JXB Advance Access originally published online on November 28, 2003
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Journal of Experimental Botany, Vol. 55, No. 394, pp. 11-25, January 1, 2004
© 2004 Oxford University Press
Plant Carbon-Nitrogen Interactions from Rhizospheres to Planet |
Global aspects of C/N interactions determining plant–environment interactions
Received 2 May 2003; Accepted 11 August 2003
1 Division of Environmental and Applied Biology, School of Life Sciences, University of Dundee, Dundee DD1 4HN, UK
2 Columbia University Biosphere 2 Center, 32540 S. Biosphere Road, PO Box 689, Oracle, AZ 85623, USA
3 Scottish Crop Research Institute, Invergowrie, Dundee DD2 5DA, UK
4 School of Sciences, University of Sunderland, Sunderland SR1 3SD, UK
* Permanent address and to whom correspondence should be sent: Division of Environmental and Applied Biology, School of Life Sciences, University of Dundee, Dundee DD1 4HN, UK. Fax: +44 (0)1382 344275. E-mail: j.a.raven{at}dundee.ac.uk Permanent address: Scottish Crop Research Institute, Invergowrie, Dundee DD2 5DA, UK.
 |
Abstract |
|---|
Top
Abstract
IntroductionThe atomic C:N ratio in photolithotrophs is a function of their content of nucleic acids, proteins, lipids, polysaccharides, and other organic materials, and varies from about 5 in some protein-rich microalgae to much higher values in macroalgae and in higher plants with relatively more structural and energy storage materials. These differences in C:N ratios among organisms means that there is more N assimilation by photosynthetic organisms in the oceans than on land despite the near equality of global photosynthetic C assimilation rates in the two environments. Aquatic organisms obtain inorganic C and inorganic N from the surrounding water. Terrestrial photolithotrophs obtain inorganic C, dinitrogen (by diazotrophy) and some combined N from the atmosphere, with the remaining combined N coming from the soil. The nitrogen cost of growth (biomass production rate per unit plant N) varies with the C:N ratio and specific growth rate of the organism. The C:N ratio of plants can be increased with no, or minimal, decrease in growth rate by switching from N-containing to N-free solutes involved in, for example, UV-B screening or by reducing the content of particular proteins. The water cost of growth (water lost per unit biomass gain) in terrestrial plants is a function of N supply and of C supply; water cost is lower with higher N and C availability. Water supply is also important in determining denitrification rates on land and on N (and C) fluxes from terrestrial to aquatic systems.
Key words: Agriculture, algae, astrochemistry, carbon, chemical defence, compatible solutes, ecology, evolution, free radical scavenging, nitrogen, UV-B screening.
|
Introduction |
|---|
Carbon and nitrogen are major components of all living organisms, and their metabolism is intimately linked. In this paper a number of questions about the interactions of carbon and nitrogen in plant metabolism are addressed. (1) What features of carbon and nitrogen relate to their biological roles? (2) What are the major carbon- and nitrogen-containing components of plants and how do they vary quantitatively among plants? (3) What possibilities are there for altering the content of these components without compromising plant growth and, for crop plants, the quality and yield of the harvested product?
These questions are considered in a wider context. In exploring the evolutionary background to the roles of carbon and nitrogen in organisms and its implications for their functioning, the genesis of carbon and nitrogen in stars and their partitioning to our solar system and, especially, to the Earth’s surface is reviewed first. The chemical properties of carbon and nitrogen in the context of what is required of life ‘as we know it’ then follows. The quantitative requirements for carbon and nitrogen in photosynthetic organisms of different life forms are then considered, and special attention is given to the extent to which the carbon:nitrogen ratio can be varied, with emphasis on economizing on the use of nitrogen and what effects such restriction of nitrogen content might have for growth under natural conditions. Finally, the requirements for carbon and nitrogen are related to the habitats of the organisms, and how the acquisition and processing of carbon and nitrogen impact on the requirements for other resources.
|
The origins and cosmic distribution of carbon and nitrogen and their compounds |
|---|
The ‘Big Bang’ only produced elements of low atomic number, mainly H and He. Nucleosynthesis of elements larger than this, including carbon and nitrogen, occurs in stars (Williams and Fraústo da Silva, 1996, 2003; Henning and Salama, 1998; Beers, 2003). Thus, the availability of carbon and nitrogen and of planets made from these and higher atomic number elements, required one cycle of now extinct stars which generated these elements. Nitrogen is the fifth most abundant element in the universe, and carbon is the sixth.
Some of the carbon and nitrogen in the universe is in stars and planets; the rest is in interstellar space. Spectroscopic studies on interstellar molecules show a large range of compounds containing carbon and/or nitrogen (Henning and Salama, 1998; Thaddeus et al., 1998; Ehrenfreund and Charnley, 2000).
Much of the interstellar organic carbon is in polyaromatic hydrocarbons and their relatives the buckminsterfullerenes. However, there are also many other molecules such as CO, CO2, CH4, CH2O, and CH2OHCOOH. Nitrogen occurs as N2, N2O and NH3. Interestingly for the subject of this paper there are also compounds containing both C and N. Examples are HCN and the homologous series of cyanopolyines HC1+2nN where n is 0 (HCN), 1 (HC3N), 2 (HC5N), 3 (HC7N), 4 (HC9N), and 5 (HC11N). HC11N is relatively abundant and has the formula H-C 
C-C C-C C-C C-C C-C N. Of more biological relevance are compounds such as CH3NH2 and, much less certainly, CH2(NH2)COOH. These data show that, even in interstellar space, photochemistry produces molecules containing both carbon and nitrogen.
