A bioreactor is a controlled environment designed to support the growth, cultivation, and manipulation of living organisms, such as cells, or microorganisms, under specific and optimal conditions. It is commonly used in various fields including biotechnology, pharmaceuticals, and research to produce biological products, conduct experiments, or study biological processes. Bioreactors provide controlled parameters such as temperature, pH, nutrient supply, and oxygen levels to optimize the growth and production of desired biological substances.

Recent advances in plant cell cultures in bioreactors encompass genetic engineering for metabolite production, microcarrier optimization for enhanced growth, or 3D culture systems mimicking natural environments (Hu, 2020; Hesami et al., 2021). Real-time monitoring and automated control of culture conditions is performed using different sensors such as pH or conductivity (Ruffoni et al., 2010; Wang et al., 2016). Precise control over environmental factors such as light intensity in the case of photobioreactors, or temperature, and dissolved oxygen has been implemented for a long time (Ruffoni et al., 2010; Zhao et al., 2015). Furthermore, synthetic biology is used for novel pathways, nutrient formulation enhancement, and integration of omics technologies with bioreactor data for a holistic understanding of cellular processes. (Ochoa-Villarreal et al., 2016; Reski et al., 2018). These technologies are collectively driving more efficient and productive plant cell culture processes with applications in pharmaceuticals, nutraceuticals, and other valuable compounds.

Scaling up plant cell cultures from laboratory volumes to industrial bioreactors has been one of the major focuses. New strategies for maintaining uniform growth include nutrient distribution, and dissolved oxygen. Plant cells are especially sensitive to mechanical shear stress due to agitation. Thus improvements are being developed, allowing for more efficient and cost-effective production of plant-derived compounds (Georgiev et al., 2013; Motolinía-Alcántara et al., 2021). One drawback associated with plant cell cultures is their extended cultivation process duration when compared to microbial cultures, primarily owing to the comparatively sluggish growth rate of plant cells. To address this challenge, the development of process strategies, such as semi-continuous and/or continuous perfusion processes, has been undertaken to bolster productivity (Zhong, 2002; Corbin et al., 2020). Unless the cell itself is the final product, cell growth is not the main outcome of the process, but a minimum growth is required to obtain any meaningful production (see below).

Researchers have been investigating strategies to induce stress responses in plant cells, which can trigger the accumulation of secondary metabolites with potential therapeutic benefits. Bioreactor conditions can be manipulated to simulate various stress factors and enhance metabolite production (Zhang et al., 2012; Narayani and Srivastava, 2017; Ashraf et al., 2018; Ho et al., 2020).

Plant cell suspension cultures, derived from undifferentiated callus cells, are established through the disaggregation of cells originating from delicate calli in liquid media. To efficiently produce recombinant proteins, an ideal plant cell suspension culture system should display rapid growth, easy genetic transformation, high protein expression capacity, minimal inherent proteolytic activity, and low levels of compounds such as phenolics or interfering phytochemicals such as oxalic acid. Additionally, the system should thrive under optimal conditions for adaptation and scalable growth within a bioreactor. While tobacco (Nicotiana tabacum) and rice (Oryza sativa) are extensively studied for generating recombinant protein through suspension cultures, other species like tomato and ginseng have also been employed in this context (Huang and McDonald, 2009).

One typical procedure of steps to produce molecules of interest in a bioreactor using plant cells is shown in Figure 1 . Thus, transgenic cells are produced to obtain active molecules for vaccine formulations, antibodies, or other proteins. Transgenic calli are analyzed for the presence of the transgene and expression of the protein or product if the system used is a classic overexpression construct. Once the best performers are screened, a second step begins whereby friable calli are converted into cell cultures. Once the small-scale flask-type screening of the genetic engineering steps has been developed, the bioreactor scale-up would be carried out to define the optimal requirements for process parameters, automation and monitoring needs, bioreactor design, and parameters for scaling up the process.

The specifications of the bioreactor used to obtain the final product may differ but the previous operations of the biological material preparation before the bioreactor step will be similar in most cases.

This section provides an in-depth exploration of various bioreactor types employed for plant cell cultures, showcasing their characteristics, applications, and contributions to advancing plant cell biotechnology. Bioreactors are used across a range of scales, including a laboratory for research and optimization ( Figure 2 ), the pilot for process validation and scaling up, and an industrial for large-scale production of plant cells and their products ( Figure 3 ).

