain the relatively low amounts of this element. It is not clear whether the absence of the circular plasmid in strain 3153A (Figure 2A) is due to a reduced amount, co-migration with chromosomes or to complete absence. However, the circular plasmid appeared to be present in a spontaneous morphological mutant 300-SG (Rustchenko-Bulgac and Howard, 1993) derived from strain 3153A (data not presented). Neither of the two plasmids separated from the bulk of the DNA by conventional electrophoresis, a result that is expected because of their large sizes. Copyright # 2000 John Wiley & Sons, Ltd.
The linear plasmids range in size from 50 to 150 kbp, sizes that are too long to be separated by conventional gels. Although large circular plasmids can be visualized on conventional gels, such as the 75 kbp rDNA plasmid from Trypanosoma cruzi (Wagner and So, 1992), the size of our plasmid, reported here, was estimated to be much larger, 1.2 Mbp (see below).
Molecular characterization of the linear and circular extrachromosomal DNAs Chromosomal blots were analysed by Southern hybridization using the following probes: rDNA; an ARS originally associated with 5S rDNA; telomeric DNA; and C. albicans mtDNA (see Experimental procedures). The telomeric probe hybridized with the bottom extrachromosomal linear band in all of the strains tested, as well as with all chromosomes, with the intensities proportional to the respective ethidium bromide staining (Figures 2A, B). However, no hybridization occurred with the upper extrachromosomal bands of CAF4-2 and WO-1, opaque or Yeast 2001; 18: 261±272.
266
white, consistent with the deduced circular nature of the molecule lacking telomeres (see above). Because both rDNA and ARS probes gave signals of similar intensity, which is consistent with the presence of an ARS in each rDNA unit, we present only the result of hybridization with the ARS probe (Figure 2C). The upper extrachromosomal band showed disproportionately stronger signals with both rDNA and ARS probes, compared to its weak ethidium bromide staining. In fact, the signal intensity in the upper band was comparable with the intensity of signal from chromosomes R which carried rDNA clusters, which implies that the circle consists of multiple tandem repeats of rRNA genes. Knowing that the total number of rDNA units in strain WO-1 is approximately 200 (Rustchenko et al., 1993), and estimating the brightness of the signal from the plasmid as approximately one-half of the two chromosome R homologues (see group T in Figure 2C), we can deduce that the number of repeats on the circular plasmid is approximately 100. Also, knowing that strain WO-1 possesses rDNA units of two sizes, 11.5 kbp and 12.5 kbp (see Rustchenko et al., 1993), and assuming that the average size of a unit is 12 kbp, we can calculate the approximate size of the circle to be 1.2 Mbp. If, however, other sequences are present, the size of the plasmid would be larger. The large size explains why the circle failed to separate by conventional electrophoresis. Contrary to the upper band, the linear lower band showed rDNA and ARS signals consistent with its ethidium bromide staining, indicating the presence of a limited number of rDNA units, perhaps one or several at most. No signals were observed with the mt DNA probe, even when the ®lters were overexposed (data not presented).
At least a portion of the linear extrachromosomal band sequence is derived from chromosome R The complete electrophoretic karyotype of strain WO-1 opaque (Figures 3A, D), was prepared using two different OFAGE running conditions favouring resolution of either chromosomal groups B and M, or T, respectively (see Experimental procedures). Two deviated homologues of chromosome R that carry rDNA clusters were identi®ed by hybridization with the rDNA-associated ARS Copyright # 2000 John Wiley & Sons, Ltd.
