3-1. Investigation of NPF genes for nitrate distribution among leaves
3-1-1. Bioinformatics research to identify potential candidates for further study.
To find out which genes might involve in regulating nitrate distribution in leaves,
firstly I used public databases to narrow down my targets. The main database used was
AtGenExpress Consortium (Arabidopsis eFP Browser,
http://bbc.botany.utoronto.ca/efp/cgi-bin/efpWeb.cgi). Data from the eFP Browser are
presented pictographically in Supplemental Figure 12. From this information, the gene
expression pattern in various tissues for each NPF family member could be found. Among
53 NPF family members, the level of gene expression in leaves can vary from 1 to 876,
including young leaves and old leaves, while 3 of the family members, At1g72120
(NPF5.14), At1g72130 (NPF5.11), and At3g54450 (NPF5.4), failed to be detected by this
method. To screen for candidate genes involved in nitrate distribution in leaves, a cutoff
level of gene expression in leaves at 100 was chosen to distinguish those genes that have
higher expression in leaves from others. Under this criterion, 18 NPF members were
chosen to be the candidates and subjected to the following nitrate content analyses (Table
1).
3-1-2. Three more NPF members were selected based on Massively Parallel Signature
Sequencing (MPSS) and RT-PCR results presented in a review article.
To further confirm the selection and furthermore to identify genes that cannot be
found in Arabidopsis eFP Browser, another public database, Massively Parallel Signature
Sequencing (MPSS, https://mpss.danforthcenter.org/dbs/index.php?SITE=at_sRNA),
and the data presented in a review article (Tsay et al., 2007) were examined.
From MPSS database, 28 out of 53 genes have signatures in leaves (Table 2). Among
the 28 genes, even all of the genes showed signatures in leaf, however, some of the
abundances of signature in leaf were too low compared to other tissues. Therefore, for
the candidates chosen, the abundances of signature in leaf at least had to be 40% of that
in the most abundant tissue. These 14 candidates genes were marked with light gray in
Table 2.
On the other hand, in our previous study (Tsay et al., 2007), the tissue-specific
expression pattern of the 53 genes was normalized with UBQ10 and shown as circles with
sizes proportional to the percentage in each tissue (Supplementary Figure 13). In
comparison of shoots and roots, the same criterion was applied that the value in shoots
had to be higher or at least 40% of that in roots to be chosen, and with this criterion, 16
genes were choseon, marked with light gray in Supplementary Figure 13.
Taken together, as e-FP Browser showed detailed information of individual leaf, the
genes selected based on e-FP Browser were all adopted, while the overlapping candidates
selected from MPSS and RT-PCR results was also adopted in the following experiments.
(Figure 1)
3-1-3. 19 candidates were chosen to measure nitrate content and biomass among leaves
According to the public databases mentioned above, we narrowed down to 21 NPF
members; however, NPF1.1 and NPF1.2 together have been reported to participate in
nitrate redistribution, so here they were excluded from the nitrate content experiment.
Mutants of the 19 candidates were randomly separated into batches and grown
hydroponically supported with 2 mM KNO3 (Supplementary Figure 14). Leaves were
collected at 17th day for all batches, except that in batch 8 were collected at 18th day. In
each batch, wild type was included as control. The medium of batch 3, 4 and 5 was
refreshed every four days, while other batches’ were refreshed twice a week, and this
resulted in the smaller plants in batch 3, batch 4 and batch 5 (Supplementary Figure 15B).
The biomass of these three batches was smaller than other batches and the seventh leaf
was not able to be collected. As for batch 8, due to the difference of collecting day, the
dry weight was higher in all the leaves, especially the big and young leaves
(Supplementary Figure 15B). Nevertheless, the pattern of nitrate content was still
comparable in all batches; the nitrate content was higher in old and big leaves (L1 to L4),
and it was lower in young leaves (L5 to L7) (Supplementary Figure 15A).
To find out if the mutants of NPFs can consistently show the defects in nitrate
distribution among leaves, the data were extracted from different batches and examined
carefully. The results of all the mutants can be divided into three categories: no difference,
slight difference, and consistent difference compared to wild type.
