WO2023102770A1 - 基于人工智能创制的高附加值代谢物底盘 - Google Patents

基于人工智能创制的高附加值代谢物底盘 Download PDF

Info

Publication number
WO2023102770A1
WO2023102770A1 PCT/CN2021/136423 CN2021136423W WO2023102770A1 WO 2023102770 A1 WO2023102770 A1 WO 2023102770A1 CN 2021136423 W CN2021136423 W CN 2021136423W WO 2023102770 A1 WO2023102770 A1 WO 2023102770A1
Authority
WO
WIPO (PCT)
Prior art keywords
added
value
metabolite
promoter
chassis
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2021/136423
Other languages
English (en)
French (fr)
Inventor
罗小舟
邓华祥
邓艳午
余函
邱玉兰
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Senris Biotechnology Shenzhen Co Ltd
Shenzhen Institute of Advanced Technology of CAS
Original Assignee
Senris Biotechnology Shenzhen Co Ltd
Shenzhen Institute of Advanced Technology of CAS
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Senris Biotechnology Shenzhen Co Ltd, Shenzhen Institute of Advanced Technology of CAS filed Critical Senris Biotechnology Shenzhen Co Ltd
Priority to PCT/CN2021/136423 priority Critical patent/WO2023102770A1/zh
Publication of WO2023102770A1 publication Critical patent/WO2023102770A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • C—CHEMISTRY; METALLURGY
    • C12—BIOCHEMISTRY; BEER; SPIRITS; WINE; VINEGAR; MICROBIOLOGY; ENZYMOLOGY; MUTATION OR GENETIC ENGINEERING
    • C12N—MICROORGANISMS OR ENZYMES; COMPOSITIONS THEREOF; PROPAGATING, PRESERVING, OR MAINTAINING MICROORGANISMS; MUTATION OR GENETIC ENGINEERING; CULTURE MEDIA
    • C12N15/00—Mutation or genetic engineering; DNA or RNA concerning genetic engineering, vectors, e.g. plasmids, or their isolation, preparation or purification; Use of hosts therefor
    • C12N15/09—Recombinant DNA-technology
    • C12N15/10—Processes for the isolation, preparation or purification of DNA or RNA

