- © 2019 American Society of Plant Biologists. All Rights Reserved.
Abstract
Cone thermogenesis is a widespread phenomenon in cycads and may function to promote volatile emissions that affect pollinator behavior. Given their large population size and intense and durable heat-producing effects, cycads are important organisms for comprehensive studies of plant thermogenesis. However, knowledge of mitochondrial morphology and function in cone thermogenesis is limited. Therefore, we investigated these mitochondrial properties in the thermogenic cycad species Cycas revoluta. Male cones generated heat even in cool weather conditions. Female cones produced heat, but to a lesser extent than male cones. Ultrastructural analyses of the two major tissues of male cones, microsporophylls and microsporangia, revealed the existence of a population of mitochondria with a distinct morphology in the microsporophylls. In these cells, we observed large mitochondria (cross-sectional area of 2 μm2 or more) with a uniform matrix density that occupied >10% of the total mitochondrial volume. Despite the size difference, many nonlarge mitochondria (cross-sectional area <2 μm2) also exhibited a shape and a matrix density similar to those of large mitochondria. Alternative oxidase (AOX) capacity and expression levels in microsporophylls were much higher than those in microsporangia. The AOX genes expressed in male cones revealed two different AOX complementary DNA sequences: CrAOX1 and CrAOX2. The expression level of CrAOX1 mRNA in the microsporophylls was 100 times greater than that of CrAOX2 mRNA. Collectively, these results suggest that distinctive mitochondrial morphology and CrAOX1-mediated respiration in microsporophylls might play a role in cycad cone thermogenesis.
Thermogenesis has been observed in the reproductive organs (inflorescences, flowers, and cones) of several primitive seed plants, from gymnosperms to angiosperms (Tang, 1987; Gibernau et al., 2005; Seymour, 2010). Many roles have been proposed for floral or cone thermogenesis, such as attracting pollinators by spreading odor (Meeuse and Raskin, 1988), providing an energetic benefit to adult insects (Seymour et al., 2003b), enhancing pollinator larval development (Terry et al., 2004), stimulating mating (Terry et al., 2004), facilitating the arrival and departure of adult insects (Terry et al., 2007), facilitating fertilization (Li and Huang, 2009), and preventing freeze damage in plants (Knutson, 1974). Of these multiple roles, the most commonly proposed role for plant thermogenesis is to promote the emission of volatiles that may serve to attract pollinators such as beetles (Kumano and Yamaoka, 2006; Gottsberger et al., 2013), weevils (Terry et al., 2004; Suinyuy et al., 2013), thrips (Terry et al., 2007), and nitidulids (Kono and Tobe, 2007).
Cone thermogenesis is a widespread phenomenon in two families of cycads (gymnosperms): Zamiaceae and Cycadaceae (Tang, 1987). In most cycad species, the male cones are thermogenic, whereas female cones are only slightly thermogenic or nonthermogenic (Table 1; Tang, 1987; Seymour et al., 2004). Thermogenic cycads are defined as those species that can increase their male cone temperature up to at least 0.5°C above ambient temperature. So far, 43 cycad species have been recognized as thermogenic (Table 1), possibly accounting for more than half of all known thermogenic gymnosperm and angiosperm plants. The levels of temperature elevation of cones vary across cycad species, but several species are able to increase their cone temperature to >10°C above ambient (Table 1). Another prominent feature of cycad thermogenesis is that cone thermogenesis can occur for long periods of time. In several Macrozamia species, thermogenic temperature peaks occur daily for up to two weeks (Terry et al., 2004; Roemer et al., 2005, 2008), and in Cycas micronesica, male cones exhibit a continuous thermogenesis for multiple weeks (Roemer et al., 2013). In contrast, floral thermogenesis in most angiosperms only occurs for one day or a few days at a time (Seymour et al., 1983, 2003a; Raskin et al., 1989; Seymour and Schultze-Motel, 1998, 1999; Miller et al., 2011). The large number of thermogenic cycad species, as well as the intense and durable heat-producing abilities of these species, makes cycads very important in comprehensive studies on plant thermogenesis.
–, no data; CR, critically endangered; EN, endangered; VU, vulnerable; NT, near threatened; LC, least concern.
Alternative oxidase (AOX), which is a terminal oxidase found in all plants, has been proposed to function to prevent reactive oxygen species production in the respiratory chain (Maxwell et al., 1999), dissipate excess cellular reducing equivalents (Vanlerberghe and McIntosh, 1997; Yoshida et al., 2008), and regulate floral thermogenesis in angiosperms (Elthon and McIntosh, 1987). In thermogenic gymnosperms, the presence of cyanide-resistant respiration was initially demonstrated in a respiratory assay using sporophylls (microsporophylls) of the male cones of Zamia pumila (Zamiaceae; Skubatz et al., 1993). However, AOX respiration was also confirmed in nonthermogenic gymnosperms, such as white spruce (Picea glauca; Johnson-Flanagan and Owens, 1986; Weger and Guy, 1991) and Araucaria angustifolia (Mariano et al., 2008; Valente et al., 2012). In addition, since the first report of molecular evidence for the existence of AOX in Pinus pinea (Frederico et al., 2009), high-throughput sequencing has revealed several deduced AOX protein sequences in gymnosperms (Neimanis et al., 2013). There is, therefore, no clear evidence that AOX-mediated respiration in sporophyll tissues is related to cone thermogenesis in gymnosperms, and if it is, it has not yet been determined which AOX genes are involved in cone thermogenesis.
The positive correlation between the energy metabolism of a tissue and the morphological features of its mitochondria, such as the number, size, and cristae densities (Ghadially, 1988), is well known. In mammals, mitochondrial morphology varies with sex (Rodriguez-Cuenca et al., 2002), type of tissue (Benard et al., 2006), and environmental conditions (Loncar et al., 1988; Cousin et al., 1992; Morroni et al., 1995). The morphology of plant mitochondria is also markedly altered during development and by cellular and environmental signals. Recent studies on mitochondrial dynamics in model plants have revealed the existence of elongated or networked mitochondria with typical small mitochondria in the chondriome (the collective mitochondria in a cell) in meristem-like cells such as shoot apical meristems (Seguí-Simarro et al., 2008), germinating cells (Paszkiewicz et al., 2017), and culture-starting protoplasts (Sheahan et al., 2005). Elevated CO2 leads to changes in the mitochondrial morphology of several plants (Griffin et al., 2001, 2004; Tissue et al., 2002; Gomez-Casanovas et al., 2007). Plant mitochondria tend to be longer in the dark in the absence of sucrose (Jaipargas et al., 2015) or under low-oxygen conditions (Ramonell et al., 2001; Van Gestel and Verbelen, 2002). Thermogenic skunk cabbage (Symplocarpus renifolius) contains abundant mitochondria in its heat-producing organs, the spadices (Ito-Inaba et al., 2009b; Ito-Inaba, 2014). The correlation between mitochondrial morphology and energy metabolism in plants is, however, largely unknown. It is therefore of great interest to elucidate whether mitochondrial morphology is related to energy demands in plants.
