Variegation mutants provide an excellent system to explore the mechanisms of chloroplast biogenesis and the nature of communication between the nucleus-cytoplasm, chloroplast, and mitochondrion (Rodermel, 2001; Yu et al., 2007). The leaves of these mutants contain green and chlorotic (white, yellow) sectors that arise as a consequence of mutations in nuclear or organelle genes (Tilney-Bassett, 1975). Whereas cells in the green sectors have morphologically normal chloroplasts, cells in the chlorotic sectors contain pigment-deficient plastids with abnormal membrane structures (Aluru et al., 2001). One common mechanism of variegation involves the induction of defective chloroplasts (or mitochondria) by mutations in nuclear genes that code for organelle proteins. In some cases the green and white cells have different genotypes. For example, transposable element activity can inactivate a gene required for normal chloroplast function in some cells (chlorotic cells) while excision can reconstitute gene function (green cells). In other cases of variegation, the two types of cells have the same (mutant) genotype, indicating that the gene product is required in some, but not all cells of the mutant. This raises the question of how the green cells escape this requirement. Is there a compensating activity? Our operating hypothesis is that answers to this question will provide insight into mechanisms of variegation, as well as into pathways and regulatory networks of chloroplast biogenesis.

The past decade has witnessed a growing interest in the use of variegation mutants to address fundamental questions in photosynthesis and chloroplast biogenesis (Yu et al., 2007; Liu et al., 2010). Yet, mechanisms of variegation are still poorly understood. Here we focus on the mechanism of variegation in immutans (im), which defines the gene for plastid terminal oxidase (PTOX), a versatile quinol oxidase in plastid membranes. Early studies led to the notion that this mechanism involves the attainment of a threshold by developing plastids (the threshold hypothesis). In this manuscript, we review the underlying assumptions and experiments in support of this hypothesis and discuss approaches and studies that have been used to gain insight into factors that compensate for a lack of PTOX.

The im variegation mutant is one of the oldest Arabidopsis mutants, reported independently by Rédei and Röbbelen in the 1960’s (Rédei, 1963; Röbbelen, 1968). Cells in the green sectors of im have morphologically normal chloroplasts, whereas cells in the white sectors are heteroplastidic and contain abnormal plastids that lack pigments and organized lamellae, as well as rare, normal-appearing chloroplasts (Wetzel et al., 1994). The extent of white sector formation in im is promoted by high light and low temperature (Rédei, 1963, 1967; Röbbelen, 1968; Wetzel et al., 1994; Rosso et al., 2009). HPLC analyses showed that the white sectors accumulate the colorless C₄₀ carotenoid intermediate, phytoene, indicating that im is impaired in the activity of phytoene desaturase (PDS), the enzyme which converts phytoene to zeta-carotene (Wetzel et al., 1994). All the steps of carotenogenesis take place in the plastid and are mediated by nuclear-encoded enzymes that are imported into the organelle post-translationally, and PDS mediates an early step of the pathway (DellaPenna and Pogson, 2006). An inhibition of PDS activity would thus result in lack of accumulation of downstream, colored (photoprotective) carotenoids, and under high light conditions, would be anticipated to give rise to white, photooxidized plastids. A defect of this sort is consistent with the light-sensitivity of im; as in other carotenoid mutants, the higher the light intensity, the greater the extent of photooxidation (white sector formation) (Oelmüller, 1989; Mayfield and Taylor, 2005). The presence of normal-appearing chloroplasts in im, on the other hand, suggests that some plastids are able to bypass the requirement for the IM gene product, likely because of compensating activities that make them less susceptible to photooxidation early in their development.