Coming down to Earth, it is found that this planet is depleted in carbon and nitrogen relative to many heavier elements when compared to cosmic abundance; even phosphorus is more abundant on Earth than are carbon and nitrogen. Attempts to simulate in the laboratory the mechanisms of prebiotic productions of carbon and nitrogen compounds go back 50 years (Miller, 1953). These experiments showed that organic molecules, including glycine and alanine, could be produced in relatively high yield when an electric discharge energized molecules in a mixture of gases which was thought to be representative of the prebiotic atmosphere of the Earth. Later experiments involved UV as the energy source, and, when HCN was a component of the gas phase, the production of adenine. However, the gas mixture used by Miller (1953), i.e. CH4, CO2, NH3, and H2O, was more reducing than what is now believed to have existed in prebiotic times on Earth 
4 Ga ago. Repetition of Miller’s experiments with these less reducing atmospheres gave much smaller yields of organic compounds. Current thinking about the origin of carbon- and nitrogen-containing compounds on Earth emphasizes inputs from comets and syntheses by chemolithotrophic processes, for example, at hydrothermal vents.
To summarize, the cosmic abundance of carbon and nitrogen is high, and, despite the Earth being depleted in these elements relative to their cosmic abundance, carbon and nitrogen are relatively abundant in the Earth’s atmosphere, hydrosphere and crust.
|
The availability of carbon and nitrogen to photosynthetic organisms over the last 3.5 Ga |
|---|
The relative availability of carbon dioxide (more generally, inorganic carbon) for photosynthesis and of combined nitrogen over the last 3.5 Ga in which photosynthetic O2-evolvers may have existed is discussed by Falkowski and Raven (1997), Falkowski (1997), Falkowski et al. (1998), Raven and Yin (1998), Stewart and Schmidt (1999), Navarro-Gonzalez et al. (2001), and Anbar and Knoll (2002); for an overview of biogeochemistry see Berner and Berner (1996). Carbon dioxide availability has decreased over the last 3.5 Ga, although the fall has not been monotonic and there have been a number of fluctuations, with low atmospheric carbon dioxide in the late Devonian–Permian and during glacial episodes in the Quaternary (Falkowski and Raven, 1997).
Combined nitrogen (in this case as NO and NO2) availability from electric storms is thought to have diminished during the period in the Archean in which the atmosphere became oxidized and carbon dioxide was decreasing (Navarro-Gonzalez et al., 2001). Anbar and Knoll (2002) suggested that a Palaeo- and Meso-Proterozoic sulphidic ocean could have limited molybdenum availability for biological nitrogen fixation, and cite 15N/14N natural abundance values to support restricted biological diazotrophy. However, the possibility of the activity of ‘alternative’ nitrogenases using vanadium or only iron, and with 15N/14N discrimination values greater than those normally cited (Rowell et al., 1998), should not be ignored.
There is a widespread view that primary productivity is restricted by the availability of combined nitrogen NH+4, NO-3, NO-2, and organic N) in many terrestrial and aquatic habitats today (Falkowski, 1997, 2000; Falkowski and Raven, 1997; Falkowski et al., 1998; Field et al., 1998). Nitrogen and carbon co-limitation can occur in terrestrial C3 plants, with increases in atmospheric carbon dioxide above the present 370 µmol mol–1 yielding increased primary productivity, provided nitrogen (or other resources, including water and light) is not in too limiting supply (Andrews et al., 2001). Geochemically, phosphorus is the element which is expected ultimately to limit primary productivity, and such limitation now occurs in some areas (Redfield, 1958; Beadle, 1966; Francis and Read, 1994; Wu et al., 2000). Furthermore, there is evidence that phosphate availability limits diazotrophy by the cyanobacterium Trichodesmium in the south-west North Atlantic (Sañudo-Wilhelmy et al., 2001). However, elsewhere in the ocean, iron limitation of diazotrophy is likely (Falkowski, 1997, 2000; Falkowski and Raven, 1997; Falkowski et al., 1998; Field et al., 1998). In some parts of the ocean, where combined nitrogen and phosphorus are relatively abundant yet primary productivity is low, it seems that iron is limiting the use of combined nitrogen in algal photosynthesis. Fe deficiency does not, however, decrease the nitrogen contents relative to the carbon contents of phytoplankton growing by N2-fixation (Berman-Frank et al., 2001) or by use of combined nitrogen (Green et al., 1991; La Roche et al., 1993; Flynn and Hipkin, 1999; De La Rocha et al., 2000; Milligan and Harrison, 2000; Davey and Geider, 2001).
The evidence that has just been discussed shows that nitrogen fixation is a process which is relatively restricted in extent, and that this results in frequent nitrogen limitation of growth of photosynthetic organisms in land and in the sea. This can be illustrated by comparing global carbon and nitrogen assimilation in net primary productivity with the assimilation of inorganic carbon and of atmospheric N2 (Table 1). Even taking into account inputs of ‘new’ combined nitrogen from processes other than biological N2 fixation, for example, lightning, ‘artificial’ fertilizers, and NOx resulting from high-temperature combustion of fossil fuels (Table 1), it can be seen that the nitrogen requirements for photosynthetic primary production are largely met by recycled combined nitrogen rather than N2 fixation. Combined nitrogen is removed by denitrification which generates N2, which can only be reconverted biologically to combined nitrogen by N2 fixation, and N2O, which is a form of combined nitrogen that is not generally available as a nitrogen source for primary producers. The loss of combined nitrogen by denitrification is generally not directly controlled by primary producers, an exception being the generation of anoxic waterlogged ecosystems in areas with high precipitation relative to evaporation. Loss of combined nitrogen by denitrification, and the cost in other resources of biological N2 fixation discussed in the preceding paragraph and by Vitousek et al. (2000), restricts the size of the combined nitrogen pool in many habitats, leading to a combined nitrogen famine despite the apparent potential for a nitrogen feast using N2.
|
Overall, it appears that nitrogen is globally ‘more limiting’ than is carbon for photosynthetic organisms, and that economizing on nitrogen use may be more significant in natural selection than is economizing in carbon use. The amounts of carbon and nitrogen that are used in biology and how the use of the elements can be substituted by other elements will be discussed next.
|
The roles of carbon and nitrogen in living organisms: background |
|---|
The properties of certain elements which have been capitalized upon by life on Earth are reviewed by Williams and Fráusto da Silva (1996, 2003). Carbon is unique in forming stable chains and rings of the element; this is the basis of organic chemistry, and nitrogen participates very strongly in both aliphatic and aromatic organic chemistry, and is also important in hydrogen bonding and in forming ligands with metals. The diversity of organic chemistry, with elements such as oxygen and nitrogen, gives further possibilities for hydrogen bonding.
