Determination of Plasmid Copy Number Reveals the Total Plasmid DNA Amount Is Greater than the Chromosomal DNA Amount in Bacillus thuringiensis YBT-1520

Bacillus thuringiensis is the most widely used bacterial bio-insecticide, and most insecticidal crystal protein-coding genes are located on plasmids. Most strains of B. thuringiensis harbor numerous diverse plasmids, although the plasmid copy numbers (PCNs) of all native plasmids in this host and the corresponding total plasmid DNA amount remains unknown. In this study, we determined the PCNs of 11 plasmids (ranging from 2 kb to 416 kb) in a sequenced B. thuringiensis subsp. kurstaki strain YBT-1520 using real-time qPCR. PCNs were found to range from 1.38 to 172, and were negatively correlated to plasmid size. The amount of total plasmid DNA (∼8.7 Mbp) was 1.62-fold greater than the amount of chromosomal DNA (∼5.4 Mbp) at the mid-exponential growth stage (OD600 = 2.0) of the organism. Furthermore, we selected three plasmids with different sizes and replication mechanisms to determine the PCNs over the entire life cycle. We found that the PCNs dynamically shifted at different stages, reaching their maximum during the mid-exponential growth or stationary phases and remaining stable and close to their minimum after the prespore formation stage. The PCN of pBMB2062, which is the smallest plasmid (2062 bp) and has the highest PCN of those tested, varied in strain YBT-1520, HD-1, and HD-136 (172, 115, and 94, respectively). These findings provide insight into both the total plasmid DNA amount of B. thuringiensis and the strong ability of the species to harbor plasmids.


Introduction
Plasmids are extra-chromosomal DNA elements that can be characterized by their copy number in the host [1,2]. Plasmids can incorporate and deliver genes by recombination or transposition, thus facilitating genetic exchanges in bacterial populations [1]. In general, plasmids encode a wide variety of valuable functional genes, such as those associated with drug resistance, pathogenicity, and many other important processes. However, plasmids are usually considered nonessential for the survival of bacteria, although they metabolically burden their hosts [2][3][4]. To minimize this metabolic load and coexist stably with their hosts, plasmids must control their replication; accordingly, the plasmid copy number (PCN) of a given plasmid is typically fixed within a given host and under defined cell growth conditions [3,5].
PCN is defined as the number of copies of a plasmid present per chromosome in a cell [6,7]. The PCN is a key feature of a plasmid: a PCN can be described either as low (1-10 copies), medium (11-20 copies), or high (.20 copies) [8]. The PCN is mostly controlled by the origin of replication, although host physiology also plays a major role in plasmid replication [5]. However, to date, the full impact of PCN on a host strain is not fully understood, but it is clear that PCN must be considered as a key feature of bacteria and requires further study [4]. The previous studies of PCN have been largely focused on modified small plasmids. There are a few studies on PCN of multiple native plasmids in the same host, but with cruder tools [9].
Bacillus thuringiensis, which is the most widely used bioinsecticide, has long been noted for its insecticidal crystal protein (ICP). Most ICP-coding genes are located in the mega-plasmid of the organism [10][11][12]. Typically, B. thuringiensis harbors numerous (i.e., [2][3][4][5][6][7][8][9][10][11][12] native plasmids, which vary in size from 2 kb to 600 kb [11,13,14]. These plasmids may constitute a substantial amount of the total genetic content of the bacterium, representing 10-20% of the genetic material of the cell [12]. The total plasmid genetic content refers to the sum of the sizes of all plasmids present in an organism. However, plasmids usually have multiple copies in their host. Thus, the total genetic content does not accurately reflect the amount of total plasmid DNA. Accordingly, it seems more appropriate to describe the total plasmid DNA amount as the value of the total (or sum) of plasmid sizes multiplied by the PCN.
The resulting total plasmid DNA amount would more accurately reflect the quantitative relationship between plasmid and chromosome in an organism. However, the total plasmid DNA amount present in B. thuringiensis remains unknown.
