【作者】 张青瑞；

【导师】 修志龙；

【作者基本信息】 大连理工大学 ， 生物化工， 2008， 博士

【摘要】 生物炼制或白色生物技术近几年受石油价格不断攀升的影响而越发受到人们的关注,作为可再生的石油替代品,生物柴油的产量近两年增长迅猛。而生物柴油副产10%的甘油,以甘油为原料生产高附加值的产品成为生物柴油行业的迫切愿望,其中用生物法将甘油转化为1,3-丙二醇是最受人们关注的方法之一。本文对克雷伯氏杆菌转化甘油为1,3-丙二醇的代谢网络进行了通量分析,并建立了基于代谢网络的动力学模型,对生物转化过程的动态行为进行了模拟,主要研究内容如下:首先,对简化的克雷伯氏杆菌代谢网络进行了代谢通量平衡分析与优化。依据甘油生物歧化为1,3-丙二醇的简化代谢网络建立了通量平衡方程,以1,3-丙二醇通量最大为目标,通过线性规划对其进行了优化分析,得到最优代谢通量分布和目标产物的最大理论得率,并分析了比生长速率、氧气消耗速率、呼吸商对1,3-丙二醇得率的影响以及各通量对1,3-丙二醇得率的限制潜能。在此基础上,利用通量平衡分析方法对连续厌氧发酵实验数据进行甘油代谢系统目标函数的识别,发现其目标函数对不同操作条件基本保持不变。细胞生长通量在目标函数中的权重系数最大,是驱动代谢的主要动力,其次是1,3-丙二醇和2,3-丁二醇通量。通过对不同目标函数下得到的通量值与实验值进行比较,发现本目标函数得到的通量值与实验值最为接近,验证了该目标函数的合理性。其次,对扩展的克雷伯氏杆菌厌氧发酵甘油代谢网络进行了代谢通量分析。利用不同操作条件下的甘油厌氧代谢的稳态数据得到胞内代谢通量分布,分析了不同操作条件对胞内通量的影响以及代谢网络中的分支点的鲁棒性。结果表明当初始甘油浓度低时,生物量随着NADH₂与NAD⁺之比的降低而增加,当甘油初始浓度高时二者变化却相反。胞内通量分布表明网络中甘油和二羟丙酮节点对不同操作条件的变化显示刚性,而丙酮酸和乙酰辅酶A节点则显示柔性,是调节细胞代谢的关键分支点。另外,在高甘油消耗速率下,乙醇途径和丙酮酸脱氢酶途径随着甘油消耗速率的变化波动比较明显。再次,对甘油兼性生物歧化代谢网络进行了基元通量分析。将氧气传感调控系统加入甘油代谢网络,运用基元模式分析了厌氧和有氧条件下甘油生物转化过程中各途径的通量分布、灵敏性和1,3-丙二醇的生产潜能。结果发现,厌氧条件下,1,3-丙二醇最大理论得率为0.844 mol/mol,此时磷酸戊糖途径和转氢酶途径是提供1,3-丙二醇生产所需NADH₂的主要途径。有氧条件下1,3-丙二醇最大理论得率为0.875 mol/mol,此时磷酸戊糖途径或三羧酸循环对提高1,3-丙二醇的得率具有重要作用。另外,分析了不同氧气水平对1,3-丙二醇和生物量产率的影响。最后,对甘油生物转化为1,3-丙二醇的动态过程进行了模拟。依据甘油代谢网络,确定了胞内代谢的基元模式,在此基础上建立了宏观反应和微观代谢的综合模型。采用经典细胞生长模型描述各基元宏观反应动力学时,根据实测的胞外代谢物流量和生物量情况识别的动力学模型能够较好地反映甘油厌氧间歇发酵的指数生长期。采用神经网络与第一工程原理结合的基于基元模式的混杂模型既能模拟甘油生物转化为1,3-丙二醇的间歇过程,又能模拟批式流加动态过程,而且还能够模拟连续振荡动态过程。由测量数据识别出的代谢网络中各基元模式动力学得到的甘油消耗速率计算值与实验值吻合较好。上述研究工作对深入理解克雷伯氏杆菌的甘油代谢途径、菌种的基因工程改造以及甘油生物转化为1,3-丙二醇过程的在线控制具有重要的参考价值和指导作用。更多还原