In this study, we investigated cycad cone thermogenesis, its underlying mechanisms, and mitochondrial function/morphology in a thermogenic cycad species, Cycas revoluta (family Cycadaceae). The categorization of the species as a least-concern species by the International Union for Conservation of Nature (Table 1) made it a suitable species for investigation. Findings of our study suggest that CrAOX1-mediated cyanide-resistant respiration in thermogenic microsporophylls and the distinctive mitochondrial population that they contain play important roles in cycad thermogenesis.
RESULTS
Heat-Producing Abilities in C. revoluta Male Cones Are Significant
Cycad thermogenesis has mostly been studied in Zamiaceae (Seymour et al., 2004; Terry et al., 2004; Roemer et al., 2005, 2008, 2012), and seldom in Cycadaceae (Roemer et al., 2013). However, even in Zamiaceae, thermal images have not been obtained. As cone thermogenesis by C. revoluta occurs during the rainy season, in which cool temperatures often last for several days, we expected that the prevalent cool-weather conditions during this season would assist in the capturing of thermal images of the cycad cones. As expected, on the day when daytime ambient temperatures did not exceed 25°C, thermal images of C. revoluta male cones (but not female cones) were successfully taken using an infrared thermal camera (Fig. 1A). The increase in surface temperature was more prominent in early- and middle-stage male cones than in late-stage male and female cones. When ambient temperature was ∼22°C, the difference between the air temperature (Tair) and surface temperature of the cones was larger in early- and middle-stage male cones (ΔT = ∼6.3°C) than in late-stage male and female cones (ΔT = ∼1.4°C; Fig. 1A). We also assessed the progressive temperature changes in both male and female cones over 10 successive days using thermal recorders (Fig. 1B; Table 2). The internal temperatures (Tcone) of both male and female cones were consistently higher than Tair, notably even when daytime temperatures did not exceed 25°C (Fig. 1B, blue background). Because male cones mature earlier than female cones, i.e. in early summer, it is more likely that they will be exposed to cooler weather conditions (as shown by the blue background in Fig. 1B). It should also be noted that the maximum temperature differences between night-time recordings of Tair and Tcone in male and female cones were 11.5°C and 8.3°C, respectively (Table 2). Collectively, the increased body temperatures observed in both male and female cones suggest that both have a heat-producing ability, although this ability is significantly intensified in male versus female cones.
Thermogenesis in cycad cones. A, Thermal images (right) of Cycas revoluta male and female cones obtained using an FLIR SC620 thermal imager (FLIR Systems) at night at ∼22°C. Photographs acquired during daytime (left). B, Internal temperature of male (left) and female (right) cones (Tcone) and ambient temperature (Tair) recorded over 10 successive days using a thermal recorder. Tcone and Tair are shown in red and gray, respectively. The blue background indicates the day when Tair did not exceed 25°C due to cloudy or rainy conditions.
Averages ± sd are presented for Tcone, Tair, and ΔT (Tcone − Tair). Maximum temperature differences are shown as ΔTmax, and the times at which differences were measured are shown as tTmax.
Ultrastructural Analyses of Individual Tissues in Thermogenic Male Cones
We analyzed the intracellular structures of two major tissues of male cones, microsporophylls and microsporangia, using transmission electron microscopy (Fig. 2) to explore how these individual tissues or intracellular structures are involved in male cone thermogenesis. The male cone consists of a central axis with many spirally attached microsporophylls, which carry dense microsporangia in which numerous microspores develop into pollen grains (Fig. 2A). Ultrastructural analyses using three male cones revealed that the microsporophylls, especially the epidermal cells, have a large number of large, round-shaped mitochondria with a cross-sectional area of 2 μm2 or more and a relatively uniform matrix density (Fig. 2B). Such enormous mitochondria, which were labeled large mitochondria (LMt), were never observed in the microsporangia. It is noteworthy that in the microsporophyll epidermal cells, many mitochondria with area <2 μm2, i.e. not categorized as LMt, also exhibited round shape and uniform matrix density similar to those of LMt (Fig. 2C). In addition, similarly shaped and nonlarge mitochondria were also observed in the microsporophyll parenchyma cells (Fig. 2D). A large number of small mitochondria of various shapes (round, oval, or a dumbbell- or rod-like configuration) were observed in the microsporangia (Fig. 2E). Quantitative analysis to estimate the ratio of LMt in each tissue revealed that in the microsporophyll cells, the appearance frequency of LMt with an area ≥2 μm2 and <4 μm2 was 8.98 ± 0.78% and that of LMt with an area of ≥4 μm2 was 4.07 ± 0.27%; however, such large mitochondria were not observed in the microsporangia (Fig. 2F, left). Similarly, mitochondria with perimeters of 6 μm or more were found in most microsporophylls but only in a few microsporangia (Fig. 2F, right). The above findings suggest that a certain proportion of mitochondria found in microsporophyll cells, especially in the epidermis, are made up of LMt. Given the structural similarities of both nonlarge Mt and large Mt in microsporophyll cells, they might exhibit similar functioning, e.g. in mitochondrial respiration. Interestingly, in microsporophylls, mitochondrial numbers in the parenchyma cells were significantly larger than those in the epidermal cells, suggesting that the density of LMt in the epidermal cells might contribute to the lower number of mitochondria in this tissue compared to the parenchymal cells (Fig. 2G). These data indicate that the chondriomes observed in both epidermis and parenchyma cells of microsporophylls are clearly different from those observed in microsporangia. The difference in the chondriome of microsporophylls and microsporangia might be correlated with their functional differences. No typical features of the central-axis mitochondrial structure were identified, because very few mitochondria were present in this tissue.
