Altered quantity of mtDNA at different stages of cotyledon

III. Results

1. Altered quantity of mtDNA at different stages of cotyledon

To check the mtDNA quantity, mitochondria were isolated from cotyledons at different stages and stained with DAPI for fluorescence microscopy observation (Zeiss Axio Imager Z1, detected by CoolSnap HQ) (Fig. 1A) and flow cytometry (Fig.

1B) to indicate the amount of DNA from isolated cotyledon mitochondria. MtDNA content increased in cotyledons, from 4 °C-12 hr samples, and peaked in 27 °C-12 hr samples. The high content was sustained in day-1 and -2 mitochondria after

germination and steeply decreased in day -3 mitochondria.

The mtDNA content in cotyledon mitochondria was analyzed by PFGE. The same amount of mitochondria in terms of protein content (0.5 mg for serial aged mitochondria) was used. Mitochondria from different stages of cotyledons were lysed and treated with protease K. MtDNA separated into two sub-populations by PFGE as previously reported (Bendich, 1996; Dai et al., 2005): one stayed in the sample loading well (well-bound, wb) and the other one migrated as a smeared zone to the area equivalent to 50 to 200 kb of linear DNA (the fast-moving component, fm). The mobility of fm mtDNA was the same among different stages of cotyledons (Fig. 2A).

The mtDNA content was detected by Southern blot analysis (right panel of Figs 2A and B). As for flow cytometry, with Southern blot analysis, 4 °C-12 hr immersed cotyledons showed less mtDNA content, especially in the wb fraction. The wb mtDNA appeared in 27 °C-12 hr mitochondria and increased in day-1 mtDNA, followed by some decrease in day-2 mtDNA. A small amount of the wb form was found in day-3 cotyledons. In 4 °C-12 hr immersed cotyledon, a large amount of fm

fluctuations of wb and fm parts indicate these two sub-populations may have a different structure and therefore different functions.

Without protease treatment in the same PFGE assay, the mtDNA–protein complex (mt-nucleoids) also separated into two sub-populations. As for protease treatment assay in 4 °C-12 hr immerse cotyledons, the wb form was not detected in dry seeds, or in 4 °C-2, -6 or -12 hr immersed cotyledons by Southern blot analysis (Fig. 2B-b, lanes 1-4). The wb component emerged unambiguously in the 27 °C-12 hr mitochondria and its relative amount increased in day-1 and -2 mtDNA (Fig. 2B-b, lanes 5-7). In addition, the fm component mobility fluctuated during cotyledon development. The retarded fm–protein complex was observed during migration as a smear zone in the area equivalent to 50 to 300 kb linear double-stranded DNA (Fig.

2B). The migration retardation of the fm mtDNA–protein complex became obvious in 27 °C-12 hr mitochondria increased in day-1 mitochondria and remained

approximately the same up to day-2 mitochondria. As a control, proteinase K treatment was applied to an extra day-1 mitochondrial sample (Fig. 2B, lanes 8); the electrophoresis mobility of the fm component was increased in the resulting mtDNA sample (Fig. 2B, lanes 6-8).

2. Physical characteristics of cotyledon mtDNA during seed germination

PFGE analysis revealed two sub-populations in mung bean mtDNA that were presumably caused by differences in DNA conformation and organization (Gerhold et al., 2010). To understand the altered mtDNA conformation/organization, different aged cotyledon mitochondria were lysed and underwent analytical ultracentrifugation (AUC) analysis in a CsCl gradient. The DNA buoyant density estimated by the

marker DNA for Micrococcus lysodeikticus is 1.731 g/cm³. Two sub-populations of mtDNA, peak a, ρ= 1.696 and peak b, ρ= 1.709 g/cm³, were observed by measuring the optical density (OD) at 260 nm after 24-hr centrifugation. Peak b was consistently detected in all mtDNA samples throughout seed germination. Peak a was considered the co-purified chloroplast DNA (cpDNA) as previously described (Wells & Ingle, 1970; Kolodner & Tewari, 1972). It was not detected in 4 ºC- 2 hr and day-3 mitochondria but was seen in other samples (Fig. 3). Southern blot hybridization analysis showed slight chloroplast DNA contamination in 27 °C-12 hr and day-1 mitochondria, but most of the DNA population in 27 °C-12 hr to day 1 mitochondria was mtDNA (Fig. 4A). The mitochondria from day 3 seedlings, as a control, showed peaks at a and b in the AUC analysis, but no chloroplast DNA was detected by Southern blot analysis. The PCR analysis shown in Fig. 4B also shows only a small amount of plastid DNA present in isolated cotyledon mtDNA. These results indicate that peak a may be the mtDNA and cpDNA mixed population. The ratio of peak a to peak b quantity decreased from day-1 to -3 mitochondria, so peak a and b may be two sub-populations of mtDNA with different physical conformation/organization and physiological function.

