, Volume 250, Issue 1, pp 317–323

An antibody against a conserved C-terminal consensus motif from plant alternative oxidase (AOX) isoforms 1 and 2 label plastids in the explosive dwarf mistletoe (Arceuthobium americanum, Santalaceae) fruit exocarp


    • Department of Biological SciencesThompson Rivers University
  • Bradford N. Ross
    • BioImaging FacilityUniversity of British Columbia
  • Garnet D. Martens
    • BioImaging FacilityUniversity of British Columbia
Original Article

DOI: 10.1007/s00709-012-0414-6

Cite this article as:
Ross Friedman, C., Ross, B.N. & Martens, G.D. Protoplasma (2013) 250: 317. doi:10.1007/s00709-012-0414-6


Dwarf mistletoes, genus Arceuthobium (Santalaceae), are parasitic angiosperms that spread their seeds by an explosive process. As gentle heating triggers discharge in the lab, we wondered if thermogenesis (endogenous heat production) is associated with dispersal. Thermogenesis occurs in many plants and is enabled by mitochondrial alternative oxidase (AOX) activity. The purpose of this study was to probe Arceuthobium americanum fruit (including seed tissues) collected over a 10-week period with an anti-AOX antibody/gold-labeled secondary antibody to determine if AOX could be localized in situ, and if so, quantitatively assess whether label distribution changed during development; immunochemical results were evaluated with Western blotting. No label could be detected in the mitochondria of any fruit or seed tissue, but was observed in fruit exocarp plastids of samples collected in the last 2 weeks of study; plastids collected in week 10 had significantly more label than week 9 (p = 0.002). Western blotting of whole fruit and mitochondrial proteins revealed a signal at 30–36 kD, suggestive of AOX, while blots of whole fruit (but not mitochondrial fraction) proteins showed a second band at 40–45 kD, in agreement with plastid terminal oxidases (PTOXs). AOX enzymes are likely present in the A. americanum fruit, even though they were not labeled in mitochondria. The results strongly indicate that the anti-AOX antibody was labeling PTOX in plastids, probably at a C-terminal region conserved in both enzymes. PTOX in plastids may be involved in fruit ripening, although a role for PTOX in thermogenesis cannot be eliminated.


Alternative oxidase (AOX)Arceuthobium americanumDwarf mistletoeFruitImmunocytochemistryPlastid terminal oxidase (PTOX)


Dwarf mistletoes, genus Arceuthobium (Santalaceae), are parasites of the coniferous families Pinaceae and Cupressaceae, and serious forest pests that reduce the growth, yield, and timber quality of the host trees (Hawksworth and Wiens 1996). Arceuthobium americanum Nutt. ex Englm., the lodgepole pine dwarf mistletoe, parasitizes Pinus contorta Dougl. ex Loud. (lodgepole pine) and Pinus banksiana Lamb. (jack pine), and is a major problem in western North American coniferous forests, causing an annual loss of several billion dollars (Ip 1992).

Arceuthobium seed dispersal occurs almost exclusively by a remarkable explosive process (Hawksworth and Wiens 1996). A mucilaginous fruit tissue called “viscin,” essentially a modified mesocarp, imbibes water during fruit development, generating a hydrostatic force that ultimately results in forcible expulsion of the seed at the end of the growing season, which, for A. americanum, is late August to early September (Ross 2006). Hawksworth (1959) as well as Hinds and Hawksworth (1965) were the first to document the discharge process in a laboratory setting, inducing seed ejection by gently heating the seeds from below. The observation that heat can induce discharge leads to an intriguing question: is heat naturally involved in the discharge process? While exogenous heat could possibly be provided by peaks in ambient temperature and/or surface warming of fruit in direct sunlight, perhaps endogenous thermogenesis has a role in explosive seed discharge.