Ultrastructural analyses of microsporophylls and microsporangia in thermogenic male cones of C. revoluta. A, Photographs of a male cone with both microsporophylls and microsporangia. Cross section of a male cone (left); a large number of microsporophylls are attached to a central axis. Numerous microsporangia were attached to the surface of a microsporophyll (middle), as evidenced in its cross section (right). B, Electron photographs showing typical LMt observed in the epidermal cells of microsporophylls. Nuc, nucleus. LMt were defined as mitochondria with a cross-sectional area of 2 μm2 or more. The size of this LMt is 5.62 μm2. C, Common structure of LMt and nonlarge mitochondria (Mt) in the microsporophyll epidermis cells. Each mitochondrial size is as follows: LMt1 = 3.41 μm2; Mt1 = 1.44 μm2; Mt2 = 1.54 μm2. Nonlarge mitochondria exhibited a uniform matrix density and round shape similar to those of the LMt. D, Mitochondria, which exhibited a roundish configuration and a uniform matrix density, in microsporophyll parenchyma cells. Although LMts were not observed in these cells, mitochondria with uniform matrix density similar to that of LMt were observed. Pt, plastid. E, Electron photographs showing typical mitochondria observed in microsporangia cells. F, The frequency of the area and perimeter of mitochondria in microsporophyll (ML) and microsporangia cells (MG). In total, 494 mitochondria in microsporophyll cells and 436 mitochondria in microsporangia cells in three male cones were analyzed. The frequency of the area and perimeter were calculated from three independent male cones (n = 3). Error bars indicate the se; *P < 0.05, as determined by Student’s t test; **P < 0.005; ns, not significant. G, Mitochondrial number in the microsporophylls and microsporangia of thermogenic male cones. Mitochondrial number in microsporophyll epidermis (ML-E) and parenchyma (ML-P) cells and in microsporangia (MG) were recorded on each ultrathin section and expressed as the mean ± se for 25 cells within the epidermis in microsporophylls, 25 cells within the parenchyma in microsporophylls, and 20 cells in MG. Two male cones were analyzed. Different letters indicate statistically significant differences between samples within each genotype by the Tukey-Kramer multiple comparison test (P < 0.05).
Cyanide-Resistant Respiration in Intact Mitochondria Isolated from Thermogenic Male Cones
Cyanide-resistant respiration via AOX seems to play a major role in angiosperm floral thermogenesis, but the role of AOX in gymnosperm cone thermogenesis is largely unknown. If cycad AOX plays a role in cone thermogenesis, it is reasonable to expect that AOX capacity would be larger in microsporophylls than in other tissues, such as in microsporangia, because microsporophylls are proposed to be the thermogenic tissues in the male cones (Skubatz et al., 1993). To determine AOX capacity in both microsporophylls and microsporangia, we first isolated intact mitochondria from both tissues. Many more mitochondria were recovered from microsporophylls than from microsporangia in a male cone (Fig. 3A). The intactness of the outer membrane of isolated mitochondria was similar in both tissues, showing that intact mitochondria were successfully isolated from the two tissues (Table 3). Using these isolated mitochondria, we conducted an oxygen consumption assay (Fig. 3B; Table 4). Under succinate oxidation conditions (complex II substrate), the addition of ADP stimulated the oxygen uptake to an average of 67.87 ± 6.19 and 67.15 ± 21.72 nmol O2 min−1 mg−1 of protein in microsporophylls and microsporangia, respectively. This indicated that state 3 respiration rates in microsporophylls and microsporangia were almost identical. However, AOX activity was affected differently in microsporophylls and microsporangia by the addition of potassium cyanide (KCN). This compound reduced state 3 respiration in microsporophylls by half, to an average of 30.84 ± 6.04 nmol O2 min−1 mg−1 of protein, whereas in microsporangia it significantly reduced AOX activity to one-tenth of the state 3 respiration, to an average of 7.20 ± 1.38 nmol O2 min−1 mg−1 of protein. In both tissues, cyanide-resistant oxygen uptake was almost completely inhibited by the AOX inhibitor n-propyl gallate. The mitochondria isolated from both tissues revealed that the cytochrome oxidase (COX) activity in microsporophylls was not significantly different from that in microsporangia (Table 3; Supplemental Fig. S1). These data indicated that the alternative respiration capacity in microsporophylls was much higher than that in microsporangia, suggesting that mitochondrial respiration via the AOX pathway in microsporophylls may play a role in the cone thermogenesis of C. revoluta.
Mitochondria in microsporophylls had higher AOX capacity and protein expression levels than those in microsporangia. A, Isolation of mitochondria from microsporophylls (ML) and microsporangia (MG) using Percoll density gradient centrifugation. Photographs (left) show the band patterns obtained following centrifugation. Arrows indicate mitochondrial fractions. The graph (right) shows the protein content in the mitochondria of microsporophylls and microsporangia. The averages were calculated from at least four different thermogenic male cones (ML: n = 4; MG: n = 6). Error bars indicate the sd. **P < 0.005 as determined by Student’s t test. B, Representative traces (left) of oxygen uptake by intact mitochondria in microsporophylls (ML) and microsporangia (MG). Changes in oxygen uptake rate by the sequential addition of succinate, ADP, KCN, and n- propyl gallate (nPG) in both mitochondria are shown. Averages (right) were calculated from at least three independent experiments from different thermogenic male cones (n = 3). Error bars indicate the sd. *P < 0.05 and **P < 0.005 as determined by Student’s t test; ns, not significant. C, Expression of AOX, UCP, and the mitochondrial marker heat shock protein 60 (HSP60) in mitochondria of microsporophylls (MLs) and microsporangia (MGs). In total, 10 µg of mitochondrial proteins were separated by SDS-PAGE and analyzed by immunoblotting using the indicated antibodies. Mitochondrial proteins were prepared from three thermogenic male cones (n = 3).
Data are the means ± sd of at least three measurements. There were no significant differences between MLs and MGs.
Oxygen uptake was measured as described in “Materials and Methods.” Data are means ± sd of at least five measurements. *P < 0.05 and **P < 0.005 as determined by Student’s t test.
