Engineering of the fast-growing cyanobacterium Synechococcus sp. PCC 11901 to synthesize astaxanthin
快速生长的蓝细菌 Synechococcus sp. PCC 11901 的工程改造以合成虾青素

Abstract  摘要

Background  背景

Astaxanthin is a red pigment required by feed, nutraceutical, and cosmetic industries for its pigmentation and antioxidant properties. This carotenoid is one of the main high-value products that can nowadays be derived from microalgae cultivation, raising important industrial interest. However, state-of-the-art astaxanthin production is the cultivation of the green alga Haematococcus pluvialis (or lacustris), which faces high costs and low production yield. Hence, alternative and efficient sources for astaxanthin need to be developed, and novel biotechnological solutions must be found. The recently discovered cyanobacterium, Synechococcus sp. PCC 11901 is a promising photosynthetic platform for the large-scale production of high-value products, but its potential has yet to be thoroughly tested.
虾青素是一种红色色素，因其色素沉着和抗氧化特性，被饲料、保健品和化妆品行业广泛需求。这种类胡萝卜素是目前可以通过微藻培养获得的主要高价值产品之一，因而引起了重要的工业兴趣。然而，当前虾青素生产的主流方法是培养绿色藻类雨生红球藻（或湖生红球藻），但其生产成本高且产量低。因此，需要开发替代且高效的虾青素来源，并寻找新的生物技术解决方案。最近发现的蓝细菌 Synechococcus sp. PCC 11901 是一种具有潜力的大规模生产高价值产品的光合平台，但其潜能尚需深入测试。

Results  结果

In this study, the cyanobacterium Synechococcus sp. PCC 11901 was engineered for the first time to our knowledge to produce astaxanthin, a high-value ketocarotenoid, by expressing recombinant β-ketolase (bKT) and a β-hydroxylase enzymes (CtrZ). During photoautotrophic growth, the bKT-CtrZ transformed strain (called BC) accumulated astaxanthin to above 80% of the total carotenoid. Moreover, BC cells grew faster than wild-type (WT) cells in high light and continuous bubbling with CO₂-enriched air. The engineered strain reached stationary phase after only 4 days of growth in an airlift 80-mL photobioreactor, producing 7 g/L of dry biomass, and accumulated ~ 10 mg/L/day of astaxanthin, which is more than other CO₂-consuming multi-engineered systems. In addition, BC cells were cultivated in a 330-L photobioreactor to link lab-scale experiments to the industrial scale-up.
在本研究中，据我们所知，首次对蓝细菌 Synechococcus sp. PCC 11901 进行了改造，使其通过表达重组 β-酮化酶（bKT）和 β-羟化酶（CtrZ）来生产高价值的酮类类胡萝卜素——虾青素。在光自养生长过程中，bKT-CtrZ 转化菌株（称为 BC）积累的虾青素占类胡萝卜素总量的 80% 以上。此外，在高光强和持续通入富含 CO₂ 的空气条件下，BC 菌株的生长速度比野生型（WT）菌株更快。在 80 mL 气升式光生物反应器中，该改造菌株仅用 4 天便达到稳定期，干重生物量达 7 g/L，虾青素的积累量约为 10 mg/L/天，超过了其他利用 CO₂ 的多重工程系统。此外，BC 菌株还在 330 L 光生物反应器中进行了培养，以将实验室规模的研究与工业化放大相结合。

Conclusions  结论

The astaxanthin volumetric productivity achieved, 10 mg/L/day, exceeds that previously reported for Haematococcus pluvialis, the standard microalgal species nowadays used at the industrial level for astaxanthin production, or for other microalgal strains engineered to produce ketocarotenoids. Overall, this work identifies a new route to produce astaxanthin on an industrial scale.
实现的虾青素体积生产率为 10 mg/L/天，超过了目前用于工业生产虾青素的标准微藻物种雨生红球藻以及其他经过工程改造以生产酮类类胡萝卜素的微藻菌株的已报道生产率。总体而言，这项研究确定了一种生产虾青素的新工业化途径。

Background  背景

Due to their ability to perform oxygenic photosynthesis, eukaryotic microalgae and prokaryotic cyanobacteria are attractive carbon-neutral biological platforms for the industrial production of bioproducts using sunlight, water, and CO₂. However, two main drawbacks limit the industrial exploitation of microalgae: low cell densities and moderate growth rates [1]. These two factors are often critical for the economic sustainability of algae production, which is already burdened by high maintenance and processing costs. Therefore, finding suitable strains for large-scale industrial use is of utmost importance.
由于能够进行产氧光合作用，真核微藻和原核蓝藻成为利用阳光、水和 CO₂生产生物产品的具有吸引力的碳中性生物平台。然而，微藻的工业开发受限于两个主要缺点：低细胞密度和适中的生长速率[1]。这两个因素通常对藻类生产的经济可持续性至关重要，而藻类生产本身已因高维护和加工成本而面临压力。因此，寻找适合大规模工业应用的菌株至关重要。

The recently isolated cyanobacterium Synechococcus sp. PCC 11901 (hereafter Syn11901) possesses suitable properties for large-scale cultivation [2]. Syn11901 has a fast growth rate, grows to high cell densities and biomass yields, tolerates high salinity, and has a range of synthetic biology tools available for genetic engineering [2,3,4,5].
最近分离出的蓝藻 **Synechococcus sp. PCC 11901**（以下简称 **Syn11901**）具有适合大规模培养的特性[2]。**Syn11901** 生长速度快，能够达到高细胞密度和高生物量产量，耐受高盐度，并且已有一系列用于基因工程的合成生物学工具可供使用[2, 3, 4, 5]。

One strategy to offset the high processing costs of growing microalgae and cyanobacteria at the industrial scale is to produce high-value products. In particular, natural products found in microalgae are an attractive target [6], especially antioxidants, which can be used in nutraceutical, cosmetic, and feed formulations [7].
在工业规模培育微藻和蓝藻时，为抵消高昂的加工成本，一种策略是生产高附加值产品。尤其是微藻中天然产物具有很高的吸引力[6]，特别是抗氧化剂，这些抗氧化剂可用于营养保健品、化妆品及饲料配方中[7]。

Among the natural antioxidants, astaxanthin (hereafter referred to as Asta) is a valuable molecule for relevant industrial applications [8] with its price reaching 60,000 €/kg. The main routes to obtain natural Asta are from bacteria such as Paracoccus carotinifaciens [9] or from photoautotrophic microalgae [10] such as Haematococcus pluvialis, also named Haematococcus lacustris [10, 11] and, to a lesser extent, Chromochloris zofingiensis [12].
在众多天然抗氧化剂中，虾青素（以下简称 **Asta**）是一种在相关工业领域中极具价值的分子[8]，其价格可高达 60,000 欧元/千克。获得天然虾青素的主要途径是从 **Paracoccus carotinifaciens** 等细菌[9]，或从光自养微藻[10]，例如 **Haematococcus pluvialis**（也称为 **Haematococcus lacustris**）[10, 11]，以及较少程度上的 **Chromochloris zofingiensis**[12]。

The use of H. pluvialis does, however, have some limitations, notably the requirement of two different growth conditions for either growing green biomass or accumulating Asta, the latter being induced by stress conditions such as high light, high/low temperature, and/or nutrient starvation [13]. Moreover, Asta accumulated in H. pluvialis has a low bio-accessibility due to the thick cell wall, making mechanical disruption necessary to release Asta for human or animal consumption [14], a process that increases production costs.
**以下是上述英文内容的简体中文翻译：** 尽管如此，使用**H. pluvialis**仍存在一些限制，尤其是其需要两种不同的生长条件：一种用于培养绿色生物质，另一种用于积累虾青素（Asta）。积累虾青素需要通过高光、高/低温和/或营养缺乏等应激条件来诱导 [13]。此外，由于**H. pluvialis**具有厚细胞壁，导致其积累的虾青素的生物可及性较低，因此需通过机械破壁来释放虾青素以供人类或动物食用 [14]，而这一过程会增加生产成本。

Advances in synthetic biology have led to the generation of genetically modified cyanobacteria and microalgae that synthesize Asta and, importantly, overcome some of the limitations of H. pluvialis strains [15,16,17,18,19,20,21]. The production of Asta in cyanobacteria [15,16,17,18, 20, 21] is also facilitated by the abundant accumulation of β-carotene and zeaxanthin (hereafter βcar and Zea, respectively), both precursors of the Asta pathway [8, 22]. Cyanobacteria have the additional advantage that carotenoids are more easily extracted from biomass than H. pluvialis [22].
合成生物学的进步已使得转基因蓝藻和微藻能够合成虾青素成为可能，而更重要的是，这些转基因生物克服了**H. pluvialis**菌株的一些局限性 [15, 16, 17, 18, 19, 20, 21]。在蓝藻中生产虾青素 [15, 16, 17, 18, 20, 21] 还得益于β-胡萝卜素和玉米黄质（以下分别简称为βcar 和 Zea）的大量积累，它们是虾青素合成途径的前体 [8, 22]。此外，与**H. pluvialis**相比，蓝藻的类胡萝卜素更容易从生物质中提取 [22]。

However, the use of cyanobacteria to produce Asta has until now been undermined by the low accumulation of biomass in lab-scale experiments [15, 21]. For all these reasons, the feasibility of producing Asta in Syn11901 was herein investigated for the first time to our knowledge. The results obtained demonstrate a novel biotechnological solution for Asta production in unicellular photosynthetic organisms: engineered Syn11901 can produce Asta at unprecedented levels in cyanobacteria, highlighting its potential route for the industrial production of Asta.
然而，迄今为止，蓝藻用于虾青素生产的一个主要问题是，其在实验室规模实验中生物质积累量较低 [15, 21]。基于以上这些原因，本研究首次探索了在**Syn11901**中生产虾青素的可行性。研究结果表明，这种新型的生物技术解决方案能够在单细胞光合生物中实现虾青素的生产：经过工程改造的**Syn11901**可以在蓝藻中实现前所未有的虾青素产量，展现出其作为虾青素工业化生产潜在途径的巨大潜力。