Definitions

  • the invention belongs to the technical field of synthetic biology and metabolic engineering, and specifically relates to a high-value-added metabolite chassis created based on artificial intelligence.
  • Terpenoids, alkaloids, polyketides, flavonoids and other natural products have physiological activities such as antibacterial, anticancer and antiviral, and have been widely used in food, agriculture, and pharmaceutical industries.
  • physiological activities such as antibacterial, anticancer and antiviral
  • These compounds have complex structures, and the chemical synthesis method also faces problems such as lengthy steps, low synthesis efficiency, and many by-products.
  • microbial synthesis has attracted widespread attention due to its advantages such as short fermentation cycle, low fermentation cost, and environmental friendliness.
  • researchers at home and abroad have systematically constructed molecular manipulation tools for various chassis (such as E. .
  • the Fowler team used the CiED model to predict the global metabolic flux, combined with rational metabolic transformation of key nodes, to achieve a naringenin production of 270mg/L. Therefore, the intelligent allocation of microbial endogenous metabolic resources is an important cornerstone for the efficient synthesis of target compounds.
  • Synthetic biology has become an effective tool for microbial design and transformation, ensuring efficient synthesis of high value-added metabolites.
  • Jae Kyung Sohng's team also systematically explained the use of synthetic biology tools to produce various natural and non-natural flavonoids.
  • Luo et al. achieved efficient synthesis of natural and non-natural cannabinoids in yeast by means of synthetic biology techniques.
  • Microbial synthesis can effectively overcome the technical bottleneck of insufficient supply of plant-derived natural products. It is worth noting that the endogenous metabolic network of microorganisms is complex, and the cellular physiological environment for metabolite synthesis changes in real time.
  • promoter engineering strategy is an effective strategy to balance the metabolic flow of high value-added substances, but the capacity of the promoter library is huge, and there is no efficient and stable molecular probe system or chemical method. Combined use and other technologies are difficult to solve the problem of balancing metabolic flow resources in a short period of time.
  • the continuous closed-loop learning of key enzyme expression and yield data of the target metabolic pathway can effectively balance the resource allocation of cell growth and core metabolite synthesis, reduce metabolic overflow, and rationally strengthen the target metabolite pathway.
  • flavonoids in plants (mainly naringenin and its derivatives) have been verified to have antioxidant activity, which can delay aging, and are widely used in the food and skin care industries; they also have physiological activities such as antibacterial, anti-inflammatory, and anti-cancer , can relieve diabetes, high blood pressure, hyperlipidemia and other diseases. To sum up, flavonoids exhibit high economic benefits and medical strategic value.
  • the synthetic pathway of naringenin requires the participation of four enzymes, including tyrosine ammonia lyase (TAL), 4-coumarate: CoA ligase (4-coumarate: CoA ligase, 4CL ), chalcone synthase (chalcone synthase, CHS), chalcone isomerase (chalcone isomerase, CHI). Therefore, the present invention takes naringenin metabolic flow as an example, continuously domesticates the gene expression of naringenin metabolic pathway with the help of artificial intelligence technology, removes its metabolic barrier, and realizes the efficient synthesis of flavonoids.
  • the present invention can provide theoretical support and technical support for the distribution of core metabolic gene expression resources of other types of high value-added metabolites; effectively solve their market supply and ensure the quality of life and health of the people.
  • the purpose of the present invention is to design and provide a high-value-added metabolite chassis created based on artificial intelligence.
  • the invention uses artificial intelligence technology to continuously domesticate the promoter library of key genes of high value-added core metabolic flow, remove metabolic barriers, optimize resource allocation of target core metabolic pathways, balance their metabolic flow, and realize efficient synthesis of target metabolites.
  • the high value-added metabolite chassis created based on artificial intelligence is characterized by the use of artificial intelligence technology to achieve design-build-test-learn closed-loop and cyclic learning of promoter and target metabolite yield data through small sample sampling, continuous domestication
  • the promoter library of key genes of high value-added core metabolic flux optimizes the resource allocation of target core metabolic pathways to obtain high-yield and high-value-added metabolite chassis.
  • the high-value-added metabolite chassis created based on artificial intelligence is characterized in that the high-value-added metabolites include flavonoid compounds, flavonoid compound derivatives, flavonoid compound modified compounds, and the flavonoid compound Including naringenin, anthocyanins, baicalein or pinogenin.
  • step (3) adopting chemical method or molecular probe method to initially screen the target metabolite output of the promoter engineering bacteria obtained in step (2) respectively in ultraviolet signal and fluorescent signal;
  • the method for creating is characterized in that the method for verifying promoters of different strengths in the step (1) is: using mKate2 fluorescent protein as a molecular probe, detecting the fluorescent signals of different promoters at 588/633nm ultraviolet wavelengths, The arabinose and T7 promoters were used as weak and strong standard promoters respectively, and the above-mentioned promoters were classified into three strengths: low, medium and high.
  • the creation method is characterized in that the establishment method of the promoter library in the step (1) is: using a high-fidelity, high-affinity linker sequence to connect the expression cassettes of key genes, using ccdB gene and/or mKate2 gene as The background plasmid was prepared by using Golden gate digestion and ligation technology to obtain the promoter library.
  • the creation method is characterized in that the conditions for the Golden gate enzyme digestion and connection in the step (1) are: connect each expression cassette with a high-fidelity, high-affinity linker sequence;
  • the second cycle is to optimize the connection conditions.
  • the invention method is characterized in that the step (2) is completed by automation technology.
  • the invention method is characterized in that the chemical method in the step (3) includes Al 3+ chemical method.
  • the creation method is characterized in that the specific operation method of the step (5) is: the promoters of different strengths of key genes are used as the model input, and the corresponding target metabolite output is used as the output of the model, based on the machine learning model Learn the mapping relationship between the two, predict all combinations, sort the output from high to low, select the sample with the highest output for experimental verification, use Golden gate technology to assemble the sample, repeat steps (2)-(4), and verify the machine learning sample Data, input the data obtained in step (4), and iterative machine learning to obtain a high-yield and high-value-added metabolite chassis.
  • Artificial intelligence strategy cyclically learns the universal technology and method of key gene expression of various high value-added compounds: small sample sampling of engineering strains with different metabolite yields, using promoters of different strengths as the input of the model, the corresponding target product
  • the output is the output of the model, learn the mapping relationship between them based on the machine learning model, predict all possible combinations, sort from high to low, and select the sample with the highest yield for verification; with the help of high-efficiency assembly technology, metabolite screening technology, HPLC quantitative technology, etc. verify machine learning sample data; iteratively learn all the above data, and then balance the expression level of each gene, enhance the metabolic flux of high added value, and achieve high yield of metabolites.
  • Machine learning strategy is a universal strategy to balance the expression of key naringenin genes: small samples are taken from engineering bacteria with different naringenin production, and the promoters with different strengths of each engineering strain are used as the input signal of the model.
  • the corresponding naringenin production is The output signal of the model, based on the machine learning model, learns the corresponding mapping relationship, and predicts all possible combinations, sorts from high to low, and selects the sample with the highest yield for inspection; with the help of high-efficiency assembly technology, metabolite screening technology, and HPLC quantitative technology Test the machine learning sample data; iteratively learn all the above data, and then balance the expression levels of key naringenin genes to achieve its high yield.
  • the present invention has the beneficial effects:
  • the present invention realizes the four-in-one autonomous learning process of "design-build-test-learn", provides new ideas for the balance of various high value-added metabolite pathways, and can effectively solve the problem of high value-added metabolism
  • the machine learning technology of the present invention adopts a closed-loop process of small sample sampling and data iterative training to balance the gene expression level of core metabolic flow and high-yield target metabolites, which can effectively reduce sample sampling and reduce the synthesis of high value-added substances by microbial methods manpower and time costs.
  • the closed-loop learning strategy of the present invention can produce high-value-added compounds such as naringenin, fundamentally solve the problem of insufficient supply of flavonoids in the market, alleviate the crisis of national health, reduce the number of cancer populations in my country, and resist the threat of the new coronavirus .
  • Fig. 1 is a flowchart of the present invention
  • Fig. 2 is the basic skeleton structure of flavonoids
  • Fig. 3 is the synthetic pathway of naringenin (Naringenin);
  • Figure 4 is a graph showing the results of metabolite production after continuous evolution of key genes of core metabolic flux by automated technology.
  • the present invention aims to use the artificial intelligence strategy to learn the core metabolic flow gene promoter library and yield data in a closed-loop manner, so as to optimally allocate key gene resources and efficiently synthesize target metabolites.
  • the present invention aims to use the artificial intelligence strategy to learn the core metabolic flow gene promoter library and yield data in a closed-loop manner, so as to optimally allocate key gene resources and efficiently synthesize target metabolites.
  • Through small sample sampling realize the "design-build-test-learn" closed-loop and cyclic learning promoter and target metabolite yield data, so as to balance the expression level of key genes in the metabolic flow of various high-value-added compounds and enhance the core metabolic flow amount, remove barriers such as intermediate products and end products inhibiting key enzymes, and efficiently synthesize high value-added metabolites.
  • the closed-loop and iterative learning of the artificial intelligence technology closed-loop and iterative learning of the key gene promoter library and yield data of naringenin from small samples is used to balance the metabolic flow and optimally distribute the naringenin
  • the expression level of key genes of flavonoids was determined to create chassis microorganisms with high yield of flavonoids.
  • the basic skeleton structure of flavonoids is shown in Figure 2, and the synthesis route of naringenin is shown in Figure 3.
  • the present invention provides 4 key enzymes of the core metabolic flow of naringenin: tyrosine ammonia lyase (tyrosine ammonia lyase, TAL; KF765779) derived from Rhodotorulaglutinis, its nucleotide sequence is as shown in SEQ ID NO.1
  • 4-coumaric acid CoA ligase (4-coumarate: CoA ligase, 4CL; KF765780)) of Petroselinum crispum origin, its nucleotide sequence is shown in SEQ ID NO.2
  • the chalcone synthase of PetuniaXhybrida origin (chalcone synthase, CHS; KF765781), its nucleotide sequence is shown in SEQ ID NO.3
  • Medicago sativa source chalcone isomerase (chalcone isomerase, CHI; KF765782), its nucleotide sequence is as shown