Instead of succinate, NAD(P)H dehydrogenase substrate, NADH, and a combination of tricarboxylic acid cycle intermediates, pyruvate and malate or glutamate and malate (Pyr + Mal and Glu + Mal, respectively; Supplemental Fig. S2), were used as respiratory substrates. However, neither NADH, pyruvate + malate, nor Glu + malate initiated oxygen uptake by mitochondria, and ADP did not stimulate their respiratory activities in either microsporophylls or microsporangia. Therefore, these substrates may not be effectively utilized for mitochondrial respiration in cycad male cones.
Protein expression was analyzed in purified mitochondria from both tissues to determine whether AOX capacities in microsporophylls and microsporangia depend on the levels of AOX proteins (Fig. 3C; Supplemental Fig. S3). An equal amount of mitochondrial protein was separated by SDS-PAGE and analyzed by Coomassie Brilliant Blue staining and immunoblotting. This staining revealed slightly different protein band profiles between the two tissues (Supplemental Fig. S3). Immunoblots detected AOX in microsporophylls but not in microsporangia. Uncoupling proteins (UCPs), which are proposed to be another type of thermogenic protein, were more abundant in microsporophylls than in microsporangia. The levels of the mitochondrial marker heat shock protein 60 (HSP60) were almost identical in the two tissues. Collectively and in combination with the results from the AOX capacity assay (Fig. 3B; Table 4), these findings indicate that the high AOX capacity may be attributable to the abundance of AOX proteins in microsporophyll mitochondria. As UCPs were also abundant in microsporophylls, we do not exclude the possibility that they are involved in thermogenesis.
Isolation and Identification of Partial CrAOX1 and CrAOX2 Genes in Male Cones
To determine the molecular characteristics of the AOX proteins involved in cyanide-resistant respiration in C. revoluta male cones, partial AOX gene sequences were obtained using two degenerate primer sets: P1 and P2 (Saisho et al., 1997) and 42AOX-F (forward) and 42AOX-R (reverse; Frederico et al., 2009). The 440 bp complementary DNA (cDNA) fragments that were amplified using each primer set were cloned independently in a pZErO vector, and the resulting nucleotide sequences of 69 independent clones (44 clones from P1 and P2 and 25 clones from 42AOX-F and 42AOX-R) enabled the identification of two different sequences, indicating that C. revoluta had at least two AOX isoforms.
The two AOX cDNA sequences, designated CrAOX1 and CrAOX2, exhibited 84% sequence identity in the C-terminal region containing ∼224 amino acid residues. The putative structural features of CrAOX1 and CrAOX2, and their amino acid sequence alignment with that of AOX from trypanosome, Arabidopsis (Arabidopsis thaliana), and other gymnosperms are shown in Supplemental Figure S4. CrAOX1 was homologous to the AOX1 of gymnosperms (89% sequence similarity to PgAOX_subtype2 and PsAOX_subtype1, and 85% sequence identity to PgAOX_subtype1), and CrAOX2 was homologous to the AOX2 of gymnosperms (81% sequence identity to PgAOX2). Although the Cys residues CysI and CysII, which correspond to Cys-127 and Cys-177, respectively, in Arabidopsis AOX1 (AtAOX1), are highly conserved in most AOX isoforms, CysI is replaced by Ser in the gymnosperm AOX1, which includes CrAOX1 (Supplemental Fig. S4). This type of AOX has been known to be stimulated by succinate, but not by pyruvate (Holtzapffel et al., 2003; Umbach et al., 2006; Grant et al., 2009; Selinski et al., 2017). Indeed, under succinate oxidation conditions, pyruvate did not stimulate the AOX capacity in microsporophyll mitochondria (Supplemental Fig. S5). The sequence of the N-terminal region of all gymnosperm AOX2 proteins containing the CysI position has not been determined yet.
The phylogenic tree based on several amino acid sequences of AOX homologs from gymnosperms and angiosperms (Fig. 4A) presented three major clades: angiosperm AOX1, gymnosperm AOX1, and AOX2. CrAOX1 was clustered with gymnosperm AOX1 and was close to Picea sp. AOX proteins (PgAOX_subtype 1, PgAOX_subtype 2, and PsAOX_subtype 2). CrAOX2 was clustered with gymnosperm and angiosperm AOX2, and it was closer to Malus domestica AOX2 proteins (MdAOX2d) than to the gymnosperm P. pinea AOX2 proteins (PpAOX2). These data indicate that CrAOX1 and CrAOX2 clearly belong to two distinct AOX subfamilies, AOX1 and AOX2, respectively.
Two cycad AOX genes, CrAOX1 and CrAOX2, showing different expression patterns in thermogenic male cones. A, Phylogenetic tree of the predicted amino acid sequences of CrAOX1, CrAOX2, and related proteins. Red letters indicate gymnosperm AOX proteins, and bold red letters indicate C. revoluta AOX proteins. The tree was created in MEGA7.0.14 based on the alignment produced in Clustal W and using the neighbor joining method. Bootstrap values >10 are shown and were determined from 1,000 replications. The scale bar represents one amino acid substitution per site. The AOX sequence from the green alga Ostreococcus lucimarinus CCE9901 was included as an outgroup. Accession numbers and species abbreviations are listed in Supplemental Table S2. B, Tissue-specific expression of CrAOX1, CrAOX2, and CrUCP transcripts in thermogenic male cones. Expression of CrAOX1, CrAOX2, and CrUCP mRNA in two major tissues of cycad male cones, ML and MG, was analyzed using real-time PCR. CrEF1γ was used as an internal control. Values are means from four different male cones, and bars indicate the sd. *P < 0.05 and **P < 0.005 as determined by Student’s t test. C, Quantitative analysis of CrAOX1 and CrAOX2 transcript levels. To estimate the copy number of each CrAOX transcript in plant mRNA, plasmids containing each CrAOX gene were used as the standard for real-time PCR. Each bar indicates the sd. The expression level of CrAOX1 in MG was set to 1.0. *P < 0.05 as determined by Student’s t test; ns, not significant.
A full-length UCP was isolated from cycad male cones and the cDNA sequence was determined. C. revoluta UCP, designated as CrUCP, was homologous to isoforms in gymnosperm UCP (85% sequence identity to Picea sitchensis UCP; Supplemental Fig. S6). A phylogenic tree for the UCP group was constructed using several amino acid sequences of other homologs from gymnosperms and angiosperms (Supplemental Fig. S7). The tree was divided into three major clades: PUMP1 and PUMP2; PUMP 3; and PUMP 4, PUMP5 and PUMP6. CrUCP was located in the gymnosperm group of the PUMP 1 subfamily, and was also located close to PsUCP. Collectively, these results show that CrUCPs are typical UCP proteins found in gymnosperms.