Our group previously isolated B. thuringiensis subsp. kurstaki strain YBT-1520, which is highly toxic to lepidopteron pests, from China [15]. We have already finished sequencing the whole genome of the strain (unpublished), and found that this strain contains 11 endogenous plasmids with vary sizes and modes of replication and its chromosome consists of 5,401,320 bp. In this study, we determined the PCNs of the 11 plasmids present in strain YBT-1520 at the mid-exponential growing stage (OD 600 = 2.0) using absolute real-time qPCR. We found that the amount of total plasmid DNA was more than the amount of chromosomal DNA in the strain. We also determined that the PCN varied in different hosts and growth stages.

Results
PCNs of 11 plasmids in B. thuringiensis strain YBT-1520 at the mid-exponential growing stage We performed real-time qPCR amplification of the total DNA samples from B. thuringiensis strain YBT-1520 harboring 11 plasmids and the plasmid DNA standard of target plasmid simultaneously. Absolute quantification was performed using the standard curves that had been constructed. All PCRs displayed efficiency between 95% and 103% and every reaction was optimized to keep the difference of the amplification efficiencies between the single special gene of the plasmid and chromosomal aiiA less than 2%. We also determined the PCN of pBMB67 by selecting three different genes on the plasmid as target fragments. Of note, we reported the sequence of the mega-plasmid pBMB67 in a previous study [16]. We designed three pairs of primers on rep (22308-22447), orf4 (41233-40373), and orf19 (62239-62365) of pBMB67 to determine the PCNs. The results from this experiment showed that the PCNs of pBMB67 targeted at three different fragments were 21.3, 20.1, and 21.0, respectively, with a relative standard deviation of 3.0%. These findings demonstrated the reliability of our experimental results.
As shown in Table 1, we calculated the PCNs of the 11 plasmids in strain YBT-1520 at the mid-exponential growing stage (OD 600 = 2.0) by the absolute quantification method. We found that the PCNs of all 11 native plasmids of strain YBT-1520 ranged from 1.38 to 172 (Table 1). Of note, the largest plasmid pBMB400 had the lowest PCN (1.38) and the smallest plasmid pBMB2062 had the highest PCN (172). The PCN reduced as the molecular weight increased, and thus the PCN negatively correlated to the molecular mass of plasmids ( Figure 1). However, the PCNs of some plasmids deviated from this trend. For example, the PCNs of pBMB7635 and pBMB8240 (136, and 139, respectively) were apparently higher than those of pBMB7921 and pBMB8513 (64.7, and 74.7, respectively), although all of these plasmids were similar in size (Table 1). Figure 1 summarizes the relationship between the PCNs of 9 of the 11 plasmids in strain YBT-1520 (except pBMB7635 and pBMB8240) and the plasmid size. We determined that regression analysis of this data presented an R 2 value of 0.9432.
In addition, we calculated the plasmid DNA amount by multiplying the plasmid size by the PCN for each of the 11 plasmids from strain YBT-1520. We found that the DNA amounts of the plasmids ranged from 355 kb to 1443 kb (Table 1). Among them, the DNA amounts of pBMB8513, pBMB95, and pBMB400, whose sizes are very different, were quite close to each other (636 kb, 612 kb, and 577 kb, respectively). However, the DNA amounts of pBMB7921 (512 kb) and pBMB8240 (1148 kb), which are similar in size, were quite different.
PCNs of pBMB2062, pBMB67, and pBMB293 dynamically shifts at different stages during the whole life cycle of strain YBT-1520 Previous studies reported that the PCNs of some modified small plasmids are different at different stages during the life cycle of the host cell [17,18]. It has been shown that measurement of the PCN must be performed over the bacteria life cycle without antibiotics in the culture media because the germination, growth, and sporulation of the bacteria are highly affected by the presence of the antibiotics [18,19]. However, the modified plasmid loss in the absence of antibiotics as a selection pressure could also have a strong effect on PCN measurement. In this study, we selected three native plasmids (pBMB2062, pBMB67, and pBMB293) with different sizes (2 kb, 67 kb, and 293 kb, respectively) and different replication mechanisms [16,20] to assay the PCNs of these plasmids over the entire life cycle (from spore germination to the spore stage) of strain YBT-1520. We found that the PCNs of all three plasmids dynamically shifted at different growth phases. Specifically, the PCNs ranged from 71 to 190 for pBMB2062, from 8 to 23 for pBMB67, and from 1 to 3 for pBMB293 ( Figure 2). The PCNs of all three plasmids were lowest during the germination phase (1-2 h), and then increased rapidly across the logarithmic phase. The PCN of pBMB2062 reached its maximum at the mid-exponential growth stage, while the values of pBMB67 and pBMB293 peaked at the early stationary stage. However, the PCNs of all plasmids decreased rapidly during the late stationary phase and remained stable and close to the minimum after presporulation (18 h) ( Figure 2).