Many plants are thermogenic, possessing the ability to raise their temperature above that of the surrounding air. To date, thermogenesis has been reported in flowers, inflorescences, and cones in several plant families including Araceae (Seymour et al. 2003, 2004; Seymour 2004; Wagner et al. 2008; Kamata et al. 2009), Cycadaceae (Tang 1987), Nelumbonaceae (Seymour and Schultze-Motel 1998; Watling et al. 2006), and Nymphaeaceae (Skubatz et al. 1990). In these plants, heat is generated in the mitochondria as a secondary process of cellular respiration and is enabled by electron transport chain enzymes called “alternative oxidases” (AOXs), which are located in plant mitochondrial cristae.

It is challenging to precisely measure or otherwise gauge the internal temperature of a mature dwarf mistletoe fruit, which with an average diameter of 2 mm, is quite small. Thus, as a first step toward assessing whether or not thermogenesis might be involved in Arceuthobium seed dispersal, we decided we would use an anti-plant AOX antibody in an immunocytochemical study along with Western blot analysis to determine if AOXs are present in the mitochondria (or other organelles) of A. americanum fruit or seed tissues, and if so, establish whether these AOXs become more numerous as explosive discharge becomes imminent. From such circumstantial evidence, we would pursue the question of thermogenesis more vigorously in subsequent research.

In this paper, we report the results of our transmission electron microscopy (TEM) study in which developing and mature A. americanum fruit were probed with an antibody against a C-terminal consensus motif from plant alternative oxidase (AOX) isoforms 1 and 2. We also show the results of our Western blot analyses on fruit proteins as well as proteins from isolated fruit mitochondria.

Materials and methods

Collection and general processing for transmission electron microscopy

Approximately five developing A. americanum fruits were collected once a week for 10 weeks (5 June 2007 to 15 August 2007) from a site containing infected lodgepole pines near Kamloops (Stake Lake, British Columbia, Canada). This 10-week period coincides with the period in which the viscin tissue and seed, particularly the embryo, undergo the bulk of their development (Ross 2006). At this site, about 10 additional fruits collected in the last week (week 10, i.e., 15 August) were set aside, while all remaining fruits were placed in a fixative consisting of 2 % paraformaldehyde and 2 % glutaraldehyde in 0.1 m phosphate buffer (pH 6.8). The material was taken immediately to the laboratory: fruits that had been set aside were frozen at −20 °C and later used in Western blotting, while fruit material in fixative was kept at 4 °C overnight, dehydrated in an ethanol series, and embedded in LR White medium grade resin. Ultrathin sections were obtained with a Reichert Ultracut E ultramicrotome (Reichert-Jung, Buffalo, NY, USA) and adhered to the dull side of formvar-coated nickel hex grids. From each fruit, approximately 30 sections were obtained, 10 containing true fruit (maternal) tissues (including the exocarp and viscin), 10 containing embryonic tissues, and 10 containing endosperm tissues; 5 of these sections from each type of tissue were distributed over two grids for further immunolabeling, and 5 sections per tissue type were distributed to two control grids.

Immunogold labeling with anti-AOX antibody

A lyophilized polyclonal antibody, anti-AOX, raised in rabbit against a synthetic peptide derived from a fully conserved C-terminal consensus motif from plant AOX isoforms 1 and 2 including Arabidopsis thaliana AOX1A (At3g22370) and AOX2 (At5g64210), Solanum lycopersicum Q7XBG9, and Oryza sativa Q7XT33 was purchased from Agrisera (Agrisera, Vännäs, Sweden). Immunolabeling with anti-AOX was facilitated by a Leica EM IGL Automated immunogold labeling system (Leica, Richmond Hill, Canada) at the University of British Columbia’s BioImaging Facility. All staining was performed at room temperature. Grids containing sectioned tissues were treated with 1 % bovine serum albumin in phosphate buffer, 0.1 m, pH 7.4, for 30 min, and then incubated with 1:50 diluted anti-AOX for 60 min. Sections were rinsed three times in phosphate buffer containing bovine serum albumin (10 min each time), and then incubated with 1:100 diluted 10-nm colloidal gold goat anti-rabbit serum (Pelco International, Redding, USA) for 60 min. Following immunolabeling, grids were routinely stained with 2 % aqueous uranyl acetate for 20 min and counterstained with Sato’s lead for contrast. Immunolabeling controls involving omission of the primary antibody were employed. Initial observations of sections were qualitative: the general binding pattern and distribution of gold label in fruit tissues (exocarp and viscin), embryo, and endosperm were noted. The sections were digitally captured with a side mounted 2K AMT camera on a Hitachi model H-7600 TEM (Hitachi, Tokyo, Japan) at an operating voltage of 80 kV, and further subjected to quantification of gold labeling. The tonal qualities of the images were adjusted, labels added, and plates assembled with Adobe Photoshop CS3.