Tissue-Specific Expression of CrAOX1 and CrAOX2 mRNAs in Male Cones
The AOX proteins accumulated differently in the mitochondria of microsporophylls and microsporangia in C. revoluta male cones (Fig. 3C; Supplemental Fig. S3B). Therefore, it was of interest to determine whether this difference was regulated at the transcriptional level and to analyze which AOX proteins, CrAOX1 or CrAOX2, were responsible for AOX capacity and expression in microsporophyll mitochondria. Real-time PCR using primers that specifically amplified CrAOX1 and CrAOX2 (Fig. 4B) revealed that the CrAOX1 transcript was more abundant in microsporophylls, whereas CrAOX2 was significantly upregulated in microsporangia. The transcript level of CrUCP (Fig. 4B) was increased in microsporophylls compared with that in microsporangia, although the difference between these two tissues was less significant than that observed for CrAOX1. We also analyzed the copy number of CrAOX1 and CrAOX2 transcripts using the plasmids pZAOX1 and pZAOX2, each of which contained one of two CrAOX genes, as the standard for real-time PCR (Fig. 4C). The transcript level of CrAOX1 was significantly higher (>100-fold) than that of CrAOX2 in microsporophylls, but the difference between these two CrAOX genes was not statistically significant in microsporangia. These results, together with the abundance of AOX proteins in microsporophyll mitochondria (Fig. 3C), suggest that the AOX proteins detected in microsporophyll mitochondria were most likely the products of CrAOX1 transcripts.
Collectively, these data suggest that CrAOX1-mediated cyanid-resistant respiration in microsporophylls with a distinctive chondriome may play a role in cycad male cone thermogenesis.
DISCUSSION
Heat-Producing Ability of C. revoluta Cones Is Comparable to That of Some Species in Zamiaceae
Tang (1987) reported that cone thermogenesis was a widespread phenomenon in Zamiaceae and Cycadaceae, and that the heat-producing abilities of thermogenic cycads in Cycadaceae were likely to be weaker than those in Zamiaceae. Several species, which belong to the genera Encephalartos and Macrozamia in Zamiaceae, are able to increase their cone temperature to >10°C above the ambient temperature, whereas in Cycadaceae, the increase in cone temperature above the ambient was at most 5°C (Table 1). Weak thermogenesis was also observed in C. revoluta male cones, and the maximum increase of the cone temperature was only 1.6°C above air temperature (Tang, 1987). Another study also reported that the temperature difference between the C. revoluta male cones and air was 5.4°C (Seymour, 2010). However, in marked contrast to these previous reports, the Tcone of male cones in our study reached 11.5°C above Tair at midnight (Tables 1 and 2), and the thermal images of the male cones were successfully taken (Fig. 1A). In addition, Tcone was consistently higher than Tair even when Tair did not exceed 25°C during the day (Fig. 1B, blue background). The paradoxical difference between the previous reports and our study may be attributable to differences in experimental conditions; for example, Tang (1987) used cut cones removed from intact plants, whereas we used intact cones attached to plants. As male cones cut from C. revoluta exhibit a gradual decline in temperature and die before pollen is released from the microsporangia, the weak thermogenesis reported in the previous study may be attributed to cone detachment. Our findings showed that the heat-producing ability of the male cones in C. revoluta may be much higher than previously believed, and it may occur at an intensity similar to that of several other Zamiaceae representatives that possess a high capacity to produce heat in their cones.
Heat production was not detected in the female cones of several cycad species and, when present, the maximum temperature was consistently lower than that detected in male cones (Tang, 1987; Seymour et al., 2004). Our findings concur with those previously reported, as both the surface temperature and the Tcone of C. revoluta female cones were lower than those of early- and middle-stage male cones (Fig. 1, A and B). However, the Tcone of female cones, even under low Tair conditions and weak sunlight, was consistently higher than Tair (Fig. 1B, blue background), and at night, female cones raised their Tcone to 8.3°C above Tair (Tables 1 and 2). Unexpectedly, 8.3°C was the highest recorded value in all cycad female cones studied to date (Table 1), and therefore, it is possible that the thermogenic abilities of female cones in other genera have been underestimated.
The Correlation between Mitochondrial Morphology and Energy Metabolism in Thermogenic Tissues
It is well known that the energy metabolism of mammalian tissue is positively correlated with the morphological features of its mitochondria (Ghadially, 1988). For example, significant differences are likely to exist between thermogenic brown adipose tissues (BATs) and nonthermogenic white adipose tissues (WATs; Cinti, 2001; Sell et al., 2004), wherein many mitochondria with tightly packed cristae are observed in BATs and very few mitochondria with sparse cristae are observed in WATs (Cinti, 2001; Sell et al., 2004). Several studies have also demonstrated that BAT mitochondria become larger in cold-acclimated rats than in control rats (Loncar et al., 1988; Morroni et al., 1995). However, in plants, only a few studies have suggested a positive correlation between heat production and mitochondrial morphology, such as size, number, and matrix density, in tissues or cells of heat-producing plants. In thermogenic skunk cabbages, a larger number of mitochondria were seen to accumulate in the tissues of both petals and pistils during the thermogenic stage than during the postthermogenic stage (Ito-Inaba et al., 2009b; Ito-Inaba, 2014), and mitochondrial protein content was much higher in thermogenic skunk cabbage than in a nonthermogenic skunk cabbage (Lysichiton camtschatcensis; Ito-Inaba et al., 2009a). In the current study, we observed large mitochondria with round shape and uniform matrix densities in the microsporophylls, especially in the epidermis, but not in the microsporangia of C. revoluta male cones. In addition, even smaller mitochondria, i.e. those not categorized as large mitochondria, in microsporophylls exhibited morphological (matrix) features similar to those of the large mitochondria (Fig. 2, C and D). Given that mitochondrial dynamics have been observed in various plants (Arimura, 2018), the morphological similarities observed in the chondriomes of both large and nonlarge mitochondria in microsporophyll cells implies that they might be derived from the consecutive fission and fusion of mitochondria, and thus, their function, including cellular respiration, might be similar. Interestingly, the shape of the mitochondria in C. revoluta microsporophylls was very similar to that observed in the appendix of Sauromatum guttatum inflorescences during its thermogenic stage (Skubatz and Kunkel, 2000). Although the large mitochondria in C. revoluta microsporophylls were not reported in S. guttatum appendices, the similarities in mitochondrial morphology, i.e. the almost spherical shape with a uniform matrix density, between these two thermogenic tissues may be correlated with the thermogenic capacity or activity, or both, in plant tissues.