D. H. Huber and E. Rustchenko
probe (Figure 3E). Detailed analysis of the WO-1electro-karyotype has been published elsewhere (Rustchenko-Bulgac and Howard, 1993; Rustchenko-Bulgac, 1991). Blots of the two gels (Figure 3A, D) were hybridized with DNA of linear extrachromosomal molecules from WO-1 opaque (Figure 3B, C, F; see Experimental procedures for the extraction and preparation of the band). Strong signals are seen in both the bottom and upper extrachromosomal bands (Figure 3B), as well as in the homologues of chromosome R (Figure 3F), which obviously can be attributed to the presence of multiple rDNAs and ARSs in the upper band and chromosome R (see above). All of the other chromosomes produced weak uniform signals only after overexposure (cf. Figure 3B, C), which is expected since they share with the linear band the telomeric DNA composed of relatively short sequences, approximately 345±690 bp (McEachern and Hicks, 1993). The regular exposure after similar hybridization with the blot optimized for the long chromosomes (Figure 3F) produced weak signals with the majority of chromosomal bands because of the compression of chromosomes of small and medium sizes, the B- and M-groups, respectively. Consistently, the smallest chromosome positioned on a top of the linear non-chromosomal band, which did not comigrate with any other chromosome, produced a barely detectable signal, which was far weaker than the signal from the linear plasmid. Although it appears that the linear plasmid is derived only from chromosome R sequences, we do not know if other sequences beside rDNA units are present.
Restriction enzyme digests of the linear extrachromosomal band The DNA of the linear plasmid was prepared (see Experimental procedures) and analysed with restriction enzymes HindIII, EcoRI and NotI (Figure 4A). Corresponding to earlier reports of two sizes of rDNA units, 11.5 and 12.5 kbp, in strain WO-1 (Rustchenko et al., 1993), digestion with NotI (lane 5) revealed two bands of appropriate sizes, which produced signals with the rDNA probe (Figure 4B). This method could not estimate the number of units, but it revealed that units retained the same length as genomic copies. It also uncovered the presence of more than one unit in at least a portion of the molecules, as well as their tandem organizaYeast 2001; 18: 261±272.
Large circular and linear rDNA plasmids in Candida albicans
267
Figure 3. Chromosomal blot analysis using linear extrachromosomal DNA as a probe. (A) The ethidium bromide stained gel demonstrating the well-separated B and M groups of chromosomes in strain WO-1 opaque (O) (see Figure 2A for explanation). (B) and (C) A Southern blot of the gel from (A) hybridized with the probe constructed from the linear plasmid, normal exposure and overexposure, respectively. The circular plasmid is denoted by arrows in (A) and (B). (D) The ethidium bromide-stained gel, demonstrating separation of the T group of strain WO-1 opaque (O). Note that the lowest band of the T group results from the co-migrating chromosomes 1 and 2, each comprised of a pair of homologues. The circular plasmid can not be detected using this condition. The linear plasmid is indicated by the arrow. The different amount of the linear plasmid in (A) and (D) is due to variations of independently-grown cells. (E) A Southern blot of the gel from (D) hybridized with the probe constructed from the ARS, revealing chromosome R. (F) The same blot hybridized with the linear plasmid probe, normal exposure. Note that the weak signals re¯ect the compression of the short and medium-sized chromosomes
Figure 4. Restriction endonuclease digestions of the linear extrachromosomal band. (A) DNA separation in a conventional gel. Lane 1, 1 kb DNA ladder; lane 2, uncut DNA of the linear plasmid; lanes 3, 4 and 5, linear plasmid DNA digested with HindIII, EcoRI and NotI, respectively. The arrows denote positions of the 3.7, 11.5 and 12.5 kbp bands. (B) A Southern blot of the gel in (A) hybridized with the rDNA probe Copyright # 2000 John Wiley & Sons, Ltd.
Yeast 2001; 18: 261±272.