3. Fluctuation of cotyledon mtDNA synthesis in organello during seed germination

The resuts from PFGE and AUC assay indicated the mtDNA in different growth stages may have different organization for different physiologic functions. To

examine the correlation between mtDNA content/mtDNA complexity and mtDNA synthesis ability during cotyledon development, ³²P-dCTP incorporation was used as described (Dai et al., 2005). At the early seed immersion stage, the newly synthesized

mtDNA was clearly in the fm form in isolated 4 °C-12 hr mitochondria after PFGE fractionation of newly synthesized mtDNA (Fig. 5A, lane 1), whereas the relative amount of the newly synthesized wb form increased gradually after an additional 2 and 6 hr incubation at 27 °C (Fig. 5A, lane 2 and lane 3). 27 °C-12 hr to day-2 cotyledon mtDNA showed a much higher DNA synthesis in both fm and wb forms (Figs 5A, lane 4 and 5B, lane 1-3). Day-3 mitochondria showed a sharp decline in mtDNA synthesis activity (Fig. 5B, lane 4). This result suggests that cotyledon mtDNA with 27 °C-12 hr immersed seed and day 1 after germination has high DNA synthesis ability.

MtDNA synthesis was inhibited with three DNA polymerase inhibitors, aphidicolin, N-ethylmaleimide (NEM), and ddCTP, to assess the nature of the polymerase responsible for the observed in organello mtDNA synthesis (Fig. 5C).

After 1-hr pulse-labeling, in organello cotyledon mtDNA synthesis was strongly inhibited by NEM and ddCTP (Fig. 5C, lane 4-8), with no effect of aphidicolin treatment, even at high concentration (100 μM; Fig. 5C lane 2, 3). Thus, the in organello DNA synthesis activity of mung bean mitochondria may be driven by an

organello localized γ-like DNA polymerase rather than a nucleus-localized α-like polymerase (Sala et al., 1980; Heinhorst & Cannon, 1993; Dai et al., 2005; Zhang et al., 2011). Furthermore, during cotyledon development, the fluctuation detected for

the in organello mtDNA synthesis was substantial, especially for the wb form, which suggests that the observed mtDNA synthetic activity reflected a replicative mtDNA synthesis not just a repair/recombination event, which may also be involved. (The DNA synthesis data were generated by Dr. Annamalai and me).

Taken together, the results indicate that cotyledon mtDNA content per

mitochondrion, the structural complexity of mtDNA, and its replication activity all

fluctuated in parallel with each other during cotyledon development. This line of evidence further suggests that the increase in structural complexity of mtDNA is linked to its replication activity as reflected by the increase in mtDNA content per mitochondrion.

4. The difference between wb and fm mtDNA in terms of buoyant density, ultrastructure and restriction enzyme patterns

To clarify the correlation of the populations on PFGE gel and AUC assay data, wb and fm 27 °C-12 hr cotyledon mtDNA was eluted from PFGE gel for AUC assay.

Both wb and fm DNA still showed low density peaks (~1.6981 and 1.7006 g/cm³, respectively) and high density peaks (~1.7094 and 1.71124 g/cm³, respectively). Both low and high density peak profiles showed as very broad peaks (Fig. 6B), which may have been caused by high fragmentation during re-purification of wb/fm DNA from PFGE gel. EM revealed a more complex DNA structure in wb than fm mtDNA (Fig.

6C). The restriction enzyme digestion pattern indicated the similarity between wb and fm mtDNA, except in some minor bands (Fig. 6D, arrows; the data were generated by Yih-Shan Lo ).