IMMUTANS was cloned in Arabidopsis by both map-based methods (Wu et al., 1999) and T-DNA tagging (Carol et al., 1999). The gene product (IM) was discovered to be a plastid membrane protein that is distantly related (37% amino acid similarity) to alternative oxidase (AOX), a mitochondrial inner membrane protein which functions in the alternative (cyanide-resistant) pathway of respiration, where it transfers electrons from ubiquinol to molecular oxygen (Wu et al., 1999). Central among its physiological functions, AOX is an important sensor of cellular redox balance (Giraud et al., 2008; McDonald, 2008). Similar to AOX, IM has quinol oxidase activity in vivo and in vitro and, consequently, it has been designated PTOX (Joët et al., 2002; Josse et al., 2003). PTOX is found in some cyanophages, which might serve as vectors for transfer of PTOX among cyanobacteria, but otherwise appears to be limited to oxygenic photosynthetic prokaryotes and eukaryotes, where it is found in all lineages (McDonald et al., 2011). IM is generally present as a single copy gene, although two copies are found in some cyanobacteria, red algae and green algae (Wang et al., 2009; Houille-Vernes et al., 2011). In chloroplasts, PTOX is bound to the stromal lamellae of thylakoids and is modeled as an interfacial membrane protein whose active site faces the stroma (Berthold et al., 2000; Joët et al., 2002; Lennon et al., 2003). It does not appear to be present in chloroplast envelope membranes.

Several functions have been ascribed to PTOX. The phytoene-accumulation phenotype of im led to the suggestion that PTOX serves as the terminal oxidase of an oxygen-dependent redox pathway that desaturates phytoene (Beyer et al., 1989; Mayer et al., 1990, 1992; Hugueney et al., 1992; Schulz et al., 1993; Nievelstein et al., 1995; Norris et al., 1995; Al-Babili et al., 1996). This pathway is thought to involve transfer of electrons from phytoene to plastoquinone (PQ) via PDS, forming ζ-carotene and plastoquinol (PQH₂), and from PQH₂ to molecular oxygen via PTOX, forming water and PQ (Carol et al., 1999; Wu et al., 1999; Rosso et al., 2009). Thylakoids of developing im plastids have over-reduced PQ pools (Rosso et al., 2009), and according to this pathway, the accumulation of phytoene in im can be explained by a decreased supply of PQ available to PDS, as suggested by Okegawa et al. (2010), and/or because electron transfer from phytoene into an over-reduced PQ pool is not energetically favorable (Rochaix, 2011). In addition to the phytoene-accumulation phenotype of mutants that lack PTOX (im in Arabidopsis and the orthologous ghost mutant in tomato) (Wetzel et al., 1994; Josse et al., 2000; Barr et al., 2004; Shahbazi et al., 2007), an involvement of PTOX in carotenogenesis is suggested by the close coordination between PTOX expression and carotenoid production in a number of systems, perhaps most strikingly in chromoplasts during the ripening of tomato, citrus, and pepper fruit (Josse et al., 2000; Barr et al., 2004). Another noteworthy example is the upregulation of PTOX expression in etiolated seedlings of Arabidopsis treated with paclobutrazol (PAC), an inhibitor of gibberellin biosynthesis; PAC causes an increase in carotenoid production (Rodríguez-Villalón et al., 2009).

It might be noted that all the enzymes of carotenogenesis are present in chloroplast envelopes, with the exception of PDS which is also found in thylakoids (Joyard et al., 2009). The involvement of PTOX in carotenogenesis is therefore puzzling given its apparent exclusive location in thylakoids (Lennon et al., 2003). This suggests that carotenoid intermediates are trafficked between the envelope and thylakoids, perhaps via vesicles, by connections between the envelope and thylakoids, or by carrier proteins (Adam et al., 2011; Ngaki et al., 2012). There might also be PTOX-independent mechanisms of plastoquinol oxidation in envelopes, such as PQH₂ auto-oxidation (Khorobrykh and Ivanov, 2002), the activity of another plastoquinol oxidase (Buchel and Garab, 1995; Lajko et al., 1997; Joët et al., 2002), or perhaps a plastoquinol peroxidase (Casano et al., 2000; Rumeau et al., 2007).