For the first category that had no difference between mutant and wild type, npf3.1-1
showed no difference in both nitrate content and dry weight compared to wild type in all
three batches, batch 1, batch 2, and batch 8 (Figure 2). Similarly, the nitrate content of
npf5.1-1 in both batches was similar and had no difference compared to wild type (Figure
3A and 3C). npf5.7-1 mutant was examined in batch 9 and 10, and in both batches, the
nitrate content and biomass were identical to those of wild type (Figure 4). Another
mutant, npf5.8-1, also showed the same pattern that the nitrate content in batch 5 and
batch 8 both showed no difference compared to wild type (Figure 5A and 5C). npf5.10-1
had smaller dry weight in the third leaf in batch 8 (Figure 6D) but had not in batch 6
(Figure 6B), and the mutant showed lower nitrate content in the oldest leaf in batch 6 but
without statistical significance (Figure 6A). Overall, npf5.10-1 had no difference
compared to wild type (Figure 6). Data from two independent experiments showed that
in nitrate content and biomass, npf5.15-1 showed similar patterns as wild type did (Figure
7). In gtr1 mutant, the patterns of nitrate content and dry weight in both batches, batch 9
and 10, were identical to wild type (Figure 8). In figure 9B, although the dry weight of
sper3-3 seemed to be bigger than wild type, there was no statistical significance in both
nitrate content and dry weight compared to wild type (Figure 9). ptr1-1 mutant was
performed twice in batch 6 and batch 10. In batch 6, there was no difference between wild
type and ptr1-1 in both nitrate content and dry weight (Figure 10A and 10B), but in batch
10, the dry weight of ptr1-1 was slightly higher than that of wild type, especially in fifth
and seventh leaf, showing statistical difference (Figure 10D). Collectively, ptr1-1 had no
difference in nitrate content compared to wild type (Figure 10).In one batch of experiment,
ptr6-1 mutant showed similar pattern compared to wild type in both nitrate content and
biomass (Figure 11).
As for mutants that had slight difference compared to wild type, nrt1.4-2 had higher
nitrate content in old leaves in batch 8 (Figure 12E), nitrate content in old leaves in batch
2 was slightly higher than wild type but had no statistical significance (Figure 12C), and
in batch 1 it showed similar level compared to wild type (Figure 12A). For dry weight,
there was no any differences compared to wild type (Figure 12B, 12D and 12F). In
summary, nrt1.4-2 overall had no consistent difference compared to wild type (Figure
12). As shown in Figure 13, npf5.11-1 was performed three times in batch 6, 7, and 10,
in which there were at least four different plants for mutants and wild type. However, the
standard deviation of old leaves of nitrate content was extremely large that showed no
statistical difference compared to wild type (Figure 13A, 13C and 13E), but the nitrate
content slightly decreased in big leaves in npf5.11-1 (Figure 13C and 13E), while there
was no difference in dry weight (Figure 13B, 13D and 13F). npf5.12-1 was examined in
three batches, batch 5, 7, and 10. For npf5.12-1, nitrate content of the 6th leaf in batch 7
was higher than that of wild type (Figure 14C); the dry weight of 1st and 3rd leaf in batch
7 and 5th leaf in batch 10 was slightly higher than that of wild type (Figure 14D and 14F).
Taken together, we concluded that npf5.12-1 had no significant difference compared to
wild type (Figure 14). As for ait1-1, the mutant had lower nitrate content in the oldest
leaf in batch 1 (Figure 15A), and this was also observed in batch 4 (Figure 15C) and batch
8 (Figure 15E) but no significant difference. In batch 8, mutant had lower dry weight in
the oldest and the third leaf (Figure 15F), and this was also shown in batch 1 but without
statistical significance (Figure 15B). Despite for the slight difference, ait1-1 overall
showed no dramatic difference compared to wild type (Figure 15). npf6.1-1 showed
higher nitrate content in the biggest leaf in batch 7 (Figure 16C), and this was also
observed in batch 1 but without significance (Figure 16A). Also in batch 8, the mutant
had higher dry weight in big and young leaves (Figure 16F), but this could not be seen in
the other two batches. Taken together, npf6.1-1 overall had no difference compared to
wild type (Figure 16). For ptr2-1, the mutant showed higher nitrate content in the second
leaf but lower nitrate content in young leaves in batch 6 (Figure 17A), while overall the
mutant had smaller dry weight compared to wild type (Figure 17B and 17C).
In the third category, three mutants, nrt1.7-2, chl1-5, and gtr2-1 showed the
difference in independent batches consistently. In two independent batches, nrt1.7-2
showed higher nitrate content (Figure 18A and 18C) and higher dry weight (Figure 18B
and 18D) in most of the leaves compared to wild type. For chl1-5, in batch 4, it showed
the minor difference of nitrate content in the first and third leaf, but there was no
difference in biomass (Figure 19A and 19B). However, in batch 8 and 10, the mutant
showed a significant decrease in both nitrate content and dry weight consistently. chl1-5
only had about two-thirds of nitrate content in wild type (Figure 19C and 19E), and
biomass was also smaller than wild type (Figure 19D and 19F), especially in batch 8 that
the biomass was almost only half of it in wild type (Figure 19D). From three independent
batches, it came to the conclusion that under this condition, chl1-5 grew smaller and had
lower nitrate content compared to wild type (Figure 19). Interestingly, gtr2-1 showed
lower nitrate content in old leaves but higher nitrate content in young leaves compared to
wild type (Figure 20A and 20C), and although it did not show significance, this pattern
could also be observed in batch 3 (Figure 20E). As for dry weight, only in batch 1 the
third leaf of gtr2-1 was smaller than it of wild type (Figure 20B), other than that there
was no significant difference compared to wild type (Figure 20D and 20F).