Quantification of gold labeling

Gold labeling in 25 randomly selected electron micrographs from the approximately 25 immunolabeled treatment sections per tissue type obtained each week (average 5 fruit × 5 treatment sections per tissue type) over the 10-week period (5 June through 15 August) was evaluated. Preliminary results were clear and guided the quantification process: essentially no gold label was present in any tissue or organelle except for plastids of the fruit exocarp, and those gold particles in plastids were only evident in samples obtained in the ninth and tenth weeks of the collection period (i.e., week 9 and week 10).

While it was easy to discern that exocarp plastids from week 8 and earlier were devoid of label, quantification and statistical analysis was required to determine if there was a statistically significant difference in the amount of label in plastids from fruit collected in week 9 compared with week 10. For each micrograph containing a gold-labeled plastid in median or near median section, the public domain NIH Image software program 1.61 (developed at the US National Institutes of Health and available on the Internet at was used to determine the plastid sectional area (density slice) and the total number of gold particles (particle count). From these data, the mean density of gold per unit area (100 × 100 nm or 0.01 μm2) ± standard error of the mean (SEM) was calculated for a set of 25 plastids in week 9 micrographs and a set of 25 in week 10 micrographs. A two-way, pairwise t test was used to compare the mean gold density per unit area in the samples from these 2 weeks (Microsoft Excel).

Total fruit membrane protein isolation, fruit mitochondrial protein isolation, and Western blots

To evaluate any positive immunolabeling observed in tissue sections, Western blots of all fruit membrane proteins as well as proteins from isolated fruit mitochondria were performed. A signal of about 30 to 35 kD would indicate that the primary anti-AOX antibody was likely recognizing an AOX homolog in the dwarf mistletoe, as the reduced protein when run on SDS-PAGE has a molecular mass of approximately 30 to 35 kD; this holds true for AOX from a number of plant species (Umbach and Siedow 1993; Day and Wiskich 1995; Hoefnagel and Wiskich 1998).

In order to carry out a Western blot on all fruit membrane proteins, we employed the method Ross Friedman et al. (2010) used to extract all membrane proteins from dwarf mistletoe fruit, which is in turn a method adapted from Bots et al. (2005). About 10 dwarf mistletoe fruits (30 g) that had been collected in week 10 (15 August) and stored at −80 °C were homogenized in a buffer consisting of a plant protease inhibitor cocktail (Sigma-Aldrich, Canada), 330 mM sucrose, 100 mM KCl, 1 mM EDTA, 50 mM TRIS, 0.05 % MES, and 5 mM DTT (pH 7.5). The homogenate was centrifuged at 1,000×g for 15 min to collect debris, and the supernatant was then centrifuged at 20,000×g for 75 min to pellet all membrane material, including mitochondrial (and plastid) membranes. The membrane pellet was dissolved in membrane buffer containing 330 mM sucrose, 200 mM DTT, and 25 mM TRIS, pH 8.5.