IM mRNAs are ubiquitously expressed in Arabidopsis, and the anatomies of various plastid types are affected in im (Aluru et al., 2001), as well as in tomato ghost (Josse et al., 2000; Barr et al., 2004). However, a positive correlation between IM mRNA levels and carotenoid accumulation does not always hold, and IM is expressed highly in some tissues with low carotenoid accumulation (Aluru et al., 2001). Considered together, these observations argue that the role of PTOX in plastid metabolism is not limited to carotenogenesis. In fact, since its discovery over 13 years ago, it has become apparent that PTOX resides at the nexus of a growing number of redox pathways in the plastid. For instance, in vitro and in vivo studies have demonstrated that PTOX is the terminal oxidase of chlororespiration (non-photochemical reduction of the PQ pool) (Peltier and Cournac, 2002; Rumeau et al., 2007). Chlororespiration involves the transfer of electrons from NAD(P)H (and/or Ferredoxin) to PQ via a thylakoid NAD(P)H dehydrogenase (NDH) complex, and from PQH₂ to molecular oxygen via PTOX (Peng et al., 2012). One purpose of chlororespiration is to help poise the redox state of the PQ pool during cyclic electron flow (CEF) around PSI (Trouillard et al., 2012). In addition to chlororespiration, PTOX mediates chromorespiration, the analogous process in chromoplasts (Josse et al., 2000; Barr et al., 2004; Rodrigo et al., 2004; Shahbazi et al., 2007).

Safety valves are photoprotective mechanisms that dissipate excess photons and electrons (Niyogi, 2000). They include non-photochemical quenching mechanisms and alternative electron acceptors that prevent over-reduction of photosynthetic electron carriers, thereby protecting PSI and PSII from photodamage (Aro et al., 1993; Allahverdiyeva et al., 2005). Soon after its discovery, Niyogi (2000) proposed that PTOX-mediated electron flow from PQH₂ to O₂ might act as a safety valve. Early support for this hypothesis came from studies showing that the green leaf sectors of im have morphological, biochemical, and molecular adaptations similar to plants acclimated to growth in high light, even when grown in permissive (low light) conditions (Aluru et al., 2001). Early studies also showed that IM mRNAs and proteins are significantly upregulated in high light conditions in antisense mutants of tobacco that lack both ascorbate peroxidase and catalase (Rizhsky et al., 2002). A number of experiments have since supported the safety valve hypothesis in systems as diverse as the alpine species Ranunculus glacialis and Soldanella alpina (Streb et al., 2005; Laureau et al., 2011); oat exposed to heat and high light (Quiles, 2006); mature green leaves and green fruit of the tomato ghost mutant subjected to high light stress (Shahbazi et al., 2007); wild species of Brassica fruticulosa versus the agricultural species Brassica oleracea adapted to both heat and high light (Díaz et al., 2007); the salt-stressed halophyte Thellungiella (Stepien and Johnson, 2009); and cold-acclimated Lodgepole pine (Savitch et al., 2010).

In addition to altered expression in response to environmental factors, PTOX is upregulated in various photosynthetic mutants, including the gun4 mutant of Chlamydomonas and the tobacco rbcL and psbA deletion mutants. The gun4 mutants lack GUN4, a regulatory subunit of Mg-chelatase, and enhanced PTOX activities in these mutants were suggested to play a physiological role in decreasing PSII excitation pressures (Formighieri et al., 2012). The rbcL mutants lack Rubisco and consequently two major electron sinks, CO₂ fixation and photorespiration; it was proposed that elevated PTOX levels in these mutants reduce oxidative pressure on both PSI and PSII (Allahverdiyeva et al., 2005). In the psbA mutants, which lack the D1 protein of PSII, it was hypothesized that enhanced PTOX levels primarily support increased carotenoid production as a way to quench singlet oxygen generated by free LHCII accumulation in the mutant (Baena-González et al., 2003). Despite the growing number of examples of PTOX upregulation, cis and trans regulatory factors have yet to be reported.

