Pyrimidines are particularly important in dividing tissues as building blocks for nucleic acids, but they are equally important for many biochemical processes, including sucrose and cell wall polysaccharide metabolism. In recent years, the molecular organization of nucleotide biosynthesis in plants has been analyzed. Here, we present a functional analysis of the pyrimidine de novo synthesis pathway. Each step in the pathway was investigated using transgenic plants with reduced expression of the corresponding gene to identify controlling steps and gain insights into the phenotypic and metabolic consequences. Inhibition of expression of 80% based on steady-state mRNA level did not lead to visible phenotypes. Stepwise reduction of protein abundance of Asp transcarbamoylase or dihydro orotase resulted in a corresponding inhibition of growth. This was not accompanied by pleiotropic effects or by changes in the developmental program. A more detailed metabolite analysis revealed slightly different responses in roots and shoots of plants with decreased abundance of proteins involved in pyrimidine de novo synthesis. Whereas in leaves the nucleotide and amino acid levels were changed only in the very strong inhibited plants, the roots show a transient increase of these metabolites in intermediate plants followed by a decrease in the strong inhibited plants. Growth analysis revealed that elongation rates and number of organs per plant were reduced, without large changes in the average cell size. It is concluded that reduced pyrimidine de novo synthesis is compensated for by reduction in growth rates, and the remaining nucleotide pools are sufficient for running basic metabolic processes.
Pyrimidine nucleotides are abundant molecules with essential functions in a multitude of biochemical processes. They are of particular importance in dividing and elongating tissues as building blocks for nucleic acid biosynthesis. In addition, as an energy source or precursors for the synthesis of primary and secondary products, they are participants in various metabolic processes. In particular, the pyrimidine nucleotides are directly involved in plant carbohydrate metabolism providing the energy-rich precursor UDP-Glc for many synthetic reactions, such as Suc and cell wall biosyntheses.
In recent years, the basic processes of plant nucleotide de novo synthesis have been analyzed in some detail (Giermann et al., 2002; Moffatt and Ashihara, 2002; Boldt and Zrenner, 2003; Stasolla et al., 2003; Kafer et al., 2004). Plants utilize the same reactions as those found in other organisms. The so-called orotate pathway of pyrimidine de novo synthesis is defined as the formation of UMP from carbamoyl phosphate, Asp, and phosphoribosyl pyrophosphate. The pathway consists of the six enzymatic reactions as shown in Figure 1. The first reaction is conducted by the Gln hydrolyzing carbamoyl phosphate synthase. This enzyme is not unique to the pyrimidine pathway but is also involved in Arg biosynthesis. It is composed of two different subunits (carbamoyl phosphate synthetase carB large subunit and Gln amidotransferase carA small subunit). The next three steps are catalyzed by three separate enzymes (Asp transcarbamoylase, dihydro orotase, and dihydro orotate dehydrogenase). The fifth and sixth steps are catalyzed by a single, bifunctional enzyme (UMP synthase).
From our molecular studies in tobacco (Nicotiana tabacum) and potato (Solanum tuberosum; Giermann et al., 2002), using exhaustive cDNA cloning and genomic Southern analysis, we showed that there is only one gene present for each single protein subunit. This is supported by analyses of the whole Arabidopsis (Arabidopsis thaliana) genome (Arabidopsis Genome Initiative, 2000) and all publicly available sequences from potato, tobacco, and Lycopersicon esculentum (http://www.ebi.ac.uk; http://www.ncbi.nlm.nih.gov), where in most cases only one gene for each single step per plant haploid genome is found. There are a few exceptions, e.g. the finding of three aspartate transcarbamoylase (ATCase) genes from Pisum sativum (Williamson and Slocum, 1994) and Glycine max or two different cDNAs for uridine monophosphate synthase (UMPSase) from Oryza sativa. From these findings, we conclude that pyrimidine de novo synthesis is an essential process and that complete loss of any one function leads to death of the plant. In order to perform a functional analysis of the whole pyrimidine de novo synthesis pathway, we therefore decided to use antisense and cosuppression strategies to create sets of plants with a progressive stepwise reduction in the expression of each single gene involved in the pathway. These transgenic plants with different levels of expression of a single protein have the advantage over complete knockout mutants with probably lethal phenotypes because analysis of consequences
of reduced expression below a certain threshold is possible.
