Effects of three-dimensional culturing on osteosarcoma cells grown in a fibrous matrix: analyses of cell morphology, cell cycle, and apoptosis

Osteosarcoma cells were cultured in stirred tank bioreactors with either a fibrous matrix or nonporous microcarriers to study the environmental effects on cell growth, morphology, cell cycle, and apoptosis. Cell cycle and apoptosis were analyzed using flow cytometry and visualized using confocal laser scanning microscopy and fluorescence microscopy. The three-dimensional (3-D) fibrous culture had better cell growth and higher metabolic rates than the two-dimensional (2-D) microcarrier culture because cells in the fibrous matrix were protected from shear stress and had lower apoptosis and cell death even under suboptimal conditions (e.g., nutrient depletion). The polyester fibrous matrix used in this study also exhibited the capability of selectively retaining viable and nonapoptotic cells and disposing apoptotic and nonviable cells. Consequently, very few apoptotic cells were found in the fibrous matrix even in the long-term (1 month) T-flask culture. In the continuous culture with packed fibrous matrixes for cell support, most cells were arrested in the G1/G0 phase after 4 days. Decreasing the dissolved oxygen level from 60 to 10% air saturation did not significantly change cell cycle and apoptosis, which remained low at approximately 15%. These results could explain why the fibrous bed bioreactor had good long-term stability and was advantageous for production of non-growth-associated proteins by animal cell cultures.     

Monitoring of autophagy in Chinese hamster ovary cells using flow cytometry

Recently, autophagy, which is a degradative process, has drawn attention as an anti-cell death engineering target in addition to apoptosis in recombinant Chinese hamster ovary (rCHO) cell cultures for enhanced production of therapeutic proteins. Appropriate autophagy monitoring methods, that are suitable for long term CHO cell cultures, are necessary in order to investigate the culture conditions that affect the autophagy pathway and to select appropriate engineering targets for autophagy control. Herein, detailed protocols for autophagy monitoring methods based on flow cytometry are provided using the GFP-LC3-overexpressing CHO DG44 host cell line or MDC-like molecules in rCHO cells grown as an adherent culture with serum-containing medium or suspension culture with serum-free medium. Furthermore, combined with the apoptosis detection based on the Annexin V-PS interaction, the simultaneous detection of autophagy and apoptosis is also described. It is anticipated that the protocols described herein will assist in the fast, high throughput monitoring of autophagy that can support other existing autophagy assays.     

Automated flow cytometry for monitoring CHO cell cultures

Flow cytometry has been used to accurately monitor cell events that indicate the spatio-temporal state of a bioreactor culture. The introduction of process analytical technology (PAT) has led to process improvements using real-time or semi real-time monitoring systems. Integration of flow cytometry into an automated scheme for improved process monitoring can benefit PAT in bioreactor-based biopharmaceutical productions by establishing optimum process conditions and better quality protocols. Herein, we provide detailed protocols for establishing an automated flow cytometry system that can be used to investigate and monitor cell growth, viability, cell size, and cell cycle data. A method is described for the use of such a system primarily focused on CHO cell culture, although it is foreseen the information gathered from automated flow cytometry can be applied to a variety of cell lines to address both PAT requirements and gain further understanding of complex biological systems.     

气升式生物反应器在杂交瘤细胞培养中的应用

前述研究工作基础上,设计开发了10L规模的动物细胞培养用气升式生物反应器。应用该生物反应器悬浮培养杂交瘤细胞．通过平行试验,考察了该反应器设计的合理性和可靠性。结果显示该反应器不存在限制细胞生长、代谢和产物生成的因素,而且细胞破损技彻底消除,表明该气升式生物反应器给细胞生长、代谢和产物生成提供了理想的培养环境,其设计是成功的。     

Regulating apoptosis in mammalian cell cultures

Cell culture technology has become a widely accepted method used to derive therapeutic and diagnostic protein products. Mammalian cells adapted to grow in bioreactors now play an integral role in the development of these biologicals. A major limiting factor determining the output efficiency of mammalian cell cultures however, is apoptosis or programmed cell death. Methods to delay apoptosis and increase the longevity of cell cultures can lead to more economical processes. Researchers have shown that both genetic and chemical strategies to block apoptotic signals can increase cell culture productivity. Here, we discuss various strategies which have been implemented to improve cellular viabilities and productivities in batch cultures.     

