Mitochondrial Transformation in Baker

The yeast Saccharomyces cerevisiae is a widely recognized model used to study mitochondrial biogenesis. Since yeast is an anaerobic, facultative organism, it is possible to extensively study the causes and consequences of introducing mutations that impair respiration. In addition, this organism possesses friendly genetic and biochemical tools to study mitochondrial pathways. However, one of the most powerful resources to explore the mechanisms of respiratory complex assembly and mitochondrial protein synthesis is the ability to transform mitochondria and modify the organelle's genome. Previously, it has been helpful to introduce in the mitochondrial DNA (mtDNA) point mutations or small deletions/insertions^1,2,3,4,5, delete genes^6,7, make gene rearrangements^7,8, add epitopes to mitochondrial proteins^9,10, relocate genes from the nucleus to the mitochondria^11,12, and introduce reporter genes like BarStar¹³, GFP^14,15, luciferase¹⁶, and the most widely used ARG8^m17,18. Mitochondrial genome modifications have allowed us to dissect and identify mechanisms that otherwise would have been difficult to comprehend. For example, the ARG8^m reporter gene inserted at the COX1 locus in the mitochondrial DNA was crucial to understanding that Mss51 has a dual role in Cox1 biogenesis. First, it is a translational activator of the COX1 mRNA, and second, it is an assembly chaperone for the newly made Cox1 protein^7,19. This work presents a detailed method to transform S. cerevisiae mitochondria. Although the mitochondrial transformation protocol was published earlier^{16,20,21,22,23}, a visual approach through a video is essential to thoroughly understand the different stages and details of the method. The method consists of various steps and is divided into four general stages:

Figure 1: Overview of the mitochondrial transformation procedure by high-velocity microprojectile bombardment: 1) The tungsten particles are sterilized and prepared before DNA coating. 2) Two different plasmids are precipitated on the surface of the WPs. One is a bacterial plasmid containing the construct that will be directed to the mitochondrial matrix. The other is a nuclear yeast expression vector carrying an auxotrophy marker. 3) A receptor yeast strain lacking mitochondrial DNA (rho⁰) is grown on a fermentative carbon source like galactose or raffinose, which does not exert any glucose repression of mitochondrial gene expression^34,35. The culture is spread on petri dishes containing the bombardment medium. 4) WPs coated with plasmids are shot to the receptor strain by high-velocity microprojectile bombardment. 5) Positive synthetic rho^- colonies containing the mitochondrial plasmid are selected by mating to a tester strain. 6) The mitochondrial construct is integrated into the mitochondrial genome at the desired locus by mating the synthetic rho^- strain (donor) with the acceptor strain by a process known as cytoduction. 7) The positive receptor, haploid strain carrying the mitochondrial construct is selected and purified in different media. Abbreviations: WPs = tungsten particles; mtDNA = mitochondrial DNA. Please click here to view a larger version of this figure.

Transformation of cells with the DNA construct intended to integrate into the mitochondrial genome (Figure 1, steps 1-4)
Although it is possible to directly transform a yeast containing a complete mitochondrial genome (rho⁺), the transformation is 10-20x more efficient if the cells lack mitochondrial DNA (rho⁰)²⁴. Tungsten particles (WPs) are coated with two different plasmids that will be co-transformed. The first one is a yeast 2 µ expression vector carrying an auxotrophy marker, such as LEU2 or URA3 (e.g., YEp351 or YEp352, respectively). If the biolistic introduction of DNA into the yeast cells is successful, the transformants will grow on the auxotrophy medium (i.e., a medium lacking leucine or uracil). This plasmid is helpful in making the first selection of the cells that acquired the nuclear plasmid; otherwise, the number of resulting colonies would saturate the plate. The second plasmid is a bacterial plasmid (such as pBluescript or similar) containing the mitochondrial construction intended to be integrated into the mitochondrial genome. The construct must contain at least 100 nt of 5' and 3' flanking mitochondrial sequences for recombination with the mitochondrial region of interest. In our experience, larger flanking sequences are more likely to successfully recombine with the target locus in the mitochondrial genome.

A specific example is shown in Figure 2²⁵. In this example, the aim was to delete the region of the mitochondrial COX1 gene coding for the last 15 amino acids of the corresponding protein (Cox1ΔC15) (Figure 2A). The bacterial plasmid carrying the COX1ΔC15 mutation contained 395 nt and 990 nt of the gene's 5' and 3' untranslated regions, respectively. The plasmid was derived from pBluescript (pXPM61) and, together with the 2 µ plasmid YEp352, they were coprecipitated on the surface of WPs (Figure 2B). The WPs were then introduced by microprojectile bombardment into the selected recipient strain (Figure 2C). This strain, named NAB69²⁶, is a MATa, rho⁰ strain bearing the kar1-1 and ade2 alleles (the importance of these characteristics is discussed below). The cells were plated on bombardment media lacking uracil to select those that acquired the 2 µ plasmid (Figure 2C). Cells were grown for 5-7 nights at 30 °C.