There are many reports on transgenic plants in which stepwise reduction of one activity has given valuable insight into the in vivo function of specific enzymes in a biochemical pathway (e.g. Stitt and Sonnewald, 1995; Haake et al., 1998). When the impact of changes at different steps within one pathway is compared, substantial differences can also be found for the susceptibility of the pathway to a change in expression at different enzymes (Stitt and Sonnewald,1995; Strand et al., 2000). In this article, we present the first functional analysis of the whole pyrimidine de novo synthesis pathway, investigating each step in the pathway by using transgenic plants with reduced expression of a single gene. Selected transformants that show a growth reduction are then analyzed to provide insights into the phenotypic and metabolic consequences of reduced pyrimidine de novo synthesis.
Production and Selection of Transgenic Plants with Reduced Expression of Steps Involved in Pyrimidine de Novo Synthesis The cloning and characterization of all cDNAs from tobacco or potato used in this analysis have been reported by Giermann et al. (2002). All constructs used to transform tobacco or potato plants were made with
the binary vector pBinAR (Ho¨fgen and Willmitzer, 1990), a derivative of the pBin19 vector containing the 35S promoter of the cauliflower mosaic virus for constitutive expression of the target gene and the octopine synthase polyadenylation signal. Sense constructs designed for cosuppression and antisense constructs designed for antisense inhibition contained at least 70% of the full-length cDNA sequences. Table I summarizes the cDNA sequences that were used to produce these constructs for plant transformation. Either N. tabacum L. cv SNN or S. tuberosum L. cv Desire´e and cv Solara or both species were used as target organisms For each targeted enzymatic step, .60 transformed plants were regenerated and analyzed. A complete summary of the expression levels found in all 538 transformants is provided in Table II. The expression levels are given in percentage of wild type in rounded numbers, and the proportion of plants showing a particular reduction is given. Plants were initially analyzed by measuring mRNA and/or protein abundance. Representative results from each analysis of differences in mRNA expression, using northern hybridizations, and protein abundance, using western analysis, are presented in Figure 2.
For screening purposes, we prefer the analysis on protein level if appropriate antibodies are available because it is faster and closer to the expected difference in protein abundance and activity. Only in case of both subunits of carbamoyl phosphate synthase (CPSase) did we initially screen mRNA and protein level. We found reduced steady-state mRNA levels of 50% to 80% in 7% to 15% of the transformants and a reduction of .80% in 0% to 2% of the transformants. In case of the RCKTs plants, where a full-length construct was used, the values for reductions on steady-state mRNA level caused by cosuppression might be underestimated because we have not analyzed the appearance of small interfering RNAs to visualize that posttranscriptional gene silencing has
already started although high levels of steady-state mRNAs are still present (Fig. 2; line RCKTs 44 has .100% mRNA and 80% protein compared to line RCKTs 31 with 30% mRNA and 80% protein). Such a scenario has been reported for other transgene expression studies (Schubert et al., 2004). Furthermore, as the analysis of the RCTs and RCTa lines revealed no clear correlation between the steady-state mRNA and the protein amounts in lines with reduced expression (Fig. 2; line RCTs s39 has 15% mRNA and 25% protein compared to line RCTa a1 with 30% mRNA and 25% protein), we opted for screening on protein abundance. Only in cases where this was not possible because no suitable antibody was available did we use the analysis on steady-state mRNA.
Sense (cosuppression) and antisense constructs led to a similar inhibition. A reduction of 50% to 80% of expression at the mRNA level was found in 16% of the 187 species-specific sense transformants (RCKTs,
RCTs, and RDTs) and in 20% of the 237 antisense transformants (RCTa, RDTa, and RUTa). A reduction of .80% of expression at the mRNA level was found in 2% of the sense and the antisense transformants. At the protein level, a reduction of 50% to 80% was found in 7% of the sense and the antisense transformants. A reduction of .80% at the protein level was only seen in 2% of the plants when antisense transformations of potato with potato cDNAs were performed. These values match results from comparable studies with enzymes in pathways (Stitt and Sonnewald, 1995; Zrenner et al., 1995, 1996; Haake et al., 1998).