5. MtDNA ultrastructural dynamics during cotyledon development

To further elucidate to the issue of mtDNA structural change during different cotyledon stages, mtDNA from different-aged cotyledons purified by CsCl-DAPI gradient ultracentrifugation were examined on EM. Before seed germination, mtDNA was observed as a short linear fragment or branch structure, most just 3-5 kb long in dry seeds (Fig. 7A and Table 2). After 4 °C-2 hr immersion, the linear fragments were slightly longer, and a high proportion was converted to a simple rosette structure in

cotyledons (89% of total length, Fig. 7B, Table 2). Most of these rosettes were 20 to 40 kb. These rosette structures maintained a similar but looser structural conformation in up to 4 °C-6 hr immersed samples; however, the rosettes plus reticular rosettes (rosettes comprise small reticular network) decreased from 91% to 69% of the total mtDNA length (Fig. 7C and Table 2). After 4 °C-12 hr imbibition, nearly all of the mtDNA was converted to a longer linear form and the rosette core disappeared simultaneously (Fig. 7D). The 4 °C-12 hr long linear mtDNA molecules were much longer than the individual rosettes for 4 °C-2 hr and 4 °C-6 hr mtDNA; only 5% of the total length of mtDNA maintained a rosette-like structure. The large linear mtDNA converted back into rosette structures when seeds were returned to 27 °C for 2 hr and 6 hr, after a 4 °C-12 hr imbibition; nearly 87% and 90% of the total length consisted of simple or regular rosette molecules, respectively (Figs. 7E, F and Table 2).

A much more complex rosette structure appeared with 27 °C-12 hr culture for day-1 and -2 mtDNA (Figs. 8-10). These complex rosette structures were often attached to linear, branched and circular mtDNA structures. Obvious core(s) were visible at the centers of rosettes or sub-rosettes. Nearly 86% of the total mtDNA length in 27 °C-12 hr and day-1 mitochondria were complex rosette plus reticular rosette structures. More reticular networks were found around the core structure, and most appeared in the early mtDNA synthesis stages (Figs. 8, 9, long dotted arrow).

The remaining parts were simple long linear or branchlinear forms. An independent circular structure was nearly absent in our observations. In day-2 cotyledon mtDNA, more than 50% of mtDNA remained as complex rosette structures (Fig. 8C). The giant rosettes were found in highly mtDNA-replicating mitochondria (Figs. 9 and 10).

In day-3 mtDNA, the compact rosette molecules mostly disappeared. Instead, linear DNA with/without attachments to looser networks and branch structures appeared;

only 19% of mtDNA maintained a very loose rosette structure in day-3 cotyledon mitochondria (Fig. 8D). These results indicate that the extent of mtDNA structural complexity is closely associated with the magnitude of synthesis activity during cotyledon development.

6. Structural characteristics of giant mtDNA rosette complex

On EM, a rosette-like structure of mtDNA was obvious in different stages. These individual rosette genome sizes were about 20 to 200 kb, much smaller than the 401 kb reported by other investigators (Alverson et al., 2011b). A giant rosette-like

complex was observed occasionally, especially at the height of the replicative mtDNA synthesis period during cotyledon development. EM revealed giant rosette mtDNA present in 27 °C-12 hr (Fig. 9A) and day-1 (Fig. 9B) mitochondria. Various-sized petal-like structures extended from the central core (arrows in Fig. 9). Around the core region, concentric-like petals were affiliated with putative “RDR forks”

(Oldenburg & Bendich, 1996; Backert & Borner, 2000; Oldenburg & Bendich, 2001;

Gerhold et al., 2010). The RDR forks are represented by the red sketches drawn in the upper left box of panels A and B in Figure 9, respectively. The satellite cores are indicated with arrowheads in panels A and B. Both the central core and satellite core presented as a single or multiple form. Petal-like structures and/or linear DNA structures extended outward from the central or satellite cores. The small reticular networks, marked by long dotted arrows, were often found in giant rosettes. A Holliday junction is indicated by a short dotted arrow in the upper left box of the figure. Generally, the size of the giant rosette structural molecule was larger than the single mtDNA genome size and could be as large as 2-6 genome units per molecule.

The DNA connection between rosettes could be visualized on EM (Fig. 10),

showing mtDNA rosette structures with similar (panel B-a) or varied (panel B-b) complexities in structure and organization linked to each other by DNA junctions.

The DNA junction connecting the rosette structure clearly shows that the connection passes from core to core between the rosettes (Fig. 10A, box b, and 10B). A putative RDR fork (red sketch), Holliday junction (short dotted arrow) and D-loop (short arrow) are also visible in Figure 10.