Differential centrifugation and non-linear sucrose ultracentrifugation as described by Antonsson et al. (2001) and Rezaul et al. (2005) were used to isolate mitochondria from dwarf mistletoe fruit in order to perform Western blotting on their mitochondrial proteins in anticipation of identifying an AOX specific to fruit mitochondria. Another 10 dwarf mistletoe fruit (30 g) that had been collected in week 10 (15 August) and stored at −80 °C were homogenized in a buffer containing the plant protease inhibitor cocktail from Sigma-Aldrich, 250 mM sucrose, 10 mM KCl, 20 mM Hepes–NaOH, 1.5 mM MgCl2, and 1 mM DTT (pH 7.5). The lysate was centrifuged at 1,000×g at 4 °C for 10 min. The resulting supernatant was centrifuged at 10,000×g at 4 °C for 25 min. The pellet, which should represent a crude preparation of mitochondria, was resuspended in a buffer containing the protease inhibitors, 250 mM sucrose, and 10 mM Tris–HCl (pH 7.5), and loaded onto a discontinuous sucrose gradient (9 ml 0.9 M sucrose, 10 ml 1.6 M sucrose, and 9 ml of 1.8 M sucrose plus 50 mM Tris–HCl 150 mM NaCl and 1 mM DTT, pH 7.5). After ultracentrifugation at 60,000 g for 20 min at 4 °C, the light-brown mitochondrial band located at the interface between 0.9 and 1.6 M sucrose was collected, washed, and resuspended in buffer containing the protease inhibitors, 20 mM Tris–HCl, and 150 mM NaCl (pH 7.5).

For Western blots of total fruit membrane proteins or those specific to the fruit mitochondrial membranes, protein samples were prepared in the presence of 50 mM ethanedithiol, a reducing agent. A set of standard molecular weight markers ranging from 6 to 205 kD with colors that transfer (Sigma-Aldrich) and 15 μg of dwarf mistletoe fruit proteins (total or mitochondrial) were separated by 12 % SDS-PAGE. The proteins were electroblotted onto a polyvinylidene difluoride (PVDF) membrane in 39 mM glycine, 48 mM TRIS base, 0.037 % SDS, and 20 % methanol (pH 8.3). For detection, the PVDF membranes was first blocked in blocking buffer (5 % non-fat dried milk in PBT) for 2 h, after which they were incubated for 2 h with the anti-AOX antibody (Agrisera) at 1:500 in blocking buffer. After incubation with the primary antibody, the PVDF membranes were rinsed and incubated with the secondary antibody, antirabbit IgG-horseradish peroxidase (Sigma-Aldrich) at 1:20,000 in blocking buffer for 1 h. All incubations were performed at room temperature. The PVDF membranes were developed using tetramethylbenzidine (TMB) Liquid Substrate (Sigma-Aldrich) according to the manufacturer’s protocol, and digitally captured with a BioRad Gel Doc system (BioRad, Canada). Adobe Photoshop CS3 was used to adjust contrast and brightness of the images, which were converted to grayscale.


Immunogold labeling

No gold label could be detected in the mitochondria of cells comprising A. americanum exocarp, viscin, embryo, or endosperm collected over the 10-week period (Fig. 1a). Gold particles were similarly absent from plastids in all tissues collected over the 10-week period (Fig. 1b), aside from cells of the exocarp collected in weeks 9 and 10, where plastids with poorly developed thylakoids and no plastoglobuli displayed obvious gold labeling (Figs. 1c and d). Fully developed chloroplasts with copious thylakoids, discrete grana, and prominent plastoglobuli, however, remained unlabeled (Fig. 1a). The average density of gold particles in week 9 exocarp plastids was 5.2 ± 0.9 particles per 0.01 μm2, while that of week 10 exocarp plastids was 11.3 ± 1.1 particles per 0.01 μm2; a two-way, pairwise t test revealed that these values are significantly different (p = 0.002). Control sections remained unlabeled, aside from the odd isolated particle.
Fig. 1

Representative transmission electron micrographs of Arceuthobium americanum fruit exocarp sections subjected to immunochemical staining with an anti-alternative oxidase (anti-AOX) antibody (Agrisera). a Mitochondria (m) and a mature chloroplast (ch) containing well-developed thylakoids/grana (arrow) as well as prominent plastoglobuli (arrowheads) in a fruit sampled later in the growing season (8 August 2007; i.e., week 9). No gold label is evident. b Plastid (p) with a few thylakoids (arrow) in a fruit sampled earlier in the growing season (5 June 2007; i.e., week 1). No gold label is evident. c Plastid (p) with a few thylakoids (arrow) in a fruit sampled later in the growing season, as in a (8 August 2007; i.e., week 9). Gold labeling is evident (arrowheads). d Plastid (p) with a few thylakoids (arrow) in a fruit sampled at the end of the collection period (15 August 2007; i.e., week 10). Gold labeling is intense (arrowheads). c cytoplasm. Bara 300 nm; b, c 100 nm; d 150 nm