In this study, nitrate content and biomass assay among leaves of most 19 npf mutants
were performed at least twice in independent experiments, except that ptr6-1 was only
performed once. Three npf mutants, nrt1.7-2, chl1-5, and gtr2-1 showed different nitrate
distribution patterns and others had similar pattern compared to wild type. It was worth
to note that gtr2-1 showed low nitrate content of mutant in old leaves while high in young
leaves compared to wild type (Figure 20), indicating that this gene might be involved in
nitrate remobilization and distribution among leaves. Therefore, it was intriguing for me
to go further and identify the function of this gene.
3-2. Functional analyses of Arabidopsis GTR2
3-2-1. GTR2 was mainly expressed in root, root-shoot junction and young leaves
In order to characterize the expression of GTR2 at vegetative stage, GUS staining of
PGTR2-GUS transgenic lines was performed. Strong staining was presented mainly in the
roots, root tips, and root-shoot junction (Figure 21A). The staining in leaves was relatively
faint and could only be observed under a microscope. The GUS signal in shoot part was
only shown in major veins in young leaves (Figure 21E to 21G), and was not shown in
cotyledon, old leaf, and big leaf (Figure 21Bto 21D).
3-2-2. More 15N was accumulated in young leaves of gtr2-1
As shown in Figure 20, when continuously grown with 2 mM KNO3, gtr2-1 mutant
showed higher nitrate content in young leaves and lower in old leaves compared to wild
type, indicating that GTR2 might be involved in nitrate distribution among leaves. To
understand how GTR2 participates in nitrate distribution, 15NO3- allocation assay was
performed by feeding 15NO3- to the root for 5 minutes, and then the 15N concentration in
individual leaves was analyzed. Compared to wild type, the 15N concentration was lower
in old and big leaves (L1 to L4) of gtr2-1 mutant, despite no statistical significance
(Figure 22). In contrast, 15N concentration in the young leaf of gtr2-1 mutant showed 1.5
times higher than that of wild type (Figure 22). Both long-term nitrate accumulation and
short-term 15N distribution are changed in the mutant, so these results suggested that
GTR2 might participate in regulating nitrate distribution among leaves, while nitrate is
transported from root via xylem.
3-2-3. Expression of GTR2 in young leaves is increased at low nitrate condition
To find out the relative expression level of GTR2 in different leaves under high and
low nitrate concentrations, leaves of wild type plants were harvested at 17th day. RNA
expression level was measured by RT-qPCR. As shown in Figure 23, GTR2 expression
level was higher under low nitrate concentration condition, while the expression level of
high nitrate concentration was lower. Under low nitrate concentration, GTR2 expression
level in the youngest leaf was almost 4-fold higher than that in the oldest leaf. As for high
nitrate concentration, GTR2 expression level had no dramatic differences in all leaves.
Taken together, the results indicated that GTR2 had higher expression level under low
nitrate concentration, participating in the young leaves.
3-2-4. Nitrate content in young leaves was higher under different nitrate
concentrations in gtr2-1
Since the expression of GTR2 is higher at low nitrate, the influence of GTR2 on the
nitrate content at low nitrate condition was examined in wild type and gtr2-1 mutant
grown with 0.2, 2, and 10 mM KNO3 mediums at pH 5.5.
As shown in Figure 24A, 24C, and 24E, for all the three concentrations tested, 0.2
mM, 2 mM, and 10 mM, the nitrate content in young leaves of gtr2-1 mutants were all
higher than wild type. Different from what we are expected from the expression pattern,
the nitrate content differences between wild type and mutant are more dramatic at higher
nitrate concentrations like 2 mM and 10 mM KNO3.
As for dry weight, under 0.2 mM KNO3 condition, it was lower in old leaves in
gtr2-1 mutant (Figure 24B); under 2 mM and gtr2-10 mM KNO3 conditions, the dry weight of young leaf was higher in gtr2-1 mutant compared to wild type (Figure 24D and 24F).
These data indicated that GTR2 might be involved in regulating nitrate distribution
among leaves under low and high nitrate concentration conditions.