The demonstration that PTOX serves as a safety valve in a particular system is not straightforward and must rely on more than gene expression data, since there are many ways of preventing over-reduction of the electron transport chain, such as upregulation of downstream electron sinks, as found in some cold-tolerant plants, or upregulation of the Mehler reaction (e.g., Gray et al., 1997; Savitch et al., 2000); alterations in PTOX expression might be a secondary effect. There are also some cases where PTOX does not act as a safety valve. For example, overexpression of Arabidopsis PTOX does not impart increased resistance to photoinhibition in mature tobacco leaves (Joët et al., 2002), nor does it regulate the redox state of the PQ pool during steady state photosynthesis in Arabidopsis (Rosso et al., 2006). These findings are consistent with the observation that the steady state flux of electrons through PTOX is relatively minor – estimated to be <2% of the flux through the photosynthetic electron transport chain (PETC) (Ort and Baker, 2002). On the other hand, these same studies showed that PTOX is important in preventing over-reduction of PSII acceptors during dark to light transients (Joët et al., 2002), and that it protects against PSI photoinhibition (Rosso et al., 2006). The latter has been confirmed in recent experiments showing that PTOX acts as a safety valve in Arabidopsis under stress conditions of low temperature and high light, where it protects against both PSII and PSI photoinhibition (Ivanov et al., 2012).

In this context it is important to note that in contrast to PSII, where there are a number of photochemical and non-photochemical mechanisms to avoid over-reduction of PSII acceptors, the repertoire is rather more limited for PSI and includes, most prominently, the water–water cycle (reviewed in Rumeau et al., 2007). PTOX might thus turn out to be an important mechanism to reduce electron pressure on PSI, as during stress when NADPH/ATP ratios are high and CEF around PSI is activated; in concert with NDH, PTOX might serve to poise the redox state of intersystem carriers by recycling electrons to PQ. This might also be the case during chloroplast biogenesis when it has been demonstrated that PTOX is a central regulator of thylakoid redox and PSII excitation pressure (Rosso et al., 2009).

Although not within the scope of this review, it might be mentioned that the functions described for PTOX in cyanobacteria and eukaryotic algae are similar to those in higher plants (reviewed by McDonald et al., 2011). These include carotenogenesis and response to diverse environmental stresses, such as high light, temperature, salt, iron, and phosphate deprivation (Moseley et al., 2006; Bailey et al., 2008; Wang et al., 2009). One difference is that cyanobacterial respiratory and PETCs share some components, which offers the possibility of PTOX control of both activities.

In summary, there is growing evidence that PTOX is a versatile terminal oxidase in chloroplast metabolism and that its primary physiological role varies from taxon to taxon. This might not be surprising, since PTOX is only one element of a large network of factors involved in stress tolerance, pigment biosynthesis, and photosynthetic control (regulation of ATP/NADPH ratios) (Kramer and Evans, 2011; Foyer et al., 2012). In such a network, it might be anticipated that the relative importance of a given element varies according to developmental and/or physiological context. For instance, PTOX might turn out to play a more significant regulatory role in PSI cyclic electron transport in bundle sheath cells of C4 plants than in mesophyll cells of C3 plants; in contrast to C3 plants, cyclic electron transport is a key regulator of ATP synthesis during C4 photosynthesis (Rumeau et al., 2007; Foyer et al., 2012). Given the ubiquity of PTOX expression, it might also be predicted that novel functions will be found for PTOX, especially in non-green plastids. A schematic diagram of PTOX in thylakoids is shown in Figure 1.

Because the green sectors of im contain morphologically normal chloroplasts, we have proposed that they arise from plastids that escape the effects of the im/im mutation early in chloroplast biogenesis, presumably by the action of compensating factors that affect excitation pressure thresholds, directly or indirectly. Two approaches have been used to gain insight into these factors: (1) molecular characterization of green versus white im sectors; and (2) identification and characterization of second-site suppressors of im variegation. These approaches will be summarized briefly.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.