In the cases where full-length constructs were used in sense orientation (RCKTs, RBPs, and ROSs) a small number of overexpressors could be found. In 8% of the species-specific RCKTs transformants, a steady-state
mRNA level of .100% was found, but there was no increase at the protein level (Table II, Fig. 2). In the line
ROSs, where the potato dihydro orotase (DHOase) is expressed in tobacco, an increase of steady-state mRNA was found in 60% of the plants (Table II, Fig.2). However, it was not possible to detect DHOasespecific proteins in the tobacco transformants (data not shown). In the RBPs transformants, where the potato ATCase is expressed in tobacco, 3% of the plants showed an increase in ATCase-specific protein (Table II, Fig. 2). All these overexpression lines are not further analyzed within this study.
As already mentioned, there is only one gene per enzymatic function involved in pyrimidine de novo synthesis present in Solanaceous species. Therefore, a reduction in expression is expected to cause some differences in plant growth and metabolism. All plants that showed ,20% of expression at the steady-state mRNA level or at the protein abundance level compared to wild-type plants were analyzed further.
Tobacco Plants with Reduced Expression of CPSase, Dihydro Orotate Dehydrogenase, and UMPSase Do Not Show Phenotypical Differences.
Primary transformant 13 of line RCKTs showed a reduction of 60% in the protein amount of the small subunit of CPSase (Fig. 2B), but no reduction in the large subunit of CPSase (data not shown) was observed. It is known from studies in Escherichia coli that the large subunit of CPSase is less stable when the small subunit is missing (Guillou et al., 1989; Cervera et al., 1993), but the coregulation of protein expression and stability of both subunits of CPSase of plants is not a subject of this study. Primary transformants RCTs 39 and RCTs 44 showed reductions of 75% and 70% in the protein amount of the large subunit of CPSase (Fig.2B), respectively, but no reduction in the small subunit of CPSase (data not shown). These lines with reduction in CPSase-specific steady-state mRNA accumulation and CPSase protein did not show any obvious phenotypic differences when compared to wild-type plants or transformants without reduction in expression (data not shown). Analysis of the offspring (12 to 16 plants) of these lines confirmed that there were no significant differences in height, leaf number, or the above ground fresh weight when grown under standardized conditions in growth chambers on sand culture for 25 d (data not shown). Thus, a reduction of .80% in mRNA expression level and 60% and 75% on protein level of the small or large subunit of CPSase does not lead to visual phenotypic changes.
Lines RDTs 69 and 41 and lines RDTa 12 and 9 showed the largest reduction (.85%) in dihydroorotate dehydrogenase (DHODH)-specific steadystate mRNA accumulation. They did not show anyobvious phenotypic differences when compared to wild-type plants or transformants without reduction in expression (data not shown). This result was confirmed by analyzing plant height, leaf number, and shoot fresh weight in the offspring of these lines.
Among all antisense transformants using the cDNA of UMP synthase, no plants with .70% of reduction in steady-state mRNA accumulation could be found. Line RUTa 31 showed the largest reduction in UMPSase- specific steady-state mRNA accumulation (.70%) and lines RUTa 24 and 25 showed intermediate reduction (60%). These lines did not show any obvious phenotypic differences when compared to wild-type plants or transformants without reduction in expression (data not shown). This result was confirmed by analyzing plant height, leaf number, and shoot fresh weight in the offspring of these lines.
Potato Plants with Reduced Expression of Either ATCase or DHOase, the Second and Third Steps of Pyrimidine de Novo Synthesis, Show Decreased Growth
As already pointed out earlier, only in the case of the antisense transformations of potato using potato cDNAs was there a reduction of .80% in protein level observed (Table II). These potato plants transformed with antisense constructs of either ATCase or DHOase were analyzed further, including lines with a reduction of .80% in protein level and plants from the same transformation series with a less severe reduction in protein amount. As explained in the beginning, there is no clear correlation of steady-state mRNA data and remaining protein amounts, dependent on specific parameters like mRNA stability, translation efficiency, or protein stability. In order to obtain a reliable picture of what amount of the specific enzyme is left in the transgenic plants, we included the results on protein
level, as these are the more significant data. A progressive reduction in the ATCase (Fig. 3A) or DHOase (Fig. 3B) protein abundance in source leaves was linked to a corresponding reduction in plant size (Fig. 3, C and D). Decreases in plant height were already apparent when 50% to 60% of the ATCase or DHOase protein was missing. Plants were usually propagated and maintained in tissue culture on Murashige and Skoog medium containing 2% Suc for 3 weeks and were than transferred to sand culture and watered daily with nutrient solution. After about 3
weeks on sand culture, when growth differences were already established, no other pleiotropic effects were visible (Fig. 3, C and D); no chlorosis or necrosis could be detected, and chlorophyll content and anthocyanin content on a leaves fresh weight basis were unchanged (data not shown; anthocyanin only increased in ROSa 40). When DHOase protein amount was ,10% of that in wild-type plants, growth was severely inhibited after transfer to soil or sand, and the transformants eventually died without flowering or tuberization.