Bioreactors can be classified based on preparation and sterilization methods. We identify three categories:

Various types of bioreactors, including Continuous Stirred-Tank Reactors (CSTR), single-use bioreactors (SUB), airlift bioreactors (ALB), and photobioreactors (PBR), offer distinct advantages and disadvantages in the production of plant cells. Table 1 illustrates the unique characteristics of each bioreactor type. The selection of the most suitable bioreactor depends on several factors, such as the specific requirements of the plant cell culture, the nature of the product, scale considerations, and available resources (Merchuk, 1990; Shukla and Gottschalk, 2013).

Ideally, a particular culture should undergo testing in multiple independent systems to determine optimal conditions and establish the production process. Economic factors and downstream processing requirements, as well as final attributes like Good Manufacturing Practices (GMP) compliance and cost per milligram, among others, significantly influence the choice of bioreactor. It is important to note that the investment in hardware necessary for setting up Sterile In-Place (SIP) bioreactors can be substantially higher, often by several orders of magnitude, compared to SUB or benchtop bioreactors. Therefore, the decision regarding the technology to use is usually driven by a combination of overall product needs and market pricing.

At present, the primary limiting factor in plant cell cultures is productivity. Consequently, larger volumes are typically favored, unless the produced molecule is required in very small quantities.

While tobacco cells have a proven track record in academic research, they show important liabilities when it comes to producing high quality pharmaceutical biomolecules. This has led to a fruitful research and development to improve it as a protein production platform. Modified glycosylation has a positive impact on the cytotoxic activity of recombinant antibodies. This has led to the engineering of the glycosylation pathway in Chinese Hamster ovary cells (Umaña et al., 1999). A similar strategy has been developed in plant cells. Efforts have included the inactivation of a fucosyl transferase and a glycosil transferase to obtain human-like glycosylation (Mercx et al., 2017; Smargiasso et al., 2019; Herman et al., 2021). The presence of endogenous proteases also decreases both quality and quantity of protein production and RNAi strategies have been used to downregulate endogenous proteases with some success (Mandal et al., 2014). These proteases as expected are found in other plant genomes such as N.benthamiana or Arabidopsis thaliana (Magy et al., 2014). Thus, knowledge obtained in N. tabacum may be useful for other plant cell platforms. Whilst the reduction of endogenous proteases by genetic means is a formal possibility, a second strategy is the use of modified culture conditions. Indeed, increased pH and the generic serine and cysteine protease inhibitor AESBF showed significant improvements on the stability of alpha antitrypsin (rATT) (Huang et al., 2009). The addition of exogenous generic protease inhibitors such as BSA, gelatin, PEG and others have been studied with mixed results as they have short life times, and they may interfere with downstream processing and have a negative effect on cell growth (Benchabane et al., 2008; Huang and McDonald, 2012). A strategy with a built in protease inhibitor has been reported with improved protein production in potato leaves (Rivard et al., 2006).

Despite being a less than perfect platform, tobacco cells have an impressive record of production of different proteins. Early proof of concepts demonstrated the ability to produce antibodies, Interleukin 2 and 4 in tobacco cells (N. tabacum) (Magnuson et al., 1996l Magnuson et al., 1998). Importantly, both IL2 and IL4 are biologically active. Another example of antibody production in tobacco cells (BY-2 cells) is the antibody M12. It binds to vitronectin, a glycoprotein found in the serum and extracellular matrix (Raven et al., 2015). It is used to fight against cancer. The bioreactor used was a 200L disposable device. The human granulocyte-macrophage colony-stimulating factor (GM-CSF) was also produced using the BY-2 cells (James et al., 2000). Next to these examples, the number of proteins produced in tobacco cell cultures goes up to several hundred. And current efforts include a holistic approach to improve growth media, bioreactors, cell properties and promoters.

The wild tobacco plant Nicotiana benthamiana has become an important model system in plant biology. The genome of N. benthamiana shows an allotetraploid structure. Several genome versions have been published (Bombarely et al., 2012; Schiavinato et al., 2019; Kurotani et al., 2023). Work performed in N.benthamiana started as a model system for plant pathogen interactions (Goodin et al., 2008). However, early work showed its potential to study gene function by Virus Induced Gene Silencing (Liu et al., 2004). In fact N.benthamiana has a defective gene silencing pathway, allowing high expression of viral vectors as it is highly susceptible to viral infection (Bally et al., 2018). This is an advantage to produce very high quantities of exogenous proteins using viral vectors.

The complete plant of N. benthamiana has been extensively used to produce many proteins and metabolites. The process has been mostly based on agroinfiltration using T-DNA constructs or via viral vector infiltration of leaf tissues. A variety of proteins have been produced in N.benthamiana leaves (Alkanaimsh et al., 2016; Jiang et al., 2020; Mateos-Fernández et al., 2021). However, like in the case of N.tabacum, the endogenous proteases are a big hurdle to obtain high productions (Grosse-Holz et al., 2018; Jutras et al., 2020).