A complicated giant rosette molecule in the form of relaxed multi-rosette units was found in day-1 mtDNA (Fig. 10A). This irregular giant rosette structure differs from that in Figure 9 and rarely appeared in the investigation. This extremely large and complicated mtDNA molecule may be relaxed or converted from the RDR intermediate shown in Figure 9 during mtDNA purification. The enlarged boxed area A of Figure 10A in the upper right corner exhibits multi-cores, replication forks, a Holliday junction (short dotted arrow), and a D-loop structure (short arrow). All features point to the RDR, for example, branching linear mtDNA (double arrowhead), and a looping circular structure (arrowhead). Some reticular networks were also found near the core (long dotted arrow). The box b of Figure 10A indicates the linkage of three sub-rosettes, from core to core by DNA junctions, similar to Figure 10B. The large complex mtDNA molecule in Figure 10A is > 1000 kb.

7. Unraveling the mtDNA rosette structure by freeze-thaw treatment

in vitro

The expansion and relaxation of the day-1 cotyledon DNA molecule was induced by a long freeze-thaw treatment in water. Day-1 rosette mtDNA (40~200 kb in general) is shown in Figure 11A-a. As compared with the regular rosette-like structure, the rosette components converted into a huge, complicated and loose

mtDNA molecule (panel A-b) after the freeze-thaw treatment in water. The b1, b2 and b3 images at the bottom of Panel A-b represent enlarged images of the corresponding boxed areas displayed in panel A-b (the red sketch presents the mtDNA EM image of Fig. 11A-b). Decomposed mtDNA rosette-like structures appear as a large linear form and are attached with various decomposed replicating structures. The sizes of these large linear structures were > 2000 kb and could not be estimated correctly because they are beyond the range of photography. Moreover, the “core” center had

disappeared. A large number of reticular small networks (~0.5–2 kb per entity, indicated by long dotted arrows in Figs. 11A-b2 and -b3) appeared in this relaxed mtDNA structure. Some large square networks (~10–20 kb per square entity, indicated by the open arrowhead) were also found (Fig. 11A-b2, 11B). Holliday

junctions were present at the corner of each large square structure (short dotted arrow), which suggests the close relationship of these large square structures with RDR.

Interestingly, the core structure also disappeared after the freeze-thaw treatment.

A large number of small reticular networks (~0.5–2 kb) formed and surrounded the core structure without the emergence of a regular/typical rosette structure at the beginning of the mtDNA synthesis (Fig. 12). This reticular network around the core appeared most frequently in the 27 °C-6 hr, 27 °C-12 hr and day-1 mitochondria. In Figure 12A, the mtDNA was purified from 27 °C-6 hr immersed cotyledons; Figures 12B, C, E, and F are from the 27 °C-12 hr immersed cotyledons, and Figure 12D is from day-1 cotyledons. These reticular small networks show a tendency to expand in size, form RDR forks and concentric petal-like structures around the core and, finally, build up a rosette structure as shown in Figure 12F.

In the foregoing study, I proposed that the “core” may represent highly

condensed mtDNA with or without the binding of RNA/protein that converted into a

long linear form with various attached sub-DNA structures in vitro. This form may comprise the complete or partial mitochondrial genome and play an initial and central role in mtDNA RDR. EtBr treatment (10-50 μg/ml) to relax the DNA helix structure before DNA spreading (Fig. 13A) resulted in incomplete decomposition of rosette structures. Therefore, the core could be loosened by EtBr.

In the RDR model, replication initiates at the D-loop formation; new DNA is synthesized on the leading strand and establishes the replication fork (Oldenburg &

Bendich, 2001; Bendich, 2010). Previous study of tobacco mitochondria showed that the wb DNA contained single-stranded mtDNA (Oldenburg & Bendich, 1996).

Single-stranded DNA was also found in the EM analysis here and was evidenced by nuclease digestion (Fig. 13B). Studies of mammalian mtDNA replication suggest that RNA is transcribed initially at the lagging strand before conversion by DNA and produces an RNA-DNA intermediate (Pohjoismaki et al., 2010). The possibility of different buoyant densities of RNA-DNA hybrid and DNA-DNA double strands resulting in the sub-populations in the AUC assay was ruled out. RNaseH treatment analysis revealed no RNA-DNA hybrid present in the mtDNA population (Fig. 14).

Heat denaturation treatment of 27 °C-12 hr mtDNA resulted in a single peak with a high buoyant density of 1.7133 g/cm³.

