In recent decades our understanding of the molecular basis of photosynthesis has increased impressively. It is increasingly evident that the fundamental structure of the photosynthetic apparatus is an example of the capacity for complex, highly sophisticated systems to evolve, since cyanobacteria, green algae and higher plants (which apparently diverged hundreds of millions of years ago) have very similar photosynthetic machineries [39, 40]. The main differences in the photosynthetic apparatus of these taxa are in the "peripheral" parts, such as the antenna systems. For example, the phycobilisomes of cyanobacteria have been replaced with the LHC proteins in green algae and higher plants, and there are wide variations in their photosynthetic pigments, many of which were previously used as key taxonomic descriptors. Much regulation is exerted by the antenna systems, where the qE type of NPQ (feedback de-excitation) and state transitions occur in conjunction with dynamic changes in antenna size in acclimation responses to, inter alia, changes in light conditions . More recently, the less abundant LHCI proteins Lhca5 and Lhca6 have been implicated in the regulation of cyclic electron transport , which is known to be subject to both evolutionary adaptation and environmental acclimation [see e.g. ]. Comparative studies of plants from diverse taxa, ecological niches or habitats (field or laboratory) show that the regulatory properties of the antenna systems typically vary more than the properties of the "core engine" of the system .
Twenty-five years ago, Arabidopsis emerged as the prime model organism for plant biology research . Its small size and rapid growth cycle has enabled photosynthesis researchers to move from experiments with "synthetic" (e.g. algal cultures) or imprecise (e.g. spinach) systems to more reproducible experiments with plants grown under highly controlled and reproducible conditions in climate chambers, growth rooms and cabinets. Although this has been important for scientific development, we believe that studies performed with plants grown under natural conditions can provide valuable complementary information. This study is a contribution to the growing body of literature describing experiments in which Arabidopsis has been exploited as a natural species rather than a "laboratory rat". In the wild, Arabidopsis grows in open, typically highly-disturbed, habitats and has significant capacity for photosynthetic acclimation . Therefore, whether or not Arabidopsis plants grown in climate chambers are like those grown in the field, which may seem trivial, is highly relevant for scientists addressing many aspects of plant biology. One aspect not covered in this work is the natural variation of the species; it is possible that our results may have been substantially different had we chosen to study a different accession rather than Colombia-0. We chose this accession because it has been used in most studies published to date; however this accession is not specifically adapted to our local study site environment in Umeå. Furthermore, as we have not compared plants grown at different sites, at different times of the year or with different photoperiods we cannot draw general conclusions about the plasticity of Arabidopsis in all environments. Additional phenotypic variations may be encountered in future experiments, and we make no claim that other field-grown Arabidopsis plants will necessarily be similar to those analyzed here. Nevertheless, we believe that the trends we have recorded are likely to represent some of the most prominent differences between indoor- and field-grown Arabidopsis plants. It is also obvious that plants grown in climate chamber under LD are better substitutes for field-grown plants than plants grown under SD--which is typically used for photosynthetic studies--although plants grown indoors under LD were still more similar in terms of photosynthetic characteristics to SD plants than to field-grown specimens (Table 3). Further studies are needed to determine if variations in LD conditions in climate chambers (e.g. 14, 16, 18 or 20 h photoperiods) significantly influence photosynthetic characteristics.
Our data show that the indoor-grown SD plants had, for example, different leaf morphology, higher levels of Lhca5, much higher levels of Lhcb1 and Lhcb2, less PsbS (and no ELIPs), and different pigment contents compared to the field-grown plants. In particular they were strongly depleted in xanthophyll cycle pigments. The differences in leaf morphology and plant stature are striking, and it is intriguing that some of the observed changes, for example in leaf size, did not follow simple patterns, notably both LL- and field-grown plants had smaller leaves than NL- and HL-grown plants. This indicates leaf developmental patterns are influenced by more than one factor. For example, several "typical photoreceptors" may respond both to differences in photoperiod and light intensity, and photosynthetic signals may influence leaf morphology. Accordingly, anatomical differences between typical sun and shade leaves seem to depend on photosynthetic signals . We also found that Lhca5, expression of which correlates well with light intensity in indoor plants, was almost undetectable in the field-grown plants. It appears, therefore, that (at least in Arabidopsis) Lhca5 is not simply a "light stress LHC", as exemplified by LI818 in Chlamydomonas , since it was down-regulated under our field conditions. It has been suggested that both the Lhca6 protein, which is present at very low levels in plants grown under most conditions, and Lhca5 regulate cyclic electron transport around PSI . However, we are not aware of any published analyses of the cyclic electron transport capacity of field-grown plants.