Landscapes

  • Genetics & Genomics (AREA)
  • Engineering & Computer Science (AREA)
  • Chemical & Material Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Zoology (AREA)
  • Wood Science & Technology (AREA)
  • Organic Chemistry (AREA)
  • Biotechnology (AREA)
  • General Engineering & Computer Science (AREA)
  • Biomedical Technology (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Physics & Mathematics (AREA)
  • Molecular Biology (AREA)
  • Microbiology (AREA)
  • Crystallography & Structural Chemistry (AREA)
  • Plant Pathology (AREA)
  • Biochemistry (AREA)
  • General Health & Medical Sciences (AREA)
  • Biophysics (AREA)
  • Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)

Abstract

基于人工智能创制的高附加值代谢物底盘，属于合成生物学与代谢工程技术领域。提供了采用人工智能技术，通过小样本采样，实现设计-构建-测试-学习闭环式、循环式学习启动子与目标代谢物产量数据，连续驯化高附加值核心代谢流关键性基因的启动子文库，优化目标核心代谢途径资源分配，得到高产高附加值代谢物底盘。还提供了基于人工智能创制的高附加值代谢物底盘的创制方法及其应用。可有效解决高附加值代谢物无高通量分子探针或者分子探针筛选不稳定的问题，减少人力成本，缩短微生物代谢调控时间，消除中间产物与终产物抑制效应，增强核心代谢通量，实现高附加值化合物高效合成。