3-2-5. Nitrate content dropped faster under starvation in gtr2-1
We are also interested to know if this gene would participate in regulating nitrate
accumulation under starvation condition. Therefore, plants were shifted to
nitrate-depleted medium after growing under 2 mM KNO3 for 17 days, and samples were
collected, before and after shifting, including 4 hours and 24 hours.
The nitrate content before shifting showed the same pattern observed before, as it
was higher in young leaves and lower of it in old leaves in gtr2-1 compared to wild type
(Figure 25A). 4 hours after shifting, the nitrate content in both wild type and gtr2-1
dropped. However, compared to wild type, nitrate content in gtr2-1 declined faster. In old
leaves, although the nitrate content was already lower in gtr2-1 than in wild type before
shifting, the nitrate content in wild type only dropped 22% to 24%, but in gtr2-1, it
dropped 27% to 36%, and this phenomenon was more dramatic in young leaves. In young
leaves, the nitrate content in the mutant was higher than in wild type before starvation,
but 4 hours after starvation, the nitrate content dropped to the same level as wild type,
and even lower than wild type in L7. 24 hours after starvation, the nitrate content of
gtr2-1 in old leaves was only half of that in wild type; while in mature leaves (L3 to L5), nitrate
content was similar in wild type and mutant; and the nitrate was almost undetectable in
young leaves (L6 and L7) in both wild type and mutant (Figure 25A).
For biomass, only in the oldest leaf and the fourth leaf, gtr2-1 was slightly lower
than wild type after 24 hours’ starvation, but overall there was not so much difference.
However, it brought our attention to that after 4 hours and 24 hours starvation, the growth
of mature leaf (L3 to L5) and young leaf (L6 and L7) in wild type was faster than gtr2-1.
In the mutant, the growth was about 0.1 to 0.2 mg in all leaves under starvation, but for
wild type, the growth, especially in mature leaf, could up to almost 0.5 mg (Figure 25B).
In summary, gtr2-1 might grow slower than wild type under starvation, and nitrate
depleted faster in gtr2-1.
3-2-6. Primary root length in gtr2-1 was longer under low nitrate condition
Although we selected GTR2 as one of our candidates based on its relatively high
expression in leaves, the highest expression tissue of GTR2 was roots; therefore, we also
examined root development to see if GTR2 played any roles in that. Seedlings were
grown on plates containing 0.2 mM or 5 mM KNO3, or 5 mM ammonium succinate. The
primary root length was measured from the 4th day after germination.
Under both high and low nitrate concentration plates, the primary root length of
gtr2-1 was shorter at the beginning than wild type, while this difference was not observed
under plates containing ammonia as nitrogen source. gtr2-1 grew faster under low nitrate
concentration plates, in which the primary root length was shorter in mutant compared to
wild type at day 4, but started from day 7, the length of primary root was longer in the
mutant (Figure 26A). On the other hand, gtr2-1 in high nitrate concentration plates did
not show longer root length at the end of this experiment but had similar length with wild
type, though the difference at day 4 was shortened (Figure 26B). As for plants grown on
ammonium plates, there was no difference between mutant and wild type in the period of
this experiment (Figure 26C). In conclusion, the primary root grew longer in gtr2-1 under
low nitrate condition in this experiment, indicating that GTR2 might take part in primary
root growth under low nitrate condition.
3-3. CLCa and CLCb are responsible for nitrate storage in vacuole and might have
impact on nitrate sensing in roots
3-3-1. clca and clcb mutants showed a reduction in nitrate content among all leaves.
It has been reported that CLCa is responsible for nitrate storage in vacuole, and the
nitrate content in clca in both shoots and roots is only 50% of that in wild type (Geelen
et al., 2000), and CLCb was reported to localized in tonoplast as well and had strong
selectivity for nitrate (von der Fecht-Bartenbach et al., 2010).
In Figure 27A, the nitrate content of clca and clcb was similar to previous research,
in which the nitrate content was dropped dramatically in clca in all leaves, and the nitrate
content of clcb was merely the same as wild type. Nevertheless, in mature leaves, the
nitrate content of clca/clcb dropped even more than clca (Figure 27A). Despite the
dramatic difference shown in nitrate content, the dry weight of wild type and mutants
were identical (Figure 27B). To summarize, CLCa and CLCb do participate in nitrate
storage in the vacuole, as the major and minor one, respectively, but the growth and
biomass of mutants were not affected.
3-3-2. clca/clcb had no dramatic difference compared to wild type in terms of primary
nitrate response under both high- and low-affinity nitrate conditions.
To investigate if internal nitrate content would affect nitrate sensing, primary nitrate
response in wild type and clca/clcb under both low- and high-nitrate conditions, 200 M
and 25 mM, respectively, was performed. In most studies, primary nitrate response only
focuses on roots, but since CLCa and CLCb are mainly expressed in shoots, both roots