Selection of the cells that acquired the bacterial plasmid containing the mitochondrial construct and the nuclear 2 µ plasmid carrying the auxotrophy marker (Figure 1, step 5)
The positive colonies will maintain many copies of the bacterial plasmid in their mitochondria²². Since the transformed cells are originally rho⁰, no mitochondrial sequences support the translation of the mitochondrial construct present in the plasmid; hence, the transformed mitochondrial gene will not be expressed. It is necessary to mate the transformants with a tester strain to detect the yeast colonies that acquired the mitochondrial plasmid. The mitochondrial genome of the tester strain contains a nonfunctional mutated version of the gene of interest. After mating, the mitochondria will fuse, and the mitochondrial sequence included in the transformed plasmid will recombine with the mutated mitochondrial gene from the tester strain; consequently, the recovery of the WT gene will reconstitute function. The resulting diploid will have a detectable positive phenotype (usually the capacity to grow in respiratory medium or a medium lacking arginine). The positive cells carrying the bacterial plasmid in mitochondria are named "synthetic rho^- cells". In the specific example of Figure 2D, synthetic rho^- cells carrying the COX1ΔC15 construct (named XPM199) were replica-plated on a medium lacking uracil (this is the master plate from which positive colonies were purified). They were also replica-plated on a tester strain L45²⁷ lawn, which contains the nonfunctional mutation cox1D369N. After mating for two nights, the mitochondrial genome of L45 and XPM199 recombined, resulting in a functional COX1 gene; therefore, the diploids recovered the ability to grow on respiratory media. From the master plate, we picked the positive colonies. They were streaked on plates lacking uracil, and the selective tests were repeated to obtain pure synthetic rho^- cells (Figure 2E). It is important to note that the testing plate is only for the identification of synthetic rho^- colonies, and mutants cannot be recovered from these plates.

Integration of the construct into the mitochondrial genome of the intended strain
This step is achieved by a process called cytoduction^28,29 (Figure 1, step 6). In this approach, the synthetic rho^- strain (donor strain) is mated to the intended acceptor strain of the opposite mating type. An essential requirement is that at least one of the two mating strains carries the kar1-1 mutation to impede nuclear fusion³⁰. Hence, mating of the two cells produces a binucleated zygote (Figure 3). The mitochondrial network from the parental cells (donor and acceptor) fuse and the mitochondrial DNA molecules recombine. The binucleated zygotes are incubated/recovered to allow budding, which will produce haploid cells. These haploids carry the nuclear background of one of the parental cells. Similarly, haploids can carry the mitochondrial DNA from either the parental cell or the recombined mitochondrial DNA of interest. However, the kar1-1 mutation is not 100% effective, and some true diploids will be formed during the mating^28,29. The synthetic rho^- strain (donor, XPM199) carried the kar1-1 allele in the example. As a selective marker, it had the ade2 allele, making it an auxotroph for adenine (Figure 4). The acceptor strain was XPM10, a rho⁺ strain bearing the cox1Δ::ARG8^m construct, where the reporter ARG8^m replaced the COX1 codons in the mitochondrial genome⁷. The mixture of both cell cultures was added to a YPD plate as a drop to allow mating. After 3-5 h, the cells were observed under a light microscope to detect the formation of shmoos (Figure 3B), a cell shape associated with mating. After the incubation/recovery time, the binuclear zygotes were plated on a medium lacking adenine to prevent the growth of the donor cells.

Selection of the haploid strain carrying the mitochondrial genome of interest (Figure 1, step 7)
A mixture of donor, acceptor, and diploid cells is present after cytoduction. Therefore, after the incubation/recovery of the binucleated zygotes, cells must be grown in different selective media to identify and purify the haploids of interest. The selective media depends on the genotypes of the donor, acceptor strains, and the intended mitochondrial construction. However, in general, the reasoning for choosing the selective media is: i) after incubation/recovery, the cytoduction mix is grown on selective medium where the donor parental strain cannot grow. In the specific example in Figure 4A,B, the donor (synthetic rho^- strain) carries the nonfunctional ade2 allele. Thus, the cytoduction mix must be incubated on a medium plate lacking adenine. This is the master plate. ii) Once the colonies grow, the master plate is replicated again on a medium lacking adenine (to generate a new master plate from where the positive colonies of interest will be purified), and next, on a medium where only diploids can grow. This is necessary to avoid further inadvertent purification of diploid colonies. In the case of the example from Figure 4C, only diploids can grow on media lacking leucine. iii) The master plate is also replicated on media where only those acceptor haploids that incorporated the mitochondrial DNA of interest will grow. In the example of Figure 4C, haploids grow on respiratory media containing ethanol/glycerol as a carbon source since the Cox1ΔC15 protein is functional. The resulting rho⁺ strain bearing the mtDNA of interest was named XPM209²⁵. The strains used in Figure 2 and Figure 4 are listed in Table 1.