268
tion, as NotI cuts only once within a unit. Thus, a unit-sized fragment can be recovered from a minimum of two tandemly positioned rDNA units. The weak ethidium bromide staining of the two bands in Figure 4A indicates the presence of a small number of tandemly arranged units carried by the linear extrachromosomal molecules, which is consistent with the Southern blot analysis of the chromosomes (see above). An unusual property of the NotI digest analysis with rDNA probe is an area of higher intensity signal, which most likely corresponds to the two rRNA bands, as well as an area of trailing signal above. It is consistent with the view that plasmid molecules are a mixed pool with various, but limited, copy number of the rDNA. If some molecules have only a single unit, then digestion of the plasmid would result in two fragments, most likely not of equal size, and positioned above the unit-sized bands. These fragments would probably create a distribution of sizes, as the plasmid is represented by a contiguous range of sizes (see above). No distinct bands were observed with the HindIII digest (Figure 4A, lane 3), a restriction endonuclease known to cut in the ¯anking regions of the chromosomal rDNA cluster (Rustchenko et al., 1993). This indicates strong sequence variability within the molecules. Because of the compression of long fragments in conventional electrophoresis, the position of signal appears similar to that of the control undigested DNA. A heavy background in the lane with EcoRI digest (Figure 4A, lane 4) supports the view that large variability in sizes and sequence occurs within plasmid molecules. Otherwise the signal with the rDNA probe in this lane corresponded to a 3.7 kbp band in the EcoRI digest from an earlier report (Magee et al., 1987). This is an expected results because the hybridization probe originated from this band (see Materials and methods), which probably explains a strong hybridization signal.
The physiological condition of the appearance of the linear band Examination of liquid cultures of C. albicans strains 3153A, WO-1 white and WO-1 opaque, showed that the amount of the linear plasmid varied with the growth cycle. As presented in Figure 5A for WO-1 opaque, logarithmically growing cells of the approximate density of 8r107 cells/ml, contained Copyright # 2000 John Wiley & Sons, Ltd.
D. H. Huber and E. Rustchenko
Figure 5. The appearance of the linear extrachromosomal band in the electrophoretic karyotype of WO-1 opaque (O) at different phases of growth in liquid cultures. For the running conditions, see Figure 1A. The electrophoretic karyotypes were performed on (A), logarithmically grown cells to an approximate density of 8r107 cells/ml; (B) cells entering in stationary phase at the approximate density of 5.75r108 cells/ml 12 h later; and (C) cells advancing to stationary phase at approximately 8.75r108 cells/ml, 50 h later. (D) The electrophoretic karyotype of WO-1 opaque (O) cells transferred from a late stationary culture (C) and grown to log phase in fresh YPD medium for 12 h. For explanation, see (A). The top and the bottom arrows denote the position of the circular and linear plasmids, respectively
an abundance of the plasmid. Cells in early and advanced stationary phase, approximately 5.75r108 cells/ml and 8.75r108 cells/ml, respectively, showed progressively diminished amounts of the plasmid (Figure 5B, C, respectively). The time of the culture growth between logarithmic and early stationary phases was relatively short, 12 h, and between early stationary and late stationary relatively long, 50 h. When cells were re-inoculated into fresh medium and were allowed to grow logarithmically for an additional 24 h, the plasmid was again found to be present in large quantities (Figure 5D). The number of cells between early and late stationary phases changed only about 1.6 times; however, the amount of linear plasmid changed dramatically. Unlike the linear plasmid, the amount of the circular plasmid remained the same in proportion to the amount of chromosomal DNA.
Discussion A sub-class of natural plasmids carrying rRNA genes (rDNA) is common. These plasmids can be Yeast 2001; 18: 261±272.