Material and methods  材料与方法

Cyanobacterial strains and cultivation
蓝藻菌株与培养

The cyanobacterium Syn11901 [2] was used as the experimental strain in this work and is referred to as the wild type (WT). Cells were maintained on AD7 solid agar medium containing 10 mM glycerol [2], whereas liquid cultures were grown in flasks using the Modified AD7 medium (hereafter named MAD medium), as previously described [2]. MAD medium used for flask cultivation was supplemented with 10 mM NaHCO₃ as a C source. High-density cultivations were conducted in batch mode in 80 mL airlift photobioreactors (PBRs) in the Multicultivator system MC-1000-OD (PSI, Czech Republic) with independent white light-emitting diodes (LEDs) illumination and temperature control at 35 °C. This device automatically monitors the optical density at 720 nm as an index of cell concentration. The optical density of the cell cultures was, in addition, measured at 720 nm upon sampling and diluting the samples to avoid the saturation of OD measurement. Before starting the growth tests, cyanobacterial cells with OD at 720 nm (OD₇₂₀) of 0.3 were acclimated for 2 days in the PBRs with illumination of 250 micromoles of photons per square meter per second (µmol/m²/s) and a constant supply of CO₂ (3% CO₂-enriched air). Each culture was then diluted into fresh MAD medium to an OD₇₂₀ of 0.1 before starting the growth experiments. Growth started with low light conditions (250 μmol photons/m²/s) for cell acclimation, then (after 12 h) shifted to higher light conditions (up to 2250 μmol/m²/s) as described in the text with a constant supply of CO₂ (3% CO₂ mixed with air). Growth in nutrient starvation was induced by growing cells in MAD medium prepared with the absence of a specific nutrient such as nitrate (MAD-N), phosphate (MAD-P) or iron (MAD-Fe). In particular, cells grown as described above at 2250 μmol/m²/s of light were harvested at stationary phase by centrifugation, washed once with the growth medium adopted for the following cultivation steps, and diluted to OD₇₂₀ 3 in MAD-N, MAD-P, MAD-Fe or MAD as control. Cells in the different growth media were then cultivated in batch in 80-ml airlift PBRs in the multicultivator system with illumination of 2250 µmol/m²/s and a constant supply of CO₂ (3% CO₂-enriched air). Large-scale growth was performed in batch mode using a 330-L PBR (Technology Farm, Italy) of diameter 50 cm, with light provided by 4 LED white light strips contained in a plastic cylinder positioned inside the PBR (~ 800 µmol/m²/s measured at the internal surface) and by six external red/white strips (~ 550 µmol/m²/s measured at the outer surface) placed outside the PBR. After 3 days of cyanobacterial cultivation, an external LED panel was placed close to the right side of the PBR to increase the light supply. Before cultivation, cells were transferred from 25-ml flasks in exponential phase in a 4-L bottle containing MAD medium obtaining as initial OD₇₂₀ of 0.02. The cells were kept at 37 °C, illuminated with a white LED lamp (~ 200 µmol/m²/s), and air-lifted with a pump. Once the culture reached an OD₇₂₀ of 2.5, it was used as an inoculum for the industrial PBR. The photobioreactor was filled with purified and UV-treated water, whereas nutrient salts were manually added inside the PBR. The final composition was the same as the MAD medium but included two modifications for economic reasons: a tenfold reduction in tris(hydroxymethyl)aminomethane (Tris) buffer concentration (1.03 mM, pH 8.2) and the use of one-third of vitamin B12 (~ 1.3 ng/L). In addition, 10 mM NaHCO₃ was added to the medium. The temperature was kept between 33 and 35 °C, and the pH was maintained between 8.2 and 8.5. Air-bubbling was supplied to mix the cells and provide some CO₂. Every hour, ~ 10 s of pure CO2 was automatically provided to restore the pH to 8.2. Biomass accumulation was followed by daily measurements of OD₇₂₀.
在本研究中，使用了蓝藻 Syn11901 [2] 作为实验菌株，并将其称为野生型（WT）。细胞在含有 10 mM 甘油的 AD7 固体琼脂培养基上维持[2]，而液体培养则在烧瓶中使用改良的 AD7 培养基（以下简称 MAD 培养基），如文献所述[2]。用于烧瓶培养的 MAD 培养基补充了 10 mM NaHCO₃作为碳源。高密度培养在 80 mL 气升式光生物反应器（PBR）中以分批模式进行，使用 Multicultivator 系统 MC-1000-OD（PSI，捷克）进行培养，配备独立的白色 LED 照明和温度控制在 35°C。该设备自动监测 720 nm 处的光密度（OD）作为细胞浓度的指标。此外，细胞培养液的光密度还通过取样和稀释样品后在 720 nm 处测量，以避免 OD 测量饱和。在开始生长测试之前，光密度为 720 nm（OD₁）为 0.3 的蓝藻细胞在 PBR 中以 250 μmol 光子/m²/s 的光照和持续供应 3% CO₂富集空气的条件下适应培养 2 天。随后，每种培养物被稀释到新鲜的 MAD 培养基中，OD₅为 0.1，开始生长实验。生长以低光条件（250 μmol 光子/m²/s）开始，以便细胞适应，然后（12 小时后）转为高光条件（高达 2250 μmol/m²/s），如文中所述，同时持续供应 3% CO₂与空气混合气体。在缺营养条件下的生长通过在制备时缺少特定营养成分（如硝酸盐（MAD-N）、磷酸盐（MAD-P）或铁（MAD-Fe））的 MAD 培养基中培养细胞来诱导。具体而言，在 2250 μmol/m²/s 光照条件下按上述方法培养的细胞在静止期通过离心收获，用随后培养步骤所采用的培养基洗涤一次，并稀释至 OD₁₁为 3 的 MAD-N、MAD-P、MAD-Fe 或作为对照的 MAD 培养基中。不同培养基中的细胞随后在 Multicultivator 系统的 80 mL 气升式 PBR 中以分批模式培养，光照强度为 2250 μmol/m²/s，并持续供应 3% CO₂富集空气。大规模培养以分批模式在直径 50 cm 的 330 L PBR（Technology Farm，意大利）中进行，光线由位于 PBR 内部的塑料圆柱内的 4 条 LED 白光灯带（在内表面测量约 800 μmol/m²/s）和安装在 PBR 外部的 6 条红/白光灯带（在外表面测量约 550 μmol/m²/s）提供。经过 3 天的蓝藻培养后，在 PBR 右侧放置了一个外部 LED 面板以增加光照供应。在培养之前，将处于指数生长期的细胞从 25 mL 烧瓶转移到含 MAD 培养基的 4 L 瓶中，初始 OD₁₇为 0.02。细胞保持在 37°C，用白光 LED 灯（约 200 μmol/m²/s）照射，并通过泵进行气升搅拌。当培养物达到 OD₁₉为 2.5 时，用作工业 PBR 的接种物。光生物反应器用纯化并经紫外线处理的水填充，同时手动向 PBR 内添加营养盐。最终成分与 MAD 培养基相同，但出于经济原因进行了以下两项修改：三（羟甲基）氨基甲烷（Tris）缓冲液浓度减少十倍（1.03 mM，pH 8.2），并使用三分之一的维生素 B12（约 1.3 ng/L）。此外，培养基中加入了 10 mM NaHCO₃。温度保持在 33 至 35°C 之间，pH 值维持在 8.2 至 8.5 之间。通过气泡搅拌细胞并提供部分 CO₂。每小时自动供应约 10 秒的纯 CO₂以将 pH 值恢复到 8.2。通过每日测量 OD₂₂跟踪生物量积累。

Generation of recombinant constructs and cyanobacterial transformation
重组构建体的生成与蓝藻转化

The gene sequences encoding the β-carotene 4-ketolase of Chlamydomonas reinhardtii (AY860820.1, bKT) and the β-carotene hydroxylase of Brevundimonas sp. SD212 (MK214313.1, crtZ) were taken from the literature [17, 19]. The sequences were codon-optimized for expression in Syn11901 using an open software system (https://www.idtdna.com/CodonOpt) with the codon usage of Synechococcus sp. PCC 7002, a close model species, as a reference [2]. Synthetic sequences were generated by Eurofins Genomics (Ebersberg, Germany). Sequences of smR and kmR selection cassettes, conferring resistance to spectinomycin and kanamycin, respectively, were amplified from pSW039 and pSW071 plasmids [2] in the order given. Such plasmids were obtained from the Addgene repository (https://www.addgene.org). Similarly, the flanking sequences required for homologous recombination in the acsA locus were taken from plasmid pSW039. P_cpt promoter and T7 terminator sequences were amplified from plasmid pSW036 and pSW071, respectively. The DNA sequences used for metabolic engineering can be found at this link: https://doi.org/10.5281/zenodo.13234572. DNA constructs were generated using Gibson Assembly by Thermo Fisher (Waltham, USA), followed by the transformation of Escherichia coli TOP10 chemically competent cells.
从文献 [17, 19] 中获取了编码**莱茵衣藻**（*Chlamydomonas reinhardtii*）β-胡萝卜素 4-酮化酶（AY860820.1，bKT）和**Brevundimonas sp. SD212** β-胡萝卜素羟化酶（MK214313.1，crtZ）的基因序列。这些序列利用开放软件系统（[https://www.idtdna.com/CodonOpt](https://www.idtdna.com/CodonOpt)）根据**Synechococcus sp. PCC 7002**（一种近缘模式物种）的密码子使用情况进行了密码子优化，以便在 Syn11901 中表达 [2]。合成的序列由 Eurofins Genomics 公司（德国埃伯斯贝格）生成。赋予光谱霉素和卡那霉素抗性的**smR**和**kmR**选择框序列分别从质粒 pSW039 和 pSW071 中扩增 [2]，顺序如文中所述。这些质粒来自 Addgene 资源库（[https://www.addgene.org](https://www.addgene.org)）。同样，从质粒 pSW039 中获取了在 acsA 位点进行同源重组所需的侧翼序列。P _cpt 启动子和 T7 终止子序列分别从质粒 pSW036 和 pSW071 中扩增。用于代谢工程的 DNA 序列可通过以下链接获取：[https://doi.org/10.5281/zenodo.13234572](https://doi.org/10.5281/zenodo.13234572)。DNA 构建通过 Thermo Fisher 公司（美国沃尔瑟姆）的 Gibson Assembly 方法生成，随后转化至 Escherichia coli TOP10 化学感受态细胞中。

Syn11901 transformations were carried out using established protocols for transforming Synechocystis sp. PCC 6803 [23, 24]. A Syn11901 culture in the exponential phase was first resuspended in fresh MAD medium supplemented with 10 mM NaHCO₃ without antibiotic selection to a final OD₇₂₀ of 2.5. Then, a 1.5-mL aliquot was transferred to a 12-well microtiter plate, and 1 µg of the desired linearized plasmid DNA was added. The plate was kept overnight in a 37 °C chamber with stirring to avoid cell precipitation and illuminated using a white light LED panel (~ 100 µmol photons/m²/s). 16 h later, cells were plated on AD7 solid medium containing the required antibiotic for selection, as indicated in previous works [2, 3].
使用已建立的转化 Synechocystis sp. PCC 6803 的协议[23, 24]进行了 Syn11901 的转化。处于指数生长期的 Syn11901 培养物首先被重悬于补充有 10 mM NaHCO₃的无抗生素选择的新鲜 MAD 培养基中，最终 OD₆₈₀调整至 2.5。随后，取 1.5 mL 样品转移至 12 孔微量滴定板中，加入 1 µg 目标线性化质粒 DNA。将板放置在 37 °C 的培养箱中，持续搅拌以防止细胞沉淀，同时使用白光 LED 面板（约 100 µmol 光子/m²/s）照明。经过 16 小时后，将细胞接种到含有所需抗生素的 AD7 固体培养基上进行选择，这些条件参考了之前的研究[2, 3]。

Genomic DNA PCR analysis of Syn11901 transformants
Syn11901 转化体的基因组 DNA PCR 分析

Genomic DNA templates were prepared as previously described [23]. Cells resuspended in 20 μL aliquot of Milli-Q water were mixed with an equal volume of 100% ethanol, followed by brief vortexing. A 200-μL aliquot of a 10% (w/v) Chelex-100 Resin (BioRad, USA) suspension in water was added to the sample before mixing and heating at 95 °C for 10 min to lyse the cells. Following centrifugation at 16,000 g for 10 min to pellet the cell debris, 5 μL of the supernatant was used as a genomic DNA template in a 25 μL PCR reaction mixture. Phusion DNA polymerase (Thermo Fisher, USA) was used for genomic DNA PCR analyses. A list of primers is given in Table 1. Transgenic DNA copy homoplasmy in Syn11901 was tested using suitable primers listed in the Supplemental Materials. The genomic DNA location of these primers is indicated in Figs. 2 and 3 for the appropriate DNA constructs. A similar approach was conducted to monitor the stability of the genetic modifications introduced, revealing stable genotypes of Syn11901 transformants for up to two years.
基因组 DNA 模板的制备按照之前描述的方法进行 [23]。将细胞重悬于 20 μL Milli-Q 水中，并与等体积的 100% 乙醇混合，随后短暂涡旋。然后向样品中加入 200 μL 10% (w/v) Chelex-100 树脂（BioRad，美国）水悬液，并在 95°C 下加热 10 分钟以裂解细胞。随后以 16,000 g 的离心力离心 10 分钟以沉淀细胞碎片，取上清液 5 μL 作为 25 μL PCR 反应混合物中的基因组 DNA 模板。基因组 DNA PCR 分析中使用的是 Phusion DNA 聚合酶（Thermo Fisher，美国）。引物列表见表 1。使用补充材料中列出的合适引物测试 Syn11901 转基因 DNA 的拷贝同质性。这些引物在基因组 DNA 中的位置在图 2 和图 3 中针对相应的 DNA 构建体进行了标注。类似方法用于监测引入的遗传修饰的稳定性，结果显示 Syn11901 转化体的基因型在长达两年的时间内保持稳定。

Table 1 Sequence of oligonucleotide primers used in the present work
表 1 本研究中使用的寡核苷酸引物序列

RNA extraction and reverse transcription
RNA 提取和反转录

RNA extraction was executed using TRIzol reagent (Thermo Fisher, USA), following instructions provided by the manufacturer. Each extraction was carried out starting from a cell pellet derived from 7 ml of a cyanobacterial culture in exponential phase brought to a final OD₇₂₀ of 3.5. The RNA concentration was determined using a NanoDrop spectrophotometer (Thermo Fisher, USA). Total RNA was reverse transcripted to cDNA using RevertAid reverse transcriptase (Thermo Fisher, USA) following the manufacturer's instructions. PCR analysis using cDNA was performed using Phusion DNA polymerase (Thermo Fisher, USA), with 30 cycles of amplification.
RNA 提取使用 TRIzol 试剂（Thermo Fisher，美国）按照制造商提供的说明进行。每次提取均以来自 7 毫升指数生长期蓝藻培养物的细胞沉淀为起点，最终 OD 值为 3.5。利用 NanoDrop 分光光度计（Thermo Fisher，美国）测定 RNA 浓度。总 RNA 按照制造商说明使用 RevertAid 逆转录酶（Thermo Fisher，美国）逆转录为 cDNA。使用 cDNA 进行 PCR 分析时，采用 Phusion DNA 聚合酶（Thermo Fisher，美国），扩增 30 个循环。

