The nitA locus of V. carteri encodes nitrate reductase (13); nitA mutants require reduced nitrogen (nitrite, ammonium, or urea) for growth and are resistant to chlorate. Successful transformation of a nitA- mutant with the cloned nitA+ gene should generate Nit+ cells that can utilize nitrate but are killed by chlorate.
Preliminary Studies. Transient Nit+ revertants, but no stable Nit+ lines, were recovered when attempts were made to transform V. carteri nitA mutants by gently agitating a cell-wall-less strain in the presence of nitA DNA and glass beads (9) or by using a UV microbeam (8) to introduce nitA DNA into plasmolyzed gonidia. The first stable Nit+ transformant of V. carteri was recovered when embryos, and hence the cytoplasmic bridges that link embryonic cells (17), were disrupted in the presence of ANR106 DNA. DNA analysis indicated that this transformant possessed integrated A DNA (data not shown); this indicated that the ANR106 insert carried a functional nitA+ gene and that Volvox cells were capable of integrating and expressing exogenous DNA. However, efforts to establish an efficient, reproducible transformation system using this method ofDNA introduction did not succeed.
Recovery of Nit+ Transformants After Bombardment with njL4+ DNA. Stable Nit+ transformants were repeatedly recovered when a device (18) that employs a rapid flow of He gas (rather than the explosive force used in other devices; ref. 7) was used to propel gold particles that had been coated with nitA DNA toward V. carteri cells carrying a stable nitAmutation. The first recipient used was the morphological double mutant, Gls/Reg (6), in which all cells reproduce and are therefore potential targets for transformation rescue. Subsequently, parallel experiments were successfully performed with gonidia derived from three morphologically wild-type siblings ofGls/Reg, strains 153-40, -45, and -48 that carry the same nitA allele as Gls/Reg.
In 17 experiments (10 with Gls/Reg and 7 with 153-45 or -48) in which variables were held within the ranges described in Materials and Methods, and in which the survival of aliquots of the bombarded cells on nonselective plates was monitored, the mean frequency of stable transformants among the survivors was 2.5 x 10-5 (SD, ± 2.4 x 10-5). No statistically significant differences were detected that could be attributed to any of the variables, including the genotype or developmental stage of the recipient cells. Under these conditions, as many as 40 stable transformants were recovered in a single series of shots that took
The first plasmid used for transformation was pVcNR4, which contains =6 kb of genomic DNA upstream of the nitA coding region (14); subsequently, pVcNR1, which contains only =1 kb of upstream DNA, was found to be as effective. Transformation efficiency was not improved when linearized plasmids were used. No transformants were recovered in experiments in which an aqueous solution of plasmid DNA was merely mixed with either the gold particles or the target cells prior to bombardment.
Stability of Transformation and Integrated Plamid Sequences. A small and variable fraction of the putative transformants that initially grew in NSVM (indicating that they possessed a functional nitA gene) stopped growing, bleached, and died after a few days or weeks, indicating that the nitA function had not been stably transmitted. Most transformants, however, have been stably maintained-in many cases for more than a year, or >1500 mitotic divisions-suggesting that plasmid DNA had been stably integrated.
Genomic DNA from recipient strains and many independent Nit+ transformants was analyzed on Southern blots for the presence of vector sequences. Sequences hybridizing to the vector were not detected in recipients but were in the transformants; moreover, different transformants often exhibited different numbers, sizes, and relative intensities of fragments hybridizing to the vector (Fig. 1A; also see Fig. 3A; other data not shown). When similar blots were hybridized with a probe detecting part of the nitA coding region, each transformant was found to possess a hybridizing fragment that comigrated with the one present in the recipient strain, plus one or more novel fragments (Fig. 1B). These data suggested that in each case one or more copies of the transforming plasmid had integrated into the genome via nonhomologous recombination at sites outside the nitA locus. Transformant cultures grown for >100 mitotic cycles under nonselective conditions retained most of the vector sequences that were present in cultures maintained under continuous selection (Fig. 2) and also retained their Nit+ phenotype when again tested. on selective medium (data not shown). This reinforced the inference that transformants contained plasmid sequences that had been stably integrated into the nuclear genome. Occasionally, however, strains that had integrated multiple copies of the plasmid were observed to have lost or gained one or two bands following prolonged cultivation (Fig. 2); we have not yet determined whether such events occur more frequently under selective or nonselective conditions.