Description

基于人工智能创制的高附加值代谢物底盘 技术领域
本发明属于合成生物学与代谢工程技术领域，具体涉及基于人工智能创制的高附加值代谢物底盘。
背景技术
萜类、生物碱、聚酮类化合物、黄酮类化合物等天然产物，具有抗菌抗癌抗病毒等生理活性，已广泛应用于食品、农业、医药行业。然而，由于环境条件复杂多变、提取成本高与纯度低等因素，严重阻碍这些高附加值代谢物的市场供给。这些化合物存在结构复杂，化学合成法也面临步骤冗长、合成效率低、副产物多等问题。相比植物提取与化学合成法，微生物合成法因发酵周期短、发酵成本低、环境友好等优势而被广泛关注。随着测序技术、基因编辑技术的跨越式发展，国内外研究者已系统性构建各类底盘(如大肠杆菌、酵母菌、放线菌)的分子操作工具，便于外源途径在微生物异源表达。
外源代谢流引入底盘微生物时，易诱发宿主内源代谢网络扰动，打破细胞氧化还原稳态，且外源途径各基因表达分配不平衡，易导致中间产物积累、代谢溢流等问题。因此，理性设计微生物内源代谢网络与外源途径，可有效扭转微生物目标代谢物产量低的现状。系统生物学技术如基因组代谢网络技术可全局分析微生物代谢网络，便于后期理性改造关键性节点，进而提高目标代谢物产量。例如，Fowler团队借助CiED模型预测全局代谢流，结合理性代谢改造关键性节点，实现柚皮素产量达到270mg/L。因此，智能分配微生物内源代谢资源为目标化合物高效合成的重要基石。
合成生物学成为微生物设计与改造的有效工具，保障高附加值代谢物高效合成。例如，Jae Kyung Sohng团队也系统性的阐释了利用合成生物学工具生产各类天然与非天然的黄酮类化合物。Luo等人借助合成生物学技术在酵母中实现天然与非天然大麻素的高效合成。微生物合成法可有效克服植物来源天然产物供应不足的技术瓶颈。值得注意的是，微生物内源代谢网络复杂，且代谢物合成的细胞生理环境实时变化。当外源高附加值化合物代谢流引入工程菌株，易诱发宿主内源代谢网络扰动，打破细胞氧化还原稳态，诱发毒理环境，限制细胞正常生理活动。此外，外源途径各基因表达分配不平衡，易导致中间产物积累、代谢溢流等问题。因此，需借助系统生物学技术，平衡细胞内源与外源代谢流的资源分配，但是系统生物学技术仅从全局水平提供可调控的关键性节点，后期仍需基因操作技术予以验证。
针对该瓶颈，近期崛起的系统基因组网络操作技术与合成生物学技术加速了微生物高附加值智能细胞工厂的进程。其中，合成生物学的人工智能技术，以“设计-构建-测试-学习” 的循环式理念，革新了微生物代谢调控进程，理性平衡核心代谢流关键性基因表达平衡，可攻克核心代谢流不平衡、效能低等壁垒，智能分配生长代谢与目标代谢流资源，保障高效合成目标代谢物。实现从上游基因元件设计、组装与下游产物测试、再设计等迭代循环回路，有效增强合成生物学理性改造微生物代谢流的深度与广度，实现高附加值代谢物高效合成。目前，运用启动子工程策略为平衡高附加值物质代谢流的有效策略，但是启动子库容量宏大、且无高效、稳定的分子探针系统或化学方法，仅借助高效液相层析、液质联用等技术，难以短时间内攻坚代谢流资源分配平衡的难题。若能借助人工智能策略，即可实现载体构建、菌体培养、小样本采样、数据学习等迭代循环式操作流程，保障高附加值目标代谢通路基因表达水平的连续循环循环过程，解决无高通量筛选方法获取突变体的瓶颈性问题，同时可解放人力，减少人为错误与时间成本，最终，跨越式提升高附加值代谢物产量，满足其市场需求，解决国民健康危机，提高国民生活品质。因此，借助人工智能技术，连续闭环式学习目标代谢通路关键性酶表达与产量数据，可有效平衡细胞生长与核心代谢物合成代谢的资源分配，减少代谢溢流，并理性强化目标代谢物通路，创制微生物智能细胞工厂，实现高附加值化合物高效合成。
植物中约有7000种黄酮类化合物(主要为柚皮素及其衍生物)被验证具有抗氧化活性，可延缓衰老，广泛应用于食品与护肤品行业；也具有抗菌消炎、抗癌症等生理活性，可缓解糖尿病、高血压、高血脂等病痛。综上所述，黄酮类化合物展现出极高的经济效益与医药战略价值。作为黄酮类基本骨架结构，柚皮素合成途径需要4个酶参与，包括酪氨酸氨裂解酶(tyrosine ammonia lyase，TAL)、4-香豆酸CoA连接酶(4-coumarate:CoA ligase，4CL)、查尔酮合成酶(chalcone synthase，CHS)、查尔酮异构酶(chalcone isomerase，CHI)。因此，本发明以柚皮素代谢流为实例，借助人工智能技术连续驯化柚皮素代谢通路基因表达，解除其代谢屏障，实现黄酮类化合物高效合成。本发明可为其余各类高附加值代谢物核心代谢基因表达资源分配，提供理论支撑与技术保障；有效解决其市场供应，保障国民生活品质与健康水准。
发明内容
针对上述现有技术中存在的问题，本发明的目的在于设计提供基于人工智能创制的高附加值代谢物底盘。本发明运用人工智能技术，连续驯化高附加值核心代谢流关键性基因的启动子文库，解除代谢屏障，优化目标核心代谢途径资源分配，平衡其代谢流，实现目标代谢物高效合成。
为了实现上述目的，本发明采用以下技术方案：