In contrast to BY2 cells derived from N.tabacum, use of N.benthamiana in liquid cultures is less extended. By using a knockdown strategy of β-1,2-xylosyltransferase and α-1,3-fucosyltransferase, in combination with the expression of an anthrax decoy fusion protein, this product was successfully generated in 1L flasks (Sukenik et al., 2018). By applying the mannosidase I inhibitor kifunensine, the anti-CD20 rituximab has been obtained in transient expression with no fucosylation, displaying a 14 fold increase of antibody-dependent cell-mediated cytotoxicity (Kommineni et al., 2019). This indicates that N. benthamiana is a good platform that may take over N. tabacum as choice for protein production.

Rice is probably the most important crop in the world. Working with rice has several advantages including a very well defined genome (Goff et al., 2002; Huang et al., 2012) with an excellent annotation (Ouyang et al., 2007; Sakai et al., 2013; Ren et al., 2019). Many rice varieties have been sequenced allowing the identification of structural variants (Fuentes et al., 2019).

Rice cell culture protocols were established more than 30 years ago (Mikami and Kinoshita, 1988; Torres et al., 1999), and ever since, they have been used for research and biotechnological purposes. Due to the importance of rice as a crop, comprehensive studies have been performed to identify genes involved in callus formation, a prerequisite for robust transformation protocols (Zhang et al., 2019). These include traits such as induction rate, induced callus weight, induced callus color, induced callus size, friability, callus proliferation ability, induction speed, and callus first appearance. Whilst plant transformation is generally genotype dependent, rice cultivars of O.sativa japonica and indica appear to have decent callus formation and regeneration capacity (Abe and Futsuhara, 1986; Binte Mostafiz and Wagiran, 2018).

Rice cell cultures have been used to produce many proteins. These include human serum albumin (Pang et al., 2020), lysozyme (Huang et al., 2002) butyrylcholinesterase (Corbin et al., 2016; Macharoen et al., 2020) or human α-1-Antitrypsin (McDonald et al., 2005).

Carrot cells have a long tradition in in vitro culture. They have pioneered studies to express foreign proteins such as chitinase, or the GLUTAMIC ACID DECARBOXYLASE autoantigen 65 (Gilbert et al., 1996; Porceddu et al., 1999). However, these studies used whole plants. Carrot cells have also played a role in uncovering the importance of cell density in cultures as a major factor impacting survival (McCabe et al., 1997). The successful production of the taligluciferase alfa with comparable activity to Cerezyme® produced in CHO cells paved the way for industrial production (Shaaltiel et al., 2007). The recombinant plant-derived GCD (prGCD) is targeted to the storage vacuoles, using a plant-specific C-terminal sorting signal from Arabidopsis endochitinase and a storage signal from tobacco endochitinase. Further development by Protalix (Tekoah et al., 2015a) allowed it to become the first plant produced biotherapeutic approved by regulatory authorities.

The moss Physcomitrium patens, previously known as Physcomitrella (Medina et al., 2019), has emerged as a valuable model system in plant science, primarily due to its significant advantages for genome manipulation. The complete genome of Physcomitrium, encompassing nuclear, plastid, and mitochondrial genomes, has already been published (Lang et al., 2018; Perroud et al., 2018).

Utilizing this moss species as a model system offers several notable benefits for plant functional genomic studies, particularly its ability to incorporate exogenous DNA into the genome through homologous recombination (Schaefer, 2002). Physcomitrium exhibits remarkable gene targeting efficiency and shares structural and functional similarities with higher plants, making it an excellent choice for studying plant biology. Being haploid, means that identifying phenotypes is straightforward (Egener et al., 2002).

Moreover, the cultivation of Physcomitrium in bioreactors is highly advantageous due to its inherent genetic stability and remarkable tolerance to fermentation parameters such as shear stress. To maximize protein production yields, it is recommended to culture the moss using filamentous protonema tissue. Maintaining the culture at pH levels within the range of 4.5-6 is crucial during this specific stage of the life cycle (Decker and Reski, 2012).

However, not all that glitters is gold for Physcomitrium cultures, as they face certain hurdles. Being autotrophic, they do not require a medium with carbon sources, but they are dependent on light. This can pose challenges in terms of bioreactor configuration and the space required to accommodate the light source. Additionally, these cultures demand extended cultivation periods of at least 20-30 days to achieve high yields, especially when operating in continuous or semi-continuous mode. Maintaining the filamentous protonema tissue in long-term cultures is crucial, but this can be effectively achieved through pH control during bioreactor cultivation (Hohe et al., 2002).