Figure 2: Diagram describing a specific example of mitochondrial transformation and selection of positive rho^- cells. (A) The intended modification of the mitochondrial genome was a deletion of the region coding for the last 15 amino acids of the Cox1 subunit (Cox1ΔC15). (B) WPs were coated with two plasmids to transform yeast cells. One was the 2 µ plasmid Yep352 that was directed and expressed in the nucleus. The other was plasmid pXPM61, which contains the COX1ΔC15 allele plus 395 nt and 990 nt of the COX1 5' and 3'-UTRs, respectively. (C) The WPs were introduced into cells from the strain NAB69, which lacks mitochondrial DNA (rho⁰ strain). Transformant colonies obtained from the bombardment plates were replicated on a medium lacking uracil, the auxotrophic marker of the nuclear plasmid YEp352. (D) To select the positive rho^- colonies, the -URA plate was replicated on a media lacking uracil. This was the master plate from which the positive colonies were picked and isolated. It was also replicated on plates with rich media (YPD) and a lawn of a tester strain. During mating, the synthetic rho^- mitochondrial DNA was recombined with the mitochondrial DNA from the tester strain, L45²⁷, which carries a mutation in the COX1 gene (D369N). The diploids that grew on respiratory media contained the transformed DNA. The corresponding colonies were picked from the master plate to restreak and purify. To help identify the positive colonies in the master plate throughout all replica plating, marks were made with a permanent marker on the edges of the plates (green lines). (E) The selected positive colonies were restreaked on a medium lacking uracil. This was the new master plate. As in D, two more rounds of testing were carried out to purify the synthetic rho^- colonies. Abbreviations: WPs = tungsten particles; mtDNA = mitochondrial DNA; -URA/Sorb = lacking uracil/ containing Sorbitol; WT = wild type. Please click here to view a larger version of this figure.

Figure 3: Diagram describing the cytoduction procedure. (A) General overview of how cells mate during cytoduction. During cytoduction, mitochondria from the donor and acceptor strains fuse so mitochondrial DNA from both strains recombine. Since nuclear fusion is reduced due to the kar1-1 mutation³⁰, binuclear zygotes are formed. After some incubation/recovery time, the binuclear zygotes bud and haploid acceptors containing the intended mitochondrial mutation are selected. (B) Mating of the synthetic rho^- strain (donor) and the acceptor strain through cytoduction allows the formation of shmoos (red arrows), a characteristic shape of mating yeast. The image was taken under a light microscope with a 100x objective. Scale bar = 10 µm. Please click here to view a larger version of this figure.

Figure 4: Diagram describing a specific example of cytoduction to integrate the mutation COX1ΔC15 in the mitochondrial genome. (A) Liquid cultures of the synthetic rho^- strain (donor, XPM199) and the strain of interest (acceptor, XPM10) were mixed to mate. After shmoo formation, the mix was recovered in a liquid culture for 2-4 h. Next, the cytoduction mix was spread on a plate with a medium lacking adenine to prevent the growth of the donor cells. (B) Schematic representation of the recombination events of the mitochondrial DNA from the donor (XPM199) and the acceptor (XPM10) during cytoduction. (C) The master plate was replicated on different selective media to identify and purify those haploids that integrated the intended mitochondrial gene in the organelle's DNA (XPM209)²⁵. Abbreviations: -ADE = lacking adenine; -LEU = lacking leucine. Please click here to view a larger version of this figure.

NOTE: We recommend making six transformations for each construct since mitochondrial transformation efficiency is usually low. The composition of the different growth media is shown in Table 2.

1. Tungsten particle preparation

1. Weigh 30 mg of 0.7 µm tungsten particles (WPs, microcarriers) in a microtube. Add 1.5 mL of 70% ethanol (EtOH) to sterilize. Vortex the WPs and let them rest for 10 min at room temperature.
2. Centrifuge at 13,200 x g for 15 min at room temperature. Wash the WPs by adding 1.5 mL of sterile water to the particles, vortexing them, and centrifuging at 13,200 × g for 15 min at room temperature. Discard the supernatant.
3. Add 500 µL of sterile 50% glycerol (final concentration is 60 µg of WPs/µL).
   NOTE: WPs can be stored at -20 °C for at least 6 months. From this stock, take the required amount for six transformations (100 µL).