For biology, organic chemistry has the advantage that most reactions do not occur at a significant rate at temperatures permitting the existence of liquid water within the range of pressures found in habitats on Earth. This means that catalysis is essential for organic biochemistry, allowing regulation of the rate and products of reactions (Williams and Fráusto da Silva, 1996).
As a background to the following discussion of the qualitative and quantitative roles of carbon and nitrogen in photosynthetic organisms, it was noted that the atomic C:N ratio varies from five in microalgae and cyanobacteria to more than 100 in woody plants (Sterner and Elser, 2002).
|
Nitrogen-containing macromolecules: DNA, RNA and protein |
|---|
The core of molecular biology involves the organic chemistry of nitrogen. Fifty years ago Watson and Crick (1953) presented the structure of the double helix of DNA, with a crucial role for hydrogen bonds involving nitrogen as well as oxygen in forming the double helix and hence in replication, transcription and repair. X-ray crystallographic studies of proteins showed the role of nitrogen-related hydrogen bonds and electrostatic interactions bound in the alpha helix and in the beta pleated sheet. Rather later, studies of the structure of rRNA, tRNA and mRNA also emphasized the role of hydrogen bonds involving nitrogen in the bases in maintaining the secondary structure and in transcription and translation.
To these qualitative needs for nitrogen and carbon in the components of the central dogma there is, in terms of overall requirement for carbon and nitrogen by organisms, the question of how many kinds of genes and hence of gene products occur in an organism, how large the genes and hence the gene products are, and how much of each gene product is produced under the growth conditions that are being considered. Determinants of these variables are known in general terms (Maynard-Smith and Szathmáry, 1995), but it is less easy to be specific. Even the determinants of the number of protein amino-acids (21, including selenocysteine) are still a matter for debate (Maynard-Smith and Szathmáry, 1995). Only perhaps four or five amino-acid residues can be directly involved in the active centre of an enzyme, and some of the rest of the amino-acid residues (of the order of 100) have regulatory functions, for example, interactions with regulatory ligands, and controlling and facilitating access of substrate(s) and removal of product(s) (Maynard-Smith and Szathmáry, 1995; Seligmann, 2003). As to genome size and gene number, the smallest fully sequenced genome of a photosynthetic O2-evolver is that of the cyanobacterium Prochlorococcus marinus, of which the MIT strain has a genome size of 1.7 Mbp with 1500 genes (Hess et al., 2001). Arabidopsis thaliana has a nuclear genome of 125 Mbp with 25 500 genes (The Arabidopsis Genome Initiative, 2000). Arabidopsis thaliana was chosen for sequencing because, inter alia, of its small genome size, and most higher plant genomes are much larger even if they do not code for more genes. Genomic obesity, for example via retrotransposons, involves additional DNA in the nucleus which is not involved in coding for genes that are required for plant function.
Turning to the quantitative requirements for RNA for plant function, most of the RNA in a cell is rRNA. In heterotrophs (non-autotrophic bacteria, fungi, metazoa) the rRNA per cell in a given organism is a linear function of the relative growth rate (Sterner and Elser, 2002). However, there is a baseline rRNA content that is needed for protein turnover even without growth. This requirement is considered in the context of mature leaves of higher plants, where the turnover of the D1 (psbA gene product) protein of photosystem II is a major component of protein turnover (Raven et al., 2002a).
For growing cells of photosynthetic O2-evolvers, for example, unicellular algae, and the meristems of higher plants, there are few data on the relationship of cellular rRNA content to growth rate. There is certainly not the monotonic, linear, increase in rRNA per cell with relative growth rate that is found in heterotrophs (Laws et al., 1983; Thomas and Carr, 1985; Binder and Liu, 1998; Binder, 2000). Using light as the limiting resource, Laws et al. (1983) and Thomas and Carr (1985) found an RNA content which was independent of growth rate for a number of eukaryotic microalgae. Binder and Liu (1998) and Worden and Binder (2003) (cf. Lepp and Schmidt, 1998; Binder, 2000) measured rRNA content in two strains of marine Synechococcus and in a strain of the marine Prochlorococcus, and found a linear relationship between rRNA content and growth rate over a restricted, intermediate, range of growth rates for these three strains of cyanobacteria. At lower growth rates the rRNA content was invariant with changes in growth rate, while at the highest growth rates the rRNA content decreased with increasing growth rate. For the relatively small range of O2-evolving organisms examined, it appears that the specific reaction rate of rRNA, rather than rRNA content, increases with growth rate, so that cells are over-provided with rRNA at low growth rates.
As for the protein content, it is clear that the specific reaction rate of enzymic, and other catalytic (e.g. light-harvesting pigment–protein complex) reactions, in photosynthetic organisms generally increases with growth rate, so that rapidly growing organisms of a given genotype have a lower protein cost of growth (biomass increase per unit protein in existing biomass per unit time: Raven, 1984a, b; Berends and Aerts, 1987; Agren and Bosatta, 1996; Andrews et al., 1999). This is despite, and sometimes because of, acclimatory changes in the content of specific proteins. Thus, acclimatory increases in the content of light-harvesting pigment–protein complexes with growth at lower photon flux densities decreases the specific reaction rate (photons absorbed per unit chromophore per unit time) because the increase in chromophores does not offset the decreased photon flux density (Raven, 1984a, b). Furthermore, if the increased chromophore content increases the ‘package effect’ (i.e. self-shading), then the rate of absorption of photons per unit chromophore is further decreased. The capacity to absorb a greater fraction of incident photons is obtained at the cost of additional investment of nitrogen (Raven, 1984a, b).