基于人工智能创制的高附加值代谢物底盘，其特征在于采用人工智能技术，通过小样本采样，实现设计-构建-测试-学习闭环式、循环式学习启动子与目标代谢物产量数据，连续驯化高附加值核心代谢流关键性基因的启动子文库，优化目标核心代谢途径资源分配，得到高产高附加值代谢物底盘。
所述的基于人工智能创制的高附加值代谢物底盘，其特征在于所述高附加值代谢物包括类黄酮化合物、类黄酮化合物衍生物、类黄酮化合物经修饰后的化合物，所述类黄酮化合物包括柚皮素、花青素、黄芩素或生松素。
所述的高附加值代谢物底盘的创制方法，其特征在于包括以下步骤：
(1)验证不同强度的启动子，建立目标代谢物的关键性基因不同表达水平的启动子文库，采用Golden gate酶切与连接技术，在抗生素平板上筛选关键性基因的启动子文库的工程菌株文库；
(2)挑取所述步骤(1)得到的工程菌株文库中的工程菌至深孔板，置于高速摇床，培养过夜，以0.5-3％接种量接种至发酵培养基中，培养1-3天，得到启动子工程菌；
(3)采用化学法或分子探针法，分别在紫外信号与荧光信号初步筛选所述步骤(2)得到的启动子工程菌的目标代谢物产量；
(4)采用Sanger测序技术测定不同工程菌的启动子序列，采用高效液相层析技术在紫外信号条件下，测定不同工程菌的目标代谢物产量；
(5)基于机器学习模型学习不同强度启动子的组合分布，得到产量最高的目标代谢物的启动子工程菌。
所述的创制方法，其特征在于所述步骤(1)中验证不同强度的启动子的方法为：以mKate2荧光蛋白为分子探针，在588/633nm紫外波长下检测不同启动子的荧光信号，分别以阿拉伯糖与T7启动子为弱与强的标准启动子，对上述启动子归类为低、中、高三种强度。
所述的创制方法，其特征在于所述步骤(1)中启动子文库的建立方法为：采用高保真、高亲和力的linker序列连接关键性基因的表达框，以ccdB基因和/或mKate2基因作为背景质粒，采用Golden gate酶切与连接技术，制备得到启动子文库。
所述的创制方法，其特征在于所述步骤(1)中Golden gate酶切与连接的条件为：以高保真、高亲和力的linker序列连接各表达框，以37℃5min，16℃5min，50次循环为优化连接条件。
所述的创制方法，其特征在于所述步骤(2)采用自动化技术完成。
所述的创制方法，其特征在于所述步骤(3)中化学法包括Al 3+化学法。
所述的创制方法，其特征在于所述步骤(5)的具体操作方法为：以关键性基因的不同强度的启动子作为模型输入，对应的目标代谢物产量作为模型的输出，基于机器学习模型学习二者的映射关系，对所有组合情况进行预测，产量从高到低排序，选取产量最高的样品实验验证，采用Golden gate技术组装样品，重复步骤(2)-(4)，验证机器学习样品数据，输入步骤(4)得到的数据，迭代机器学习，得到高产高附加值代谢物底盘。
所述的基于人工智能创制的高附加值代谢物底盘在高效合成黄酮类化合物上的应用。
人工智能策略循环式学习各类高附加值化合物关键性基因表达的普适性技术与方法：小样本采样不同代谢物产量的工程菌株，以不同强度的启动子作为模型的输入，对应的目标产物产量为模型的输出，基于机器学习模型学习它们之间的映射关系，预测所有可能的组合情况，从高到底排序，选取产量最高的样本进行验证；借助高效率组装技术、代谢物初筛技术、HPLC定量技术等验证机器学习样品数据；迭代学习以上所有数据，进而平衡各基因表达水平，增强高附加值代谢通量，实现代谢物高产。
机器学习策略平衡柚皮素关键性基因表达的普适性策略：小样本采样不同柚皮素产量的工程菌，以各工程菌株不同强度的启动子作为模型的输入信号，对应柚皮素产量为模型的输出信号，基于机器学习模型学习对应映射关系，并预测所有可能的组合情况，从高到底排序，选取产量最高的样本进行检验；借助高效率组装技术、代谢物初筛技术、HPLC定量技术等检验机器学习样品数据；迭代学习以上所有数据，进而平衡柚皮素各关键性基因表达水平，实现其高产。
与现有技术相比，本发明具有的有益效果：
(1)本发明借助人工智能技术，实现“设计-构建-测试-学习”四位一体的自主学习过程，为各类高附加值代谢物途径平衡，提供新思路，可有效解决高附加值代谢物无高通量分子探针或者分子探针筛选不稳定的问题，减少人力成本，缩短微生物代谢调控时间，消除中间产物与终产物抑制效应，增强核心代谢通量，实现高附加值化合物高效合成。
(2)本发明的机器学习技术，采用小样本采样及数据迭代训练的闭环式流程，平衡核心代谢流基因表达水平，高产目标代谢物，可有效减少样品采样，降低微生物法合成高附加值物质的人力与时间成本。
(3)本发明的闭环式学习策略可高产柚皮素等高附加值化合物，根本性解决黄酮类化合物市场供应不足的问题，缓解全民健康的危机，减少我国癌症人口数量，抵御新冠病毒的威胁。
附图说明
图1为本发明流程图；
图2为黄酮类化合物基本骨架结构；
图3为柚皮素(Naringenin)合成路径；
图4为自动化技术连续进化核心代谢流关键性基因后代谢物产量结果图。
具体实施方式
以下将通过附图和实施例对本发明作进一步说明。
实施例1：