Total plasmid DNA amount is greater than chromosomal DNA amount in strain YBT-1520 To examine the relationship between the total plasmid DNA amount and the chromosomal DNA amount, we calculated the total plasmid DNA amount (Table 1). These data showed that the total DNA amount of the 11 plasmids was equivalent to 8746 kb at the mid-exponential growing stage (OD 600 = 2.0) ( Table 1). To our surprise, the calculated total plasmid DNA amount was 1.62-fold greater than the chromosomal DNA amount (5401 kb).
The data described above showed that the PCNs of all three tested plasmids (pBMB2062, pBMB67, and pBMB293) were close to their respective minimums in the spore stage and remained steady ( Figure 2). To further analyze the relationship of total plasmid DNA and chromosomal DNA amounts in the entire life cycle of strain YBT-1520, we calculated the total plasmid DNA amount in the spore stage. The PCNs of the three tested plasmids (pBMB2062, pBMB67, and pBMB293) at this stage were determined to be 100, 14.1, and 1.95, respectively (Table 2), which are close to the corresponding integral numbers (100, 14, and 2, respectively). Accordingly, we deduced that the PCN of pBMB400 was 1 at the spore stage because the PCN at the spore stage was an integral number and was lower than that of the midexponential growth stage (1.38). We subsequently compared the PCNs of the four plasmids (pBMB2062, pBMB67, pBMB293, and pBMB400) at the mid-exponential growth stage (OD 600 = 2.0) to those at the spore stage. These data showed that the ratios (spore stage: mid-exponential stage) of the PCNs of the four plasmids at the different stages were 1.72, 1.51, 1.39, and 1.38, respectively ( Table 2). This finding indicates that the ratio decreased slightly with the increase of plasmid size.
According to the ratios of the PCNs of the four plasmids at the mid-exponential growth (OD 600 = 2.0) and spore stages (Table 2) and the logged sizes, we fit a linear regression equation (Figure 3). This equation reflected the relationship between plasmid size and the PCN ratio of the two different stages. Therefore, we were able to derive the PCN ratios of the other seven plasmids at the two different stages and their DNA amounts in the spore stage ( Table 2). The total plasmid DNA amount of strain YBT-1520 in the spore stage was calculated to be 5675 kb, which was slightly more than the chromosomal DNA amount (5401 kb).

PCNs of a single plasmid vary in different B. thuringiensis hosts
A previous study showed that the PCN of modified plasmids varied from host to host [18]. Furthermore, it was determined that such modified plasmids grew with selection pressure and that antibiotics affected the PCN [18,19]. Therefore, we measured the PCN of the native plasmid pBMB2062 in three different hosts to further investigate this matter. Of note, pBMB2062 is a widespread cryptic small plasmid. The DNA sequence of this plasmid in strain YBT-1520, HD-1, and HD-136 showed 100% identity. The PCN of pBMB2062 was determined to be the highest (172) among all 11 plasmids in the strain YBT-1520. Therefore, we selected pBMB2062 to monitor the PCN of the same plasmid in different hosts. The PCNs of pBMB2062 in strain YBT-1520, HD-1, and HD-136 were 172, 115, and 94 at the midexponential growth stage (OD 600 = 2.0), respectively. The PCN of pBMB2062 in strain YBT-1520 was almost 1.83-fold of that in strain HD-136. These results indicate that the PCN of a plasmid varies in different hosts, at least for the plasmid included in this study.