Large circular and linear rDNA plasmids in Candida albicans
found in both pathogens and non-pathogenic organisms, making their role in pathogenesis dubious. For example, the presence of a large rDNA containing circle in distant protozoa of the family Trypanosomatidae prompted the authors to speculate on its important role in the biology of these parasites (Wagner and So, 1992). Another prominent example is the pathogen Plasmodium falciparum, which has two coexisting rDNA plasmids, linear and circular (Feagin et al., 1992). On the other hand, the yeast S. cerevisiae carries circles with various numbers of rDNA repeats (Olson, 1991), as well as organisms such as slime moulds, Dictyostelium discoideum and Physarum polycephalum, which carry rDNA linear plasmids (Cole and Williams, 1992; Ferris, 1985). Although rDNA plasmid implication in pathogenesis is doubtful, it is quite possible that it might provide some physiological advantage for survival in a host, especially in the situation where an increase in synthetic activity in the pathogen is desirable. It is important for the purposes of this paper to mention that, in a number of organisms, all rRNA genes are located on plasmids. This is in contrast to the majority of living organisms, which have DNA units arrayed as multiple tandem repeats on a single chromosome. For example, in D. discoideum and P. polycephalum, all of the rRNA genes are carried on large linear plasmids, 88 kbp and 60 kbp respectively, and in the protozoan parasite Entamoeba histolytica they are on a circular plasmid of 24.5 kbp (Dhar et al., 1996). In addition, the soil amoebae of the genus Naegleria contain rRNA genes on different circular plasmids (De Jonckheere, 1989). Perhaps even more interesting are examples in which a unique copy or a few copies of the rRNA genes are stored in the genome, and multiple copies exist on plasmids. For example, the alga Euglena gracilis has about four copies of rRNA genes integrated in the genome, and between 800 and 4000 as extrachromosomal DNA (Ravel-Chapuis, 1988). Similarly, in the protozoan Tetrahymena a single rDNA copy is stored in the micronucleus. However, unlike E. gracilis, in Tetrahymena the extrachromosomal copies are periodically created de novo. During the sexual cycle a single copy ampli®es into up to 200 copies, organized as a single palindromic repeat on 22 kbp linear plasmids in the macronucleus, which is responsible for all transcriptional activity (Yao and Gall, 1978). These examples clearly indicate that rRNA genes are Copyright # 2000 John Wiley & Sons, Ltd.
269
required to be present in multiple copies, and their transfer to the plasmid is biologically functional and even advantageous. In this study, we have identi®ed two natural rDNA plasmids in different C. albicans laboratory strains; a linear form, which is represented by a pool of molecules of 50±150 kbp in size, carrying a limited number of rDNA units, and a large circular form of an unknown size with multiple rDNA repeats. The functional importance of these plasmids will be a subject of further study, although we have shown that the linear plasmid is growth-cycle dependent, accumulating in large amounts in actively dividing cells. Both plasmids can be visualized as ethidium bromide-stained bands at the top and the bottom of electrophoretic karyotypes of different laboratory strains (Figures 1, 2A). As discussed in Results, the intensity of their staining is inconsistent with the staining of nearby chromosomal bands, thus indicating autonomous replication, which, in its turn, implies the lack of a centromere. Further evidence for autonomous replication comes from the fact that both plasmids carry ARS sequences (Figure 2C). In unrelated experiments with mutants of strains CAF4-2 and 3153A, we observed the irreversible loss of the linear plasmid (data not presented), a ®nding that is consistent with the view that the linear plasmid replicates autonomously, and is not, for example, being continuously derived from chromosome R. The presence of telomeres on the lower but not the upper band (Figure 2B) indicates that the bands have two different forms, linear and circular, respectively. The speci®c migration pattern of each band, as well as a peculiar shape of the upper band, support this conclusion (see Results). The absence of hybridization of the mitochondrial probe to either of the plasmids indicates that these plasmids did not correspond to mitochondrial DNA. Both plasmids, however, hybridized with a chromosomal rDNA probe, which places them into the group of rDNA plasmids. Although we did not investigate the cellular location of the plasmids, it is reasonable to suggest that the circular plasmid is located in the nucleoli or nucleus by the analogy with rDNA circles in S. cerevisiae and other circular rDNA plasmids in different organisms (see e.g. Dhar et al., 1996). Due to the dif®culties in preparing suf®cient quantities, the circular molecule was not analysed with restriction endonucleases. Yeast 2001; 18: 261±272.