Protein analysis  蛋白质分析

Cells in the mid-exponential growth phase (OD₇₂₀ ~ 1) were harvested by centrifugation at 4,000 g for 10 min. The pellet was resuspended in a solution buffered with 25 mM Tris–HCl, pH 8.2, containing 1 mM benzamidine, 5 mM ε-aminocaproic acid, and 1 mM phenylmethylsulfonyl fluoride (PMSF) as protease inhibitors. Cells were lysed with a Cell Disruptor (Constant System Limited, UK), set at 37,500 kpsi. A slow-speed centrifugation (350xg for 5 min) was applied to remove unbroken cells. For protein electrophoretic analysis, sample extracts were solubilized upon incubation for 1 h at room temperature in the presence of 125 mM Tris–HCl, pH 6.8, 3.5% (w/v) sodium dodecyl sulfate (SDS), 10% (w/v) glycerol, 2 M urea, and 5% β-mercaptoethanol. Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) was performed using a Mini-PROTEAN gel system (BIORAD, USA). Western blot analysis entailed the transfer of separated proteins to a 0.45-μm pore size PVDF membrane (Life Technologies, USA), followed by immunoblotting with rabbit anti-His-tag specific polyclonal antibodies (Sigma-Aldrich, USA) and detection with chemiluminescent western blotting kit (Sigma-Aldrich, USA).
在中指数生长期（OD ₇₂₀ ~ 1）的细胞通过离心（4,000 g，10 分钟）收集。沉淀重新悬浮于含有 25 mM Tris-HCl（pH 8.2）、1 mM 苯甲胺、5 mM ε-氨基己酸和 1 mM 苯甲基磺酰氟（PMSF）作为蛋白酶抑制剂的缓冲溶液中。细胞使用细胞破碎仪（Constant System Limited, 英国）在 37,500 kpsi 的设置下破碎。通过低速离心（350xg，5 分钟）去除未破碎的细胞。用于蛋白电泳分析的样品提取液在含有 125 mM Tris-HCl（pH 6.8）、3.5%（w/v）十二烷基硫酸钠（SDS）、10%（w/v）甘油、2 M 尿素和 5% β-巯基乙醇的溶液中，在室温下孵育 1 小时后溶解。十二烷基硫酸钠-聚丙烯酰胺凝胶电泳（SDS-PAGE）使用 Mini-PROTEAN 凝胶系统（BIORAD，美国）进行。Western blot 分析通过将分离的蛋白转移到孔径为 0.45 μm 的 PVDF 膜（Life Technologies，美国），随后使用兔抗 His 标签特异性多克隆抗体（Sigma-Aldrich，美国）进行免疫印迹，并通过化学发光 Western blot 试剂盒（Sigma-Aldrich，美国）进行检测。

Pigment extraction and absorption spectra and HPLC analyses
色素提取与吸收光谱及高效液相色谱分析

Cyanobacterial cultures were centrifuged at 10,000 × g for 5 min, and pigments were extracted using dimethyl sulfoxide (DMSO). An incubation time of at least 90 min in a rotating mixer facilitated complete pigment extraction. Extracts were then diluted in acetone 95%, with the latter previously buffered with Na₂CO₃, to a final acetone concentration of 80%. The absorption spectra of the pigment extracts were measured with the Jasco V-730 spectrophotometer in the visible range (350–750 nm). Then, DMSO-acetone spectra were fitted with the different pigment absorption spectral forms [25]. Reversed-phase HPLC analysis was conducted as previously described [19].
蓝藻培养物以 10,000 × g 离心 5 分钟，用二甲基亚砜（DMSO）提取色素。在旋转混合器中至少孵育 90 分钟，以促进色素的完全提取。随后，将提取物稀释到最终丙酮浓度为 80%，稀释时使用预先用 Na ₂ CO ₃ 缓冲的 95%丙酮。色素提取物的吸收光谱在 Jasco V-730 分光光度计上测量，范围为可见光区域（350–750 nm）。之后，DMSO-丙酮光谱与不同色素吸收光谱形式进行拟合[25]。反相 HPLC 分析按照之前描述的方法进行[19]。

Thin-layer chromatography analysis
薄层色谱分析

Pigments were extracted in isopropanol from a cell pellet of transformed cyanobacteria, and then they were separated by thin-layer chromatography (TLC) on Silica Gel 60 F254 Coated Aluminum-Backed TLC Sheets (Sigma-Aldrich, USA) using an appropriate mobile phase (60% hexane, 20% chloroform, 20% acetone). The TLC fraction predicted as Asta was collected and eluted in isopropanol. Upon elution, the sample was dried in a speed vacuum system.
从转化蓝细菌的细胞沉淀中使用异丙醇提取色素，随后通过薄层色谱（TLC）在硅胶 60 F254 涂层铝背板 TLC 板（Sigma-Aldrich，美国）上分离，使用适当的流动相（60%己烷、20%氯仿、20%丙酮）。预测为虾青素（Asta）的 TLC 组分被收集并用异丙醇洗脱。洗脱后，样品在真空离心系统中干燥。

Mass spectroscopy analysis
质谱分析

The dried sample obtained by TLC was resuspended in methanol and loaded on HPLC in tandem with an Orbitrap mass spectrometer. Data acquisition was performed in full scan mode in the mass range of 50–800 m/z. The full description of the MS results can be found at this link: https://doi.org/10.5281/zenodo.14283202.
通过薄层色谱法获得的干燥样品重新悬浮于甲醇中，并加载到与 Orbitrap 质谱仪串联的高效液相色谱仪（HPLC）上。在质量范围为 50–800 m/z 的全扫描模式下进行数据采集。质谱结果的完整描述可通过以下链接查看：https://doi.org/10.5281/zenodo.14283202。

Dry weight measurement  干重测量

For dry biomass weight measurement, 30 mL of culture from each airlift PBR in the multicultivator system was collected into pre-weighed tubes. For the 330-L PBR, 50 mL of culture was taken for the dry weight analysis. Cells were then pelleted by centrifugation (4500 g, 15 min), and the supernatants were discarded. Pellets were washed once with milliQ water before transferring the tubes to a lyophilization system for freeze drying under vacuum. Samples were weighed again to obtain the dry weight.
为了测量干生物量重量，从多培养器系统中的每个气升式光生物反应器（PBR）中收集 30 mL 培养液至预先称重的试管中。而对于 330 升的光生物反应器，取 50 mL 培养液用于干重分析。随后通过离心（4500 g，15 分钟）将细胞沉淀，并弃去上清液。将沉淀用 MilliQ 水清洗一次后，转移至冷冻干燥系统，在真空条件下进行冻干。然后再次称重样品以获得干重。

Statistical analysis  统计分析

The statistical significance was evaluated by comparing results obtained in the same experiment running Tukey–Kramer multiple comparison tests. Statistically significant variations with a p < 0.05 are marked with different letters.
统计显著性通过运行 Tukey-Kramer 多重比较检验的同一实验结果进行评估。具有统计显著性差异（p < 0.05）的结果用不同字母标注。

Results  结果

Heterologous expression of beta-carotene 4-ketolase in Synechococcus sp. PCC 11901
**在 Synechococcus sp. PCC 11901 中异源表达 β-胡萝卜素 4-酮化酶**

Previous work has shown that the synthesis of Asta in cyanobacteria can be achieved by heterologous expression of prokaryotic ketolase and hydroxylase enzymes, as reported in Fig. 1 [15, 17, 18, 21]. In the case of the green alga, Chlamydomonas reinhardtii, the constitutive overexpression of just the endogenous β-carotene 4-ketolase gene (Uniprot Q4VKB4), hereafter referred to as Cr-bKT was sufficient to convert both βcar and Zea to Asta (Fig. 1), achieving high accumulation of the latter [19].
先前的研究表明，通过异源表达原核的酮化酶和羟化酶，可以在蓝藻中实现虾青素（Asta）的合成，如图 1 所示 [15, 17, 18, 21]。对于绿色藻类**衣藻（Chlamydomonas reinhardtii）**来说，仅通过持续过表达其内源 β-胡萝卜素 4-酮化酶基因（Uniprot Q4VKB4，以下简称 Cr-bKT），就足以将 β-胡萝卜素（βcar）和玉米黄质（Zea）转化为虾青素（Asta）（如图 1），并实现后者的高积累 [19]。

Fig. 1  图 1

Schematic view of astaxanthin biosynthesis. Ketocarotenoids are reported in red. The enzymes required for conversion of βcar into ketocarotenoids are reported. G3P glyceraldehyde 3-phosphate, IPP isopentenyl diphosphate, DMPP dimethylallyl diphosphate
虾青素生物合成示意图。酮类类胡萝卜素以红色表示。将β-胡萝卜素转化为酮类类胡萝卜素所需的酶已标注。G3P：甘油醛-3-磷酸，IPP：异戊二烯焦磷酸，DMPP：二甲基丙烯焦磷酸。

To evaluate the ability of the Cr-bKT enzyme alone to drive the synthesis of Asta in Syn11901 as in the case of Chlamydomonas reinhardtii, a DNA construct was designed for the transformation of wild-type Syn11901 (WT) through double homologous DNA recombination at the acsA locus, identified previously as a suitable site for insertion of foreign DNA [2]. Construct p-bKT (Fig. 2a) was codon-optimized for expression in Syn11901 and designed to replace the acsA gene. To achieve this, the sequence encoding the native chloroplast transit peptide sequence was deleted from Cr-bKT, and a kanamycin-resistance cassette (kmR) was inserted downstream in the same operon. Expression of the two heterologous genes was controlled by the constitutive P_cpt promoter [2, 26]. The nucleotide sequence of this construct is described in the DNA constructs section of Supplementary Information.
为了评估仅由 Cr-bKT 酶驱动在 Syn11901 中合成虾青素（Asta）的能力（如在衣藻 Chlamydomonas reinhardtii 中的情况），设计了一个用于通过双同源 DNA 重组改造野生型 Syn11901（WT）的 DNA 构建体，插入位点为先前确定适合外源 DNA 插入的 acsA 位点[2]。构建体 p-bKT（图 2a）经过密码子优化，以适应在 Syn11901 中的表达，并设计用于替代 acsA 基因。为此，从 Cr-bKT 中删除了编码原生叶绿体过渡肽序列的片段，并在同一操纵子下游插入了卡那霉素抗性盒（kmR）。两个外源基因的表达由组成型 P _cpt 启动子控制[2, 26]。该构建体的核苷酸序列在补充信息的 DNA 构建体部分中进行了描述。

Fig. 2  图 2

Generation of the bKT transformant and pigment analysis. a Schematic of the p-bKT DNA construct, harboring flanking sequences for the replacement of the acsA locus by homologous recombination [2]. The two heterologous genes, namely the bKT gene from Chlamydomonas reinhardtii [19] and the kanamycin-resistance cassette (kmR), were in an operon configuration and under the control of the P_cpt constitutive promoter. b Genomic DNA PCR analysis using primers acsA-5’ fw and acsA-3’ rv (left panel). The expected size of the PCR products was 3767 and 3820 bp in WT and bKT lines, respectively. Genomic DNA PCR analysis using primers acsA-5’ fw and bKT rv (middle panel). The expected size of the PCR product in bKT samples was 1168 bp, whereas no amplification was expected in the WT sample. A faint byproduct of 750 bp was also present in the bKT samples. Genomic DNA PCR analysis using primers acsA-5’ fw and acsA rv (right panel). The expected size of the PCR product in the WT sample was 1107 bp, whereas no amplification should occur in the bKT sample. c RT-PCR analysis for the verification of the expression of transgenic bKT gene and constitutive rpnA gene transcripts in WT and bKT cell cultures. The PCR products were separated on 1.5% agarose gel. The expected size of the PCR products was 187 and 231 bp for rpnA and bKT gene transcripts, respectively. d Pigmentation of WT and bKT photoautotrophic cultures grown in flasks. e Absorption spectra in the visible light range (350–750 nm) of pigment extracts from WT and bKT cultures shown in C. f HPLC analyses of microalgal pigments from WT and bKT cultures shown in d. 1, Mixoxanthophyll; 2, zeaxanthin; 3, chlorophyll a; 4, echinenone; 5, β-carotene; 6, 3S,3’S trans-astaxanthin; 7, 3S,3’S 9-cis-astaxanthin; 8, canthaxanthin
bKT 转化体的生成及色素分析。 a. p-bKT DNA 构建体示意图，包含用于通过同源重组替换 acsA 位点的侧翼序列 [2]。两个外源基因，即来自衣藻（Chlamydomonas reinhardtii）的 bKT 基因 [19] 和卡那霉素抗性盒（kmR），以操纵子形式配置，并由 P _cpt 组成型启动子驱动。 b. 使用引物 acsA-5’ fw 和 acsA-3’ rv 进行基因组 DNA PCR 分析（左图）。在 WT 和 bKT 株中的预期 PCR 产物大小分别为 3767 bp 和 3820 bp。使用引物 acsA-5’ fw 和 bKT rv 进行基因组 DNA PCR 分析（中图）。bKT 样本中预期的 PCR 产物大小为 1168 bp，而 WT 样本中不应有扩增。在 bKT 样本中还出现了一个 750 bp 的弱副产物。使用引物 acsA-5’ fw 和 acsA rv 进行基因组 DNA PCR 分析（右图）。在 WT 样本中的预期 PCR 产物大小为 1107 bp，而 bKT 样本中不应有扩增。 c. RT-PCR 验证 WT 和 bKT 细胞培养物中转基因 bKT 基因和组成型 rpnA 基因转录本的表达。PCR 产物在 1.5% 琼脂糖凝胶上分离。rpnA 和 bKT 基因转录本的预期 PCR 产物大小分别为 187 bp 和 231 bp。 d. 在烧瓶中培养的 WT 和 bKT 自养培养物的色素表现。 e. WT 和 bKT 培养物色素提取物的可见光范围（350–750 nm）吸收光谱（如 c 所示）。 f. 对 d 中 WT 和 bKT 培养物的微藻色素进行 HPLC 分析。1，混合黄藻黄素；2，玉米黄素；3，叶绿素 a；4，刺红素；5，β-胡萝卜素；6，3S,3’S 反式虾青素；7，3S,3’S 9-顺式虾青素；8，斑蝥黄素。