2. WPs DNA coating

1. Add 5 µg of the 2 µ plasmid and 15 µg of the bacterial plasmid containing the mitochondrial sequence to a microtube on ice.
   NOTE: The final volume must not exceed 50 µL to avoid further diluting the reagents needed for DNA coating.
2. Add 100 µL of the WP suspension and vortex for 5 s (6 mg of WPs).
3. Add 4 µL of 1 M spermidine. Briefly vortex.
4. Add 100 µL of ice-cold 2.5 M CaCl₂, briefly vortex. Incubate on ice for 10 min; briefly vortex the mix every 3 min.
   NOTE: The CaCl₂ solution is filter-sterilized and must not be autoclaved.
5. Centrifuge the WPs for 30 s at 13,200 × g at 4 °C and discard the supernatant. Wash the particles by gently resuspending them in 200 µL of 100% EtOH at -20 °C.
6. Centrifuge the WPs for 30 s at 13,200 × g at 4 °C and discard the supernatant. Repeat the wash.
7. Resuspend the WPs in 60-70 µL of room temperature 100% EtOH using a pipette tip; do not vortex.
   NOTE: To ensure the quality of the plasmids, DNA coating should be performed on the same day as biolistic transformation.

3. Biolistic material preparation

1. Before bombardment, sterilize six macrocarrier holders (Figure 5) in an autoclave and thoroughly dry them. Alternatively, soak them in EtOH and sterilize them using a flame with the help of forceps. Place them on a sterile glass plate.
2. During DNA precipitation (10 min incubation on ice in step 2.4), insert six macrocarriers (or flying discs) in each one of the six macrocarrier holders with the help of sterile forceps.
   NOTE: Do not sterilize the macrocarriers.
3. Mix the DNA-coated WPs with the pipette. Add 10 µL of the WPs to each macrocarrier surface and distribute over the entire surface with the pipette tip. If necessary, add a few extra microliters of 100% EtOH to the macrocarrier surface to make an even spreading of the WPs. Let the suspension dry at room temperature for ~20 min.

4. Preparation of yeast cells and bombardment plates

1. Inoculate the rho⁰ receptor strain on 2 mL of YPR medium (Table 2). Grow for one night at 30 °C in a rotating wheel.
2. Dilute 300 µL of the culture in 30 mL of fresh YPR medium. Grow at 30 °C and 200 rpm on a rotary shaker for two nights until the culture is saturated.
3. Centrifuge the yeast cells at 600 × g for 5 min at room temperature. Discard the supernatant and resuspend the yeast cells in 600 µL of YPD medium (Table 2).
4. Using a sterile glass handle, spread 100 µL of the yeast cell suspension on a solid bombardment medium plate.
   NOTE: The bombardment medium must lack the selection marker of the 2 µ plasmid that will be used for transformation.
5. Repeat step 4.4 for the other five plates.
6. Incubate for 1-3 h at room temperature before bombardment to ensure the liquid medium is completely absorbed on the plate's surface.

5. Generation of a synthetic rho^- strain by biolistic transformation

NOTE: Before beginning, be sure to read the user's manual for the biolistic equipment. We will describe the protocol of the referenced equipment (Table of Materials and Figure 5).

1. Clean the chamber of the biolistic particle delivery system with 70% ethanol to prevent contamination and let it stand for at least 10 min. Just before bombardment, wipe the surface with a clean paper towel.
2. Open the helium tank valves to set around 2,000 psi.
3. Power on the equipment and turn on the vacuum source.
4. Using forceps, place the macrocarrier facing down on the macrocarrier holder.
   NOTE: Do not use the stopping screen that comes with the equipment.
5. Screw the macrocarrier lid on top of the macrocarrier. All parts together constitute the microcarrier launch system (Figure 5).
6. Place the microcarrier launch system inside the vacuum chamber at the fifth level from the bottom.
7. Place one of the plates containing the rho⁰ strain lawn on the second level from the bottom of the chamber. Ensure that the plate faces up and is without the lid.
8. Using sterile forceps, put one rupture disk on the retaining cap. Screw the retaining cap on the acceleration tube; make sure that the retaining cap is snug.
   NOTE: Do not treat the ruptured discs with isopropanol as suggested in the manual since we have observed that transformation efficiency decreases.
9. Close the door of the vacuum chamber.
10. Put the VAC/VENT/HOLD switch in the VAC position. When 25-28 inHg vacuum is reached, change the VAC/VENT/HOLD switch to the HOLD position.
    NOTE: Be aware that achievable vacuum can vary according to sea level. We found success reaching 25 inHg.
11. Press the FIRE switch to shoot the WPs. Wait for the pressure to build in the acceleration tube until it reaches 1,300 psi when the rupture disk will break, making a pop sound.
12. Immediately after the rupture disk breaks, the FIRE switch and the vacuum are released by changing the VAC/VENT/HOLD switch to the VENT position.
13. Carefully, with sterile forceps, remove any debris of the rupture disk present on the surface of the plate and cover the plate with its lid.
14. Repeat the procedure for the remaining five plates, placing a new rupture disk and macrocarrier in each round.
15. At the end, close the main valve of the helium tank.
16. Close the chamber and build a vacuum by taking the VAC/VENT/HOLD switch to the VAC position. Release the system from any helium pressure by pressing and releasing the FIRE switch a few times. Repeat with the door open.
17. Turn off the vacuum pump and the bombardment equipment.
18. Clean the delivery system with 70% EtOH. Wash the macrocarrier supports with a 0.1% SDS solution; rinse with sterile distilled water and 100% EtOH.