Western blots

Western blot analysis of total fruit membrane proteins yielded two signals; a strong one at approximately 40–45 kD as well as a slightly weaker one at approximately 30–36 kD (Fig. 2). In the Western blot of the mitochondrial fraction, only the one smaller 30–36 kD band was present (Fig. 3), which suggests that the 40–45 kD band is in the plastid fraction as opposed to representing an AOX heterodimer. This immunowork was carried out with a polyclonal antiserum raised against a conserved peptide, and so there are limitations regarding the conclusions regarding the link between any AOX and cellular signal, but the Western blot clearly shows the disappearance of the larger band from the mitochondrial fraction.
Fig. 2

Western blot of total membrane proteins from fruit collected on 15 August 2007 (week 10) using the anti-AOX antibody. The first lane shows standard molecular weight markers in kD. In the second lane (labeled “Fruit”), the primary antibody detected two proteins: one of approximately 40–45 kD (stronger band) and one at approximately 30–36 kD (weaker signal)
Fig. 3

Western blot of mitochondrial proteins from fruit collected on 15 August 2007 (week 10) using the anti-AOX antibody. The first lane shows standard molecular weight markers in kD. In the second lane (labeled “Mito”), the primary antibody detected only one protein of approximately 30–36 kD


The anti-AOX antibody failed to label mitochondria in any cells of the exocarp, viscin, embryo, or endosperm in A. americanum fruit collected from 5 June through 15 August 2007. However, the 30–36 kD signal on the two types of Western blots suggests that AOX proteins do exist in the A. americanum fruit, as reduced AOX proteins from other plant species have a molecular mass of approximately 30 to 35 kD when run on SDS-PAGE (Umbach and Siedow 1993; Day and Wiskich 1995; Hoefnagel and Wiskich 1998). Likewise, the presence of a single 30–36 kD band from the mitochondrial protein fraction suggests that the AOX protein is present in fruit mitochondria. Future work can focus on verifying the identity of this band (e.g., depletion of the antiserum by fixed cells, outcompeting of the bands and immunolocalization signal with the peptides, etc.).

It is difficult to determine why the signal could not be detected in A. americanum fruit in situ, although perhaps the C-terminal epitope of the native or oxidized form of a putative AOX in A. americanum cannot be readily accessed by the antibody, which was raised against a fully conserved C-terminal consensus motif from plant AOX isoforms 1 and 2. Andersson and Nordlund (1999) have proposed that plant AOX may be interfacial rather than true transmembrane proteins, and such a conformation might render the C-terminal epitope unavailable. Additionally, the glutaraldehyde fixative in combination with the mitochondrial environment could have possibly impeded binding. Indeed, there are some limitations in this study, primarily due to the use of a polyclonal antiserum raised against a conserved peptide in the immunolocalization process. More immunocytochemical work is needed to potentially detect AOX in mitochondria, perhaps with even older A. americanum fruit in which explosive discharge is actively ongoing, and, as mentioned in the previous paragraph, the identity of the smaller band can be verified by a variety of approaches. The possibility that mitochondrial AOX are involved in thermogenesis and explosive discharge in A. americanum cannot be ruled out.