PCR analysis confirmed the correct genomic insertion of the exogenous DNA construct into the acsA locus [2, 27] of the antibiotic-resistant strains. Primers acsA-5’ fw and acsA-3’ rv (Table 1) were designed according to the flanking regions of the selected locus (Fig. 2a). PCR amplification using WT genomic DNA as a template generated a product of 3,767 bp, while it generated a 3,820 bp amplicon in the bKT lines. Due to the similar sizes of the PCR products from WT and transformant lines, it was impossible to distinguish them clearly on agarose gel (Fig. 2b, left). Thus, the recombinant DNA sequence insertion in the bKT transformants genome was verified using primers acsA-5’ fw and bKT-rv, with the latter annealing specifically to the recombinant bKT gene. The expected size of the PCR product was 1168 bp (Fig. 2b, middle). No PCR product was obtained using WT DNA as a template, whereas bKT samples generated a product of the expected size. PCR amplification using primers acsA-5’ fw and acsA rv was conducted to verify the absence of WT sequences in the bKT samples. The predicted WT sequence of 1107 bp was amplified in the WT sample. In contrast, it was absent in the bKT samples (Fig. 2b, right), suggesting complete segregation of the transgenes in the transformants. A reverse transcriptase PCR (RT-PCR) analysis was performed to confirm the expression of the bKT gene transcript. Specific primers (Table 1) were designed to amplify sequences from bKT and rpnA gene transcripts: the latter was used as a reference, according to previous literature [28]. The predicted 187-bp DNA product from rpnA transcript was amplified in both WT and bKT samples. Conversely, only bKT strain expressed the mRNA encoding the bKT gene product (Fig. 2c), as evidenced by the amplification of the predicted DNA sequence of 231 bp.
PCR 分析确认了外源 DNA 构建体正确插入抗生素抗性菌株的 acsA 位点[2, 27]。引物 acsA-5' fw 和 acsA-3' rv（表 1）根据所选位点的侧翼区域设计（图 2a）。以野生型（WT）基因组 DNA 为模板的 PCR 扩增生成了 3,767 bp 的产物，而在 bKT 菌株中则生成了 3,820 bp 的扩增子。由于 WT 和转化菌株的 PCR 产物大小相似，无法在琼脂糖凝胶上清晰区分（图 2b，左）。因此，使用引物 acsA-5' fw 和 bKT-rv 验证了 bKT 转化菌株基因组中重组 DNA 序列的插入，其中 bKT-rv 特异性结合于重组 bKT 基因。预期的 PCR 产物大小为 1,168 bp（图 2b，中）。以 WT DNA 为模板未获得 PCR 产物，而 bKT 样本生成了预期大小的产物。使用引物 acsA-5' fw 和 acsA rv 进行 PCR 扩增，以验证 bKT 样本中是否缺失 WT 序列。预测的 WT 序列为 1,107 bp，在 WT 样本中被扩增出来，而在 bKT 样本中未检测到（图 2b，右），表明转基因在转化菌株中完全分离。通过逆转录 PCR（RT-PCR）分析确认了 bKT 基因转录本的表达。设计了特异引物（表 1）用于扩增 bKT 和 rpnA 基因转录本序列，后者作为参考基因，依据已有文献[28]。rpnA 转录本的预测产物为 187 bp，在 WT 和 bKT 样本中均被扩增。而 bKT 菌株特异表达编码 bKT 基因产物的 mRNA（图 2c），通过扩增预测的 231 bp DNA 序列得以证实。

WT and bKT cultures were grown photoautotrophically in shake flasks at an irradiance of ~ 100 µmol/m²/s of white light at 37 °C. Liquid MAD medium, specially formulated for growing Syn11901 cultures [2], was supplemented with 100 mM NaHCO₃ as a C source. A characteristic feature of WT and bKT cultures was their different pigmentation upon reaching the stationary phase (Fig. 2d): WT was green, whereas bKT had a darker/brownish hue. Absorption spectra of chlorophyll and carotenoids extracted from cells revealed a shoulder between 500 and 550 nm in bKT that was absent in WT samples (Fig. 2e), consistent with the accumulation of ketocarotenoids [15, 18, 19].
野生型（WT）和 bKT 菌株在摇瓶中以光自养方式培养，光照强度约为 100 µmol/m²/s 的白光，温度为 37°C。液体 MAD 培养基经过专门配制，用于培养 Syn11901 菌株 [2]，并补充了 100 mM NaHCO₃作为碳源。WT 和 bKT 菌株的一个典型特征是在进入稳定期时表现出不同的色素特征（图 2d）：WT 呈绿色，而 bKT 呈较深的棕褐色。从细胞中提取的叶绿素和类胡萝卜素的吸收光谱显示，bKT 在 500-550 nm 之间出现了一个肩峰，而 WT 样品中没有这一特征（图 2e），这与酮类胡萝卜素的积累一致 [15, 18, 19]。

To confirm the presence of ketocarotenoids, HPLC analyses comparing WT and bKT extracts were conducted (Fig. 2f). The attribution of the pigments was based on evaluating the absorption spectra of eluted fractions obtained by reversed-phase HPLC (Supplementary Figure S1, Additional File 1), integrated with previous analyses conducted in the lab with engineered microalgal strains [19]. The WT extract comprised three primary pigments: chlorophyll a (Chla), Zea, and βcar. Traces of mixoxanthophyll (Mixo), a carotenoid glycoside possibly involved in cell-membrane structure and thylakoid organization in cyanobacteria [29], were also observed at ~ 8 min. By contrast, bKT samples showed a very different profile. No Zea was observed; instead, a large amount of canthaxanthin (Cantha) was present, resulting from the conversion of βcar by the heterologous bKT catalytic activity. Consequently, the βcar/Chla ratio was lower in the bKT line compared to WT (Fig. 2f). 3S,3’S trans-astaxanthin (Asta), Mixo, and 3S,3’S 9-cis-astaxanthin (cisAsta) were also synthesized, although in minor amounts compared to Cantha.
为了确认酮类类胡萝卜素的存在，进行了 WT 和 bKT 提取物的 HPLC 分析比较（图 2f）。色素的归属基于通过反相 HPLC 获得的洗脱组分的吸收光谱评估（补充图 S1，附加文件 1），并结合实验室之前对工程化微藻菌株的分析结果[19]。WT 提取物包含三种主要色素：叶绿素 a（Chla）、玉米黄质（Zea）和β-胡萝卜素（βcar）。此外，在约 8 分钟处还观察到微量的混合黄素（Mixo），这是一种可能与蓝藻细胞膜结构和类囊体组织相关的类胡萝卜素糖苷[29]。相比之下，bKT 样品显示出完全不同的色谱特征。未检测到 Zea，而是观察到大量的角黄素（Cantha），这是通过异源 bKT 催化活性将βcar 转化而来的结果。因此，bKT 株系中的βcar/Chla 比值比 WT 低（图 2f）。此外，还合成了 3S,3’S 反式虾青素（Asta）、Mixo 和 3S,3’S 9-顺式虾青素（cisAsta），但相较于 Cantha，它们的含量较少。

Previous literature showed that levels of Zea increased in cells grown under high irradiance compared to cells grown under low irradiance [30, 31]. Since Zea is the substrate of bKT for the synthesis of Asta [8, 22], the level of Asta was investigated in bKT cells grown at high light intensity. To this aim, bKT cultures were grown in airlift 80-mL photobioreactors (PBRs) exposed to a higher light intensity than flask cultures described above, 1500 µmol/m²/s, and a 3% CO₂-enriched air supply. Samples were taken at day 3 and day 7 (Supplementary Figure S2, Additional File 1), followed by pigment extraction and analysis by HPLC. Compared to the previous growth condition in shake flasks, bKT strains accumulated similar amounts of Asta (Supplementary Figure S2, Additional File 1), confirming that this was an inherent feature of the transformant since Asta content did not considerably change in cultures grown under high irradiance, where the metabolic flux toward terpenoids was possibly enhanced. Conversely, Cantha abundantly accumulated, indicating that bKT was acting prevalently on βcar and poorly on Zea.
先前的研究表明，与在低光强下生长的细胞相比，在高光强下生长的细胞中 Zea 的水平有所增加[30, 31]。由于 Zea 是 bKT 合成 Asta 的底物[8, 22]，因此研究了在高光强下生长的 bKT 细胞中 Asta 的水平。为此，将 bKT 培养物在空气升液式 80 mL 光生物反应器（PBR）中培养，其光强高于上述烧瓶培养条件（1500 µmol/m²/s），并提供 3% CO₂富集空气。分别在第 3 天和第 7 天取样（补充图 S2，附加文件 1），随后通过 HPLC 进行色素提取和分析。与之前在摇瓶培养条件下相比，bKT 菌株积累了相似量的 Asta（补充图 S2，附加文件 1），这表明这是转化体的固有特性，因为在高光强下培养的情况下 Asta 的含量没有显著变化，而此时向萜类化合物的代谢通量可能有所增强。相反，Cantha 大量积累，这表明 bKT 主要作用于βcar，而对 Zea 的作用较弱。

Overall, these data suggested that the expression of the heterologous bKT from C. reinhardtii alone in Syn11901 was insufficient to promote substantial remodeling of the terpenoid pathway toward the synthesis of Asta. The reason is probably due to differences in the activities of the endogenous hydroxylases: in C. reinhardtii this is sufficient to convert Cantha to Asta, whereas in cyanobacteria the hydroxylase activity is insufficient. It is worth noting that even in C. reinhardtii overexpression of a β-carotene hydroxylase enzyme strongly increased the Asta yield at the expense of Cantha [32].
总体而言，这些数据表明，仅在 Syn11901 中表达来自 C. reinhardtii 的异源 bKT 不足以显著重塑类萜途径以合成虾青素。原因可能是内源性羟化酶活性差异所致：在 C. reinhardtii 中，这种活性足以将玉米黄质转化为虾青素，而在蓝藻中，羟化酶活性不足。值得注意的是，即使在 C. reinhardtii 中，过表达一种β-胡萝卜素羟化酶酶也显著提高了虾青素的产量，但以玉米黄质为代价[32]。

Heterologous expression of bKT and CrtZ in Synechococcus sp. PCC 11901
在聚球藻属 PCC 11901 中异源表达 bKT 和 CrtZ

Previous work has shown that heterologous expression of crtZ gene from Brevundimonas sp. SD-212 in Synechocystis sp. PCC 6803 successfully converts Cantha to Asta [17, 33]. Based on this, the crtZ gene was codon-optimized for expression in Syn11901 (see Materials and methods section) and inserted into the genome of Syn11901 using a second DNA construct, named p-BC, which was designed to enhance Asta yield in Syn11901 by co-expressing the bKT and CrtZ recombinant enzymes. In more detail, the p-BC construct harbored the same genetic elements present in the p-bKT construct but the kmR cassette was replaced by crtZ and a spectinomycin-resistance cassette (smR) in a single operon (Fig. 3a). In parallel, another construct was designed, harboring only the kmR cassette as a recombinant gene. The use of this last construct, referred to as p-KmR, was intended as a control for the replacement of acsA locus in the Syn11901 genome.
先前的研究表明，通过在聚球藻（Synechocystis sp. PCC 6803）中异源表达来自 Brevundimonas sp. SD-212 的 crtZ 基因，可以成功地将坎塔克桑素（Cantha）转化为虾青素（Asta）[17, 33]。基于此，crtZ 基因经过密码子优化以便在 Syn11901 中表达（详见材料与方法部分），并通过一个名为 p-BC 的第二个 DNA 构建体插入到 Syn11901 的基因组中。该构建体旨在通过共同表达 bKT 和 CrtZ 重组酶来提高 Syn11901 中虾青素的产量。具体来说，p-BC 构建体包含与 p-bKT 构建体相同的遗传元件，但将 kmR 盒替换为 crtZ 和一个在单一操纵子中的链霉素抗性盒（smR）（图 3a）。同时，设计了另一个仅携带 kmR 盒作为重组基因的构建体。这个构建体被命名为 p-KmR，用作替代 Syn11901 基因组中 acsA 位点的对照。