Another aspect of variation in proteins related to light harvesting is phylogenetic rather than acclimatory. Raven (1984a, b) notes the larger protein content per unit chromophore for the phycobilin light-harvesting pigment–protein complexes of most cyanobacteria, all red and glaucophyte algae, and cryptophyte algae. By contrast, the protein per chromophore is lower in the chlorophyll- and carotenoid-containing pigment–protein complexes of some cyanobacteria and all cryptophytes in addition to their phycobilins, and in all other O2-evolvers (Raven, 1984a, b). In organisms adapted or acclimated to low photon flux densities, the differences in protein allocation per unit chromophore can make a significant (up to 10%) difference in overall protein content for an organism with a phycobilin-based rather than a chlorophyll-based light-harvesting apparatus in an otherwise similar organism (Raven, 1984a, b; MacIntyre et al., 2002; Ting et al., 2002; contrast Sterner and Elser, 2002).
Better-known phylogenetic variations in protein allocation in photosynthetic organisms concern the different quantities of Rubisco protein needed to sustain a given rate of CO2 fixation per unit leaf area in C3 plants and C4 plants in air. The C4 plants need, and have, less Rubisco per unit area of leaf, and attain the same or higher CO2 fixation rates per unit leaf area in air levels of CO2 because the CO2 concentration, and CO2/O2 ratio, at the site of Rubisco activity is higher in C4 than in C3 plants. Even when the extra protein needed in the C3+C1 carboxylation, and C4–C1 decarboxylation cycles is accounted for, the C4 plants still need less leaf N to attain a given rate of CO2 fixation, with the decreased content of enzymes of the photorespiratory carbon oxidation cycle as a contributory factor.
In the present context of C/N interactions, it is of interest that transgenic rice (Orzya) and tobacco (Nicotiana) with decreased content of Rubisco show altered N allocation in the plants (Makino et al., 2000a, b; Matt et al., 2002; see also Parry et al., 2003). Makino et al. (2000a, b) found that the lower leaf N content when non-transformed rice plants were grown in high, rather than ambient, CO2 concentrations did not show optimal N allocation, in that Rubisco did not decrease relative to the decrease in total leaf N. Optimal allocation requires that Rubisco content should fall in C3 plants grown at high CO2, since the same carboxylation rate can be sustained at high CO2 with a lower Rubisco content (assuming equal activation states under the two growth conditions). Rubisco-antisense rice with 65% of wild-type Rubisco had 20% lower photosynthesis in normal atmospheric CO2 levels, but 5–15% higher rates at saturated CO2. However, while the transgenic plants had a higher N use efficiency for growth than did controls after the seedling stage, in terms of overall biomass production this was offset by slower growth in the early stages of growth.
Matt et al. (2002) used Rubisco antisense technology to decrease Rubisco expression in tobacco. The decreased photosynthetic carbon assimilation caused an inhibition of nitrogen assimilation, and altered primary and secondary (chlorogenic acid; nicotine) metabolite levels for both nitrogen-containing and nitrogen-free compounds.
Raven et al. (2002b) discuss other work showing that down-regulation of a number of photosynthetic enzymes other than Rubisco by 25% (or, in some cases, much more) does not cause a decrease in the rates of photosynthesis and growth under the conditions tested. However, such down-regulations do not always lead to increased biomass production rates because compensatory increases in other proteins occur, for example, an increase of Rubisco when Rubisco activase is down-regulated. Furthermore, there is always the possibility that growth conditions which do produce a phenotype of inhibited photosynthesis and growth exist, but have not yet been tested. However, in some cases there are large effects of small decreases in activity of certain photosynthetic enzymes other than Rubisco. The work of Henkes et al. (2001) shows for antisense Nicotiana transformants that a small decrease in plastid transketolase, which is involved in the oxidative as well as the reductive pentose phosphate pathway, causes not only decreases in the rate of photosynthesis over a wide range of photon flux densities but also decreased synthesis of phenylpropanoids.
Raven et al. (2002c) also consider the possibility of economizing on nitrogen by genetic manipulation of rRNA. They conclude that there is an apparent overprovision of rRNA in mature leaves relative to what is apparently needed for the turnover of D1 and of other proteins.
More radical attempts to modify plant performance by manipulating protein content is the expression of genes from other organisms in a plant. One such procedure is the expression of the bifunctional cyanobacterial enzyme fructose-1,6-bisphosphate-1-phosphatase/sedoheptulose-1,7-bisphosphate-1-phosphatase in higher plant plastids which already possess separate enzymes with these two phosphatase activities (Miyagawa et al., 2001). This procedure increases the growth and photosynthetic rates of Nicotiana at least under optimal conditions, although effects on expression of the native phosphatases and on the rate of biomass increase per unit plant nitrogen is not clear.
Another procedure involved the expression of algal Rubiscos with a higher CO2/O2 selectivity (e.g. from red algae or diatoms) in higher plants. Complete replacement of the native Rubisco with such algal enzymes could increase the photosynthetic rate per unit nitrogen over the entire light and CO2 range (Whitney and Andrews, 2001). However, while the transcripts are produced in the transformed plants, the translation products are not properly assembled in the stroma and do not function (Whitney et al., 2001). By contrast, the gene for Form 1 Rubisco from an alpha-proteobacterium which was transferred to the plastid genome of Nicotiana was expressed and the product assembled into a functional enzyme which could support photosynthesis and growth. However, the bacterial Rubisco has a very low selectivity for CO2 over O2, and photosynthesis and growth of Nicotiana using this Rubisco requires high CO2 concentrations (Whitney and Andrews, 2001).