Discussion
Among the 11 plasmids we examined, the PCNs of pBMB7635 and pBMB8240 were very high (Table 1). We previously reported that only the two plasmids of all the 11 plasmids in strain YBT-1520 harbored toxin-antitoxin (TA) systems [21]. TA systems are known to maintain the stability of plasmid in bacteria as addiction systems [22,23]. Up to now, there is no report about the relationship between the PCN and TA systems. The PCNs of pBMB7635 and pBMB8240 showed high-copy and deviated from the trend that the PCN negatively correlated to the molecular mass of plasmids in this study (Table 1). Whether the TA system will affect the PCNs needs to be validated in the further work.
A previous study indicated that the PCNs of foreign recombinant plasmids were at their minimum during germination and at their maximum during the stationary growth phase during the growth life cycle of Bacillus cereus [18]. In the present study, we measured the PCNs of three native plasmids (pBMB2062, pBMB67, and pBMB293) over the entire life cycle of the B. thuringiensis strain YBT-1520 ( Figure 2). We found that the PCNs were at their minimum during the germination stage for all three native plasmids, and were at their maximum during the stationary growth phase for both pBMB67 and pBMB293. These results are consistent with the previous study [18]. However, the PCN of  Figure 3. Relationship between the ratio of PCN and plasmid size. The ratio denotes the PCN of pBMB2062, pBMB67, pBMB293, and pBMB400 at the mid-exponential growth stage to that at the spore stage. The ratios of PCNs are also shown in Table 2 pBMB2062 reached its maximum at the middle of the logarithmic phase. The PCN of different plasmids reach their maximum at different phase may be due to different plasmid replication regulations or unknown PCN control mechanism on individual plasmid. In addition, there is a dip during germination (1-2 h, actually cell outgrowth within 0.5 h) which is found in previous study also [18]. During this period, the outgrowth cells were adjusting to the changes in the environment; there would be minimal cell growth [24]. The possible reasons for the dip existence might be that the chromosomal DNA replication is earlier than plasmid DNA replication. The pBMB293 has a pathogenicity island, which contains four ICP genes, cry1Aa, cry1Ia, cry2Aa, and cry2Ab. It is interesting that the PCN of pBMB293 reached its maximum at the early stationary growth phase (10 h) and was subsequently reduced to its approximate minimum and stably maintained at 18 h ( Figure 2). However, the crystal protein initiate expression at 11 h and reached its maximum at 18 h [25,26]. Most ICP coding genes are located on low-copy large plasmids [10]. High expression of an ICP should be controlled by its strong promoters, stable mRNA, coexpression of accessory protein, and protein crystallization [26,27]. In this study, ICPs reach their highest expression level when the PCN is reduced to a level similar to the minimum, indicating that ICP expression may be increased by gene dosage (PCN) but does not have to be proportional. This case was similar as previously reports [26].
In general, bacterial plasmids are considered as ''selfish'' elements [28]. In other words, their presence is normally a burden for host cells, and thus the host would benefit from elimination of the plasmid intruder [3,4,28]. To coexist stably with their hosts, plasmids must control their replication to minimize the metabolic load they exert of their hosts [3,29]. Plasmids control their PCNs by negative regulatory systems that adjust the rate of replication per plasmid copy in response to fluctuations in the PCN [3]. These regulatory mechanisms have previously been studied in depth [30][31][32]. B. thuringiensis strain YBT-1520 contains 11 plasmids; the total plasmid genetic content (988 kb) was determined to be equal to 18.3% of the chromosomal DNA amount (5401 kb), and the total DNA amount of the plasmids (8746 kb) was 1.62-fold more than the total chromosomal DNA amount (5401 kb). It is established that B. thuringiensis harbors numerous native plasmids, varying in number from 2 to 12, with sizes from 2 kb to 600 kb [11,13,14], and we have shown that the plasmid DNA amount may be more than the total chromosomal DNA amount in this species. These indicate that B. thuringiensis strains have a strong ability to harbor plasmids, and that many foreign plasmids could be transferred into B. thuringiensis for recombination.
