Peering into the separate roles of root and shoot cytosolic glutamine synthetase 1;2 by use of grafting experiments in Arabidopsis

ABSTRACT

Cytosolic glutamine synthetase 1;2 plays an important role in the primary nitrogen assimilation in roots. Based on characterization of the knockout mutant gln1;2 we have recently demonstrated that Gln1;2 is also essential for ammonium handling in shoots. Here we built reciprocally grafted plants between wild type (Wt) and gln1;2 in order to separate the root and shoot roles of Gln1;2. Significant reduction in silique number and seed yield were observed in the grafted plants 1;2_shoot/Wt_root relative to Wt_shoot/1;2_root and Wt_shoot/Wt_root. Shoot Gln1;2 thus played a crucial role for seed production. Tracing experiments with ¹⁵N showed that the relative nitrogen remobilization from vegetative organs to seeds in gln1;2 was just as efficient as in the Wt plants. This was the case although the total quantity of nitrogen in gln1;2 was significantly lower compared to that in the Wt. We conclude that the functions of shoot Gln1;2 are primarily associated with internal N signaling for establishment of seed yield capacity rather than with nitrogen remobilization.

Introduction

Cytosolic glutamine synthetase (GS1) is responsible for the primary assimilation of ammonium generated via nitrate reduction in plant roots.⁹ GS1 is also expressed in shoots and may during vegetative growth stages be up-regulated to accommodate requirements for ammonium assimilation during periods with high ammonium supply.³ The most dramatic up-regulation of GS1 in shoots takes place during reproductive growth stages and is considered to play an important role for nitrogen remobilization from senescing tissues to new sinks such as seeds.¹ Arabidopsis has 5 isogenes for GS1 (Gln1;1-5) and among these Gln1;2 seems to play particular roles for ammonium assimilation in roots as well as for nitrogen remobilization during seed germination and seed production.² However, the separate importance of the root and shoot functions of Gln1;2 and their mutual interdependence are not known.

Characterization of grafted Arabidopsis plants reveals a relatively greater importance of Gln1;2 in shoots than in roots

In order to separate the function of Gln1;2 in N shoots from its effects in roots, grafted plants with a conjunction of the root from wild type (Wt) plants and the shoot from Gln1;2 knock-out mutant plants, and vice-versa, were built (Fig. 1). This way, the shoot function of Gln1;2 could be further explored without biases from impaired N assimilation in roots. Compared to Wt_shoot/Wt_root, the Wt_shoot/1;2_root grafted plants behaved markedly better than the 1;2_shoot/Wt_root plants, showing that Gln1;2 was more important in shoots than in roots. A relatively greater importance of Gln1;2 in shoots than in roots was also suggested by the fact that both silique number and seed yield were significantly lower in the 1;2_shoot/Wt_root grafted plants relative to the Wt_shoot/Wt_root plants, while Wt_shoot/1;2_root plants did not differ significantly from the Wt_shoot/Wt_root control with respect to these parameters (Fig. 1c, d).

Figure 1.

Characterization of grafted plants with Arabidopsis Wt shoot + Wt root (Wt_shoot/Wt_root), gln1;2 shoot + gln1;2 root (1;2_shoot/1;2_root), Wt shoot + gln1;2 root (Wt_shoot/1;2_root), and gln1;2 shoot + Wt root (1;2_shoot/Wt_root). Wt_shoot/Wt_root and 1;2_shoot/1;2_root were controls. (a) Schematic of grafting to separate Gln1;2 shoot function from its root function. (b) Images of grafted plants. Note that the grafted plants were relatively small and developed relatively weak inflorescences. (c) Number of siliques per plant (n = 4-6). (d) Seed yield (n = 3-6). Arabidopsis seedlings were reciprocally grafted between Wt and gln1;2, kept vertically in long-day conditions for about 7 days, then transferred to pots with sand.^10,12 The surface of the sand was covered by black vermiculate in order to avoid algae. Pots were watered 3 times per week over 24 h by immersion of the base of the pot in nutrient solutions containing 2 mM NH₄NO₃together with all other essential plant nutrients as specified in Guan et al. (2016).³ Results are means ±SE. Asterisks indicate statistically significant differences between Wt_shoot/Wt_root and grafted plants, determined using Student’s t-test: *, P < 0.05; **, P < 0.01. Values above the columns are expressed relative to Wt_shoot/Wt_root set to 100. Two independent experiments were conducted with similar results.