270
However, the deduced high number of the rDNA repeats indicated a size close to 1 Mbp (for details, see Results), which is consistent with the failure of separation by conventional electrophoresis. If the circular plasmid has the proposed large size, then the weak ethidium bromide staining, in comparison with the nearest chromosomal band (Figure 2A), points to a low copy number of plasmid per cell, possibly even one or no molecules in some cells. This explains the decreased staining in strain CAF4-2, as well as the absence of a visible band in 3153A. Perhaps strain-dependent differences in the rate of replication of the large circle are responsible for the visibly different amount of circular DNA. On the other hand, the consistency between the ethidium bromide staining of the linear plasmid and the hybridization signals for rDNA and ARS (Figure 2A, C), and the weak appearance of the two rDNA units obtained with NotI digest (Figure 4), indicate a limited number of rDNA repeats in the linear plasmid. Although similar ®ndings of a limited number of one or two rRNA transcriptional units on either circular or linear plasmids in different lower eukaryotes, including various number of units in circles of S. cerevisiae, have been reported previously (Ferris, 1985; Dhar et al., 1996; Johansen et al., 1992; Sinclair and Guarente, 1997), there have been no reports on an excessive number of rDNA copies in circular plasmids, comparable with our deduced result. We uncovered a prominent variability within the linear molecules, manifested by a range of sizes of 50±150 kbp (Figure 1), no distinct bands in HindIII digest, and a heavy background in the EcoRI digest (Figure 4). Several regions in the molecule could be responsible for this variability. Telomeres and subtelomeric regions in C. albicans WO-1 were previously sequenced and analysed (McEachern and Hicks, 1993). The sequences were reported to consist of 15±30 copies of 23 bp repeats, which would give change in length of about 345±690 bp. The sub-telomeres also vary in sequence content. A different, although limited, number of rDNA units, which are about 12 kbp in some strains and are represented by two sizes of 11.5 kbp and 12.5 kbp in strain WO-1 (Rustchenko et al., 1993; see also Figure 4), may also contribute to the variable size. If we assume that the smallest plasmid of 50 kbp contains three rDNA units and two telomeres of maximum size, a segment of at least 12 kbp remains to be accounted for. Copyright # 2000 John Wiley & Sons, Ltd.
D. H. Huber and E. Rustchenko
It is not known whether the two C. albicans plasmids are related to each other. In P. falciparum, two rDNA plasmids, a 6 kbp linear and a 35 kbp circular, coexist in the cell but are not closely related (Feagin et al., 1992). Also, the mechanism by which the plasmids reported here arise is not clear. It is reasonable to suggest that the circular plasmid originally formed as a looped out structure from a cluster of tandem rDNA repeats on chromosome R, presumably as a result of recombinational events, and is subsequently maintained by autonomous replication. It is doubtful that the linear plasmid was initially formed, for example, from a broken end of the chromosome R right arm carrying the rDNA cluster, which included one or two rDNA units, and which then acquired a telomere at the site of breakage. As the distance from the end of chromosome R to the rDNA cluster is at least 400 kbp (Chu et al., 1993), which is about eight times larger than the basic plasmid size of 50 kbp, this mechanism would require two events, the break of the end of the chromosome and a large deletion. Taking into account that the plasmid hybridized strongly with only chromosome R (see Results and Figure 3F), another hypothesis would be that the plasmid is assembled from a few rDNA units, telomeres and an unknown segment of DNA of approximately 12 kbp, which may or may not be derived from chromosome R, but which was not derived from any of the other chromosomes. The ability of fungal cells to add telomeres to linear DNA has been shown previously with transformed DNA (Long et al., 1998). The unknown DNA segment could contain at least 10 genes or may consist of some repetitive sequences, possibly unique to the plasmid, as observed in the other systems (Kempken et al., 1992). Perez-Martin et al. (1999) previously reported an extrachromosomal band of approximately 50±100 kbp in the electrophoretic karyotypes of SIR2 null mutants, but not in the parental strain CAI4. After sequencing several fragments, the authors concluded that the entire band consisted of rDNA. In the absence of more detailed analysis, it is dif®cult to relate these data to the two plasmids presented here. Although plasmids are commonly found in fungi, including a substantial number of yeast species, their functions have not been completely elucidated. The cytoplasmic linear plasmids encoding toxins, in particular two killer plasmids, pGKL1 and pGKL2, Yeast 2001; 18: 261±272.