One previous report showed that Borrelia burgdorferi B31 contains a linear chromosome of 910,725 bp and at least 17 linear and circular plasmids with a combined size of more than 533,000 bp [33]. The total plasmid genetic content of B. burgdorferi B31 (533,000 bp) was found to be 58.5% of the chromosomal DNA amount (910,725 bp). It is reasonable to suggest that the total plasmid DNA amount of B. burgdorferi B31 is more than that of its chromosome, also. However, an enormous plasmid DNA amount would inevitably cause great metabolic burden on a host. Thus, one important issue that remains to be resolved is why the total DNA amount of plasmids, which represents an accessory genetic resource, can exceed the chromosomal DNA amount. We have attempted to cure the plasmids in strain YBT-1520 by heat elimination, although we have found that certain plasmids could not be removed (unpublished). In addition, B. thuringiensis, especially those strains that are highly toxic to pests, such as YBT-1520, HD-1, H14, etc, contains numerous plasmids [34][35][36].
These information suggest that some plasmids are valuable to permit their hosts to survive better in an adverse environment or to compete better with other microorganisms occupying the same ecological niche.
Bioinformatics analysis of the 11 plasmids in strain YBT-1520 showed that there were many genes present that were likely beneficial to the growth and survival of the organism. For example, pBMB67 may play a role in cell-cell signaling and transcriptional regulation through Rap-Phr cassettes [16]. We speculate that plasmids are more than a burden on their host, and that there may be coevolution and/or a mutualistic relationship present between plasmids and their hosts. This is supported by the ideas that the host spends a lot of energy maintaining the existence of many plasmids and that most plasmids that we examined could not be removed from bacteria.
In conclusion, we demonstrated that the total plasmid DNA amount was greater than the chromosomal DNA amount in the B. thuringiensis strain YBT-1520. In addition, we found that the PCN varied at different growth stages and in different hosts. These findings provide us with a new understanding both of the total plasmid DNA amount present and of the strong ability of B. thuringiensis to harbor plasmids. Furthermore, our data provide insight into the biological characteristics of plasmids in terms of their variation in both different hosts and different growth stages.

Bacterial strains, plasmids, and cultural conditions
The bacterial strains and plasmids used in this study are listed in Table 3. The B. thuringiensis strain YBT-1520 was isolated in China and the whole-genome was sequenced by a shotgun approach previously (with 2 gaps in chromosome, unpublished data) [15,37]. All B. thuringiensis strains were inoculated from freshly streaked Luria Bertani (LB) agar plates (1% tryptone, 0.5% yeast extract, 0.5% NaCl, and 1.5% agar) cultured overnight into 10 mL of LB media. After incubation at 28uC for 8 h, we transferred 0.5 mL of the culture into 25 mL of LB medium and allowed the bacteria to grow in an incubator shaker (220 rpm) at 28uC until its absorbance (OD 600 ) was 2.0. We also used E. coli DH5a as the host strain for plasmid amplification. Recombinant E. coli was cultured in LB medium (containing 100 mg/mL ampicillin) at 37uC.
The growth curve during B. thuringiensis life cycle B. thuringiensis strain YBT-1520 spores were inoculated into a 500-mL Erlenmeyer flask containing 40 mL of LB medium to a final concentration of 10 6 spores per mL of LB medium, in order to obtain a growth curve of the organism. The spores were prepared using the previously reported method of Turgeon et al. [18]. Before spores were used in experiments, they were incubated at 70uC for 10 min to eliminate remaining vegetative cells. During the strain YBT-1520 life cycle, we collected samples at different times and measured their absorbance at OD 600 using a Spectrum Lab 23A (LengGuang Tech, Shanghai, China). We also observed sporulation synchronously using an optical microscope. The growth curve includes a plot of absorbance (OD 600 ) versus time in culture.