Detailed Growth Analysis of Potato Plants with Reduced Expression of Either ATCase or DHOase.
More information about the inhibition of growth in lines with reduced expression of either ATCase or DHOase is provided in Figure 4.With reduced protein abundance of ATCase (or DHOase) the total fresh weight and also the plant dry weight was reduced by the 23rd day (or 19th day for DHOase) after the plants were transferred to sand and watered daily with nutrient solution (Fig. 4, A and B). Under the applied conditions, root growth on a fresh weight basis was even more reduced than shoot growth (data not shown). At the indicated time of harvest, the total leaf area per plant decreased with abundance of ATCase or DHOase protein, and the leaf number was also reduced, although to a lesser extent (Fig. 4, C and D).
Depending on the growth conditions, there were no obvious differences in the development of plants with or without reduction in the amount of ATCase or DHOase protein. The plants started to tuberize at approximately the same time, but with decreasing ATCase or DHOase protein the tuber yield per plant was reduced at the end of the growing period. Although line RBPa 51, with 40% of ATCase protein remaining, showed no reduction in tuber yield or tuber number per plant, lines RBPa 8, 53, and 19 showed a decrease in yield of approximately 15%, 20%, and 85% and
a concomitant reduction in tuber number per plant of approximately 15%, 25%, and 75%, respectively, when compared to wild-type plants grown for 15 weeks in 1.5-L pots in sand culture (Fig. 4E). Comparable results
were found for the ROSa transformants (Fig. 4F) and also for RBPa and ROSa transformants grown in bigger pots in soil in the greenhouse during summer or winter (data not shown).
Metabolite Analysis Reveals Different Responses in Roots and Shoots of Plants with Decreased Abundance of Proteins Involved in Pyrimidine de Novo Synthesis
We investigated whether reduced growth rates of plants with decreased abundance of enzymes involved in pyrimidine de novo synthesis were accompanied by changes in free nucleotide levels. The nucleotide content and nucleotide ratios of leaves and roots were measured in RBPa plants after 6 h of illumination, 23 d after transfer to sand culture. Table III shows a summary of these results. Although there were no striking changes, a reduction in the nucleotide content of leaves from plants showing strongest reduction of ATCase protein was observed, but without any change in the purine-to-pyrimidine ratio. In roots, there was a slight increase in nucleotide levels in the intermediate lines and a decrease in the strongest inhibited line. This change occurs for both adenine and uridine nucleotides. Comparable results were found in leaves and roots for ROSa transformants (data not shown; roots of intermediate lines were analyzed). In general, there were no specific effects measurable on pyrimidine nucleotide
levels, despite the plants having reduced amounts of either ATCase or DHOase protein, which are involved in pyrimidine de novo synthesis. In the few examples where a slight change of pyrimidines was observed, the purine levels were changed in the same way.
We further investigated whether reduced growth rates of plants with decreased abundance of enzymes involved in pyrimidine de novo synthesis were accompanied by changes in carbohydrate levels. The Glc, Fru, Suc, and starch content was measured in leaves and roots of ROSa plants at the beginning and end of the day, in plants 19 d after transfer to sand culture. The results from the end of the day are shown in Figure 5, A and C. There was an increase in both soluble sugars and starch in the leaves, whereas in the roots, the content of all carbohydrates was reduced at every time point analyzed (Fig. 5; data not shown). Essentially the same results were obtained for RBPa plants harvested after 6 h of illumination, 23 d after transfer to sand culture (Fig. 5, B and D).
In order to analyze if reduced growth rates are caused by altered nitrogen and amino acid availability, the metabolite analysis was extended to include the steady-state amino acid content of leaves and roots of RBPa plants harvested after 6 h of illumination, 23 d after transfer to sand culture. The results are summarized in Table IV. No large overall change in amino acid content of leaves was apparent when compared to the untransformed control plants, and only a reduction in the amino acid content of leaves from plants showing strongest reduction of ATCase protein was observed. Although there were no striking changes in amino acid levels from roots, a slight increase in intermediate lines and a decrease in roots from plants showing strongest reduction of ATCase protein was observed. Comparable results were found in leaves and roots for ROSa transformants (data not shown; only roots of intermediate lines were analyzed).