Anti-AOX antibody labeling of “non-chloroplast” Arceuthobium plastids was unexpected. However, plastids can contain “plastid terminal oxidases” (PTOX), which are plastoquinol oxidases located in thylakoid membranes (Cournac et al. 2000; Shahbazi et al. 2007). The identity between PTOX and AOX polypeptides is approximately 25 % (Josse et al. 2000). Alignments of plant PTOX sequences have revealed that the C-terminal domain contains a conserved motif that matches a putative iron-binding site also conserved in plant AOXs. As the antibody used in this study had been raised against a C-terminal consensus motif from plant AOXs, it is possible that the anti-AOX antibody was localizing a PTOX in A. americanum fruit “non-chloroplast” plastids. The C-terminal epitope of an A. americanum PTOX in plastids might simply be more accessible to the antibody than that of an AOX. The presence of a 40–45-kD signal on the Western blot provides further evidence that the antibody was indeed localizing PTOX in plastids, as reduced PTOX in produce a band of 41 kD in, e.g., red pepper, Capsicum annuum L. (Josse et al. 2000). Furthermore, the disappearance of the 40–45 kD band from the mitochondrial fraction suggests that this larger band was indeed in the plastid fraction. It would be useful to perform Western blots on fruit collected earlier and later in the growing season, particularly to determine if the PTOX can be detected on gels when immunocytochemical results are negative. Similarly, if AOX is involved in thermogenesis to trigger explosive discharge, the amount of AOX in tissues should increase considerably, since the expression of AOX during typical fruit development and in non-thermogenic tissues would likely not be adequate to generate heat.

PTOX are proposed to have roles in chlororespiration when found in green chloroplasts, and in carotenoid pigment biosynthesis, particularly when found in chromoplasts (Josse et al. 2000; Carol and Kuntz 2001; Shahbazi et al. 2007). PTOX in chromoplasts become more active throughout fruit maturation in many plants, and the chromoplasts themselves become more numerous during fruit ripening (Josse et al. 2000). Visibly, A. americanum fruits do become slightly more yellow upon ripening, which suggests an increased presence of carotenoids. Moreover, the lack of well-developed thylakoids in the gold-labeled A. americanum plastids coupled with the fact these labeled plastids were only found in the fruit exocarp indicate that the plastids were probably chromoplasts, and that their PTOX activity might be involved with fruit ripening and color change. The significantly higher degree of gold labeling in week 10 plastids compared with week 9 plastids is likely correlated with increased PTOX activity, although more conclusive evidence requires an expression study. Even though chromoplasts have less obvious thlyakoid, thylakoids are still present, as thus could easily contain PTOX.

The potential function of these putative A. americanum PTOX needs further investigation, though. Certainly a role in fruit maturation and ripening through color change cannot be eliminated. However, the phenomenon of a slight color change in fruit that forcibly discharges its seeds is a little perplexing, though, as color change during ripening is thought to be a visual cue for seed dispersers (Van der Pijl 1966). The slight color change in A. americanum fruit might be a vestigial character, perhaps from an ancestor that was animal-dispersed, as endozoochory is not believed to occur in A. americanum (Hawksworth and Wiens 1996). A very interesting question thus arises from this study: could PTOX play a role in thermogenesis? No evidence exists suggesting this is true in any plants, but this question is certainly worth further pursuit, especially as a fruit ripening/color change role for PTOX in A. americanum seems limited.

In conclusion, while an anti-AOX antibody did not label AOX in the mitochondria of developing A. americanum fruit, the antibody did bind to plastids (likely chromoplasts) in the fruit exocarp, and was likely labeling PTOX at its C-terminal region, which shares homology with the C-terminal region of AOX. Western blot analysis strongly indicated that AOX proteins were present in the mitochondria, and that the antibody was likely binding to a plastid-fraction protein, most probably PTOX. The plastid labeling only became evident in fruit collected in the final 2 weeks of our study, suggesting that PTOX units become more numerous as the fruit matures. The function of a putative PTOX in the A. americanum fruit remains obscure, but these enzymes in this species might be involved in a color change related to fruit ripening, and could conceivably have a role analogous to the thermogenic capacity normally attributed to AOX.


This work was supported by a Natural Sciences and Engineering Research Council of Canada Discovery Grant [grant number 164375 provided to CRF]. We are very grateful to Ronald G. Smith of Thompson Rivers University for postulating that AOXs might be involved in thermogenesis-driven explosive seed discharge in the dwarf mistletoes. We also thank two anonymous reviewers who helped us improve the quality of the original study.

Conflict of interest

The authors declare that they have no conflict of interest.

Copyright information

© Springer-Verlag 2012