6. Synthetic rho^- colony selection

1. Incubate the bombardment plates for 5-7 nights at 30 °C.
   NOTE: It is important to check the plates daily for contamination. If necessary, the contaminated region can be cut out with a sterile spatula to avoid further contamination.
2. Inoculate the tester strain of the opposite mating type in 2 mL of YPD medium overnight (see the specific example in Figure 2D,E).
3. Inoculate 150 µL of the tester strain culture on a YPD plate. Using a sterile glass handle, spread the tester evenly on the plate surface. Let the plate dry for at least 5 min.
4. Make replicas of the bombardment plates using velvet replicating pads and a replicator stamp.
   1. Replicate on a dropout medium plate corresponding to the transformed 2 µ plasmid auxotrophy marker. Growth takes one overnight incubation at 30 °C. Afterward, store at 4 °C until further use. This plate is the master plate.
   2. Replicate on the YPD plate containing the tester´s lawn from step 6.3. Incubate at 30 °C for two nights. Afterward, replicate this plate on the selective medium to detect the presence of the mitochondrial gene of interest, for example, respiratory medium or a medium lacking arginine, and incubate overnight at 30 °C. Only the cells that harbored the synthetic bacterial plasmid in mitochondria will grow on the selective medium after mating with the tester strain.
5. Identify on the master plate (step 6.4.1) the corresponding colonies that grew in step 6.4.2.
   NOTE: Since the bombardment plate (and thus, the master plate) is crowded with transformant colonies, it can be difficult to identify those positive colonies observed on the tester's mating plate. For this reason, it is highly recommended to draw a couple of small lines on each side of all plates with a permanent marker. Draw similar lines on the ring of the replicating stamp. These lines will be helpful in orienting and locating positive colonies on the master plate (see the example in Figure 2D).
6. Purify the positive colonies by streaking them from the master plate to a second selective dropout plate, as in step 6.4.1. Incubate for two nights at 30 °C.
   NOTE: It is essential that streaking allows the growth of single colonies to ensure a pure strain.
7. Repeat the selection process (steps 6.2-6.5) to confirm the presence of the bacterial plasmid in mitochondria.
   NOTE: Usually, this process can be repeated two more times to select stable synthetic rho^- strains.
8. Once the synthetic rho^- strain stably maintains the plasmid DNA in mitochondria, select 3-5 different positive colonies and inoculate them in 2 mL of YPD. Use this culture to prepare a cryovial stock (by mixing 600 μL of culture and 600 μL of 50% glycerol to preserve at -70 °C) and to insert the plasmid sequence into the mtDNA of interest by cytoduction (step 6.7).
   NOTE: Since there is no selective pressure to maintain the transformed plasmid in the mitochondria from the synthetic rho^- strain, the plasmid can be easily lost. Therefore, we recommend making the cytoduction as soon as possible.

7. Integration of the construction present in the synthetic rho^- strain into the target site in the mitochondrial genome by cytoduction