Large circular and linear rDNA plasmids in Candida albicans
in Kluyveromyces lactis, are by far the best analysed (reviewed in Fukuhara et al., 1997; Hermanns and Osiewacz, 1992). Another relatively well-studied function is the senescence associated with the linear mitochondrial plasmids kalilo and marahar in fungi of the genus Neurospora (reviewed in Hermanns et al., 1995). Recently, opposite to senescence, longevity has been found to be related to a linear mitochondrial plasmid, pAL2-1, in a long-lived mutant of Podospora anserina (Hermanns et al., 1994). Speci®cally in yeasts, plasmid functions have been assigned to only two groups, killer plasmids (Fukuhara, 1995), and ageing-associated rDNA circles (Sinclair and Guarente, 1997). Although the ®rst mitochondria-associated plasmid pPK2 was recently reported (Blaisonneau et al., 1999), its function is still unknown. In this study, a newly identi®ed linear rDNA plasmid in C. albicans was found to increase in number in actively dividing cells, but to diminish dramatically when cells ceased to grow (Figure 5). It is not clear whether change in the amount of the plasmid is a result of a simple selection or is under a controlling mechanism. The strong decrease of the plasmid in late stationary phase, with a very small increase in the total cell number, is indicative of control. The ®nding of the physiological condition controlling the amount of the linear plasmid in Candida is reminiscent of the well-known cases of physiologically related ampli®cation of the cellular copy number of rDNA in other organisms. For example, in higher eukaryotes autonomous replication of extachromosomal rDNA occurs in the oocytes of amphibians, mammals and insects, and seems to be related to the high level of ribosomal synthesis in these cells (reviewed in Yu and Blackburn, 1990; Spear and Gall, 1973). A similar example in a lower eukaryote is the extrachromosomal rDNA copies ampli®ed de novo during the formation of the macronucleus in Tetrahymena (see above), which is responsible for all transcriptional processes. Because the rDNA copies organized as a circular plasmid did not seem to vary with the growth cycle, we assumed that the copy number of the linear form is more regulated, despite a limited number of rDNA units in each molecule. We suggest that the increase in rDNA in the opportunistic pathogen C. albicans is required during the active growth of cells, a condition that might be advantageous under certain situations in the host. In summary, we have reported for the ®rst time a Copyright # 2000 John Wiley & Sons, Ltd.
271
rDNA type of linear plasmid in yeast. This is also the ®rst report of circular and linear plasmids coexisting in a yeast species. Furthermore, this is the ®rst report on plasmids in C. albicans.
Acknowledgements We thank F. Sherman, Y.-K. Wang, E. Phizicky and G. Janbon for valuable discussions, as well as J. Greenberg, C. Haidaris, N. Panahian and M. Wellington for reading and commenting on the manuscript. We also thank J. B. Hicks, M. J. McEachern and W. S. Riggsby for plasmids. This investigation was supported by US Public Health Science Research Grants Nos AI29433 and GM12702 from the National Institutes of Health.