Other roles for proteins include functions in the cytoskeleton, including molecular motors. There does not seem to be much scope for reducing the cellular content of these protein components. Membrane lipids contain nitrogen in the polar head-group in the case of some, mainly non-thylakoid, constituents.
|
Extracellular structural materials |
|---|
Extracellular structural materials in photosynthetic organisms include some proteins, as well as other nitrogen-containing polymers, including peptidoglycan in cyanobacteria, and chitin in some diatoms and (in symbiosis) the cell walls of fungal associates in lichens and mycorrhizas. However, in general, photosynthetic organisms have nitrogen-free carbohydrate polymers as their turgor (tension)-resisting components and invariably have (nitrogen-free) lignin as the component which resists compression (Raven, 1993, 1995b, 2000).
|
Macromolecular and other insoluble compounds storing energy and organic carbon |
|---|
Storage polymers related to energy and organic carbon storage include polysaccharides and triglycerides. Storage proteins are found in many seeds, and the aspartate–arginine polypeptide cyanophycin in cyanobacteria. Attempts to find specific vegetative storage proteins in several plants have failed; there seem to be no specific proteins used during vegetative dormancy.
|
Substitution of other elements for carbon and nitrogen in macromolecules |
|---|
Extracellular structural constituents, and intracellular storage polysaccharides, differ from nucleic acids and proteins in that elements other than carbon (and nitrogen) can, in principle, be substituted for the carbon (and nitrogen)-containing component.
One example is energy storage as polysaccharides or triglycerides. In ATP units, each (CH2O) unit in a polysaccharide can yield up to 6 ATP, depending on assumptions made about stoichiometries of protons in mitochondrial electron transport and ADP phosphorylation. Energy (but not organic carbon) can also be stored as polyphosphate, where all but one of the phosphate units in a polyphosphate needs 1 ATP converted to 1 ADP to add that unit and can subsequently be used to phosphorylate 1 ADP to form 1 ATP. While such energy storage economizes on carbon it is very expensive in terms of phosphate, and polyphosphates usually function as osmotically (relatively) inert stores for phosphate. Even when they may serve as energy stores (Lambert et al., 2002) under conditions of ample phosphate supply the maximum quantity that can be stored in a cell can only energize a growing cell for minutes rather than the hours or days in the case of polysaccharides.
For extracellular structural components there is the possibility of substituting the rigid organic components (mainly lignin) with inorganic components such as polymeric silica or crystalline calcium carbonate. Lignin resists mechanical forces of compression caused by gravity in above-ground parts of vascular plants which are too large to be supported by a hydrostatic skeleton, and by tension of the water column in the walls of conducting conduits of xylem (Raven, 1993, 1995b). Higher (vascular) plants with high contents of polymeric silica include Equisetum, lycopods, members of the Urticales and Cucurbitales, members of the Eriocaulaceae, some sedges, and all grasses (Raven, 1983, 2001b; Epstein, 2001). In some cases (Equisitum, lycopods) the high silica content is correlated with a low lignin content (Raven, 1983). However, in Cucurbita fruits the Hard Rind (Hr) gene controls lignin deposition, and, in parallel, silica deposition (Piperno et al., 2002). Accordingly, the replacement of lignin as a structural component of vascular plants with silica is not a universal aspect of silica deposition in these plants (Bonilla, 2001; Datnoff et al., 2001). However, silica in higher plants clearly has an important role in limiting grazing and parasitism (Epstein, 2001; Kim et al., 2002).
Some work has been performed on the mechanical properties of higher plant silica in situ (Speck et al., 1998), but more is needed. Recent work demonstrated, quantitatively, the mechanical properties of silica in situ in diatoms (Hamm et al., 2003). The silicified cell walls (frustules) are, as is universal for silicified structures, not extensible, so they cannot show plastic extension during growth, or elastic changes with variations in external osmolarity, in the manner of typical plant or algal cells. The increase in volume of growing cells occurs by a decreasing overlap of the two halves of the wall (valve). Hamm et al. (2003) showed that the mechanical properties of silicified cell walls of diatoms were able to account for the inability of certain crustacean grazers to crush the cells.
By contrast to silica, calcium carbonate seems to have a much less important mechanical role in algae and plants, and does not substitute significantly for structural polysaccharides. However, it can have anti-herbivore functions (Meyer and Paul, 1995).
|
Carbon and nitrogen in low relative molecular mass (Mr) organic compounds |
|---|
While macromolecules account for most of the carbon and nitrogen in algae and plants (Gnaiger and Bitterlich, 1984; Handley et al., 1989; Lourenço et al., 1998; Falkowski, 2000; Raven, 2001a; Geider and La Roche, 2002) there are significant quantities of lower Mr, soluble compounds which have important roles and can account for quite a large fraction of total cellular carbon and/or nitrogen, for example, in compatible solutes in organisms growing in habitats of very high osmolarity.
