Plants are considered biological factories with their ability of converting solar energy into chemical energy in the form of various commercially valuable products, such as food, biofuel and pharmaceuticals. The yields of these products are directly influenced by the level of nitrogen nutrient supply. However, both biological and industrial nitrogen fixation are energetically expensive and thus managing the nitrogen cycle has been identified as one of the 14 grand challenges by the National Academy of Engineering (NAE). Therefore it is desirable to investigate how plants themselves adapt to nitrogen deficient environment and improve their nitrogen use efficiency (NUE). A powerful tool to study metabolism is isotope-assisted metabolic flux analysis (isotopic MFA), which quantifies intracellular chemical reaction rates (fluxes) via isotopic labeling experiments (ILEs) and subsequent mathematical modeling. In ILEs the labeling patterns of the metabolites can be measured at either isotopic steady state or isotopic instationary state. Between these two methods, collecting data during isotopic instationary state saves experimental time, but is computationally more challenging due to that instationary MFA involves solving ordinary differential equations (ODEs). In this study, we firstly developed an approach that combined the concept of "originomer" with an analytical based solution method to improve computational efficiency of instationary MFA. Simulation results showed that this approach reduced computational time by 23-fold for certain realistic metabolic network. Secondly, we managed to solve an intrinsic problem that affect steady state MFA in fed-batch cell culture environment - the influence of unlabeled biomass that are present before applying isotopic tracers in an ILE. We proposed a full "reflux" metabolic network model that significantly improved the accuracy of evaluated fluxes when compared to the models without "reflux". Finally, we investigated the ability of adapting nutrient deficiencies and the NUE-improving mechanisms in suspension cells of poplar, a woody perennial tree capable of efficiently managing its nitrogen reserves. Through (i) steady-state 13C MFA and (ii) transcripomic profiling via microarray on poplar cells growing under different carbon (C) and nitrogen (N) supply levels, we found a plastidic localization of oxidative pentose phosphate pathway (oxPPP), as well as a lower oxPPP flux under low nitrogen supply. Gene expression data also points to possible NUE improving mechanisms employed by poplar cells. We hope this study will shed light on potential metabolic engineering directions to improve NUE in plants.