References Beverley S. 1988. Characterization of the `unusual' mobility of large circular DNAs in pulsed ®eld-gradient electrophoresis. Nucleic Acids Res 16: 925±939. Blaisonneau J, Nosek J, Fukuhara N. 1999. Linear DNA plasmid pPK2 of Pichia kluyveri: distinction between cytoplasmic and mitochondrial linear plasmids in yeasts. Yeast 15: 781±791. Cannon RD, Jenkinson HF, Shepherd MG. 1990. Isolation and nucleotide sequence of an autonomously replicating sequence (ARS) element functional in Candida albicans and Saccharomyces cerevisiae. Mol Gen Genet 221: 210±218. Carle GF, Olson MV. 1984. Separation of chromosomal DNA molecules from yeast by orthogonal-®eld-alternation gel electrophoresis. Nucleic Acids Res 12: 5647±5664. Chu W-S, Magee BB, Magee PT. 1993. Construction of an S®I macrorestriction map of the Candida albicans genome. J Bacteriol 175: 6637±6651. Cole RA, Williams KL. 1992. Tandem repeats in extrachromosomal ribosomal DNA of Dictyostelium discoideum, resulting from chromosomal mutations. Genetics 130: 757±769. De Jonckheere JF. 1989. Variation of electrophoretic karyotypes among Naegleria spp. Parasitol Res 76: 55±62. Dhar SK, Choudhury NR, Mittal V, Bhattacharya A, Bhattacharya S. 1996. Replication initiates at multiple dispersed sites in the ribosomal DNA plasmid of the protozoan parasite Entamoeba histolytica. Mol Cell Biol 16: 2314±2324. Feagin JE, Werner E, Gardner MJ, Williamson DH, Wilson RJM. 1992. Homologies between the contiguous and fragmented rRNAs of the two Plasmodium falciparum extrachromosomal DNAs are limited to core sequences. Nucleic Acids Res 4: 879±887. Ferris PJ. 1985. Nucleotide sequence of the central nontranscribed spacer region of Physarum polycephalum rDNA. Gene 39: 203±211. Fukuda K, Maebuchi M, Takata H, Gunge N. 1997. The linear plasmid pDHL1 from Debaryomyces hansenii encodes a protein highly homologous to the pGKL1 plasmid DNA polymerase. Yeast 13: 613±620. Fukuhara H. 1995. Linear DNA plasmids of yeasts. FEMS Microbiol Letters 131: 1±9. Yeast 2001; 18: 261±272.
272
Grif®ths AJ. 1995. Natural plasmids of ®lamentous fungi. Microbiol Rev 59: 673±685. Hermanns J, Osiewacz HD. 1992. The linear mitochondrial plasmid pAL2±1 of a long-lived Podospora anserina mutant is an invertron encoding a DNA and RNA polymerase. Curr Genet 22: 491±500. Hermanns J, Asseburg A, Osiewacz HD. 1994. Evidence for a life span-prolonging effect of a linear plasmid in a longevity mutant of Podospora anserina. Mol Gen Genet 243: 297±307. Hermanns J, Asseburg A, Osiewacz HD. 1995. Evidence for giant linear plasmids in the ascomycete Podospora anserina. Curr Genet 27: 379±386. Iwaguchi S-I, Homma M, Tanaka K. 1992. Clonal variation of chromosome size derived from the rDNA cluster region of Candida albicans. J Gen Mocrobiol 138: 1177±1184. Janbon G, Sherman F, Rustchenko E. 1998. Monosomy of a speci®c chromosome determines sorbose utilization in Candida albicans. Proc Natl Acad Sci U S A 95: 5150±5155. Johansen S, Johansen T, Haugli F. 1992. Extrachromosomal ribosomal DNA of Didymium iridis: sequence analysis of the large subunit ribosomal RNA gene and sub-telomeric region. Curr Genet 22: 305±312. Kempken F, Hermanns J, Osiewacz HD. 1992. Evolution of linear plasmids. J Mol Evol 35: 502±513. Long DM, Smidansky ED, Archer AJ, Strobel GA. 1998. In vivo addition of telomeric repeats to foreign DNA generates extrachromosomal DNAs in the taxol-producing fungus Pestalotiopsis microspora. Fungal Gene Biol 24: 335±344. Magee BB, D'Souza TD, Magee PT. 1987. Strain and species identi®cation by restriction fragment length polymorphisms in the ribosomal DNA repeat of Candida species. J Bacteriol 169: 1639±1643. McEachern MJ, Hicks JB. 1993. Unusually large telomeric repeats in the yeast Candida albicans. Mol Cell Biol 13: 551±560. Meinhardt F, Kempken F, Kamper JK, Esser K. 1990. Linear plasmids among eukaryotes: fundamentals and application. Curr Genet 17: 89±95. Olson MV. 1991. Genome structure and organization in Saccharomyces cerevisiae. In The Molecular and Cellular Biology of the Yeast Saccharomyces, Vol. 1. Genome Dynamics, Protein Synthesis, and Energetics, Broach JR, Jones EW, Pringle JR (eds). Cold Spring Harbor Laboratory Press: New York. Perepnikhatka V, Fisher FJ, Niimi M, et al. 1999. Speci®c chromosomal alterations in ¯uconazole-resistant mutants of Candida albicans. J Bacteriol 181: 4041±4049. Perez-Martin J, Uria JA, Johnson AD. 1999. Phenotypic switching in Candida albicans is controlled by a SIR2 gene. EMBO J 18: 2580±2592.