The functions of low Mr organic compounds
In addition to low Mr organic compounds which are parts of energy transformation pathways, or of biosynthetic pathways leading to macromolecules, the functions of low Mr compounds can be categorized as follows (Table 2). (1) Storage of energy/reduced carbon and nitrogen. (2) Accumulation of organic salts related to pH regulation (Raven and Smith, 1976; Raven, 1985; Andrews, 1986; Andrews et al., 1995a, b; Yin and Raven, 1997). (3) Compatible solutes and osmoregulation (Raven, 1985; Oren, 1999). (4) UV-B screening and/or visual signalling to animals (Gronquist et al., 2001). (5) Restriction of grazing or parasitism, and establishment of mutualisms (Harborne, 1993). (6) Free radical scavenging (Smirnoff and Cumbes, 1988; Sunda et al., 2002). (7) The organic C which is cycled in the carboxylation–decarboxylation cycle of Crassulacean Acid Metabolism, involving much more carbon than in the case of C4 photosynthesis (Raven and Spicer, 1996).
|
Nitrogen-containing and nitrogen-free low Mr organic compounds
Some nitrogen-containing solutes fulfil more than one of these functions, for example, proline, which is a protein amino acid, can also act as a compatible solute, and scavenges free radicals (hydroxyl radicals) (Smirnoff and Cumbes, 1988). Citrulline is not a protein amino acid, but acts as a nitrogen storage and long-distance transport compound and as a hydroxyl radical scavenger (Akashi et al., 2001). Glycinebetaine is a compatible solute with very little capacity to scavenge hydroxyl radicals (Smirnoff and Cumbes, 1988). Mycosporine-like amino acids (MAAs) are important UV-B-screening compounds in many aquatic organisms, with high trophic levels deriving their MAAs from the primary producers (e.g. dinoflagellates, red algae) which synthesize them (Shick and Dunlap, 2002). The MAAs have little capacity to scavenge hydroxyl radicals, unlike the parent, nitrogen-free, gadusols (Shick and Dunlap, 2002). The MAAs are often present at high concentrations in compartments with a high concentration and diversity of proteins, and so must be compatible solutes. Nitrogen-containing compounds also function as relatively grazer-specific anti-grazer agents in higher plants (alkaloids) and as flower, fruit and seed pigments in the Chenopodiales (Centrospermae) (as betalains) (Harborne, 1993).
With the obvious exception of nitrogen storage, these functions can also be performed by nitrogen-free compounds (Table 2). Compatible solutes lacking nitrogen include sugar alcohols, non-reducing sugars, and the sulphur-containing dimethylsulphoniopropionate (DMSP). These compounds also act as effective free radical scavengers and, when stored in vacuoles where they can be replaced by inorganic solutes, the sugar derivatives can act as organic carbon and energy stores. DMSP is broken down using a lyase enzyme (with replacement by de novo synthesis), resulting in acrylate and dimethylsulphide (DMS), both of which are very effective hydroxyl radical scavengers (Sunda et al., 2002). Oxidation products of DMS, first dimethylsulphoxide (DMSO) and then methanesulphinic acid, are also very effective hydroxyl radical scavengers (Sunda et al., 2002).
Ascorbate is another sugar derivative that is an effective hydroxyl radical scavenger, as well as being involved as an enzyme substrate in peroxidases. Ascorbate is often present at high enough concentrations in plastid stroma to qualify as a compatible solute. The sugar derivatives and DMSP (and derivatives) are not UV-B absorbers, but there are a range of nitrogen-free compounds which act as UV-B absorbers. Examples are phlorotannins in the brown algae (which lack MAAs) (Schoenwalder, 2002), coumarins in some green marine macroalgae which have very low contents of MAAs (Gomez et al., 1998; Perez-Rodriguez et al., 1998, 2001) and various soluble phenylpropanoids (mainly flavonoids) in higher plants.
Chemical categories which contain UV-B-absorbing nitrogen-free compounds can also function in optical signalling to animals (anthocyanins in flowers, fruits and seeds) and as grazing deterrents (tannins in higher plants; phlorotannins in brown algae). Terpenoids are also involved as feeding deterrents in algae and higher plants (Paul et al., 1993; Meyer and Paul, 1995).
The two preceding paragraphs show that, with the exception of storage of nitrogen as low Mr organic nitrogen compounds, essentially all of the functions considered can be performed by nitrogen-containing or by nitrogen-free low Mr organic compounds. The functions are storage, compatible solute, UV-B screening and optical signalling, defence against grazers and parasites, and free radical scavenging (Table 2).
Taxonomic distribution of nitrogen-containing and nitrogen-free low Mr organic compounds
There are considerable differences among taxa with respect to the use of nitrogen-free and nitrogen-containing low Mr organic compounds. Marine red algae use MAAs as their UV-B screening compounds, but otherwise use nitrogen-free low Mr organic compounds (Karsten et al., 1994; Bischoff et al., 2000). However, marine red algae also use phycobilins in their light-harvesting complexes; these are more nitrogen-costly than chlorophyll-based light-harvesting complexes. Furthermore, perhaps a third of marine macroalgae rely on diffusive CO2 entry rather than a carbon concentrating mechanism (Raven et al., 2002b, c). Diffusive CO2 entry could involve a greater nitrogen investment in Rubisco plus photorespiratory enzymes than would be involved in a carbon-concentrating mechanism. The red algae have a very low content of nitrogen-containing carbohydrates, or of proteins, in their cell walls. The brown algae are even less involved in nitrogen-containing low Mr organic solutes, since they use phlorotannins rather than MAAs as their UV-B screens. Green algae may have low concentrations of MAAs, and can use proline as a compatible solute, but also operate a relatively low-nitrogen economy for low Mr solutes. Neither brown nor red algae use the nitrogen-costly phycobilins in light-harvesting, and generally have the (probably) nitrogen-thrifty carbon concentrating mechanisms rather than diffusive CO2 entry (Raven et al., 2002b c).
For diatoms the compatible solutes of marine representatives include nitrogenous compounds such as glycinebetaine and homarine, as well as DMSP (Keller et al., 1999a, b). Other clades of algae generally have some nitrogenous compounds (e.g. glycinebetaine) and nitrogen-free compatible solutes such as polyols, glucosyl glycerides or galactosyl glycerides, or DMSP (Keller et al., 1999a, b). Diatoms (even in high osmolarity environments) also lack high concentrations of soluble UV-B absorbing compounds (Beardall and Raven, 2004; Tartarotti and Sommaruga, 2002).
Higher plants lack nitrogen-containing UV-B screening compounds (other than betalains), but some have nitrogen-containing compatible solutes such as proline, glycinebetaine and citrulline.