本发明旨在运用人工智能策略，闭环式学习核心代谢流基因启动子文库与产量数据，借以最优分配关键性基因资源，高效合成目标代谢物。通过小样本采样，实现“设计-构建-测试-学习”闭环式、循环式学习启动子与目标代谢物产量数据，借以平衡各类高附加值化合物代谢流关键性基因表达水平，增强核心代谢流通量，解除中间产物、终产物对关键性酶抑制等屏障，高效合成高附加值代谢物。
详细阐述机器学习闭环式、迭代式学习各关键性基因表达水平的普适性策略，流程图如图1所示，其具体技术方案如下所示：
(1)验证不同启动子表达强度。以mKate2荧光蛋白为分子探针，在588/633nm，检测不同启动子荧光信号，并将备选启动子分为低、中、高三种不同强度。
(2)制备关键性基因不同表达水平的启动子文库。以高保真性、高亲和力的linker序列连接各关键性基因的表达框；背景质粒选取ccdB基因与mKate2基因；优化Golden gate酶切与连接条件，在相应抗生素平板筛选不同启动子工程菌文库。
(3)提供基因突变文库挑取与培养技术，即挑取克隆子至96深孔板，放置高速摇床培养过夜。以1％接种量，转移至发酵培养基，培养2天。
(4)提供小样本采样的初筛方法。借助化学法或分子探针法，根据特定荧光与紫外信号，筛选出不同产量的启动子工程菌。
(5)提供不同工程菌株各基因启动子与目标代谢物产量测定的方法。运用Sanger测序技术测定不同工程菌株启动子序列信息；借助高效液相层析技术测定目标高附加值代谢物产量。
(6)基于机器学习模型学习不同强度启动子的组合分布，寻找高附加值物质代谢通路平衡最优解的方法。以各个关键性基因工程菌对应的不同强度的启动子作为模型的输入，对应的目标产物产量为模型的输出，基于机器学习模型学习它们之间的映射关系，并对所有可能的组合情况进行预测，从高到底排序，选取产量最高的样本实验验证。借助Golden gate技术 组装各个样品，根据本发明第三到五方面培养及测定方法，验证机器学习样品数据。重新输入本发明第六方面数据，迭代机器学习，最终，增强高附加值代谢通量，实现代谢物高产。
实施例2：
以柚皮素高附加值代谢物为实例，举例说明人工智能技术闭环式、迭代式学习小样本来源的柚皮素关键性基因启动子文库与产量数据，借以平衡代谢流，最优分配柚皮素关键性基因表达水平，创制高产黄酮类化合物的底盘微生物。黄酮类化合物基本骨架结构如图2所示，柚皮素(Naringenin)合成路径如图3所示。
详细说明机器学习技术连续驯化各类关键性代谢流基因表达的普适性策略，以高附加值的黄酮类化合物(柚皮素)为实例，详实说明机器学习技术闭环式、迭代式学习核心代谢流基因表达文库的流程。
具体包括如下步骤：
(1)本发明提供柚皮素核心代谢流的4个关键性酶：Rhodotorulaglutinis来源的酪氨酸氨裂解酶(tyrosine ammonia lyase，TAL；KF765779)，其核苷酸序列如SEQ ID NO.1所示；Petroselinum crispum来源的4-香豆酸CoA连接酶(4-coumarate:CoA ligase，4CL；KF765780))，其核苷酸序列如SEQ ID NO.2所示；PetuniaXhybrida来源的查尔酮合成酶(chalcone synthase，CHS；KF765781)，其核苷酸序列如SEQ ID NO.3所示；Medicago sativa来源的查尔酮异构酶(chalcone isomerase，CHI；KF765782)，其核苷酸序列如SEQ ID NO.4所示。
(2)制备柚皮素关键性基因不同表达水平的启动子文库。以高保真性、高亲和力的linker序列连接各关键性基因的表达框；背景质粒选取ccdB基因与mKate2基因；运用Golden gate技术，制备启动子文库，并在相应抗生素平板筛选柚皮素关键性基因不同启动子工程菌株文库。
(3)提供基因突变文库挑取与培养技术，即挑取克隆子至96深孔板，并放置高速摇床培养过夜。以1％接种量，转移至发酵培养基，培养2天。
(4)提供小样本采样的初筛方法。借助Al 3+化学法，分别在373nm紫外信号与382/505nm的荧光信号筛选不同柚皮素产量的启动子工程菌。
(5)提供不同工程菌株各基因启动子与目标代谢物产量测定的方法。运用Sanger测序技术测定不同工程菌株启动子序列信息；借助高效液相层析技术，在290nm紫外信号条件下，测定不同工程菌柚皮素产量。
(6)基于机器学习模型学习不同强度启动子的组合分布，寻找柚皮素代谢通路平衡最优解的方法。以柚皮素4个关键性基因工程菌对应的不同强度的启动子作为模型的输入，对应的目标产物产量为模型的输出，基于机器学习模型学习它们之间的映射关系，并对所有可能 的组合情况进行预测，从高到底排序，选取产量最高的样本实验验证。借助Golden gate技术组装各个样品，根据本发明第九到十一方面培养及测定方法，验证机器学习样品数据。重新输入本发明第十二方面数据，迭代机器学习，最终，增强高附加值代谢通量，实现代谢物高产。
实验结果证明：人工智能技术，使得柚皮素产量提高至1.32g/L，如图4所示，为目前文献报道的最高水平。