Fig. 3  图 3

Generation of BC and KmR transformants and pigment analysis. a Schematic of p-BC and p-KmR DNA constructs. The first was designed for the replacement of the kmR gene in the bKT recipient strain by homologous recombination. The recombinant crtZ gene and the spectinomycin-resistance cassette (smR) were in an operon configuration, with the bKT gene as the leader gene, and under the control of the P_cpt constitutive promoter. The p-kmR construct was designed for replacing the acsA locus with the kmR gene only. b Genomic DNA PCR analysis of BC transformants and bKT recipient strain using primers bKT fw and acsA-fl3’ rv. The expected size of the PCR products was 1677 and 2181 bp in bKT and BC lines, respectively. c Genomic DNA PCR analysis of KmR lines and WT recipient strain using primers acsA-5’ fw and acsA-3’ rv. The expected sizes of the PCR products were 3767 and 2862 bp in WT and KmR lines, respectively. d Comparison of the pigmentation of WT, KmR, bKT, and BC cultures photoautotrophically grown in flasks. BC culture showed a brownish coloration, different from the blue-green WT and KmR cultures. e Absorption spectra in the visible range of pigment extracts from WT, KmR, bKT and BC lines. Spectra were normalized to the absorbance of chlorophyll a. f Absorption spectra in the visible range of pigment extracts from WT, KmR, bKT and BC lines. Spectra were normalized to the absorbance of carotenoids. Arrow indicates the shift of absorption attributed to the accumulation of Asta in the BC line. g Representative HPLC profiles of WT, bKT, and BC pigment extracts. 1, Mixoxanthophyll; 2, zeaxanthin; 3, chlorophyll a; 4, echinenone; 5, βcarotene; 6, 3S,3’S trans-astaxanthin; 7, mixoxanthophyll; 8, 3S,3’S 9-cis-astaxanthin
BC 和 KmR 转化体的生成及色素分析。 a. p-BC 和 p-KmR DNA 构建的示意图。第一个构建旨在通过同源重组替换 bKT 受体菌株中的 kmR 基因。重组的 crtZ 基因和链霉素抗性盒（smR）以操纵子形式配置，bKT 基因为启动基因，并受 P _cpt 组成型启动子的控制。p-KmR 构建仅用于用 kmR 基因替换 acsA 位点。 b. 使用引物 bKT fw 和 acsA-fl3’ rv 对 BC 转化体和 bKT 受体菌株进行基因组 DNA PCR 分析。PCR 产物的预期大小分别为 bKT 和 BC 菌株中的 1677 bp 和 2181 bp。 c. 使用引物 acsA-5’ fw 和 acsA-3’ rv 对 KmR 菌株和 WT 受体菌株进行基因组 DNA PCR 分析。PCR 产物的预期大小分别为 WT 菌株和 KmR 菌株中的 3767 bp 和 2862 bp。 d. 比较 WT、KmR、bKT 和 BC 培养物在光自养条件下于烧瓶中生长的色素表现。BC 培养物表现为棕色，与蓝绿色的 WT 和 KmR 培养物不同。 e. WT、KmR、bKT 和 BC 菌株色素提取物在可见光范围内的吸收光谱。光谱以叶绿素 a 的吸光度进行了归一化。 f. WT、KmR、bKT 和 BC 菌株色素提取物在可见光范围内的吸收光谱。光谱以类胡萝卜素的吸光度进行了归一化。箭头指示吸收峰的移动，与 BC 菌株中 Asta（虾青素）积累有关。 g. WT、bKT 和 BC 色素提取物的代表性 HPLC 色谱图。1，混合黄嘌呤；2，玉米黄素；3，叶绿素 a；4，依克红酮；5，β胡萝卜素；6，3S,3’S 反式虾青素；7，混合黄嘌呤；8，3S,3’S 9-顺式虾青素。

Upon cyanobacterial transformation and growth in selective media, the genotype of the transformants was tested by PCR analysis (Fig. 3b, c). Genomic DNAs from three BC independent cultures were analyzed using bKT fw and acs-fl3’ rv primers. DNA from the bKT culture, the recipient strain of the p-BC construct, was used as a control. The latter strain generated a PCR product of 1677 bp, whereas the BC genomic DNAs generated a single PCR product of 2181 bp, confirming the correct replacement of the p-bKT construct with the p-BC construct (Fig. 3b). Similarly, PCR analyses were conducted to evaluate KmR transformants generated PCR products of 3767 bp and 2862 bp in WT control and KmR lines, respectively (Fig. 3c). This result confirmed complete segregation of the transgenes in the KmR transformants.
在蓝藻转化和选择性培养基中生长后，通过 PCR 分析测试了转化体的基因型（图 3b, c）。使用 bKT fw 和 acs-fl3' rv 引物分析了来自三个 BC 独立培养物的基因组 DNA。来自 bKT 培养物（p-BC 构建体的受体菌株）的 DNA 用作对照。后者菌株生成了一个 1677 bp 的 PCR 产物，而 BC 基因组 DNA 生成了一个 2181 bp 的单一 PCR 产物，证实了 p-bKT 构建体被 p-BC 构建体正确替换（图 3b）。同样，PCR 分析用于评估 KmR 转化体，在野生型（WT）对照和 KmR 菌株中分别生成了 3767 bp 和 2862 bp 的 PCR 产物（图 3c）。这一结果证实了 KmR 转化体中转基因的完全分离。

WT, KmR, bKT and BC photoautotrophic cultures were grown in shake flasks using ~ 100 µmol/m²/s of white light at 37 °C. The first two evidenced a similar green coloration, whereas BC liquid culture showed a brownish hue, more evident than the bKT strain (Fig. 3d), suggesting a different pigment composition.
WT、KmR、bKT 和 BC 自养光合作用培养物在摇瓶中培养，使用约 100 µmol/m²/s 的白光，培养温度为 37 °C。前两者表现出相似的绿色，而 BC 液体培养物显示出棕色调，比 bKT 菌株更为明显（图 3d），这表明其色素组成存在差异。

Astaxanthin productivity by photoautotrophic cultivation of engineered Synechococcus sp. PCC 11901
通过工程化改造的聚球藻属 PCC 11901 在光自养培养中的虾青素生产力

Samples from WT, KmR, bKT, and BC cultures were collected during exponential phase for pigment analysis. Absorption spectra in the visible region of pigment extracts from WT and the KmR control did not show major differences (Fig. 3e), as expected given the similar coloration of the liquid cultures (Fig. 3d). The spectrum from the BC strain showed a clear distinctive shoulder peaking at ~ 480 nm, in agreement with the accumulation of ketocarotenoids (Fig. 2e). This shoulder was even higher than the one observed in the bKT spectrum, possibly related to a greater value for the ratio of carotenoids to chlorophyll. Of interest was the comparison of the spectra normalized to the carotenoid content (Fig. 3f), which indicated that the absorption shoulder attributed to carotenoids was slightly shifted towards longer wavelengths in the BC extract compared to the case of bKT. This was corroborated by the evidence that, after pigment separation by HPLC, the maximum absorption peak of Cantha was 474 nm, whereas the maximum of Asta was red-shifted to 478 nm of (Supplementary Figure S1, Additional File 1).
从 WT、KmR、bKT 和 BC 培养物中采集了处于指数生长期的样品用于色素分析。WT 和 KmR 对照组色素提取物在可见光区域的吸收光谱未显示出明显差异（图 3e），这与液体培养物显示的相似颜色一致（图 3d）。BC 菌株的光谱在约 480 nm 处显示出一个明显的肩峰，与酮类类胡萝卜素的积累相符（图 2e）。这一肩峰甚至高于 bKT 光谱中观察到的肩峰，这可能与类胡萝卜素与叶绿素的比例更高有关。值得注意的是，将光谱归一化至类胡萝卜素含量后进行比较（图 3f），结果表明，BC 提取物中归因于类胡萝卜素的吸收肩峰相比 bKT 略微偏向更长的波长。这一点通过 HPLC 色素分离后的证据得到了验证：Cantha 的最大吸收峰为 474 nm，而 Asta 的最大吸收峰红移至 478 nm（补充图 S1，附加文件 1）。

Consistent with the absorption spectra, HPLC analyses confirmed that 3S,3’S trans-astaxanthin was the most abundant carotenoid in the BC line (Fig. 3g) and represented more than 70% of total carotenoid. No Zea was detected and only traces of Cantha were identified. βcar was poorly accumulated in both bKT and BC samples compared to WT. Traces of 3S,3’S 9-cis-astaxanthin were observed in the BC extract, probably as a side product of the Asta biosynthesis pathway.
与吸收光谱一致，HPLC 分析证实 3S,3'S 反式虾青素是 BC 品系中最丰富的类胡萝卜素（图 3g），占总类胡萝卜素的 70%以上。未检测到玉米黄质（Zea），仅检测到微量的胡萝卜素酮（Cantha）。与 WT 相比，β-胡萝卜素（βcar）在 bKT 和 BC 样品中的积累较少。在 BC 提取物中观察到微量的 3S,3'S 9-顺式虾青素，这可能是虾青素生物合成途径的副产物。

To assess protein expression, total cell extracts from Syn11901 WT and transformants were analyzed by SDS-PAGE followed by Coomassie blue staining or Western blot (Supplementary Figure S3a and b, Additional File 1). The most abundant polypeptides observed for all extracts were the α and β subunits of phycocyanin (~ 15 kDa), in accordance with other cyanobacteria [34, 35]. Heterologous bKT, CrtZ, KmR, and SmR have predicted molecular masses of ~ 31, ~ 17, ~ 28 and ~ 26 kDa, respectively. However, there was no evidence in the gel for their overexpression using the P_cpt promoter. This suggested low accumulation of the heterologous enzymes under the control of Pcpt promoter [2, 26]. Nevertheless, the strong redirection of the carotenoid biosynthetic pathway toward astaxanthin in the BC strain (Fig. 3g) confirms that the enzymatic activities of bKT and CrtZ were sufficient to achieve the desired metabolic engineering, with no need to further increase the accumulation of bKT and CtrZ enzymes to boost ketocarotenoid accumulation. To prove the stable expression of the BC DNA operon under the control of the P_cpt promoter, RT-PCR analyses were conducted using the cDNA sequences derived from the BC cultures grown for different days in airlift PBRs as templates. In more detail, algal cultures were grown for 3, 4, 7 and 10 days under high-light condition (see also Fig. 5a). The endogenous constitutive rpnA gene transcript was expressed in each sample, as expected. Similarly, BC operon expression, evaluated through the amplification of crtZ gene transcript, was observed in each tested condition (Supplementary Figure S3c, Additional File 1). The primers used for the amplification of the crtZ gene transcript are listed in Table 1.
为了评估蛋白质的表达，从 Syn11901 野生型（WT）和转化株中提取的总细胞裂解物通过 SDS-PAGE 分析，并使用考马斯亮蓝染色或 Western blot 检测（补充图 S3a 和 b，附加文件 1）。在所有提取物中观察到的最丰富的多肽是藻胆蛋白的α和β亚基（约 15 kDa），这与其他蓝藻的研究一致[34, 35]。异源 bKT、CrtZ、KmR 和 SmR 的预测分子量分别为约 31、约 17、约 28 和约 26 kDa。然而，在凝胶中未发现使用 P _cpt 启动子导致其过表达的证据。这表明 Pcpt 启动子控制下的异源酶积累较低[2, 26]。尽管如此，在 BC 菌株中类胡萝卜素生物合成途径向虾青素强烈重定向（图 3g）表明，bKT 和 CrtZ 的酶活性足以实现所需的代谢工程，无需进一步增加 bKT 和 CrtZ 酶的积累来提高酮类胡萝卜素的积累。为了证明在 P _cpt 启动子控制下 BC DNA 操纵子的稳定表达，使用从不同培养天数的 BC 培养物中提取的 cDNA 序列作为模板进行了 RT-PCR 分析。更具体地说，藻类培养物在高光条件下培养 3、4、7 和 10 天（另见图 5a）。如预期的那样，每个样本中均检测到内源性组成型 rpnA 基因转录本。同样，通过 crtZ 基因转录本的扩增评估的 BC 操纵子表达在每个测试条件下均被观察到（补充图 S3c，附加文件 1）。用于扩增 crtZ 基因转录本的引物列于表 1 中。