Copyright # 2000 John Wiley & Sons, Ltd.
D. H. Huber and E. Rustchenko
Ravel-Chapuis P. 1988. Nuclear rDNA in Euglena gracilis: paucity of chromosomal units and replication of extrachromosomal units. Nucleic Acids Res 16: 4801±4810. Rustchenko EP, Curran TM, Sherman F. 1993. Variations in the number of ribosomal DNA units in morphological mutants and normal strains of Candida albicans and in normal strains of Saccharomyces cerevisiae. J Bacteriol 175: 7189±7199. Rustchenko-Bulgac EP, Howard DH. 1993. Multiple chromosomal and phenotypic changes in spontaneous mutants of Candida albicans. J Gen Microbiol 139: 1195±1207. Rustchenko-Bulgac EP. 1991. Variations of Candida albicans electrophoretic karyotypes. J Bacteriol 173: 6586±6596. Rustchenko-Bulgac EP, Sherman F, Hicks JB. 1990. Chromosomal rearrangements associated with morphological mutants provide a means for genetic variation of Candida albicans. J Bacteriol 172: 1276±1283. Sadhu C, McEachern MJ, Rustchenko-Bulgac EP, Schmid J, Soll DR, Hicks JB. 1991. Telomeric and dispersed repeat sequences in Candida yeasts and their use in strain identi®cation. J Bacteriol 173: 842±850. Sambrook J, Fritsch EF, Maniatis T. 1989. Molecular Cloning. A Laboratory Manual. Cold Spring Harbor Laboratory Press: New York. Sherman F, Fink GK, Hicks JB. 1982. Methods in Yeast Genetics. Cold Spring Harbor Laboratory Press: New York. Sinclair DA, Guarente L. 1997. Extrachromosomal rDNA circles ± a cause of aging in yeast. Cell 91: 1033±1042. Spear BB, Gall JG. 1973. Independent control of ribosomal gene replication in polytene chromosomes of Drosophila melanogaster. Proc Natl Acad Sci U S A 70: 1359±1363. Wagner W, So M. 1992. Identi®cation of a novel large extrachromosomal DNA (LED) in the Trypanosomatidae. Mol Microbiol 6: 2299±2308. Wickes B, Staudinger J, Magee BB, Kwon-Chung K-J, Magee PT, Scherer S. 1991. Physical and genetic mapping of Candida albicans: several genes previously assigned to chromosome 1 map to chromosome R, the rDNA-containing linkage group. Infect Immun 59: 2480±2484. Wills JW, Lasker BA, Sirotkin K, Riggsby WS. 1984. Repetitive DNA of Candida albicans: nuclear and mitochondrial components. J Bacteriol 157: 918±924. Yao M-C, Gal JG. 1977. A single integrated gene for ribosomal RNA in a eukaryote, Tetrahymena pyriformis. Cell 12: 121±132. Yu G-L, Blackburn EH. 1990. Ampli®cation of tandemly repeated origin control sequences confers a replication advantage on rDNA replicons in Tetrahymena thermophila. Mol Cell Biol 10: 2070±2080.
Yeast 2001; 18: 261±272.