1. Inoculate two tubes with 3 mL of YPD media, one with the donor cells (synthetic rho^- cells) and one with the acceptor cells (yeast cell where the final construction is expected). Incubate overnight at 30 °C.
2. Mix 750 µL of the donor strain and 250 µL of the receptor strain in a sterile microtube (3:1 ratio).
3. Centrifuge the mixture for 1 min at 13,200 × g; discard the supernatant. Using a 1 mL pipette tip with the end cut off, take one drop of the cell pellet and place it on a YPD plate. Incubate at 30 °C for 2-4 h.
   NOTE: To avoid cross-contamination, do not put more than two cell suspensions in a single plate (see the example in Figure 4A).
4. Take a small amount of the cell mixture, resuspend it in 15 µL of sterile water on a slide, and place a cover slip. Check under a light microscope for the presence of shmoos with a 40x or 100x objective; this characteristic shape shows up when two mating types are present (Figure 3B).
5. If shmoos are present, resuspend a small amount (~0.5 mm³) of the cell mass in 2 mL of YPD medium. Incubate for 2 h at 30 °C in a rotating wheel to allow the binuclear zygotes to recover.
6. Vortex the mix and dilute 10 µL of the cell suspension in 990 µL of sterile water (1:100 dilution). Use glass beads or a sterile glass handle to spread 100 µL of the cell dilution on a dropout medium where only the acceptor strain (and sporadic diploids) grow. Incubate for 2-3 nights at 30 °C.
7. Replicate the resulting colonies on the necessary selective media as explained in the introduction (stage 4; see Figure 4C for a specific example).
8. After identification of positive colonies, restreak them on the same selective media to purify the haploid strain carrying the mtDNA of interest.
9. Select at least two positive colonies and grow them on 2 mL of YPD medium at 30 °C overnight. Make a cryogenic backup by mixing 600 µL of 50% glycerol with 600 µL of the saturated YPD culture. Store at -70 °C.
10. Isolate DNA of the resulting strains^31,32. Amplify the desired region by PCR and sequence to confirm the presence of the mitochondrial mutation.
    NOTE: Each positive colony is a technical replicate that must be confirmed independently to ensure the correct sequence in the mitochondrial genome.

This section presents some representative results from the different stages of mitochondrial transformation. Figure 6 shows a bombardment procedure. The synthetic rho^- cells carried a bacterial plasmid with the reporter gene ARG8^m, which will replace the coding sequence of a mitochondrial gene (Figure 6A). After bombardment, the plate was replicated on a medium lacking uracil (-URA); this is the master plate (Figure 6B). The growing colonies have the yeast expression vector with the URA3 auxotrophy marker. According to protocol section 6, the bombardment plate was also replicated on a YPD medium containing a lawn of the tester strain. In this case, the tester strain harbored a defective allele of the ARG8^m gene in the mitochondrial genome (ARG8^m-1)¹. After 2 nights of incubation, the tester lawn and the transformant cells mated, and the mitochondria fused. This plate was replicated on a medium lacking arginine. Only those diploids containing the mitochondrial plasmid were able to grow on this selective medium because the ARG8^m gene present in the synthetic rho^- strain replaced the defective ARG8^m-1 from the mitochondrial genome of the tester strain.

The number of positive colonies generally varies from zero to 20 per plate. In the present example, four positive synthetic rho^- colonies were obtained, which were purified by streaking them on medium lacking uracil. In the first round of purification, many negative colonies were obtained along with the positive ones. However, after two or three rounds of purification, it is possible to obtain the pure positive synthetic rho^- strain. In Figure 6C, we showed a couple of positive synthetic rho^- colonies after three rounds of purification and testing. Most colonies from the -URA master plate are positive for the tester probe. Interestingly, not all the cells of a positive colony contain the mitochondrial plasmid, indicating that the transformed plasmid was lost in a population of the colony. This result is expected since there is no selective pressure to keep the mitochondrial plasmid, making it unstable. However, a synthetic rho^- strain like those shown in Figure 6C can be taken forward to the next step, cytoduction (protocol section 7), and used to prepare stock to be stored at-70 °C. Figure 6D also shows the third round of purification of a synthetic rho^- strain. However, in this case, as the transformants were practically unable to maintain the mitochondrial plasmid, they were discarded.

We described one general strategy to select positive synthetic rho^- colonies, which is presented in Figure 7A. In this case, the mitochondrial construct present in the synthetic rho^- cells has several hundreds of nucleotides to recombine with the mitochondrial DNA from the tester strain (for example, the construct COX1ΔC15 in the synthetic rho^- strain recombines with the mutated gene cox1D369N on the tester's mitochondrial DNA (Figure 2D); or the cox1Δ::ARG8^m construct in the synthetic rho^- strain with the mutated gene ARG8^m-1 in the tester's mitochondrial DNA (Figure 6A). However, if there is no availability of a tester strain to select the positive synthetic rho^- cells of interest, there is an indirect approach to select them³³. In this case, the bacterial plasmid must also contain a fragment of a mitochondrial gene, like COX2 or COX3. The tester strain contains a nonfunctional copy of the mitochondrial cox2 or cox3 gene. In the given example, the sequence of the cox3 fragment present in the bacterial plasmid can complement the cox3 sequence present in the tester's mitochondria and, after recombination, provide a functional COX3 gene. After crossing the synthetic rho^- cells with the tester strain, diploids can respire only if the synthetic rho^- cells contain the bacterial plasmid inside mitochondria (Figure 7B).