B. thuringiensis total DNA isolation and purification
Total DNA of B. thuringiensis vegetative cells was extracted according to the method of Kalman et al. [38]. Extracting spores DNA is more difficult than from vegetative, to ensure spores are disrupted, the pellet of spores was resuspended with solution I (25% sucrose, 25 mM Tris-HCl (pH 8.0), 25 mM EDTA), and then ground in liquid nitrogen in a prechilled mortar for 15 min with the constant addition of liquid nitrogen as described by Yoshikawa et al. [39] and Rowland et al. [40]. Then, we extracted the total DNA from disrupted spores according to the method of extracting vegetative cells DNA. Three separate samples were prepared to reduce the bias of the difference of extraction efficiency between the plasmids and chromosomal DNA. The concentration and quality of extracted total DNA was measured using an UV spectrophotometer DU800 (Beckman Coulter, Fullerton, CA, USA).

Preparation of plasmid DNA standards for absolute realtime qPCR
We analyzed the strain YBT-1520 genome (unpublished data) and selected a single-copy specific gene as the target gene on the chromosome and 11 plasmids. The chromosomal aiiA gene encodes N-acyl homoserine lactone-degrading enzyme, which confers resistance in plants to soft rot disease and other diseases [41,42]. It was previously determined that chromosomal aiiA is a single-copy gene harbored on the chromosome of strain YBT-1520 and many other genomes of sequenced strains of the Bacillus cereus species (NC 014171, NC005957). We used chromosomal aiiA as a reference gene to detect strain YBT-1520 chromosomal DNA. Most of the plasmid primer pairs were targeted at Rep genes and the other plasmids that lacked a specific single-copy Rep gene were targeted at other single-copy orfs. The primers used in this study are listed in Table 4. All these PCR products were shown to be a single-copy in the strain YBT-1520 genome by a BLAST search against the strain YBT-1520 genome local database. The amplified product of the chromosomal aiiA gene was cloned into a pGEM-T Easy vector (Promega, Madison, WI, USA) to produce pGEM-T-aiiA. Subsequently, the 11 single-copy fragment of the target plasmid was cloned into pGEM-T-aiiA (utilizing restriction endonuclease site for PstI on pGEM-T Easy) to produce 11 plasmid DNA standards, respectively (Table 3). Accordingly, the plasmid DNA standard can be detected using either chromosomal aiiA primer pairs or target plasmid primer pairs.  A 10-fold serial dilution series of the plasmid DNA standard, ranging from 1610 4 to 1610 8 copies/mL, was used to construct the standard curves for both chromosomal aiiA and the single-copy fragment of the target plasmid. The concentration of the plasmid DNA standard was measured using a fluorometer and the corresponding PCN was calculated using the following equation [43].
PCN determination by quantitative real-time PCR As described previously [6,44], we calculated the PCN via dividing the target plasmid absolute quantity by the chromosomal absolute quantity in the template total DNA. The IQ5 instrument (Bio-Rad, Hercules, CA, USA) was used for qPCR amplification and detection. Real-time qPCR reactions were performed in triplicate using 25 mL mixtures. The mixture for one reaction contained 2.5 mL of 106 SYBRH Green PCR Master Mix (Toyobo, Kyoto, Japan), 5 mL of template DNA, and 12.5 pmol of each primer. The PCR reaction was conducted for all amplicons with the following cycling conditions: 2 min at 95uC, followed by 40 cycles of 95uC for 15 s, 60uC for 20 s, and 72uC for 30 s. Upon completion of 40 cycles of PCR amplification, a dissociation step of ramping the temperature from 55uC to 95uC steady for 20 min was performed, while the fluorescence signal was continually monitored for melting curve analysis. We determined the cycle threshold (Ct) values after automatic adjustment of the baseline and manual adjustment of the threshold using IQ5 software.
The plasmid DNA standard curve was established according to the method of Lee et al. [45]. Briefly, the standard curve includes a plot of the Ct values versus the log concentration of the plasmid DNA standard. For any unknown total DNA sample, the absolute quantity of both its plasmid and chromosomal DNA were obtained by interpolating its Ct value against the standard curve. Subsequently, the PCN was calculated by dividing the copy number of the special gene of a plasmid by the copy number of chromosomal aiiA.