To evaluate the impact of Asta accumulation on Syn11901 growth rate, WT, KmR and BC strains were cultivated in the airlift PBRs using the same conditions previously adopted for bKT lines using 1500 μmol/m²/s of light and 3% (v/v) CO₂ (Supplementary Figure S4a, Additional File 1). BC lines were characterized by a faster growth rate than WT and KmR lines (Fig. 4a, b), with stationary phase reached after ~ 50 h of growth compared to more than 80 h for WT and KmR. KmR cultures showed slightly delayed growth compared to WT, in agreement with previous literature showing that the replacement of the acsA locus causes a reduction in growth rate [27], although in stationary phase WT and KmR cultures reached similar ODs.
为了评估 Asta 累积对 Syn11901 生长速率的影响，将 WT、KmR 和 BC 菌株在气升式光生物反应器（PBRs）中培养，采用与之前用于 bKT 菌株相同的条件，即 1500 μmol/m²/s 的光照强度和 3%（v/v）的 CO₂（补充图 S4a，附加文件 1）。结果显示，BC 菌株的生长速率比 WT 和 KmR 菌株更快（图 4a、b），在约 50 小时后进入稳定期，而 WT 和 KmR 则需要超过 80 小时。KmR 培养物的生长相比 WT 略有延迟，这与之前的文献一致，表明替换 acsA 位点会导致生长速率的降低[27]，尽管在稳定期 WT 和 KmR 培养物达到了相似的光密度（OD）。

Fig. 4  图 4

Effect of the accumulation of non-native astaxanthin on the microalgal growth. a Growth curves of WT, KmR and BC lines cultivated in batch in the 80-ml airlift PBRs, using OD₇₂₀ as an index of cell density. Cultures were exposed to 300 µmol/m²/s photons of white light for the first 8 h, then light was increased from 300 to 1500 µmol/m²/s in 12 h. Eventually, light was kept at 1500 µmol/m²/s until the end of the experiment. WT and KmR lines served as controls. Two independent cultures per condition were evaluated. Error bars represented standard deviation (n = 4). b Manual OD₇₂₀ measurements, conducted on a daily basis, using diluted samples to avoid saturation. This analysis was in agreement with automatic OD-monitoring of the device used for cyanobacterial cultivations (a). Two independent cultures per condition were evaluated. Error bars represented standard deviation (n = 4)
外源虾青素积累对微藻生长的影响。 a. 在 80 毫升气升式光生物反应器中批量培养 WT、KmR 和 BC 菌株的生长曲线，使用 OD ₇₂₀ 作为细胞密度的指标。培养物在前 8 小时暴露于 300 µmol/m ² /s 的白光照射下，然后光强在 12 小时内从 300 µmol/m ² /s 增加到 1500 µmol/m ² /s。最后，光强保持在 1500 µmol/m ² /s 直至实验结束。WT 和 KmR 菌株作为对照。每种条件下均评估了两组独立培养。误差条表示标准偏差（n = 4）。 b. 每日通过手动 OD ₇₂₀ 测量进行分析，使用稀释样品以避免饱和。该分析结果与用于蓝细菌培养的设备的自动 OD 监测结果（a）一致。每种条件下均评估了两组独立培养。误差条表示标准偏差（n = 4）。

Table 2 reports the dry biomass accumulated by the different cultures at days 4 and 7: the highest biomass accumulation was obtained for the WT and KmR lines, which reached ~ 3.4 g/L at day 4 and ~ 7.8 g/L. BC cultures accumulated less biomass at day 7 than the other two strains. Absorption spectra of acetone pigment extracts obtained from BC cells harvested on day 4 or 7 indicated enhanced absorption at ~ 500 nm attributable to ketocarotenoids (Supplementary Figure S4b, Additional File 1). HPLC analysis confirmed the Asta accumulated to ~ 0.4% of total cell biomass in BC cultures on both day 4 and 7 (Supplementary Figure S4c, Additional File 1). As a further confirmation of the synthesis of Asta in the BC strain, a preparative thin-layer chromatography (TLC) was performed. The TLC profile in the BC strain was in agreement with the HPLC analysis, considering the abundant red band present only in the BC extract which possibly represented Asta (Supplementary Figure S4d, Additional File 1). The eluted pigment in acetone 80% had a maximum of absorption at 480 nm, same as previously described for astaxanthin (Supplementary Figure S4e, Additional File 1). In addition, the eluted fraction was analyzed by mass spectrometry. Interestingly, two peaks referring to Asta were identified, the protonated molecule [M + H]⁺ at m/z 597 and the metal adduct ion [M + Na]⁺ at m/z 619 (Supplementary Figure S4f, Additional File 1), as described in literature [36].
表 2 显示了不同培养物在第 4 天和第 7 天积累的干生物量：WT 和 KmR 菌株的生物量积累最高，第 4 天达到约 3.4 g/L，第 7 天达到约 7.8 g/L。相比之下，BC 培养物在第 7 天的生物量积累少于其他两种菌株。从第 4 天或第 7 天采集的 BC 细胞中提取的丙酮色素的吸收光谱显示，在约 500 nm 处的吸收增强，这归因于酮类胡萝卜素（补充图 S4b，附加文件 1）。HPLC 分析确认 BC 培养物在第 4 天和第 7 天积累的 Asta 约占总细胞生物量的 0.4%（补充图 S4c，附加文件 1）。为了进一步确认 BC 菌株中 Asta 的合成，进行了制备型薄层色谱（TLC）分析。BC 菌株的 TLC 图谱与 HPLC 分析一致，BC 提取物中仅存在的丰富红色条带可能代表 Asta（补充图 S4d，附加文件 1）。丙酮（80%）中洗脱的色素在 480 nm 处显示吸收峰，与之前对虾青素的描述一致（补充图 S4e，附加文件 1）。此外，洗脱的色素组分通过质谱分析。令人感兴趣的是，发现两个与 Asta 相关的峰，即质子化分子[M + H] ⁺ ，m/z 为 597，以及金属加合离子[M + Na] ⁺ ，m/z 为 619（补充图 S4f，附加文件 1），与文献[36]描述一致。

Table 2 Biomass and Asta accumulation in WT, KmR and BC cultures grown in the airlift PBRs
表 2 在气升式光生物反应器中生长的 WT、KmR 和 BC 培养物的生物质和虾青素积累情况

Differently from Asta, Chla content was decreased in the BC line compared to both WT and KmR strains (Table 2), which is consistent with the phenotype of C. reinhardtii strains engineered to accumulate Asta [19]. This evidence was corroborated by whole-cell absorption spectra taken and normalized to the chlorophyll content (Supplementary Figure S4g, Additional File 1). WT and KmR spectra were, as expected, similar, whereas BC samples showed enhanced absorption, especially in the blue region of visible light. This is attributable to greater light scattering, a proxy of a greater cell density, in BC cultures compared to those WT and KmR, containing the same chlorophyll content.
与 Asta 不同，BC 株系的 Chla 含量相比 WT 和 KmR 株系有所下降（表 2），这与通过基因工程积累 Asta 的 C. reinhardtii 株系的表型一致[19]。这一证据通过对全细胞吸收光谱进行测量并归一化至叶绿素含量得以证实（补充图 S4g，附加文件 1）。如预期所料，WT 和 KmR 的光谱相似，而 BC 样本的吸收增强，尤其是在可见光的蓝光区域。这归因于 BC 培养物中较大的光散射，相比 WT 和 KmR，其细胞密度更高，但叶绿素含量相同。

Accumulation of Asta was monitored in a batch culture of BC cells grown in airlift PBRs with a final light intensity of 1500 μmol/m²/s, and cells were sampled at days 3, 4, 7 and 10 of growth. The Asta content (mg/L) increased over-time (Fig. 5a) and % Asta/car was higher than 80% in each condition. Most importantly, % Asta/dcw did not substantially vary in the evaluated phases of the cyanobacterial cultivation.
在气升式光生物反应器中，以最终光强度为 1500 μmol/m²/s 培养的 BC 细胞的分批培养中监测了 Asta 的积累情况，并在培养的第 3 天、第 4 天、第 7 天和第 10 天取样。Asta 含量（mg/L）随时间增加（图 5a），且 Asta/car 的百分比在每种条件下均高于 80%。最重要的是，% Asta/dcw 在蓝藻培养的各评估阶段并未发生显著变化。

Fig. 5  图 5

Astaxanthin and biomass accumulation of cyanobacterial transformant. a Accumulation of astaxanthin in BC cultures grown in batch mode for 10 days in 80-ml airlift PBRs. Samples were collected on days 3, 4, 7, and 10. Cells were exposed to an initial gradient of white light, reaching 1500 µmol/m²/s at day 2. Statistical significance is expressed by different letters according to Tukey–Kramer test (n = 4). b Growth curves of BC lines cultivated in batch mode in 80-ml airlift PBRs, using OD₇₂₀ as an index of cell density. Cultures were exposed to 250 µmol/m²/s of white light for the first 6 h. The light was then increased following a stepwise gradient (1 h to increase light intensity to the next level of light intensity, which was kept for 6 h, and then this whole step was further replicated until reaching the desired light intensity). Final light intensities reached in the 4 tested conditions were 250 µmol/m²/s, 750 µmol/m²/s, 1500 µmol/m²/s, and 2250 µmol/m²/s. Two independent cultures per condition were evaluated. Error bars represented standard deviation (n = 4). c OD₇₂₀ of samples described in a: samples were diluted accordingly below 1 before measurements to avoid saturation. Error bars represented standard deviation (n = 4)
蓝藻转化株的虾青素和生物量积累情况。 a. 在 80 毫升气升式光生物反应器（PBR）中，以分批模式培养 10 天后，BC 培养物中虾青素的积累情况。在第 3、4、7 和 10 天采集样品。细胞在培养初期暴露于白光梯度，第 2 天达到 1500 µmol/m²/s。统计显著性通过 Tukey-Kramer 检验以不同字母表示（n = 4）。 b. BC 株系在 80 毫升气升式 PBR 中以分批模式培养的生长曲线，以 OD 作为细胞密度的指标。培养物在最初 6 小时暴露于 250 µmol/m²/s 的白光，然后按照逐步梯度增加光强（每小时增加至下一级光强，并保持 6 小时，整个步骤重复，直到达到目标光强）。4 种测试条件下的最终光强分别为 250 µmol/m²/s、750 µmol/m²/s、1500 µmol/m²/s 和 2250 µmol/m²/s。每种条件评估了两组独立培养样品。误差线表示标准差（n = 4）。 c. a 中描述的样品的 OD：测量前样品根据需要稀释至低于 1，以避免饱和。误差线表示标准差（n = 4）。

To better understand the relationship between the accumulation of biomass and Asta, a BC culture was grown under stepwise light gradients in airlift PBRs, reaching final light intensities of 250, 750, 1500, or 2250 µmol photons/m²/s (Fig. 5b). These experiments revealed a correlation between light intensity, cell growth (Fig. 5b, c) and biomass accumulation (Table 3): 4 days growth at 250 µmol/m²/s resulted in ~ 2.2 g/L of dry cyanobacterial biomass (dry cell weight, dcw), whereas the 2250 µmol/m²/s culture generated ~ 7.5 g dcw/L. Accordingly, the biomass surface productivity, calculated based on the surface area of PBRs exposed to light, increased from 9 to 31.3 g/m²/day (Table 3) upon increasing the light intensity. The areal productivity recorded was higher compared to other microalgal records reported in the literature [37], even considering optimized condition in a semicontinuous cultivation system [38]. Potentially, higher biomass productivity could be obtained upon continuous cultivation [39].
为了更好地理解生物质累积与虾青素（Asta）之间的关系，在气升式光生物反应器（PBRs）中，通过逐步增加光照梯度培养 BC 菌株，最终达到 250、750、1500 或 2250 µmol 光子/m²/s 的光强（图 5b）。这些实验揭示了光强、细胞生长（图 5b、5c）与生物质累积（表 3）之间的相关性：在 250 µmol/m²/s 的光强下培养 4 天，得到约 2.2 g/L 的干蓝藻生物质（干细胞重量，dcw），而在 2250 µmol/m²/s 的培养条件下，生成了约 7.5 g dcw/L。相应地，根据光照面积计算的 PBRs 表面生物质生产力从 9 g/m²/天增加到了 31.3 g/m²/天（表 3）。记录的面积生产力相比文献中其他微藻的报道更高[37]，即使是在半连续培养系统的优化条件下[38]。在连续培养条件下，可能会获得更高的生物质生产力[39]。

Table 3 Relation among light intensity, surface biomass productivity, biomass and Asta accumulation in BC cultures grown in the airlift PBRs
表 3 光强度、表面生物量生产率、生物量和虾青素积累在空气升液管光生物反应器中 BC 培养的关系