Mitochondria with Distinctive Morphology in More General Plants and Green Algae
There has been a long-standing interest in mitochondrial pleomorphy, and giant or enlarged mitochondria have been observed in various plants and green algae. There are a few reports addressing giant mitochondria in egg cells of Pelargonium zonale (Kuroiwa and Kuroiwa, 1992) and in synchronized cells of Chlamydomonas reinhardtii (Ehara et al., 1995) and Euglena gracilis (Osafune et al., 1975). Many of these giant mitochondria contain a large electron-lucent area with a few peripherally localized cristae. The structure of these previously reported giant mitochondria is significantly different from that of the large mitochondria seen in the cycad microsporophylls in the current study (Fig. 2B), as the latter have a uniform matrix density. The notable difference in the mitochondrial structure between the three previously reported noncycad species and the cycads suggests that the respiratory function of the large mitochondria in cycads likely differs from that of the giant or enlarged mitochondria observed in the three noncycad species. Actually, the formation of giant mitochondria in C. reinhardtii induced a temporary reduction in respiratory function (Ehara et al., 1995). Enlarged mitochondria have also been observed in cultured cells of tobacco (Van Gestel and Verbelen, 2002) and Arabidopsis leaves (Ramonell et al., 2001). However, the formation of those mitochondria was induced by low oxygen pressure. Despite the inward parenchyma cells in cycad microsporophylls likely being under lower oxygen partial pressure compared with the epidermal cells, they contained few large mitochondria, and it is therefore improbable that the formation of large mitochondria in cycads was induced by hypoxia. In addition, although elongated or networked mitochondria have also been observed in germinating seeds, shoot apical meristems, and culture-starting protoplasts in several angiosperm plants (Sheahan et al., 2005; Seguí-Simarro et al., 2008; Paszkiewicz et al., 2017), these mitochondria markedly differ in shape from the spherical mitochondria observed in cycad microsporophylls.
In nonthermogenic plants or tissues, energy metabolism and mitochondrial morphology seem to be not always correlated each other. The number of mitochondria increases in the leaves of several plants under conditions of elevated CO2 levels, but in many cases, irrespective of the effects on mitochondrial number, the elevated CO2 decreases the rate of mass-based dark respiration (Griffin et al., 2001, 2004; Tissue et al., 2002; Gomez-Casanovas et al., 2007). In the shoot apical meristem of Arabidopsis, very large and tentaculate/cage-like mitochondria have been located in close proximity to the nucleus, but such a distinctive mitochondrial structure has not been observed in the root apical meristem cells, suggesting that such a unique type of mitochondrion may not be required to satisfy the energy needs of meristematic cells (Seguí-Simarro et al., 2008). Such an elongated or enlarged mitochondrial structure in those plants might be related to mitochondrial DNA replication/recombination rather than energy demand, to provide a kind of quality control for mitochondrial populations in the new generation (Rose and McCurdy, 2017; Arimura, 2018).
AOX Respiration, via CrAOX1, in Microsporophylls Plays a Role in Male Cone Thermogenesis
Although the presence of cyanide-resistant respiration in the microsporophylls of male cones of the thermogenic cycad Z. pumila (Zamiaceae) was first demonstrated in a respiratory assay using the tissue itself (Skubatz et al., 1993), it had not been explored whether such AOX respiration is involved in cycad thermogenesis. Thus, the current study examined the AOX capacities using intact mitochondria isolated from microsporophylls (proposed to be thermogenic) and microsporangia (possibly nonthermogenic) of C. revoluta male cones. The findings clearly indicated that the capacity of cyanide-resistant respiration was significantly higher in microsporophylls than in microsporangia (Fig. 3B). The AOX capacities in mitochondria isolated from cell cultures of the possibly nonthermogenic gymnosperm A. angustifolia (Mariano et al., 2008; Valente et al., 2012) were similar to those in C. revoluta microsporangia obtained in the current study (Fig. 3B; Table 4). The tissue-specific AOX capacity in the C. revoluta microsporophylls, therefore, suggested that AOX respiration in microsporophylls and that in microsporangia may play a major and a minor role, respectively, in cycad thermogenesis. In a male cone, the total weight of microsporophylls was about 10-fold greater than that of microsporangia, and thus, intact mitochondria recovered from microsporophylls were >10-fold larger than those from microsporangia (Fig. 3A, right). The higher AOX capacities and greater weight of microsporophylls suggests that most of the heat in a male cone is produced in microsporophylls rather than in microsporangia.
Recent increases in gymnosperm genomic and Expression sequence tag data revealed that the gymnosperm AOX is encoded by a multigene family (Frederico et al., 2009; Neimanis et al., 2013). However, in gymnosperms, the AOX2 subtypes are limited to two species, P. pinea (Frederico et al., 2009) and Cryptomeria japonica (Neimanis et al., 2013), both belonging to Coniferophyta. In the current study, two AOX genes, CrAOX1 and CrAOX2, were identified in the gymnosperm C. revoluta (Cycadophyta). CrAOX1 mRNA was highly expressed in the microsporophylls, whereas CrAOX2 mRNA was highly expressed in the microsporangia. The transcript level of CrAOX1 was 100-fold greater than that of CrAOX2 in the microsporophylls. In addition, because the mitochondria in microsporophylls had higher AOX capacity and protein expression levels than those in microsporangia (Fig. 3, B and C), CrAOX1 is proposed to play a major role in C. revoluta male cone thermogenesis. This study provides firm experimental evidence that cycad CrAOX1, but not CrAOX2, is involved in cone thermogenesis.
MATERIALS AND METHODS
Plant Materials
Cycads (Cycas revoluta) used in the experiments were grown at the University of Miyazaki and along prefectural road no. 367 in Tsunehisa, Miyazaki-city, Japan.
Surface and Internal Temperature Measurements
The surface temperature of cycad cones was analyzed using images from a thermal imager, FLIR SC620 (FLIR Systems). To eliminate the influence of solar radiation during daytime, thermal images were taken at night when the heat acquired from sun exposure is negligible. The internal temperatures of cycad cones were measured using a thermal recorder (TR-52, T&D).