【Abstract】 Bio-refinery or white-biotechnology has been paid much attention due to the high price of petroleum.As an alternative fuel,biodiesel is increasedly produced and 10 percent of glycerol is accompanied as a by-product.Much attention has been paid on the bioconversion of glycerol into a high valuable chemical such as 1,3-propanediol（1,3-PD）.In this paper,the process of glycerol bioconversion into 1,3-PD in Klebsiella pneumoniae was analyzed by metabolic flux analysis,and the dynamic process was simulated by a hybrid model based on metabolic elementary modes.The main work of this paper was summarized as followed:Firstly,simplifed metabolic network was constructed according to the glycerol metabolism in K.pneumoniae.A model of metabolic flux balance was developed for 1,3-PD production by K.pneumoniae under anaerobic and microaerobic conditions based on the metabolic network.The maximum theoretical yield of 1,3-PD to glycerol and the optimal flux distributions were obtained by flux balance analysis（FBA） with the maximum flux of 1,3-PD as the objective function.Furthermore,the influences of the specific growth rates,the molar fraction of NADH₂ oxidized by molecular oxygen in TCA（tricarboxylic acid ） cycle and the respiratory quotient（RQ） on the maximum theoretical yield of 1,3-PD to glycerol were analyzed.The limitation potential of various fluxes,such as ethanol,acetic,lactic,H₂, 2,3-butanediol,for the maximum theoretical yield of 1,3-PD were calculated.Additionally, the objective function driving the cellular metabolism was identified by a nonlinear bilevel programming（NBP） model.The results showed that the almost same objective functions in term of different experimental data were obtained and the important coefficient of biomass growth flux which was the main power of driving the cellular metabolism was the largest in the fitness function and then was 1,3-PD and 2,3-butaniol in turn.Secondly,the metabolic network of glycerol metabolism in K.pneumoniae under anaerobic conditions was improved and extended according to the KEGG database and related references.The measured fluxes obtained under steady-state conditions were used to estimate intracellular fluxes and identify the robustness of branch points of the anaerobic glycerol metabolism in K.pneumoniae for the production of 1,3-PD by metabolic flux analysis（MFA）. The biomass concentration increased as NADH₂/NAD⁺ decreased at low initial glycerol concentrations but inversed at high initial glycerol concentrations.The absolute flux distribution revealed that the branch points of glycerol and dihydroxyacetonephosphate （DHAP） were rigid to the environmental conditions.However,the pyruvate and acetyl coenzyme A（ACCOA） metabolisms gave cells the flexibility to regulate the energy and intermediate fluxes under various environmental conditions.Additionally,it was found that the fluxes of ethanol and pyruvate formate lyase（PFL） appeared visible fluctuations at high glycerol uptake rates.Thirdly,a model that utilized existing knowledge of oxygen and redox sensing/regulatory system to assist elementary flux modes（EFMs）,was developed and was carried out to predict the metabolic potential of K.pneumoniae for the production of 1,3-propanediol（1,3-PD） under anaerobic and aerobic conditions.It was found that the theoretical optimal 1,3-PD yield could reach to 0.844 mol mol^-1 if the pentose phosphate pathway（PPP） and transhydrogenase had a high flux under anaerobic conditions.However,PPP had little influence on the theoretical 1,3-PD yield and the flux through tricarboxylic acid（TCA） cycle was high under aerobic conditions.Both conditions exhibited a different distribution of NADH₂ supplied during maximum 1,3-PD production.Under anaerobic conditions,66.67%of NADH₂ was generated via the PPP.Whereas,NADH2 supply relied on the TCA-cycle in addition to the PPP under aerobic conditions.Additionally,the effect of oxygen level on the 1,3-PD and biomass was further analyzed.Finally,on the basis of glycerol metabolic network in K.penumoniae,the elementary flux modes were computed and translated into a set of macro-reactions connecting the extracellular substrates and products.A model combining macroscopical with microcosmic metabolic network was developed.As the elementary macroscopical dynamics were described by a classical dynamic model under balanced growth conditions,the model was suitable for describing the dynamic process during the exponential growth phase in batch fermentations. As the elementary macroscopical dynamics were described by a bioreactor dynamical hybrid model which combined first principles modeling with artificial neural networks（ANNs）,the results showed that the model could describe the dynamic processes of batch,fed-batch and oscillation simultaneously.This approach allowed the quantification of fluxes carried by individual elementary modes which was of great help to understand the dynamics properties.更多还原