The absorption spectra of acetonic extracts from samples collected after 4 days of cultivation were taken. Such spectra, after normalization to the Chla content, showed an absorption at 480 nm, attributable to carotenoids, that correlated with light intensity (Supplementary Figure S5, Additional File 1). Conversely, spectra normalized to the carotenoid content evidenced a very similar absorption attributable to carotenoids (Supplementary Figure S5, Additional File 1). This suggested a similar carotenoid composition of carotenoids accumulating in the photosynthetic membranes of BC cells grown under the different light conditions.
对从培养 4 天后采集的样品的丙酮提取物进行了吸收光谱测定。将这些光谱归一化为叶绿素 a（Chla）含量后显示，在 480 nm 处有一个吸收峰，可归因于类胡萝卜素，并与光强呈相关性（补充图 S5，附加文件 1）。相反，将光谱归一化为类胡萝卜素含量后，显示出一个非常相似的吸收峰，也归因于类胡萝卜素（补充图 S5，附加文件 1）。这表明，在不同光照条件下生长的 BC 细胞光合膜中积累的类胡萝卜素成分相似。

Astaxanthin accumulation on a volume basis was greater at high light intensities, reaching ~ 38.4 mg/L in 2250 µmol/m²/s samples after 4 days of growth. Moreover, Asta represented more than 80% of total carotenoid (Table 3). The Chla/biomass ratio content was lower in the culture grown at higher light intensities, suggesting a reduction of photosystems per cell in response to the greater photon availability. The increased growths at higher light intensities, in the presence of a non-limiting amount of CO₂, showed that Syn11901 could tolerate very high light with a surface productivity that, under the conditions tested, was close to 30 g/m²/day. From the productivity analysis (Table 3), it emerged that photosynthetic active cells of Syn11901 produce up to ~ 0.6% of astaxanthin on cell dry weight against the 3 mg/g (0.3%) reported for the closely related Synechococcus PCC 7002 [15].
在高光强条件下，按体积计算的虾青素积累量更高，在 2250 µmol/m²/s 光照条件下培养 4 天后达到约 38.4 mg/L。此外，虾青素占总类胡萝卜素的比例超过 80%（表 3）。在高光强条件下培养的藻类，其叶绿素 a 与生物量的比值较低，表明细胞为了适应更高的光子供应量，减少了每个细胞的光合系统数量。在 CO₂供应不受限制的情况下，高光强条件下的生长表明 Syn11901 能够耐受极高的光强，其表面生产率在测试条件下接近 30 g/m²/天。从生产率分析（表 3）得出，Syn11901 的光合作用活性细胞的虾青素产量可达细胞干重的约 0.6%，相比之下，与其密切相关的 Synechococcus PCC 7002 报告的虾青素产量为每克干重 3 mg（0.3%）[15]。

Astaxanthin accumulation is not triggered by nutrient starvation
虾青素的积累不是由营养匮乏引发的

Nutrient starvation is known to boost astaxanthin accumulation in several algal species, such as H. lacustris [10] or C. zofingiensis [12].
营养缺乏已知能够促进多种藻类物种中虾青素的积累，例如 **H. lacustris** [10] 或 **C. zofingiensis** [12]。

To evaluate the effect of nutrient starvation on astaxanthin productivity in the engineered Syn11901, the BC strain was cultivated in replete MAD medium or MAD medium deprived of a specific nutrient inducing nitrogen, phosphorus or iron starvation (see Materials and methods section). After 3 days of growth at 2250 µmol photons/m²/s, nitrogen starvation caused a decrease in biomass productivity compared to the nutrient-replete condition (Fig. 6a, b). However, the Asta concentration per dry weight was similar or even lower compared to the control condition (Table 4). The negative effect on biomass accumulation was less evident in cultures depleted of phosphorus or iron, suggesting a sufficient endogenous pool of the two elements.
为了评估营养缺乏对工程化藻株 **Syn11901** 中虾青素生产力的影响，将 BC 菌株培养在充足的 MAD 培养基中，或在缺乏某种特定营养（如氮、磷或铁）以诱导营养缺乏的 MAD 培养基中（参见材料与方法部分）。在 2250 µmol 光子/m²/s 的光强下生长 3 天后，氮缺乏导致与营养充足条件相比生物量生产力下降（图 6a, b）。然而，干重中的虾青素浓度与对照条件相比相似甚至更低（表 4）。而在缺磷或缺铁的培养中，对生物量积累的负面影响较不明显，这表明细胞内这两种元素的内源储备可能是足够的。

Fig. 6  图 6

Nutrient starvation of BC cultures in airlift PBRs. a Growth of BC cultures cultivated in batch mode in 80-ml airlift PBRs in the absence of specific nutrients, such as iron (Fe), nitrogen (N) and phosphorus (P). BC cells were previously grown in MAD medium reaching stationary phase, and then cells were washed and resuspended to OD₇₂₀ of 3 in the different media. MAD medium was used as a control. Pictures were taken on days 0, 1, and 2. b OD₇₂₀ of samples described in A: samples were diluted accordingly before measurements to avoid saturation. Error bars represented standard deviation (n = 4)
BC 培养物在气升式光生物反应器中的营养匮乏实验。 a. 在 80 毫升气升式光生物反应器中批量培养 BC 培养物，缺乏特定营养物质（如铁（Fe）、氮（N）和磷（P））。BC 细胞先在 MAD 培养基中生长至稳定期，然后将细胞洗涤后重悬于不同培养基中，初始 OD ₇₂₀ 为 3。MAD 培养基作为对照。照片分别拍摄于第 0 天、第 1 天和第 2 天。 b. 图 a 中所描述样品的 OD ₇₂₀ 值：在测量前样品根据需要稀释以避免饱和。误差线表示标准偏差（n = 4）。

Table 4 Evaluation of the effect of nutrient starvation on Asta accumulation in BC cultures grown in the airlift PBRs shown in Fig. 6
表 4 营养缺乏对空气升液式光生物反应器（图 6 所示）中 BC 培养物中虾青素积累影响的评估

Exposure to nutritional stressing conditions is thus not an advantageous strategy to boost Asta productivity in Syn11901 BC strain. Differently, in the case of endogenous Asta accumulation observed in H. lacustris [10] or C. zofingiensis [12], Asta accumulation is specifically triggered to overcome oxidative stress in conditions where the photosynthetic activity is strongly downregulated. Optimization of Asta production by Syn11901 BC strain is thus related to improving carbon fixation and biomass productivity and continuing the metabolic engineering of Syn11901.
因此，将营养胁迫条件作为提高 Syn11901 BC 菌株 Asta 产量的策略并不可取。不同的是，在 H. lacustris [10]或 C. zofingiensis [12]中观察到的内源性 Asta 积累，Asta 的积累是为了应对强光合作用下调条件下的氧化胁迫。优化 Syn11901 BC 菌株的 Asta 产量应侧重于提高碳固固定和生物质生产力，并继续对 Syn11901 进行代谢工程改造。

Up-scaling of cyanobacterial cultivation
蓝藻培养的规模化

Next, cultivation of the BC strain was scaled up to a 330-L PBR. The PBR was air-bubbled for cell mixing and supplied with pure CO₂ on demand to maintain the desired pH (see Material and methods). Cells were inoculated to an OD₇₂₀ of 0.033, a lower value compared to the previous experiments conducted in the 80-ml PBRs. Despite the low cyanobacterial inoculum, the color of the culture already shifted from a pale yellow to a more vivid green after 1 day of cultivation, indicating photosynthetic activity (Fig. 7a). After 4 days of cultivation, the culture had a dark coloration, suggesting a continuous growth of the BC strain. No clear visible differences in pigmentation of the culture were present from 4 to 8 days of growth, although cell density continuously increased during this time span, as evidenced by OD₇₂₀ measurements (Fig. 7b). However, the measured OD₇₂₀ was lower compared to what observed in the 80-ml PBRs (Fig. 5c). In the latter, OD₇₂₀ ranged between 5 and 15 after 4 days of growth, depending on the light irradiance, whereas, in the case of the 330-L PBR, the increase of biomass accumulation slightly slowed down after 4 days of growth, reaching an OD₇₂₀ of ~ 3.8 after 8 days of cultivation (Fig. 7b).
接下来，将 BC 菌株的培养扩大到 330 升的光生物反应器（PBR）。PBR 通过气泡供气进行细胞混合，并根据需要供应纯 CO₂ 以维持所需的 pH 值（参见材料与方法）。细胞以 OD ₇₂₀ 为 0.033 的值接种，比之前在 80 毫升 PBR 中的实验值更低。尽管蓝细菌接种量较低，但培养液的颜色在培养 1 天后已从浅黄色变为更鲜艳的绿色，表明发生了光合作用（图 7a）。在培养 4 天后，培养液呈现出深色，表明 BC 菌株的持续生长。从培养第 4 天到第 8 天，培养液的色素变化没有明显差异，但在此期间，细胞密度持续增加，这通过 OD ₇₂₀ 测量得到了证明（图 7b）。然而，测得的 OD ₇₂₀ 值低于 80 毫升 PBR 中观察到的值（图 5c）。在后者中，根据光照强度的不同，培养 4 天后的 OD ₇₂₀ 范围在 5 到 15 之间，而在 330 升 PBR 的情况下，生物量的积累在培养 4 天后略有减缓，在培养 8 天后达到约 3.8 的 OD ₇₂₀ （图 7b）。

Fig. 7  图 7

Cultivation of BC cells in an industrial PBR. a BC strain was photoautotrophically grown in batch mode in a 330L-PBR with limited sterility. The numbers on top of each picture indicate the day of growth. b Biomass accumulation curve of BC strain, as measured from the optical density of the cultures at 720 nm. Cultures were inoculated to an initial cell concentration corresponding to OD₇₂₀ 0.037 ± 0.003. Error bars represent standard deviation (n = 4)
在工业级光生物反应器（PBR）中培养 BC 细胞。 a. 在一个 330L-PBR 中，以有限的无菌条件下，采用分批模式光自养培养了一种 BC 菌株。每张图片上方的数字表示培养的天数。 b. BC 菌株的生物量累积曲线，通过测量培养液在 720 nm 处的光密度（OD）获得。培养物的初始细胞浓度对应于 OD ₇₂₀ 0.037 ± 0.003。误差线表示标准偏差（n = 4）。

After 8 days of growth, the cyanobacterial biomass was ~ 1 g dcw/L with a biomass surface productivity of 16.7 g/m²/day and a final Asta accumulation of 0.4% of total dry biomass (Table 3). Despite the lower biomass concentration obtained compared to the 80-mL PBR growth reported above, the surface productivity was in the range of the surface productivity obtained upon illumination with 750 µmol/m²/s, indicating that light penetration in the large-scale 330-L PBR was the limiting factor for biomass accumulation.
经过 8 天的生长，蓝藻生物质约为 1 g dcw/L，生物质表面生产率为 16.7 g/m²/天，最终虾青素（Asta）积累量为总干生物质的 0.4%（表 3）。尽管与上述 80 毫升光生物反应器（PBR）生长实验相比，获得的生物质浓度较低，但其表面生产率与在 750 µmol/m²/s 光照条件下获得的表面生产率相当，这表明在大型 330 升 PBR 中，光渗透是限制生物质积累的主要因素。

Discussion  讨论

The advancements in synthetic biology promoted the generation of alternative microalgal strains [15,16,17,18,19,20,21, 40, 41], which were genetically modified with the aims of synthesizing non-native Asta and, most importantly, overcoming limitations of H. lacustris. Cyanobacterial platforms had the advantage over eukaryotic microalgae of efficiently synthesizing a smaller pool of carotenoids, mainly βcar and Zea, the precursors of microalgal Asta [22].
**翻译为简体中文：** 合成生物学的进步促进了替代微藻菌株的产生[15, 16, 17, 18, 19, 20, 21, 40, 41]，这些菌株经过基因改造，旨在合成非天然的虾青素（Asta），更重要的是克服**H. lacustris**的局限性。与真核微藻相比，蓝藻平台具有优势，因为它们能更高效地合成更少种类的类胡萝卜素，主要是β-胡萝卜素（βcar）和玉米黄质（Zea），这些是微藻虾青素的前体[22]。