One company that specializes in producing pharmaceuticals and therapies based on Physcomitrium cultures is Eleva GmbH, formerly known as Greenovation Biotech GmbH. They have successfully developed treatments for various diseases, such as C3 Glomerulopathy, Fabry disease, and Pompe disease (Häffner et al., 2017; Hennermann et al., 2019; Hintze et al., 2020; Hintze et al., 2021). Their production process is carried out in single-use wave bioreactors of up to 200L capacity (Niederkrüger et al., 2019).

The plant chemodiversity is being exploited to produce many chemicals with a wide range of industrial purposes. While traditional industrial processes relayed on extraction from cultivated or wild species, this practice is not sustainable (Payne and Shuler, 1988). Especially when endangered species with high value secondary metabolites are not well-established culture plants.

The use of elicitors as molecules that activate secondary metabolite pathways, has been a long-standing practice in plant biotechnology since 1992 (Gundlach et al., 1992). Subsequent research has demonstrated their efficacy in various plant types and tissues, leading to enhanced secondary metabolite production in cell cultures for biotechnological applications (Sohn et al., 2022). Particularly, the combined application of jasmonates and cyclodextrins has shown synergistic effects in activating these pathways (Lijavetzky et al., 2008), resulting in elevated levels of anthocyanins and stilbenes (Lijavetzky et al., 2008; Belchí-Navarro et al., 2012; Almagro et al., 2015a).

Amongst the secondary metabolites of interest, anthocyanins and resveratrol/stilbenes have been studied in many systems, especially cells from grape berries. The use of grape cell cultures has made it a model system to obtain and improve resveratrol and anthocyanin biosynthesis for biotechnological production (Hirasuna et al., 1991; Belhadj et al., 2008; D’Onofrio et al., 2009; Martínez-Márquez et al., 2016; Jeong et al., 2020). Moreover, other elicitors, such as coronatine, a fungal elicitor that binds to the JA receptor COI-1, have also been employed to improve secondary metabolite production (Almagro et al., 2015a; Lee et al., 2018).

However other biotic elicitors include acetyl salicylic acid, salicylic acid, yeast extract, chitosan or pectin see (Alcalde et al., 2022) for a recent review. Amongst the abiotic elicitors UV radiation has been extensively used to increase anthocyanin production in planta and catharantin in Catharantus roseus cell cultures (Ramani and Chelliah, 2007; Liu et al., 2017; Castillejo et al., 2021; Tripathi et al., 2021).

Methyl Jasmonate (MeJa), initially identified in jasmine flowers by Demole et al., 1962, serves as a pivotal phytohormone alongside its free-acid counterpart, jasmonic acid, participating in various plant processes. The primary function of MeJa and jasmonic acid lies in activating plant defense mechanisms against both biotic and abiotic stresses (Wasternack and Parthier, 1997). Notably, the jasmonates receptor is shared with the phytotoxin coronatine, the COI1 (Feys et al., 1994; Xie et al., 1998). Both coronatine and jasmonates initiate similar molecular pathways, such as anthocyanin accumulation, illustrating the versatility of these compounds beyond biotic stress responses. Abiotic stresses, also trigger secondary metabolites production pathways like anthocyanin biosynthesis (Christie et al., 1994; Rudolf and Resurreccion, 2005; D’Alessandro et al., 2022).

Salicylic acid, identified as another elicitor by Angelova et al., 2006, plays a significant role in the plant defense pathway (Chen et al., 1995; Ding and Ding, 2020). No other phytohormones are known to act as elicitors, since they do not participate in the same way in plant defense as they do.

Expanding on this notion, molecules beyond traditional phytohormones can act as elicitors to trigger plant defense pathways. In recent years, there has been a notable shift towards using nanoparticles (NPs) instead of biotic elicitors, owing to the customizable physicochemical properties of NPs. Metallic, metal oxide, combinations of metallic NPs, magnetic, silica, and biopolymer NPs represent diverse categories that have demonstrated promising elicitation results and increased production yields in in vitro cultures (Fazal et al., 2016; Jamshidi and Ghanati, 2017; Javed et al., 2018; Nourozi et al., 2019; Khan et al., 2021; Arya et al., 2022).

Although using JA as a method is a simple and effective way to increase secondary metabolite production, it has some important caveats. First, JA is a stress signal and as such, it is not a good procedure for fed-batch cultures as stressing the cells causes a decrease in overall growth capacity. This can be overcome in some cases by increasing sugar input (Kümmritz et al., 2016). In general, elicitors activate a stress response that causes an inhibition of cell division (Logemann et al., 1995). Thus, when elicitors are used, we create a two-step production protocol of growth followed by activation of secondary metabolite production at a cost of increased biomass.