Claims (10)

  1. 基于人工智能创制的高附加值代谢物底盘，其特征在于采用人工智能技术，通过小样本采样，实现设计-构建-测试-学习闭环式、循环式学习启动子与目标代谢物产量数据，连续驯化高附加值核心代谢流关键性基因的启动子文库，优化目标核心代谢途径资源分配，得到高产高附加值代谢物底盘。
  2. 如权利要求1所述的基于人工智能创制的高附加值代谢物底盘，其特征在于所述高附加值代谢物包括类黄酮化合物、类黄酮化合物衍生物、类黄酮化合物经修饰后的化合物，所述类黄酮化合物包括柚皮素、花青素、黄芩素或生松素。
  3. 如权利要求1或2所述的高附加值代谢物底盘的创制方法，其特征在于包括以下步骤：
    (1)验证不同强度的启动子，建立目标代谢物的关键性基因不同表达水平的启动子文库，采用Golden gate酶切与连接技术，在抗生素平板上筛选关键性基因的启动子文库的工程菌株文库；
    (2)挑取所述步骤(1)得到的工程菌株文库中的工程菌至深孔板，置于高速摇床，培养过夜，以0.5-3％接种量接种至发酵培养基中，培养1-3天，得到启动子工程菌；
    (3)采用化学法或分子探针法，分别在紫外信号与荧光信号初步筛选所述步骤(2)得到的启动子工程菌的目标代谢物产量；
    (4)采用Sanger测序技术测定不同工程菌的启动子序列，采用高效液相层析技术在紫外信号条件下，测定不同工程菌的目标代谢物产量；
    (5)基于机器学习模型学习不同强度启动子的组合分布，得到产量最高的目标代谢物的启动子工程菌。
  4. 如权利要求3所述的创制方法，其特征在于所述步骤(1)中验证不同强度的启动子的方法为：以mKate2荧光蛋白为分子探针，在588/633nm紫外波长下检测不同启动子的荧光信号，分别以阿拉伯糖与T7启动子为弱与强的标准启动子，对上述启动子归类为低、中、高三种强度。
  5. 如权利要求3所述的创制方法，其特征在于所述步骤(1)中启动子文库的建立方法为：采用高保真、高亲和力的linker序列连接关键性基因的表达框，以ccdB基因和/或mKate2基因作为背景质粒，采用Golden gate酶切与连接技术，制备得到启动子文库。
  6. 如权利要求3所述的创制方法，其特征在于所述步骤(1)中Golden gate酶切与连接的条件为：以高保真、高亲和力的linker序列连接启动子的表达框，以37℃ 5min，16℃ 5min，50次循环为优化连接条件。
  7. 如权利要求3所述的创制方法，其特征在于所述步骤(2)采用自动化技术完成。
  8. 如权利要求3所述的创制方法，其特征在于所述步骤(3)中化学法包括Al 3+化学法。
  9. 如权利要求3所述的创制方法，其特征在于所述步骤(5)的具体操作方法为：以关键性基因的不同强度的启动子作为模型输入，对应的目标代谢物产量作为模型的输出，基于机器学习模型学习二者的映射关系，对所有组合情况进行预测，产量从高到低排序，选取产量最高的样品实验验证，采用Golden gate技术组装样品，重复步骤(2)-(4)，验证机器学习样品数据，输入步骤(4)得到的数据，迭代机器学习，得到高产高附加值代谢物底盘。
  10. 如权利要求1或2所述的基于人工智能创制的高附加值代谢物底盘在高效合成黄酮类化合物上的应用。
PCT/CN2021/136423 2021-12-08 2021-12-08 基于人工智能创制的高附加值代谢物底盘 Ceased WO2023102770A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/136423 WO2023102770A1 (zh) 2021-12-08 2021-12-08 基于人工智能创制的高附加值代谢物底盘

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/136423 WO2023102770A1 (zh) 2021-12-08 2021-12-08 基于人工智能创制的高附加值代谢物底盘

Publications (1)

Publication Number Publication Date
WO2023102770A1 true WO2023102770A1 (zh) 2023-06-15

Family

ID=86729523

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/136423 Ceased WO2023102770A1 (zh) 2021-12-08 2021-12-08 基于人工智能创制的高附加值代谢物底盘

Country Status (1)