The first step of the experimental effort described in this manuscript was inserting of the bKT construct (Fig. 2a) in the acsA locus, whose sequences for HR were already available [2]. The obtained transformant was characterized by a brownish pigmentation (Fig. 3a), similarly to other engineered microalgae accumulating ketocarotenoids [15,16,17,18,19,20,21, 40,41,42]: accordingly, HPLC analysis demonstrated the main accumulation of Cantha in the bKT transformant, while Asta was found only as a minor fraction of total carotenoids. This finding demonstrates that the activity of the endogenous hydroxylation activity of CrtZ enzyme was limiting the conversion of Cantha to Asta. Thus, a second round of transformation was conducted, replacing the kanamycin-resistance cassette of bKT transformant with crtZ from Brevundimonas sp. SD-212 [17] and smR genes, with the latter conferring resistance to spectinomycin, in an operon configuration. The choice for the use of crtZ from Brevundimonas sp. SD-212 over other possible βcar hydroxylases was due to previous literature about the efficiency of this enzyme in converting Cantha to Asta [17, 33]. Moreover, the prokaryotic origin of the crtZ gene herein adopted mitigate the risk of inefficient heterologous gene expression observed in some cases expressing eukaryotic genes in cyanobacteria [43, 44]. HPLC analysis confirmed that Cantha was almost absent in BC extract, being entirely replaced by Asta (Fig. 3g). Remarkable was that the SDS-PAGE analysis of total protein extracts from the evaluated lines (Supplementary Figure S3, Additional File 1) showed no bands attributable to the recombinant bKT and CrtZ enzymes. This suggested low accumulation of the heterologous enzymes under the control of Pcpt promoter [2, 26]. Nevertheless, transcripts analysis demonstrated that the enzymes were expressed (Supplementary Figure S3c, Additional File 1) successfully redirecting carotenoid biosynthesis. Anyway, other genetic tools could be evaluated in Syn11901 to further boost recombinant enzymes expression because the success of synthetic biology approaches generally requires true overexpression of pathway enzymes and proteins of interest to attain higher yields and lower costs [44].
本文所描述实验工作的第一步是将**bKT**构建体（图 2a）插入到**acsA**基因位点，该位点的同源重组序列已被确定[2]。获得的转化体表现出棕褐色的色素沉积（图 3a），这与其他能够积累酮类胡萝卜素的工程化微藻相似[15, 16, 17, 18, 19, 20, 21, 40, 41, 42]。相应地，HPLC 分析显示，**bKT**转化体主要积累了坎托黄质（Cantha），而虾青素（Asta）仅占类胡萝卜素总量的一小部分。这一发现表明，内源性**CrtZ**酶的羟化活性限制了坎托黄质向虾青素的转化。因此，进行了第二轮转化，用来自**Brevundimonas sp. SD-212**的**crtZ**基因和**smR**基因替换了**bKT**转化体中的卡那霉素抗性盒，后者能够以操纵子配置提供抗壮观霉素的能力。选择使用**Brevundimonas sp. SD-212**的**crtZ**基因而非其他可能的β-胡萝卜素羟化酶，是基于已有文献中关于该酶高效将坎托黄质转化为虾青素的报道[17, 33]。此外，此处采用的**crtZ**基因来源于原核生物，这减少了在蓝藻中表达真核基因时可能出现的异源基因表达效率低下的风险[43, 44]。HPLC 分析确认，在**BC**提取物中，坎托黄质几乎完全消失，被虾青素完全取代（图 3g）。值得注意的是，对评估菌株的总蛋白提取物进行 SDS-PAGE 分析（补充图 S3，附加文件 1）后，没有发现任何条带可归因于重组的**bKT**和**CrtZ**酶。这表明，在**Pcpt**启动子的控制下，异源酶的积累水平较低[2, 26]。尽管如此，转录分析显示，这些酶成功表达（补充图 S3c，附加文件 1），并有效地重新导向了类胡萝卜素的合成途径。不管怎样，可以在**Syn11901**中评估其他基因工具，以进一步提高重组酶的表达，因为合成生物学方法的成功通常需要目标途径酶和蛋白质的真正过表达，以获得更高的产量和更低的成本[44]。

Importantly, the BC line displayed the fastest growth in exponential phase, despite replacement of acsA (Fig. 4). Thus, the presence of astaxanthin did not negatively impact growth of the BC line despite the strong reduction of Zea and βcar in the Syn11901 BC strain. Zea is usually found in lipid membranes having a role in photoprotection in cyanobacteria [45], but this role is likely complemented by Asta in Syn11901 BC strain. βcar is essential for photosystems assembly, but the residual βcar is likely sufficient for ensuring the accumulation of the photosynthetic complex required for photoautotrophic growth of Syn11901 BC strain. Rather, Syn11901 BC was characterized by a ‘boost’ in the initial stages of growth, with a faster growth rate in the exponential phase compared to WT. This effect is consistent with previous growth data for C. reinhardtii engineered to accumulate Asta [46] and could be due to: (1) the antioxidant properties of astaxanthin protecting cells and photosystems in the early exponential phase, where the culture is relatively dilute [46, 47] and/or (2) the reduced amount of chlorophyll per cell (Table 2), conferring a pale-green-like phenotype, which favors greater light penetration, thereby enhancing photosynthetic efficiency [34].
值得注意的是，尽管替换了 *acsA*（图 4），BC 菌株在指数生长期表现出最快的生长速度。因此，尽管 Syn11901 BC 菌株中的 Zea 和 β-胡萝卜素显著减少，虾青素的存在并未对 BC 菌株的生长产生负面影响。Zea 通常存在于脂质膜中，在蓝细菌中起到光保护作用 [45]，但这个作用可能在 Syn11901 BC 菌株中由虾青素取代。β-胡萝卜素对于光合系统的组装至关重要，但残留的 β-胡萝卜素可能已经足以确保光合复合物的积累，从而支持 Syn11901 BC 菌株的自养光合作用生长。相反，Syn11901 BC 在生长初期表现出“加速”，在指数生长期的生长速率比野生型更快。这一现象与之前针对工程化累积虾青素的莱茵衣藻 (*C. reinhardtii*) 的生长数据一致 [46]，可能是由于以下原因： (1) 虾青素的抗氧化特性在早期指数生长期（此时培养物相对稀释）保护细胞和光合系统 [46, 47]； (2) 每个细胞的叶绿素含量减少（表 2），呈现出类似浅绿色的表型，从而有利于更好的光穿透，增强光合作用效率 [34]。

The yield of Asta produced under the best conditions tested resulted in 38.4 mg/L after 4 days of cultivation (Table 3) at an average rate of production of ~ 9.6 mg/L/day, which exceeds that for H. pluvialis, with reported yields in the range of 0.12–4.4 mg/L/day using tubular or bubble columns [48]. Even if a higher production yield could be obtained for H. pluvialis in more complex cultivation systems [49, 50], increasing light availability could further increase the Asta production yield even in the case of the BC strain herein reported. Asta production in Syn11901 is also substantially higher than Asta heterologously synthesized in the cyanobacterium Synechocystis sp. PCC 6803 (2.8 mg/L/day [21]) and the green alga C. reinhardtii (6.96 mg/L/day; [32]). Engineered yeasts produce Asta at rates of 37.5 mg/L/day [51], but growth is heterotrophic and relies on adding glucose to the media. A summary of the production yield obtained in this work compared to other systems is reported in Table 5. One of the advantage in using Syn11901 compared to other systems is the fact that most carotenoids are present in this species as zeaxanthin and βcar. These molecules are the substrates of the bKT and CtrZ enzymes introduced, allowing for efficient production of Asta (Fig. 1). Other carotenoids are strongly accumulated in other systems, such as Chlamydomonas reinhardtii or Nannochloropsis, providing competitive sinks for the precursors needed for astaxanthin biosynthesis. As reported in Table 3, the lowest Asta content per dry weight was observed in cells grown at the lowest irradiances in 80-ml photobioreactors or in cells grown on a 330-L scale, where light penetration is strongly limited (Table 3): exposure to sufficient light is thus needed to boost astaxanthin content per dry weight. Even if nutritional stress did not provide any improvement in Asta titer (Table 4), we cannot exclude that other stressing conditions might somehow improve carotenoids biosynthesis.
在最佳测试条件下，Asta 的产量在培养 4 天后达到 38.4 mg/L（表 3），平均生产速率约为 9.6 mg/L/天，这超过了 H. pluvialis 的产量，其在管状或气泡柱中报告的产量范围为 0.12–4.4 mg/L/天 [48]。即使在更复杂的培养系统中，H. pluvialis 的产量可能更高 [49, 50]，增加光照强度仍能进一步提高本文所述 BC 菌株的 Asta 产量。相比之下，Syn11901 的 Asta 产量也显著高于在蓝藻 Synechocystis sp. PCC 6803（2.8 mg/L/天 [21]）和绿藻 C. reinhardtii（6.96 mg/L/天 [32]）中异源合成的 Asta。工程化酵母的 Asta 产量为 37.5 mg/L/天 [51]，但其生长为异养型，需要在培养基中添加葡萄糖。本文研究中获得的产量与其他系统的对比总结见表 5。与其他系统相比，使用 Syn11901 的优势之一在于，该物种的大部分类胡萝卜素以玉米黄质和 β-胡萝卜素的形式存在。这些分子是引入的 bKT 和 CtrZ 酶的底物，从而实现了 Asta 的高效生产（图 1）。在其他系统中，例如 Chlamydomonas reinhardtii 或 Nannochloropsis，其他类胡萝卜素大量积累，为 Asta 生物合成所需的前体提供了竞争性分流。如表 3 所示，在 80 毫升光生物反应器中低光照强度条件下或在 330 升规模下光穿透显著受限的细胞中，最低的干重 Asta 含量被观察到：因此，足够的光照暴露是提高单位干重 Asta 含量的必要条件。尽管营养胁迫未能提高 Asta 产量（表 4），我们不能排除其他胁迫条件可能在某种程度上改善类胡萝卜素的生物合成。

Table 5 Summary of the astaxanthin productivity and content per dry weight in different production systems
表 5 不同生产系统中每克干重虾青素生产率和含量的汇总

There is also scope to improve the production of Asta in the BC strain through additional metabolic engineering and improved PBR design. For instance, levels of isopentenyl pyrophosphate (IPP) and dimethylallyl pyrophosphate (DMAPP) which are the precursors of carotenoids and other terpenoids could be enhanced [21, 32, 44, 52,53,54]. Given that deletion of acsA might have a potential negative impact on the growth of the BC strain new loci for insertion of the bKT and crtZ genes could also be tested [4]. Regarding the cultivation of the engineered strain, the scaling-up costs can be reduced thanks to the recent isolation of a cobalamin-independent strain of Syn11901 that does not require the addition of vitamin B12 [3].
还可以通过进一步的代谢工程和改进光生物反应器（PBR）的设计来提高 BC 菌株中虾青素（Asta）的生产。例如，可以提高类胡萝卜素和其他萜类化合物前体异戊烯基焦磷酸（IPP）和二甲基烯丙基焦磷酸（DMAPP）的水平[21, 32, 44, 52, 53, 54]。鉴于删除 acsA 可能对 BC 菌株的生长产生潜在的负面影响，可以尝试在新的基因座插入 bKT 和 crtZ 基因[4]。关于工程菌株的培养，由于最近分离出一种不依赖维生素 B12 的 Syn11901 菌株，无需添加维生素 B12，因此可以降低规模化生产的成本[3]。

Possible cultivation in outdoor systems using natural sunlight could also be considered as a possible strategy to further reduce the production costs, even if the cultivation of GMO strains, such as Syn11901 BC is subject to strict regulation.
在户外系统中利用自然阳光进行种植也可以被视为进一步降低生产成本的潜在策略，即使转基因菌株（如 Syn11901 BC）的种植受到严格监管。

The high salinity of the MAD medium also represents a barrier for contamination by bacteria, weedy algae/cyanobacteria and other organisms, during the cultivation process. However, this barrier can be further strengthened by introducing the PtxD/phosphite-utilizing system [55] in Syn11901, which was shown to allow cyanobacterial productions in non-sterile outdoor reactors, reducing costs. In terms of PBR design, improving light penetration, for example by using a tubular PBR, will improve photosynthetic performances and, consequently, biomass and Asta yields.
MAD 培养基的高盐度在培养过程中对细菌、杂藻/蓝细菌及其他生物的污染起到了屏障作用。然而，通过在 Syn11901 中引入 PtxD/磷酸盐利用系统 [55]，可以进一步加强这一屏障，该系统已被证明能够在非无菌的室外反应器中实现蓝细菌的生产，从而降低成本。在光生物反应器（PBR）设计方面，例如通过使用管状 PBR 改善光穿透性能，将提升光合作用效率，从而提高生物质和虾青素的产量。

Conclusions  结论

The metabolic engineering approach herein reported in Syn11901 lead to Asta production at rates of ~ 10 mg/L/day under photoautotrophic growth conditions without the need for stress conditions such as nutrient starvation. Moreover, Syn11901 produces phycocyanin, another industrial-relevant product with different applications in the food and cosmetics sectors. Thus, a biorefinery process to produce ketocarotenoids and phycocyanin in the Syn11901 BC strain is a promising industrial strategy. Furthermore, considering the high photosynthetic efficiency of this fast-growing strain, its cultivation could be integrated with CO₂-emitting processes for carbon sequestration and conversion into high-value products such as Asta.
本文中报道的代谢工程方法使 Syn11901 在光自养生长条件下无需营养匮乏等应激条件即可实现约 10 mg/L/天的虾青素生产。此外，Syn11901 还能生产藻蓝蛋白，这是一种在食品和化妆品领域具有多种应用的工业相关产品。因此，在 Syn11901 BC 菌株中生产酮类类胡萝卜素和藻蓝蛋白的生物精炼工艺是一种极具潜力的工业策略。此外，考虑到该快速生长菌株的高光合作用效率，其培养过程可与二氧化碳排放工艺相结合，用于碳捕集并转化为虾青素等高价值产品。