Sites of Recombination Within the Transforming Plasmid. To substantiate the conclusion that stable transformants contained integrated plasmid DNA, and to determine where within each plasmid recombination had occurred during integration, a more detailed Southern blot analysis was performed. First, blots containing transformant DNAs digested by BamHI, EcoRI, or BamHI plus EcoRI were hybridized with a probe that detects the vector region of the transforming plasmid (Fig. 3A and C). Free pVcNR1 plasmid yields a 9.6-kb EcoRJ, a 4.8-kb BamHI, and a 3.2-kb EcoRIBamHI fragment that hybridize to this probe (Fig. 3A and C). In a plasmid that underwent recombination near the 5' end of the nitA insert (region I, Fig. 3C), the 9.6-kb EcoRI fragment should be replaced by a novel EcoRI fragment containing genomic DNAflanking the integration site; in copies in which recombination occurred within the 3' region of the insert (region III, Fig. 3C), a novel BamHI fragment should be present; and in copies that had recombined within the vector (region II, Fig. 3C), six novel fragments should be present in place ofthe three fragments characteristic offree plasmid. In the second part of this analysis, Southern blots of HindEIdigestedDNAfrom the same transformants were probed with a fragment representing the 3' end of the nitA gene (Fig. 3 B and C). This permitted us to define a fourth region (region IV, Fig. 3C) and thereby subdivide region I into two sections, since only plasmids that had recombined within region IV should retain the 8.3-kb HindIII fragment of pVcNR1 but lack the 9.6-kb EcoPJ fragment.
In each transformant analyzed in the above manner, hybridizing fragments different from those offree pVcNR1 were observed, indicating that every transformant contained at least one recombinant plasmid. But many hybridization patterns were more complex than could be accounted for by a single plasmid integrant. It was often possible to account for all of the bands in a transformant as a reflection of a small, integral number of plasmids that had integrated by some combination of the types of recombination events described above. For example, one may infer from the data in Fig. 3A that transformant Hill 143 arose via integration of two copies of the plasmid: one that recombined in region III (leaving its 9.6-kb EcoRI fragment and 3.2-kb EcoRI-BamHI fragments intact, but replacing its 4.8-kb BamHI fragment with a novel one) and one that recombined in region III (replacing all three of the pVcNR1 BamHI/EcoRI fragments with novel ones). The pattern obtained with a HindIII digest of Hill 143 (Fig. 3B) is consistent with this interpretation: in addition to the 6.5-kb fragment derived from the mutant nitA gene of the recipient strain (which is retained by all transformants yet studied), Hill 143 has two novel HindIII fragments, as expected for a strain possessing two copies of the plasmid that recombined outside region IV. Similarly, one can infer from the more complex pattern of fragments generated from Hill 142 DNA (Fig. 3A) that this strain resulted from four separate plasmid-integration events, two in region I, one in region II, and one in region III. Data from the HindIII digest (Fig. 3B) confirm and extend this interpretation: Hill 142 has four HindIII fragments (in addition to the 6.5-kb mutant-gene fragment), but the fact that one of these is of the same size as that present in pVcNR1 indicates that of the two plasmids that recombined within region I, one recombined within region IV, and the other recombined outside that region. In some transformants the numbers and intensities of BamHI, EcoRI, BamHI-EcoRI, and HindIII fragments indicated that as many as 10 copies of the plasmid had integrated, and in some cases fragments were present that could only be accounted for by assuming that certain plasmids had rearranged before or during integration (data not shown). Cot formatlon of an Unselected Marker. Ultimately, the utility of a nitA-based transformation system for Volvox investigators will depend on the ease with which cotransformation can be achieved with an unselected marker, such as one of the genes believed to regulate cellular differentiation in Volvox (2). In the related unicellular organism, C. reinhardtii, cotransformation frequencies of -50% were achieved when the selected and unselected markers were delivered on separate plasmids (20). To determine whether the same might be true of V. carteri, transformation was performed with a mixture of DNAs from pVcNR1 or pVcNR4 and another plasmid (pLV13-6, pLV131-3, or pLV1314) that contained a A DNA marker that does not cross-hybridize to Volvox DNA or to either of the nitA plasmids. All Nit+ transformants recovered were analyzed to determine whether they contained the A DNA marker. In the first experiment in which Gls/Reg cells were bombarded with a mixture of pVcNR1 and pLV1314, 16 of 20 Nit+ transformants that were examined were found to contain the A DNA marker and the nitA plasmid (representative data in Fig. 4). Other experiments involving different recipient strains (153-40, -45, and -48), a different transforming plasmid (pVcNR4), and/or different A-marked plasmids have all yielded cotransformants with frequencies of -40-80%.
Answers to all your Biology Questions