Ultrastructural Analyses
Ultrastructural analyses were conducted as previously described (Toyooka et al., 2000; Ito-Inaba et al., 2009b), with modifications. Briefly, the microsporophylls, microsporangia, and central axis were separated from each other using a knife and clamping forceps and cut into 1- to 3-mm cubes. Each tissue type was fixed with 4% (w/v) paraformaldehyde and 2% (v/v) glutaraldehyde in 50 mm sodium cacodylate buffer (pH 7.4) overnight at 4°C. Tissues were postfixed with 1% (v/v) osmium tetroxide in 50 mM cacodylate buffer (pH 7.4) for 3 h at room temperature. After dehydration in a graded methanol series, tissue cubes were further substituted in methanol:propylene oxide (1:1), 100% (v/v) propylene oxide, and propylene oxide:Epon812 resin (Taab; 3:1, 1:1, 1:3), and finally embedded in 100% (v/v) Epon812 resin. Ultrathin sections (60–80 nm) were obtained using a diamond knife on ultramicrotome (EM UC7; Leica) and transferred to Formvar-coated grids to be stained with 4% (w/v) uranyl acetate for 12 min and with lead citrate solution for 2 min. Stained sections were examined under a transmission electron microscope (JEM-1400; JEOL). Images were acquired using a charge-coupled device camera. Mitochondria and plastids were distinguished based on the membrane structure inside each organelle. Mitochondrial size, perimeter, and number were quantified using Photoshop CS3 extended software (Adobe).
Mitochondrial Isolation from C. revoluta Male Cones
Mitochondria were isolated using differential centrifugation and Percoll density gradient centrifugation as described by Ito-Inaba et al. (2008a))), with minor modifications of the cell disruption procedures. Briefly, the microsporophylls, microsporangia, and central axis of male cones were separated with a knife and a microspatula and immediately soaked in grinding medium (0.4 M mannitol, 25 mM MOPS-KOH, 2 mM EDTA, 10 mm KH2PO4, 1% [w/v] PVP-40, 20 mM ascorbic acid, 4 mM Cys, 2 mM pyruvate, 1% [w/v] bovine serum albumin [fatty acid free], 2% [w/v] poly[vinylpolypyrrolidone]; pH 7.2). Microsporophylls and central axis tissues were then disrupted in a Mixer HR-9391 (EÜPA) for 10 s. Microsporangia were ground in a Polytron PT 2500E (EYELA) for 5 s. These processes were performed at 4°C. Mitochondria were successfully isolated from microsporophylls and microsporangia, but not from the central axes.
Oxygen Consumption Assay
Oxygen consumption was monitored using an Oxytherm-1 oxygen electrode (Hansatech) at 25°C. Mitochondria (300 µg) were added to 1.2 mL assay buffer (0.3 M sucrose, 20 mM MOPS, 1 mM MgCl2/6H2O, 10 mM KCl, 5 mM KH2PO4, 0.1% (w/v) bovine serum albumin; pH 7.2). Respiratory substrates NADH (1 mm), succinate (5 mm), or a mixture of pyruvate (10 mm) and malate (10 mm) were then added to the assay buffer. When a mixture of Glu (10 mm) and malate (10 mm) was used as the respiratory substrate, the assay buffer was prepared to pH 6.8 (Jacoby et al., 2015). After the addition of the substrate, ADP (0.5 mm), KCN (1 mm), and n-propyl gallate (0.1 mm) were added to the assay buffer, in this order, as previously described (Ito-Inaba et al., 2009a). When succinate was used as the respiratory substrate, ATP (0.1 mm) was included in the assay buffer to activate succinate dehydrogenase. Prior to the addition of pyruvate + malate or Glu + malate, a cofactor mix that included NAD+ (2 mm), CoA (12 μm), and thiamine pyrophosphate (200 μm) was added. A mixture of pyruvate (10 mm) and dithiothreitol (5 mm) was added after the addition of KCN (1 mm) to analyze the cycad AOX sensitivity to pyruvate.
Measurements of COX Activity and Outer Membrane Integrity
Cytochrome c oxidase (COX) activity was measured using two methods: ascorbate-N,N,N′,N′-tetramethyl-p-phenylenediamine (TMPD) assay (Jacoby et al., 2015) and CYTOCOX1 assay kit (Sigma-Aldrich). In the ascorbate-TMPD assay, mitochondria (300 μg) were added to 1.2 mL assay buffer (the same buffer used in the oxygen consumption assay; pH 7.2), and ascorbate and TMPD were sequentially added. The CYTOCOX1 assay was carried out according to the manufacturer’s instructions. Outer membrane integrity was measured as previously described (Ito-Inaba et al., 2008a).
SDS-PAGE and Immunoblotting
The SDS-PAGE and immunoblotting procedures were conducted as described by Ito-Inaba et al. (2008a).
Antibodies
Polyclonal antibodies against AOX (Ito-Inaba et al., 2009a) and against UCP (Ito-Inaba et al., 2008b) were prepared as previously described. Anti-HSP60 (Stressgen) antibodies were purchased.
Isolation and Sequencing of CrAOX1 and CrAOX2 cDNAs
To isolate CrAOX transcripts, total RNA was first extracted from male cones using the RNeasy Mini Kit (QIAGEN). First-strand cDNAs were generated with a PrimeScript RT Reagent Kit (Takara) using oligo(dT) primers and random hexamers. Reverse transcription PCR was successively performed using the degenerate primer pairs, P1 and P2, which were used to amplify Arabidopsis (Arabidopsis thaliana) AOX genes (Saisho et al., 1997), and 42AOX-F and 42AOX-R, which were used to amplify Pinus pinea AOX genes (Frederico et al., 2009). The obtained fragments were cloned into the EcoRV site of pZErO-2 vector (Invitrogen) and sequenced. The two partial cDNA fragments obtained had high levels of sequence similarity to plant AOX1 and AOX2, respectively.