Quantitative aspects of low Mr organic compounds
Discussion thus far of the occurrence of nitrogen-containing and nitrogen-free low Mr organic solute has been qualitative, insofar as the concentration of each of the often multifunctional solutes in the cells is concerned. Before considering specific cases of the quantity of the solutes in organisms, some generalities can be discussed. The largest relative and absolute range of concentrations of low Mr organic solutes relates to the genotypic and phenotypic variation in total low Mr organic solutes as a function of the 
component of external w. The water depth component of w, i.e. 1 MPa per 100 m, does not require balancing intracellular osmolytes.
Raven (1982, 1995a, 2003) shows that the low intracellular osmolarity of effectively wall-less freshwater cells can be related to the continuous energy expenditure for active water efflux. ‘Effectively wall-less’ means cells with exposed plasmalemma, i.e. all flagellates regardless of whether the cell body has a ‘wall’ or not, and amoeboid cells (Raven, 1982). Raven (1982, 1995a, 2003) points out that the energy cost of active water efflux is, for a given hydraulic conductivity of the plasmalemma, directly proportional to the square of difference in water potential between the cytosol and the medium. Clearly, minimizing the energy cost of volume regulation by active water efflux need not be the prime metabolic priority under all growth conditions, but low osmolarity (as low as 40 osmol m–3) does characterize many freshwater effectively wall-less cells (Raven, 1982). Low intracellular osmolarity means that no compatible solutes can be accommodated, and there are significant restrictions on the concentration of soluble UV-B absorbers, free radical scavengers, and anti-biophage compounds. These constraints on the quantity per unit cell volume of compounds which meet abiotic and biotic challenges have not thus far been characterized as decreasing inclusive fitness relative to walled cells of higher osmolarity in the same habitat. Furthermore, there do not seem to be data on the fraction of total cell carbon and nitrogen in these low-osmolarity cells which is in low Mr compounds rather than macromolecules. The prediction for low osmolarity cells is that a smaller fraction of cell carbon and nitrogen is in low Mr solutes. Inter alia, this would mean that close predictions of cell protein could be obtained from cell nitrogen and a factor derived from the protein:nitrogen mass ratio for phototroph protein (corrected for other nitrogen-containing macromolecules, i.e. RNA and DNA) (Handley et al., 1989; Lourenço et al., 1998). By contrast, a less close prediction would occur for cells with a higher intracellular osmolarity and hence potentially greater concentrations of low Mr storage nitrogen compounds, and nitrogen-containing UV-B screens, free radical scavengers and anti-biophage compounds.
Even when nitrogen in macromolecular nitrogen compounds, and in major categories of low Mr nitrogenous compounds, is accounted for, there can be a significant shortfall of nitrogen in measured components of a cell relative to the total nitrogen in the cell (Keller et al., 1999a, b; Geider and La Roche, 2002).
At the other extreme are photosynthetic cells in very high osmolarity environments (e.g. the Dead Sea; the Great Salt Lake). The only phototrophs able to grow in these habitats are Dunaliella species; these green flagellates are iso-osmotic with the medium, so at least they do not maintain a higher internal than external osmolarity. Nevertheless, these cells have to maintain compatible solute at a concentration sufficient to equal, with other intracellular solutes, the osmolarity outside, involving a concentration of compatible solute of up to 5 kmol m–3. The compatible solute in Dunaliella is glycerol, which, with only three carbon atoms, is the cheapest (in energy and carbon terms) compatible solute to produce; it is still not clear how leakage of glycerol to the medium is maintained at the low level which is observed (Raven, 1984b). Even using this low-carbon (and no nitrogen) compatible solute, the concentration of carbon in glycerol in Dunaliella comprises 60% of the total cell carbon. Other things being equal, the C:N ratio of a freshwater Chlamydomonas (with no compatible solute) should only be half that of its close relative Dunaliella growing in very high osmolarity environments. None of the organisms able to grow at the highest osmolarities tolerated by photolithotrophs has a nitrogen-containing compatible solute. In addition to any shortage of nitrogen in hyper saline waters, there is also the question of the higher energy cost of producing nitrogenous compatible solutes, none of which have fewer than five carbon atoms and one nitrogen atom per molecule, especially if nitrate is the nitrogen source (Raven, 1985).
|
Extracellular low Mr carbon and nitrogen compounds |
|---|
Thus far only the materials retained within the alga or plant have been considered. Organic materials lost by the alga or plant must also have been processed through the resource acquisition and metabolism pathways, and have energetic and catalytic requirements for their synthesis, in addition to those of the organic carbon and nitrogen that are retained by the organism. While both nitrogen-free and nitrogen-containing organic compounds are lost from algae and plants, it is not clear if the ratio in which the elements are lost is less than, equal to, or greater than the ratio in the organism. Some genotypically and/or phenotypically determined losses of organic compounds from algae and plants are now considered.
(1) Secretion of organic anions and protons in relation to acquisition of nutrients by roots of higher plants. Secretion of malate or citrate with protons facilitates the release of phosphate from solid mineral forms (apatite; ferric iron complexes) (Dakora and Phillips, 2002; Lambers et al., 2002; Roelefs et al., 2001; Ryan et al., 2001; Penaloza et al., 2002; Sas et al., 2002). The most extreme example of this is in plants with cluster roots (Skene, 2003), where up to a quarter of the carbon assimilated in net primary productivity can be secreted as citrate (Dinkelaker et al., 1989). Here there is a clear nutritional role in the excretion of organic anions.
(2) Secretion of malate anion by root tips of A1-tolerant genotypes of Triticum aestivum has a role in decreasing A1 toxicity by chelating toxic A1 species (Ryan et al., 2001).
(3) While the role is unknown, Mahmood et al. (2002) showed that sugars were excreted by roots of ammonium-grown Leptochloa fusca to a much greater extent than by roots of nitrate-grown plants. This difference could, if generalizable, help to explain the lower than expected energetic and growth rate advantages of growth on ammonium rather than on nitrate (Britto et al., 2001).