The Elongation Rate and Number of Roots per Plant Is Reduced in Plants with Decreased Protein Abundance of Enzymes in the Pathway of Pyrimidine de Novo Synthesis
One of the marked changes in plants with reduced amounts of protein of enzymes involved in pyrimidine de novo synthesis was a reduction in growth (Fig. 4) and the decreased levels of sugars (Fig. 5) and almost unchanged levels of amino acids (Table IV) in roots.
Figure 6 shows root growth in more detail in these transformants. Shoot cuttings (homogeneous shoot tips of approximately 1 to 1.5 cm length with four leaves) were grown for 7 d on vertical plates in tissue culture on Murashige and Skoog medium containing 2% Suc. Root number and root elongation rate were monitored every 12 h for at least 3 d after the first two roots had been established. A summary of the averaged elongation rates and root number at the end of the observation period is shown in Figure 6 for RBPa and ROSa transformants. Significant decreases in root elongation rates were found when ,25% of the ATCase or DHOase protein remained in the transformants. When elongation rates were reduced, the total amount of roots per plant also decreased. Significant reduction in the number of roots was observed when the root elongation rate was reduced by .50%. As already found for the above-ground organs, where a reduction of ATCase or DHOase protein led to a decrease in the overall plant height, accompanied by a reduction in total area and number of leaves (Fig. 4), the size of the roots was also strongly reduced, followed by a decrease in total root number per plant.
Detailed Analysis of Leaves Reveals That Organ Size and Number Were Reduced Due to a Lower Cell Number but without a Decrease in Cell Size
In order to look for growth differences at a cellular level, the total cell numbers and cell sizes were analyzed in the tips and bases of source leaves of ROSa transformants. To obtain standardized material, the plants were cultivated in a growth chamber on sand and watered daily with nutrient solution. After 3 weeks, when growth differences were fully established (see above), the terminal leaflet of the fourth leaf from the tip of the plant (first leaf is defined as
being longer than 1 cm) was harvested, and discs were cut from the apical and basal part. As shown in Figure 7, the cell number per square centimeter was higher in the leaf base than in the leaf tip, as expected because the leaf base is a growing zone while the leaf tip is more expanded. When DHOase protein was reduced to approximately 20% of wild-type levels, the cell sizes in the tip and the base of the leaves increased, which means that the cells were slightly bigger in both regions. Together with the finding that total leaf size was also reduced (Fig. 4D), growth reduction was also due to smaller organ size because of fewer cells per organ. Comparable results were found for RBPa plants (data not shown).
Plants Expressing Antisense- or Sense-Specific cDNAs Are an Appropriate Tool for a Functional Analysis of Nucleotide de Novo Synthesis
A functional analysis of the whole pyrimidine de novo synthesis pathway was performed by using antisense and cosuppression strategies to create plants with a stepwise reduction in the expression of each single gene involved in this pathway. As discussed in the introduction, each step in this pathway is encoded by a single gene. This may explain why no homozygous Arabidopsis T-DNA insertion line in pyrimidine de novo synthesis has been described yet. So far, only the establishment of tobacco cell cultures with reduced UMPSase activity has been reported (Santoso and Thornburg, 1992, 2000). We conclude that pyrimidine de novo synthesis is an essential process, where
a complete loss of one function will lead to death of the plant. For this reason, sense or antisense constructs were stably transformed into tobacco or potato plants to obtain a decrease in expression of each gene required for the pathway.
We found reduced steady-state mRNA levels of 50% to 80% in 16% to 20% of the transformants and a reduction of .80% in 2% of the transformants. Similar results were obtained, irrespective of whether the sense or antisense strategy was used. In cases where full-length constructs were used, the values for reductions caused by cosuppression using the sense strategy might be underestimated because we have not analyzed the appearance of small interfering RNAs to visualize that posttranscriptional gene silencing has already started although high levels of
steady-state mRNAs are still present. The analysis of inhibition on mRNA level is also imprecise as no clear correlation of the remainder of steady-state mRNA and the remaining protein amounts exist when posttranscriptional gene silencing has started (Schubert et al., 2004). At the protein level, a reduction of 50% to 80% was found in 7% of all transformants, and in 2% of the plants, a reduction in protein level of .80% was found. These values match results from comparable studies with enzymes in pathways (Stitt and Sonnewald, 1995; Zrenner et al., 1995, 1996; Haake et al., 1998). In the case of the enzymes of the pyrimidine pathway, the relatively small reduction
of transcript and protein is clearly due to an incomplete inhibition of expression because each of these genes is encoded by a single gene. The antisense strategy gave us a whole range of transformants, including both moderately and severely affected lines, and was therefore more useful than RNA interference techniques (Waterhouse et al., 1998), which usually only give plants with strong repressed expression. Plants showing intermediate reduction in expression with a progressive increase in phenotypic differences are a better tool for a more detailed analysis of physiological and developmental consequences of reduced nucleotide biosynthesis because fewer secondary or pleiotropic responses are expected to occur than in severely affected plants.