Some current studies support that the core consists of a DNA–protein complex (Dudareva et al., 1988; Backert & Borner, 2000; Kaufman et al., 2007; Junier et al., 2010; Rebelo et al., 2011). However, mtDNA from CsCl gradient LC-MS/MS did not reveal a significant candidate for the possible proteins bound with the mtDNA rosette core, perhaps because of the low protein level in the DNA–protein complex at the core region in CsCl gradient purified DNA.

8. RDR recombination assessed by real-time PCR

To assess the extent of recombination activity at different stages, two

representative repeats, “E” and “F”, in the mitochondrial genome of mung bean used in recombination activity testing by real-time PCR were selected according to

previous study of the mung bean mitochondrial genome (Table 1A) (Alverson et al., 2011b). The diagram of the main genome real-time PCR product E1/E2 and F1/F2 and corresponding recombination forms E1’/E2’ and F1’/ F2’ is in Figure 15A. The analysis of the repeat’s flanking sequences (a, b, c and d), direct repeat (E), and invert repeat (e, f, g and h) (F) are in Table 1B and Figure 15A. Primers for real-time PCR assay are in Table 1B (cox3 was a reference sequence).

The relative quantification of the real-time PCR products for the main genome are in Figures 15B-a (E1 and E2) and C-a (F1 and F2). The recombination products are in Figures 15B-b (E1’ and E2’) and C-b (F1’ and F2’). The relative recombination activity between E1 and E2/F1 and F2 is shown in Figures 15B-c and C-c,

respectively. The short repeat regions E1 and E2/F1 and F2 may undergo low recombination activity (~3.02-4.79x 10^-4 between E1 and E2; 1.22-1.79- x10^-4 between F1 and F2). Short repeats in cotyledon mtDNA with recombination activity may be involved in RDR. Recombination activity did not differ among different-aged mtDNA even if RDR activity differed among different-aged mitochondria.

9. Mitochondrial nucleoproteins present in cotyledon mitochondria with cotyledon aging during mung bean seed germination

To determine the possible function in different-stage mtDNA, native mt-nucleoids were purified by PFGE fractionation to obtain wb and fm

mitochondria moved in front of fm mtDNA–protein complexes (supporting figure in Table 3B-a).

Elongation factor 1- appeared in the wb DNA–protein complex of 27 °C-12 hr to day-3 cotyledon, but only 27 °C-12 hr and day-1 fm DNA–protein complex showed the same elongation factor. Chaperonin protein CPN 60 and heat shock protein 70 kDa were visible in both wb and fm DNA–protein complexes harvested from 27 °C-12 hr to day-3 cotyledon mitochondria. Some membrane binding proteins, such as ATPase , ATPase, and ANT1, appeared in fm and wb DNA–protein

complexes. The outer membrane protein Porin was also found in every stage. Actin is associated with wb DNA–protein complex of cotyledon mitochondria, from our lab’s previous data (Lo et al., 2011). As the marker protein of mt-nucleoid and joined with mtDNA replication (Edmondson et al., 2005), single-stranded DNA binding protein (mtSSB) was not found in either fm or wb DNA–protein complexes. According to our lab’s unpublished data, mtSSB can be found in day-3 seedling mt-nucleoids isolated by sucrose gradient on LC-MS/MS.

I also detected mtSSB protein in cotyledon mitochondria from all stages by western blot analysis (Table 3B-b). The absence of ssDNA binding protein and the other proteins associated with mtDNA replication may be due to low recovery of the wb/fm protein complex from PFGE gel, and hence, functional proteins associated with mtDNA replication with cotyledon aging cannot be shown (Table 3A).

IV. Discussion

In this thesis, I studied mtDNA replication in mung bean cotyledons to understand the association with the dynamics and changes of physical

structure/conformation during cotyledon development under specific experimental conditions. Combining flow cytometry, PFGE, EM, real-time PCR and biochemical analyses, the results support the following conclusions. (1) MtDNA replication in cotyledon proceeds by a multiple recombination-dependent initiation process generating an organizationally complex physical structure that mediates

recombination and replication. (2) The “rosette core” might contain highly condensed mtDNA with a replication origin and plays an initial and central role in mtDNA replication. (3) A freeze-thaw treatment on mtDNA opened up the unusually large and complex physical structure observed during the active phase of mtDNA replication to reveal a collection of conventional recombination markers, which further confirmed its functional role as an RDR intermediate. (4) Short repeats are actively involved in mtDNA recombination-dependent initiation and likely also in formation of

在文檔中綠豆發芽過程中子葉粒線體DNA之複雜性和DNA複製型式相關性之探討 (頁 27-0)