Country Link
WO (1) WO2023102770A1 (zh)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120034661A1 (en) * 2010-05-07 2012-02-09 The Research Foundation Of State University Of New York Strains for the production of flavonoids from glucose
CN109554415A (zh) * 2017-09-27 2019-04-02 中国科学院微生物研究所 提高thaxtomin发酵产量的方法
CN110144626A (zh) * 2019-03-29 2019-08-20 浙江工业大学 一种启动子文库的构建方法
CN110869502A (zh) * 2017-06-06 2020-03-06 齐默尔根公司 高通量转座子诱变
CN110878262A (zh) * 2019-12-04 2020-03-13 郑州大学 一种柚皮素产物成瘾装置及其应用
US20200291445A1 (en) * 2015-06-10 2020-09-17 Northwestern University Cell-free protein synthesis driven metabolic engineering

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20120034661A1 (en) * 2010-05-07 2012-02-09 The Research Foundation Of State University Of New York Strains for the production of flavonoids from glucose
US20200291445A1 (en) * 2015-06-10 2020-09-17 Northwestern University Cell-free protein synthesis driven metabolic engineering
CN110869502A (zh) * 2017-06-06 2020-03-06 齐默尔根公司 高通量转座子诱变
CN109554415A (zh) * 2017-09-27 2019-04-02 中国科学院微生物研究所 提高thaxtomin发酵产量的方法
CN110144626A (zh) * 2019-03-29 2019-08-20 浙江工业大学 一种启动子文库的构建方法
CN110878262A (zh) * 2019-12-04 2020-03-13 郑州大学 一种柚皮素产物成瘾装置及其应用

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
DONG, JUNKAI: "The Establishment and Applications of the Pathway Assembly Method Based on Standard Biological Elements in Saccharomyces Cerevisiae", DOCTORAL DISSERTATION, no. 04, 1 July 2018 (2018-07-01), CN, pages 1 - 110, XP009546188, DOI: 10.27266/d.cnki.gqhau.2018.000179 *
ZHOU, SHENGHU: "Static and Dynamic Regulation Strategies Drive (2S)-Naringenin Biosynthesis", DOCTORAL DISSERTATION, no. 05, 1 December 2018 (2018-12-01), CN, pages 1 - 108, XP009546189, DOI: 10.27169/d.cnki.gwqgu.2018.000001 *

Similar Documents

Publication Publication Date Title
Naseri et al. Application of combinatorial optimization strategies in synthetic biology
Czajka et al. Tuning a high performing multiplexed-CRISPRi Pseudomonas putida strain to further enhance indigoidine production
Liu et al. Rapidly moving new bacteria to model-organism status
Guo et al. Strategies to improve the efficiency and quality of mutant breeding using heavy-ion beam irradiation
WO2023104046A1 (zh) 含有烯醇激酶基因和异戊烯基磷酸激酶基因的重组菌株及其在萜类化合物生产中的应用
CN116312766A (zh) 基于人工智能创制的高附加值代谢物底盘
Cao et al. MYC: orchestrating secondary metabolism and glandular trichome formation
WO2023102770A1 (zh) 基于人工智能创制的高附加值代谢物底盘
CN110544507B (zh) 基于组学整合技术调控固态发酵的方法及系统和应用
Zuo et al. Automated genetic manipulation for the construction of Pichia pastoris cell factories
Oosthuizen et al. Evolution of mutualistic behaviour between Chlorella sorokiniana and Saccharomyces cerevisiae within a synthetic environment
CN119431555A (zh) 基因工程技术的rhFN重组人源化纤连蛋白工艺
WO2023092547A1 (zh) 基于自动化技术构建的高附加值代谢物智能细胞工厂
CN116064340B (zh) nagB基因改造的弗式链霉菌重组菌及制备方法和应用
CN102533722B (zh) 一种通过温度突变提高哺乳动物细胞重组蛋白瞬时表达的方法
Komives et al. Bioprocessing technology for production of biopharmaceuticals and bioproducts
CN105018609A (zh) 一种甲烷氧化菌丰度的自动化检测方法
CN116179450A (zh) 基于自动化技术构建的高附加值代谢物智能细胞工厂
CN115851606A (zh) Btk突变细胞株及其构建方法和应用
Guo et al. Transcriptome-Based Mining of the Strong Promoters for Hyperproduction of Gibberellin GA3 in Fusarium fujikuroi
CN106636151A (zh) 一株过表达糖基转移酶基因的工程菌及构建方法
CN104342483A (zh) 一种卤醇脱卤酶的快速筛选方法
Ahlawat et al. Novel Bioproducts Through Metabolic Engineering of Secondary Metabolites
Cui et al. Metabolic Engineering of Yarrowia lipolytica with Massive Gene Assembly and Genomic Integration
CN121495929A (zh) 一种增强运动发酵单胞菌上游基因表达的3'utr元件及其应用

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21966694

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21966694

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 11/04/2025)

122 Ep: pct application non-entry in european phase

Ref document number: 21966694

Country of ref document: EP

Kind code of ref document: A1