Reduced Steady-State mRNA Levels Do Not Necessarily Lead to Phenotypic Differences When Compared to Plants without Reduced Expression
A strong reduction in steady-state mRNA level could be found for all of the genes, with the exception of UMPSase. However, the impact on the plant phenotype depended upon which enzyme was inhibited. Plants with 80% reduction of transcript for both CPSase subunits and DHODH did not show any obvious phenotypic change with respect to plant growth and development. Although reduced steadystate mRNA levels are not necessarily reflected in lower protein amounts, in the case of both subunits of CPSase, reduced expression could also be monitored at the protein level. When ,20% of the CPSase-specific steady-state mRNAwas present, there was a reduction of 60% to 75% in CPSase-specific protein, without any visible phenotypic differences. It has been argued that expression of enzymes that are subject to complex posttranslational or allosteric regulation can often be strongly reduced without a change in pathway flux because feedback loops that decrease in vivo enzyme activity are relaxed when less protein is present (Stitt and Sonnewald, 1995; Stitt, 1999).
It is known that CPSase enzymes from various sources are regulated allosterically (Delannay et al., 1999; Simmons et al., 1999), and one might speculate that relaxation of this feedback loop compensates for the reduced amount of CPSase protein. In the case of plant DHODH, nothing is known about posttranscriptional regulation. The plant enzyme has only recently become accessible for in vitro characterization, and a significant structural difference to the animal enzymes has been reported (Ullrich et al., 2002). A suitable antibody was not available to further characterize the reduction on protein level. When 90% of the DHODH expression was abolished, no phenotypic consequences were obvious at the whole plant level.
Since our previous studies strongly suggest the absence of isoforms with high homology (Giermannet al., 2002), we conclude that only 10% or less of the steady-state mRNA levels of DHODH are necessary for adequate pyrimidine de novo biosynthesis in plants grown under standard conditions. This might be due to posttranscriptional or posttranslational adjustments or to the presence of isoforms with a related dehydrogenase activity but with low sequence homology.
A 50% to 60% reduction in protein amount of the enzymes catalyzing the second and third steps in the pyrimidine de novo synthesis pathway (ATCase and DHOase reaction) led to a clearly visible inhibition of plant growth. Transformants were usually maintained in tissue culture for up to 3 weeks, before transferring plantlets of near identical size and appearance to sand or soil culture. After a further 3 weeks on sand culture, lines with reduced ATCase and DHOase protein abundance showed reduced fresh and dry weight accumulation, reduced height, reduced number and size of leaves, and reduced tuber yield. The qualitative effects of reduced ATCase and DHOase were essentially the same. For both enzymes, a stepwise decrease in protein level was accompanied by a concomitant appearance of phenotypical differences. When the growth rates were plotted against ATCase or DHOase protein amount, a steeper decrease in growth parameters was found for ATCase (Fig. 8). This is the enzyme that catalyzes the first unique step for pyrimidine synthesis. Interestingly, decreased expression of the enzyme that catalyzes the next step, DHOase, led to almost the same consequences for plant growth. This is another example of a protein lacking strong regulatory properties (Washabaugh and Collins, 1984) but not being expressed in large excess (Haake et al., 1998) and exerting high control over the pathway when expression is slightly reduced.
Based on biochemical assays and mutant cell lines, it has previously been proposed that ATCase and UMPSase
are the rate limiting steps for pyrimidine biosynthesis (Lovatt and Cheng, 1984; Santoso and Thornburg, 1998). Our results demonstrate that ATCase is not expressed in large excess, as halving of the amount of protein already limits flux to nucleotide biosynthesis, resulting in a progressive decrease in growth rates. No phenotypical differences were
achieved for UMPSase, but the changes in transcript levels were also smaller for this gene than the others targeted in this study. An RNA interference strategy might be useful to obtain larger reductions in UMPSase steady-state mRNA or protein levels.