(4) Organic matter losses from shoots include gaseous losses of methanol and of terpenoids (Monson and Holland, 2001; Sharkey and Yeh, 2001; Rosenstiel et al., 2003). These losses can be significant fractions of net primary productivity, with wide genotypic variability. Older ideas that such losses economize on water loss in transpiration by diverting energy dissipation as the latent heat of evaporation of organic materials rather than water are negated by considerations of the large number (hundreds) of water molecules lost in transpiration to fix the carbon needed to produce even a one-carbon volatile organic compound, and the lower latent heat of evaporation per molecule of methanol, and of terpenoids, than of water.
A role for volatile and inflammable organic material can be seen in fire-tolerant plants in fire-prone habitats. Production of an explosive mixture of organic volatiles and air can promote the spread of fire initiated by lightning, destroying fire-intolerant competitors of the fire-tolerant, fire-promoting plants (Gill et al., 1981).
(5) Losses of non-volatile organic solutes from shoots include organic anions with and without protons, e.g. malic acid from shoot glands of Cicer (Raven, 1985). Restriction of losses of inorganic and organic nutrients from shoots may be a function of the thick, waxy cuticle of schlerophyllous/xeromorphic plants (Beadle, 1966; Wright and Westoby, 2003; cf. Lamont et al., 2002).
In addition to these losses of nitrogen-free organic matter, there are also losses of nitrogen-containing organic compounds. As well as amino acid losses from roots with no obvious function, there are also losses of other organic nitrogen compounds which may act as siderophores in cyanobacteria and grasses. Losses of volatile organic nitrogen from land plant shoots include the indoles involved in attracting carrion-eating pollinating insects to flowers and inflorescences, especially the thermogenic influorescences of aroids (Kite, 1995; Thien et al., 2002).
|
Losses of inorganic carbon and nitrogen compounds |
|---|
Major contributions to the overall carbon, but not generally the nitrogen, balance of plants and algae are losses of inorganic compounds. Respiratory and photorespiratory (C3 plants) losses of carbon dioxide account for half or more of gross fixation of carbon dioxide in photosynthesis. Net loss of gaseous ammonia from plants to low-ammonia atmospheres are a small component of the overall nitrogen flux in plant metabolism. Raven et al. (1992a) cite data suggesting that global gaseous ammonia losses from vegetation (natural and managed) do not exceed 2.3 Tmol N per year, i.e. only about 1% of the total primary assimilation of nitrogen by land plants of 209 Tmol N per year (Table 1). In high-ammonia atmospheres (e.g. near seabird colonies, or intensive pig or poultry production units) there can be net uptake of ammonia by plant shoots (Raven et al., 1992a, 1993; Hill et al., 2001). Global gaseous losses of nitrogen from vegetation as nitric oxide are about two orders of magnitude lower than losses as ammonia (Wildt et al., 1997).
|
Interactions of nitrogen nutrition with water relations and carbon dioxide supply |
|---|
For terrestrial plants, water availability can be a significant constraint on growth. CO2 fixation and growth necessarily involves very significant water loss in transpiration. For vascular plants the magnitude of water loss per unit dry matter increases from CAM (50–100 g g–1) to C4 (200–300 g g–1) and to C3 (400–500 g g–1). The extent of water loss per unit dry matter gain can, in some plants, vary with the form and concentration of nitrogen sources. Most of the experimental observations and quantitative modelling has involved C3 plants. Here there is some mis-match between theoretical predictions and experimental observations. Theoretical considerations suggest that the order of increasing water cost of dry matter gain should be ammonium < nitrate < dinitrogen (Raven, 1985). However, measurements of water loss per unit dry matter gain rarely show lower values for ammonium than for nitrate as the nitrogen source (Raven et al., 1992b, 1993; Yin and Raven, 1998). For a number of dicotyledonous crops the water lost per unit dry matter gain is greater for plants grown on ammonium than on nitrate; however, at least part of the greater water cost for growth on ammonium could be related to slower growth on the reduced than on the oxidized nitrogen source (Raven et al., 1992a, b, 1993; Yin and Raven, 1998). For dinitrogen fixation, fewer data are available; while the available information suggests that the water cost of dry matter accumulation is greater for diazotrophically growing Phaseolus vulgaris than when this plant is growing on combined nitrogen (Allen et al., 1988). Data on the
13C of plant organic matter, as an indicator of water loss in transpiration per unit carbon gain in gross photosynthesis, suggests that there is also a higher water cost in diazotrophic plants of Casuarina equisetifolia than in otherwise similar plants growing on combined nitrogen (Martinez-Carrasco et al., 1998) The comparisons of theoretical and observed values considered all involve laboratory studies, and mainly concern different forms of nitrogen. Field observations are also available (Patterson et al., 1997) which indicate the occurrence, genotypic and phenotypic, of trade-offs between water costs of growth and nitrogen costs of growth.
The hydrological cycle has an important role in determining nitrogen supply to photosynthetic organisms. Waterlogging of soils leads to nitrate loss by denitrification, while through-flow removes nitrate. Non-water-logged soils promote nitrification of mineralized organic matter. The hydrological cycle also removes organic and inorganic carbon and nitrogen from terrestrial habitats to inland waters and, ultimately, marine habitats (Berner and Berner, 1996; Falkowski and Raven, 1997; Raven and Falkowski, 1999). The atmosphere also provides a conduit, albeit minor, for combined nitrogen exchange between terrestrial and aquatic habitats (Raven et al., 1992a, 1993). In the ocean, the thermohaline circulation is important in determining the supply of nitrate to surface waters from deep waters where nitrate is generated from sedimented organic nitrogen by mineralization and nitrification (Falkowski and Raven, 1997).
An unequiv