To isolate CrAOX1 cDNA, primers for 3′-RACE reactions were designed based on the partial sequences derived from fragments amplified using P1 and P2 primers; 3′-RACE reactions were performed using the RNA Ligase-Mediated kit (Applied Biosystems) and the primers CrAOX-3′RACE-OP1 for the first PCR and CrAOX-3′RACE-IP1 for the second PCR. Obtained products were cloned into the EcoRV site of pZErO-2 vector (Invitrogen) and sequenced. Unfortunately, 5′-RACE reactions were unsuccessful. To extend the 5′-cDNA ends of the obtained partial fragments, conserved nucleotide sequences were identified by aligning the following AOX1 sequences: AtAOX1a (Arabidopsis AOX1a; NM_113135); PgAOX_s1 (Picea glauca AOX subtype 1; BT119347); PgAOX_s2 (P. glauca AOX subtype 2; BT104039); and PsAOX_s2 (Picea sitchensis AOX subtype 2; EF084004). Based on conserved portions of these sequences, a forward primer (CrAOX-degF3) was designed and used to amplify cDNA fragments together with primer CrAOX-R1N. The obtained fragments were cloned into the EcoRV site of pZErO-2 vector (Invitrogen), resulting in pZAOX1, and sequenced.
To isolate CrAOX2 cDNA, primers for 3′-RACE reactions were designed as described above and reactions were performed using the RNA Ligase-Mediated RACE kit (Applied Biosystems) and the primers CrAOX2-3′RACE-OP1 for the first PCR and CrAOX2-3′RACE-IP1 for the second PCR. The products obtained were treated by ExoSAP (Thermo Fisher Scientific) and directly sequenced. The 5′-RACE reactions were unsuccessful. The 5′-cDNA ends of the partial fragments obtained were, therefore, also extended based on the alignment of AOX2 sequences from several plants: AtAOX2 (Arabidopsis AOX2; NP_201226); NnAOX2d (Nelumbo nucifera AOX2d; gi|719971362); GmAOX2a and GmAOX2d (Glycine max AOX2a [U87906] and AOX2d [U87907], respectively); ClAOX2 (Citrullus lanatus AOX2; ADD84880); VuAOX2a and VuAOX2d (Vigna unguiculata AOX2a [AJ319899] and AOX2d [AJ421015], respectively); and DcAOX2a and DcAOX2b (Daucus carota AOX2a [EU286575] and AOX2b [EU286576], respectively). Based on the conserved portions of these sequences, the forward primer CrAOX2-degF2 was designed and used to amplify cDNA fragments together with the primer CrAOX2-SP-R2. The obtained fragments were cloned into the EcoRV site of pZErO-2 vector (Invitrogen), resulting in pZAOX2, and sequenced. The primers designed in the current study are listed in Supplemental Table S1.
Molecular Phylogenetic Analysis of AOX Proteins
Amino acid sequences of the AOX proteins were retrieved from the DNA Data Bank of Japan/EMBL/GenBank and RefSeq databases. In addition, short-read data from the Sequence Read Archive (SRA; available at the National Center for Biotechnology Information; https://www.ncbi.nlm.nih.gov/sra) were used to retrieve AOX genes from magnoliids and basal Magnoliophyta. The SRA data were assembled into contigs using the CAP3 Sequence Assembly Program (http://doua.prabi.fr/software/cap3). Deduced cDNAs from the SRA data were translated into amino acid sequences using the translate tool at the ExPASy web server (http://web.expasy.org/translate/). All deduced proteins were checked against the AOX protein sequences using the BLAST BLASTp, available at the National Center for Biotechnology Information. The phylogenetic analysis was performed in MEGA version 7.0.14 (Kumar et al., 2016) using 146 residues within the conserved regions of AOX proteins, which corresponded to positions 181 to 326 of SrAOX (AB183695). Sixty-five AOX protein sequences were aligned using CLUSTAL W (Larkin et al., 2007), and the phylogenetic tree was constructed using the neighbor-joining method; bootstrap values presented on the consensus tree were based on 1,000 replications.
Expression Analyses of CrAOX1 and CrAOX2 Transcripts by Real-Time PCR
Total RNA was extracted from microsporangia and microsporophylls using the RNeasy Mini Kit (QIAGEN). First-strand cDNAs were generated with a PrimeScript RT Reagent Kit (Takara) using oligo(dT) primers and random hexamers. Real-time PCR was performed on a Thermal Cycler Dice Real-Time System (Takara) using TB Green Premix ExTaq II (Takara). All the real-time PCR analyses were performed using biological quadruple samples. Primers used for real-time PCR are listed in Supplemental Table S1. Transcript levels of each gene were normalized to that of CrEF1γ.
Statistical Analysis
Data were statistically analyzed by Student’s t test or Turkey-Kramer multiple comparison test, as described in the figure legends for each experiment.
Accession Numbers
The nucleotide sequences reported in this paper were submitted to the DNA Data Bank of Japan under accession numbers LC081345 for CrAOX1, LC229699 for CrAOX2, and LC081346 for CrUCP.
Supplemental Data
The following supplemental materials are available.
Supplemental Figure S1. Assay of cytochrome pathway respiration and COX (Complex IV) rate in microsporophyll (ML) and microsporangia (MG) mitochondria.
Supplemental Figure S2. Oxygen uptake measured in microsporophylls (left) and microsporangia (right) mitochondria using various respiratory substrates.
Supplemental Figure S3. Expression of AOX proteins in the microsporophylls and microsporangia mitochondria.
Supplemental Figure S4. Deduced amino acid sequences of CrAOX1 and CrAOX2 aligned with trypanosomal AOX (XP_822944) and AOX isoforms in Arabidopsis and other gymnosperms.
Supplemental Figure S5. Oxygen uptake rate in intact mitochondria isolated from microsporophylls.
Supplemental Figure S6. Deduced amino acid sequences of CrUCP aligned with UCP isoforms in other gymnosperms and thermogenic skunk cabbage.
Supplemental Figure S7. Phylogenic relationships of CrUCP with UCP isoforms in other plants.
Supplemental Table S1. Gene-specific primers used in PCR.
Supplemental Table S2. List of AOX genes used in phylogenetic analysis.
Supplemental Table S3. List of UCP genes used in phylogenetic analysis.
Acknowledgments
We thank Mayumi Wakazaki and Kei Hashimoto (RIKEN Center for Sustainable Resource Science) for their assistance in experiments with electron microscopy. We also thank Yoko Katayama and Saori Hamada (University of Miyazaki) for their assistance in molecular biological experiments. We are grateful to Takehiko Ito (Tokyo Institute of Technology) for valuable advice.
Footnotes
- Received February 4, 2019.
- Accepted March 12, 2019.
- Published March 27, 2019.