Reduced Growth Is Established as Slower Growth Rates and Is Neither Accompanied by Pleiotropic Effects nor by Changes in the Developmental Program
A gradual reduction in the ATCase or DHOase protein abundance below 50% in source leaves was linked to a consecutive reduction in plant size, total plant fresh and dry weight, and leaf area. When growth rates were calculated as fresh weight increase per day and leaf area increase per day and were normalized as percentage of the wild type, a steeper decrease in growth parameters was found with reduced abundance of ATCase protein. Tuberization of potato plants started at approximately the same time, but tuber yield per plant was reduced with decreasing abundance of ATCase or DHOase protein. The reduced growth rates with reduced final organ sizes were accompanied by a decrease in organ number but no big changes in cell size. It is known from cellular studies of tobacco that the final organ and cell size is dependent on nutrient availability (Walter and Schurr, 2005). In contrast to our plants, the reduction of final leaf size is always pronounced as a reduction of final cell size and, depending on the stage of development when the nitrogen limitation occurs, a concomitant reduction of the numbers of cells (Roggatz et al., 1999).
At the metabolite level, there were no major changes, except that carbohydrate levels in leaves were increased and carbohydrate levels in roots decreased with reduction in ATCase or DHOase protein abundance. Despite this, no obvious pleiotropic effects like chlorosis or necrosis were visible. The extractable amounts of free nucleotides in mature leaves were almost unchanged, with only a small decrease in the leaves from the strongest inhibited plants. In roots, there was an increase in nucleotide levels in the intermediate lines and a decrease in the strongest inhibited line. This change occurs for both adenine and uridine nucleotides. As purines and pyrimidines are affected in the same way, it is presumably not a direct effect of the inhibition of pyrimidine nucleotide biosynthesis but an indirect effect due to adjustments in growth and physiology. This explanation is further supported by comparable changes that were also found for amino acid levels in leaves and roots, excluding a general nitrogen limitation for growth.
The rather unaffected free nucleotide pools in fully grown tissues clearly shows that nucleotide levels are maintained in mature tissues. This finding leads to the suggestion that in mature tissue, nucleotide recycling is sufficient to run basic metabolism. It is already known that the expression of genes involved in nucleotide de novo synthesis is under a certain developmental control (Giermann et al., 2002). In addition, under standardized growth conditions, the nucleotide pools seem to be coordinated because changes occur for both adenine and uridine nucleotides. In contrast, this adjustment can be disturbed when feeding high concentrations of nucleotide precursors to mature potato tuber slices (Loef et al., 1999), but these changes might be caused by the rather unphysiological conditions.
It was reported for Arabidopsis seedlings grown in the presence of the potent ATCase transition-state analog inhibitor PALA (Bassett et al., 2003) that there is a general reduction in growth and development, accompanied by chlorosis. These findings were linked to a strong decrease in the extractable amount of total RNA and protein. In our plants with a strong reduction in the amount of ATCase protein, we did not find comparable effects on total extractable RNA or protein amounts (data not shown) and did not observe any chlorosis. Whether these differences are due to the different plant material or due to additional effects of the PALA treatment needs to be investigated further.
In summary, root growth seems to be limited by the decreased supply of carbohydrates rather than nucleotides, whereas in leaves, a small decrease in nucleotide pools is accompanied by an increase in carbohydrate content. Leaves are the organs where carbohydrate synthesis occurs and the decreased leaf growth rate might be limited by nucleotide supply. To further test this hypothesis, a more cell-specific determination of nucleotide pools is needed, especially from dividing rather than elongating or fully grown tissue. The approach should be further confirmed by
looking at expression patterns of cell cycle markers, for example, but this more detailed analysis will be further complicated by the fact that nucleotide pools are strongly dependent on the developmental stage of
the tissue (Meyer and Wagner, 1986).
Each single step of the pyrimidine de novo synthesis pathway exerts different control strength over the whole process. Plants with reduced pyrimidine de novo synthesis compensate for the restricted nucleotide availability with reduced growth rates, organ number, and size. In fully grown tissues, almost no reduction in nucleotide pools could be detected, indicating that nucleotide recycling is sufficient to run basic metabolism.
Tell us your learning needs in detail and get immediate response from qualified tutors