Surprisingly, the 1;2_shoot/1;2_root grafted plants produced more siliques and higher seed yield than did the 1;2_shoot/Wt_root plants (Fig. 1c, d). A possible explanation may be that the presence of the Wt root with the Gln1;2 isozyme caused excessive N transport from roots to shoots, thereby overloading the shoot and impeding silique development (Fig. 1c, d). In the grafted 1;2_shoot/1;2_root plants, less N may have been translocated from the roots to the shoots, imposing a less severe stress on the shoots. It is well known that excessive N promotes cytokinin biosynthesis and vegetative growth, thereby delaying the development of reproductive organs.⁵

The important function of Gln1;2 in promoting shoot growth is not limited to generative growth stages as shown by the fact that the knockout mutant gln1;2 produced significantly lower mass of rosette leaves and other shoot components than Wt plants during vegetative growth.³ The role of Gln1;2 may also be affected by the form of nitrogen supplied to plants because Gln1;2 expression is enhanced by ammonium supply relative to nitrate as the sole N source.^3,4

¹⁵N tracing experiments with the gln1;2 mutant to investigate the role of Gln1;2 for nitrogen remobilization

The results obtained in the above grafting experiments show that shoot Gln1;2 plays a crucial role for seed production in Arabidopsis. This essential shoot function of Gln1;2 is corroborated by previous results using non-grafted single and double mutant plants.² However, it is not known if the critical functions of Gln1;2 are primarily associated with nitrogen remobilization or with N signaling. A high flux of N to juvenile inflorescences is essential for ensuring that a high number of the initiated primordia develop into fertile siliques with a high number of seeds.^6,11 Gln1;2 may be critical for controlling the flux of N to the developing inflorescences and siliques within a fairly narrow time window in which the final yield components are established.

In order to study the role of Gln1;2 in N remobilization, we labeled wild type and gln1;2 mutants with ¹⁵N in the vegetative growth stage, after which flowering was induced and ¹⁵N labeling stopped. The amount of ¹⁵N contained in each tissue that emerged during the subsequent chase period was considered as the quantity of N that became remobilized N during reproductive growth stages.⁸ Wild type plants were much larger and contained significantly more ¹⁵N than the gln1;2 mutant at maturity (Fig. 2). However, the proportion of ¹⁵N in different organs of gln1;2 and Wt plants was nevertheless similar (Fig. 2). Thus, Gln1;2 did not affect the relative pattern of N distribution which is in agreement with Lothier et al.⁷ This suggests that the primary role of Gln1;2 may be related to internal N signaling and establishment of the actual yield capacity via ensuring sufficient N fluxes during critical growth stages rather than to the efficiency of N remobilization during generative growth stages.

Figure 2.

Plant dry weight, nitrogen content and ¹⁵N content in gln1;2 Arabidopsis plants expressed relative to the wild type plants (column pairs 1, 3 and 5), the proportion of total plant dry matter and nitrogen content allocated to seeds (column pairs 2 and 4) and the relative ¹⁵N distribution in different tissues (column pairs 6 to10). Plant DW: 100 corresponds to 5.6 g per plant (n = 3); total plant N content: 100 corresponds to 142 mg per plant (n = 3); Total plant ¹⁵N content: 100 corresponds to 594 µg ¹⁵N per plant (n = 3). Results are means ±SE (data re-calculated from Guan et al. 2015).²

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Funding

This work was carried out within the ERA-NET Coordinating Action in Plant Sciences (ERA-CAPS) program, project ERACAPS13.089 Root Barriers, with financial support from Innovation Fund Denmark 4084-00001B to J.K.S. M.G. received additional support from the National Natural Science Foundation of China (NSFC), project 31600251.