Figure 8 exemplifies the cytoduction process. The donor and acceptor strains have leucine and adenine auxotrophy, respectively. Hence, after cytoduction, the mix was plated on a medium lacking leucine, preventing the growth of the donor cells. This is the master plate. At this point, only haploids with the nuclear background of the acceptor strain and sporadic diploids will grow. To rule out diploids, the master plate was replicated on media lacking adenine, where only diploids can grow. In this example, the resulting mitochondrial DNA after cytoduction is mutated, and the intended haploids cannot grow on the respiratory medium of ethanol/glycerol. Therefore, the master plate needs to be replicated on YPD plates with a lawn of a tester strain. The position of positive respiring diploids indicates which colonies to purify from the master plate. Two rounds of purification are necessary to obtain the final, pure strain. The mitochondrial integration is usually corroborated by PCR amplification of the surrounding region of interest and sequencing.

Finally, genes can be inserted in the specific locus in the mitochondrial genome. For example, reporter genes can be inserted at the 3' end of a mitochondrial gene's coding sequence or replace the mitochondrial gene's coding region, among other options (Figure 9A). However, genes can also be inserted at ectopic sites within noncoding regions of the mitochondrial genome. The commonly used site is 295 nt upstream of the start codon of the COX2 gene. The alteration at this noncoding region is innocuous¹³. To successfully express the gene in the ectopic site, it must be flanked by a mitochondrial promoter and terminator. The COX2 flanking sequences are the most optimal since COX2 is the mitochondria gene with the shortest 5' and 3' untranslated sequences (73 nt of the 5' UTR and 119 nt of the 3' UTR) (Figure 9B). For this construction, the bacterial plasmid has the following characteristics: the template is a plasmid that contains the COX2 coding region, with 1120 nt downstream of the stop codon and 670 nt upstream of the AUG start codon, named pXPM50²⁵. The gene of interest is engineered to be flanked by 73 nt of the COX2 5'-UTR and 119 nt of the COX2 3'-UTR. Afterward, it was inserted, as previously indicated, at 295 nt upstream of the COX2 AUG start codon, which is a noncoding region (Figure 9B). Note that this plasmid, and consequently, in the resulting mitochondrial genome, the COX2 5' and 3' UTRs are duplicated. After the bombardment of this plasmid, the synthetic rho^- cells are mated for cytoduction with a receptor strain that has an extensive deletion of the COX2 region (cox2-62 allele)¹. During cytoduction, the mitochondrial plasmid of the synthetic rho^- strain recombines with the mitochondrial DNA from the acceptor. The resulting mitochondrial genome has the ectopic construction and a WT COX2 gene. Therefore, the positive colonies will be able to grow in a respiratory medium (Figure 9B).

Figure 5: Schematic representation of the high-velocity microprojectile bombardment equipment. Our lab uses the Biolistic-PDS-1000/He Particle Delivery System (Table of Materials). Abbreviation: WPs = tungsten particles. Please click here to view a larger version of this figure.

Figure 6: Example of a high-velocity microprojectile transformation plate. (A,B) After bombardment, plates were incubated for 5-7 nights at 30 °C. If a contaminant grows (indicated by a blue arrow), it can be cut out with a sterile spatula. After replicating the plates on the selection marker medium (-URA in this case), the master plate was also tested for the presence of the mitochondrial plasmid by mating to a tester strain. In this case, the tester strain has the non-functional allele ARG8^{m-1 1}. After mating for 2 nights and replicating the plate on a medium lacking arginine (-ARG), four positive colonies arose (indicated with red arrows). (C) The positive synthetic rho^- colonies were purified by restreaking them on -URA medium, and replica testing was carried out as previously described. These plates represent the third round of purification. (D) Some positive synthetic rho^- colonies lost the mitochondrial construct. Abbreviations: mtDNA = mitochondrial DNA; -URA = lacking uracil; -ARG = lacking arginine. Please click here to view a larger version of this figure.

Figure 7: Diagram showing how to select positive synthetic rho^- colonies after bombardment. (A) The bacterial plasmid contains a nonfunctional mutated version of a mitochondrial gene. For detection, the synthetic rho^- colonies are mated to a tester strain containing a different nonfunctional mutation of the same gene in mitochondrial DNA. After mating, the mutant mtDNAs recombine, and a functional gene can be recovered by complementation. Here, the diploid grows on a selective medium where neither the synthetic rho^- strain nor the tester can grow. (B) In this case, the plasmid contains an additional sequence of a mitochondrial gene (i.e., a piece of COX3). After transformation, the synthetic rho^- strain is selected by its ability to complement a COX3 mutation in the tester strain, so only the diploids can grow on a selective medium (i.e., respiratory medium). This method is handy if a tester strain for the gene of interest is unavailable. Detection of positive synthetic rho^- colonies in this case is indirect. Abbreviation: mtDNA = mitochondrial DNA. Please click here to view a larger version of this figure.