Most or all PSI and PSII core proteins are present in unit stoichiometry and this also probably applies to the PSI antenna proteins Lhca1-4 and the minor Lhcb antenna proteins Lhcb3-6. Our data show that the PSI antenna in the plants grown indoors was similar to that of the plants grown under field conditions but--as we have noted before--the PSII antenna may be more flexible. On a PSII basis, the levels of Lhcb5 (CP26) and, in particular, Lhcb6 (CP24) were lower in indoor plants, raising questions whether PSII centers lacked these proteins in the indoor plants, or a fraction of the proteins was present, but they were not bound in their "normal positions" in PSII in the field-grown plants (or both). Our results relating to the major LHCII proteins (Lhcb1, Lhcb2 and Lhcb3) are particularly intriguing. Taking known pigment and protein stoichiometries into account, there may have been three to four LHCII trimers per PSII monomer in the LL plants. The supermolecular structure of PSII has been studied extensively, and it is known that up to three LHCII trimers, denoted S, M and L, can associate with each PSII complex in a dimer . S, M and L refer to strongly, medium and loosely bound trimers, respectively. It is possible that the M trimer is composed of Lhcb3 and two Lhcb1 subunits . It is not known if there is any specificity for Lhcb1 and Lhcb2 at any position in the S and L trimers. It is conceivable that other LHCII trimers may aggregate in "LHCII-only domains", which must be attached to the photosystems, since energy transfer from all parts of the LHCII antenna into the photosystems is very efficient. Naïvely, the S, M and L trimers plus trimers found in LHCII-only domains may account for three to four trimers/PSII in LL plants. However, the field-grown plants contained only ca. a third of this amount of LHCII, i.e. one or at most two trimers/PSII. Lhcb3 was present in approximately equal amounts in field-grown and indoor plants, suggesting that M trimers were present in most or all of their PSII centers. Our data show that plants with only small amounts of LHCII trimers are perfectly capable of performing state transitions, consistent with the finding that the fitness of the Stn7 mutant grown under field conditions deviates from that of wild-type counterparts . However, since the M trimer--at least Lhcb3--is not believed to participate in state transition , Lhcb1 and Lhcb2 in S trimers are likely to be efficiently phosphorylated and participate in state transitions in field-grown plants. Alternatively, M trimers may become phosphorylated and detach from PSII. There are insufficient data from our study to enable us to confirm this possibility, but a more detailed study of PSII in Arabidopsis grown under field conditions may show which PSII supercomplexes are most abundant when Arabidopsis is exposed to its naturally-adapted light regimes. Taken together, although the LHCII content is much lower in field grown plants, antenna function is not much affected.
ELIPs, most likely involved in pigment metabolism in plastids, were originally identified as proteins that transiently accumulate during early plastid development, but subsequent studies have shown that they also accumulate under diverse stress conditions . ELIPs play an important protective role under light-saturated conditions, such as may occur in the field and, except in some artificially-controlled growth conditions in climate chambers; they are likely to be abundant thylakoid proteins. Nevertheless, our results indicate that the plants lacking ELIPs were well adapted to their growth conditions and had high levels of fitness; our 2-year study of double ELIP mutants suggests that ELIP functions in mature leaves may be redundant or of low importance. However, ELIPs may be more important in early developmental stages and it is also possible that they play crucial roles under conditions that the plants did not encounter during these 2 years.
Xanthophyll cycle pigments and PsbS are typically involved in photoprotective processes. In our experiments these factors were found at very low levels in indoor plants compared with field-grown samples. This is consistent with the view that under natural conditions photoprotection by NPQ and other mechanisms is of vital importance for the fitness of the plant . We have also shown that the level of NPQ is balanced and there is some evidence that selective forces act to reduce the level of photoprotection . Finally, our comparison of NPQ levels in a set of Arabidopsis accessions grown in the lab and the field illustrates how conclusions drawn from studies in the lab may be invalid for field-grown plants, due to phenotypic plasticity.
Plants have evolved many mechanisms that are involved in responses to changes in their growth conditions, ranging from long-term developmental processes that affect the morphology or physiology of the whole plant or individual leaves [25, 49], to adjustments in the functioning of individual proteins within the photosynthetic apparatus, operating on timescales ranging from seconds to hours . We have studied some of these adjustments, in particular relating to the functions of the photosynthetic light harvesting apparatus. In addition, adjustments to PSI/PSII ratios, variations in components of the inter-photosystem energy flow apparatus, and rates of cyclic electron transport, ATP generation and the photosynthetic dark reactions may be as important as those investigated here. We anticipate that other studies will focus on comparisons of photosynthetic properties that vary between and within species, or in single genotypes, as a result of phenotypic plasticity.