Figure 8: Example of cytoduction to integrate a construct into the target locus in mitochondrial DNA. After mating and recovery, the mixed cells and binuclear zygotes were plated on a medium lacking leucine. In this case, as the donor strain has the leu2,3-112 allele, this medium prevents donor growth. This plate was replicated on the same medium (master plate), in a medium where only diploids can grow (-ADE) and in a medium where the acceptor with the intended mutation can grow (red arrows). In this case, neither the donor nor receptor cells can grow on respiratory media. Only the resulting diploid after mating with a lawn of a tester strain can grow on a respiratory medium. These respiring diploid colonies are a guide to selecting and purifying the positive colonies from the master plate. Abbreviations: -LEU = lacking leucine; -ADE = lacking adenine. Please click here to view a larger version of this figure.

Figure 9: Diagram showing how to integrate the mutated mitochondrial construct (present in the synthetic rho^- strain) into the acceptor strain. (A) In this example, the acceptor strain carries the ARG8^m reporter gene that replaces the codons from the gene of interest. The synthetic rho^- DNA contains a mutated version of this gene. The plasmid and mitochondrial DNA recombine through the 5' and 3' flanking regions, and the ARG8^m reporter is replaced with the mutated sequence of the gene of interest. (B) Foreign genes can be ectopically inserted in specific regions of the mitochondrial genome^7,13,36. In this example, a gene can be introduced at a position 295 nt upstream of the COX2 gene. This is a noncoding region of mtDNA. A foreign gene of interest is flanked by 73 nt and 119 nt of the COX2 5' UTR and 3' UTR, respectively. This construction is inserted in a plasmid containing the COX2 gene plus 670 nt and 1120 nt of the 5' and 3' UTR regions. The synthetic rho^- strain (donor) carries this construct in mitochondria. In this case, the acceptor strain (XPM13)⁷ has a significant deletion of the cox2 gene (allele cox2-62)¹. After cytoduction, mitochondrial DNA from the donor and acceptor recombine. The COX2 gene is restored, and the gene of interest is inserted at the ectopic site (295 nt upstream of the COX2 gene). Abbreviations: mtDNA = mitochondrial DNA; nt = nucleotides. Please click here to view a larger version of this figure.

Table 1: Yeast strain examples used in biolistic transformation and cytoduction.

Table 2: Yeast growth media used in this protocol.

The present work described how to transform mitochondria from the yeast S. cerevisiae successfully. The process, from high-velocity microprojectile bombardment until purification of the intended yeast strain, takes ~8-12 weeks, depending on how many rounds of purification of the synthetic rho^- strain are necessary. Some of the critical steps of the method are as follows. First, the larger the flanking regions added around the mutation site in the mitochondrial gene construct, the higher the probability of successful integration of the target mutation. The minimum size of the flanking regions is ~100 nt; however, using more than 500 nt on each side of the mutation is highly recommended. Second, when selecting the synthetic rho^-colonies after transformation, it is recommended that no more than 3-4 rounds of purification take place; otherwise, there is a high risk that eventually the mitochondrial construct will be lost, as there is no selective pressure to keep the plasmid in mitochondria. Third, to avoid harming the user or the equipment, you must carefully read the bombardment machine's manual before starting the procedure. This equipment is delicate and uses high pressure and vacuum.

One limitation of this procedure is that so far, only yeast and the green algae C. reinhardtii are candidates to transform mitochondria with this method. It would be relevant if we could introduce site-directed mutations on the mitochondrial genome from other organisms. Another limitation is that it is challenging to transform many different constructs in the same session. It is recommended to make six transformations for each construction, each of which must be tested, selected, and purified. This results in over 50 plates of media only to obtain pure rho^- colonies of a single mitochondrial mutation.

Despite these limitations, the ability to manipulate the yeast mitochondrial genome has significantly contributed to understanding many pathways of respiratory complexes biogenesis. This approach can be used to introduce reporter genes to monitor and quantify mitochondrial translation^1,7,36,37, interrupt mitochondrial genes³⁸, identify assembly chaperones of respiratory complexes³⁹, better understand the mechanisms of coupled mitochondrial translation and assembly^7,11,12,36, and look for suppressors when following the expression of a mutated mitochondrial gene⁴⁰. So far, although certain approaches to manipulate the mitochondrial genome in mammals have been reported earlier⁴¹, no widespread method to transform mitochondrial DNA exists for these organisms.