Impact of aeration strategy on CHO cell performance during antibody production

Stirred tank bioreactors using suspension adapted mammalian cells are typically used for the production of complex therapeutic proteins. The hydrodynamic conditions experienced by cells within this environment have been shown to directly impact growth, productivity, and product quality and therefore an improved understanding of the cellular response is critical. Here we investigate the sub‐lethal effects of different aeration strategies on Chinese hamster ovary cells during monoclonal antibody production. Two gas delivery systems were employed to study the presence and absence of the air–liquid interface: bubbled direct gas sparging and a non‐bubbled diffusive silicone membrane system. Additionally, the effect of higher gas flow rate in the sparged bioreactor was examined. Both aeration systems were run using chemically defined media with and without the shear protectant Pluronic F‐68 (PF‐68). Cells were unable to grow with direct gas sparging without PF‐68; however, when a silicone membrane aeration system was implemented growth was comparable to the sparged bioreactor with PF‐68, indicating the necessity of shear protectants in the presence of bubbles. The cultures exposed to increased hydrodynamic stress were shown by flow cytometry to have decreased F‐actin intensity within the cytoskeleton and enter apoptosis earlier. This indicates that these conditions elicit a sub‐lethal physiological change in cells that would not be detected by the at‐line assays which are normally implemented during cell culture. These physiological changes only result in a difference in continuous centrifugation performance under high flow rate conditions. Product quality was more strongly affected by culture age than the hydrodynamic conditions tested. © 2012 American Institute of Chemical Engineers Biotechnol. Prog., 2013.     

Single-cell analysis demonstrates how nutrient deprivation creates apoptotic and quiescent cell populations in tumor cylindroids

Understanding how quiescent and apoptotic populations form in tumors is necessary because these cell types can considerably diminish therapeutic efficacy. Most cancer therapeutics are ineffective against quiescent cells because they target rapidly proliferating cells. Distinguishing apoptosis is important because apoptotic cells are committed to death and do not require treatment. Regrowth of quiescent cell can lead to tumor re-occurrence and metastasis, which are the leading causes of cancer mortality. We hypothesized that cylindroid cultures and acridine orange staining could be used to determine how nutrient diffusion creates apoptotic and quiescent regions in tumors. To test this hypothesis we developed a microscopy technique to measure cellular DNA and RNA content in single cells using thin cylindroids and acridine orange staining. Cell classification was compared to flow cytometry of cells grown in defined monolayer cultures. The presence of apoptosis was confirmed by morphological nuclear analysis. The effect of diffusion was determined by varying incubation time, cylindroid size, and exposing cylindroids to nutrient-deficient media. Four overlapping regions were identified as a function of cylindroid radius: an outer viable/quiescent region; a second quiescent/apoptotic region; a third late-stage apoptotic region; and an inner dead region. In monolayer cultures the absence of glutamine and growth factors induced apoptosis and hypoxia induced quiescence. Treating with nutrient-deficient media suggested that cells became quiescent near the periphery because of glucose and oxygen limitations, and became apoptotic and died further from the edge because of glutamine and growth factor limitations. These results show that cellular microenvironments can be identified in cylindroids using simple acridine orange staining and that single cell fluorescence can be measured in three-dimensional culture. The developed techniques will be useful for developing cancer therapies and determining how cell death and apoptosis are induced in three-dimensional tumor tissue.     

Cell cycle kinetics of expanding populations of neural stem and progenitor cells in vitro

Neural stem cells (NSCs) are undifferentiated, primitive cells with important potential applications including the replacement of neural tissue lost due to neurodegenerative diseases, including Parkinson's disease, as well as brain and spinal cord injuries, including stroke. We have developed methods to rapidly expand populations of mammalian stem and progenitor cells in neurosphere cultures. In the present study, flow cytometry was used in order to understand cell cycle activation and proliferation of neural stem and progenitor cells in suspension bioreactors. First, a protocol was developed to analyze the cell cycle kinetics of NSCs. As expected, neurosphere cells were found to cycle slowly, with a very small proportion of the cell population undergoing mitosis at any time. Large fractions (65-70%) of the cells were detected in G1, even in rapidly proliferating cultures, and significant fractions (20%) of the cells were in G0. Second, it was observed that different culturing methods influence both the proportion of neurosphere cells in each phase of the cell cycle and the fraction of actively proliferating cells. The results show that suspension culture does not significantly alter the cell cycle progression of neurosphere cells, while long-term culture (>60 days) results in significant changes in cell cycle kinetics. This suggests that when developing a process to produce neural stem cells for clinical applications, it is imperative to track the cell cycle kinetics, and that a short-term suspension bioreactor process can be used to successfully expand neurosphere cells.     

Cell cycle model for growth rate and death rate in continuous suspension hybridoma cultures

As a result of recent advances in flow cytometry, renewed interest is shown in modeling the kinetic behavior of cells in culture on the basis of cell cycle parameters. An important but often overlooked kinetic variable in hybridoma cultures is the cell death rate. Not only the overall cell growth but also the kinetics of nutrient metabolism and monoclonal antibody production have been shown to depend on the cell death rate in continuous suspension hybridoma cultures. The present study shows that the death rate in hybridoma cultures is proportional to the fraction of cells arrested in the G(1) phase of the cell cycle. The steady-state cell age distributions in the various phases of the division cycle have been calculated analytically. A simple mathematical model has been used to produce the profiles of the cycling and arrested cell fractions with respect to the dilution rate. The calculated steady-state growth rate, death rate, and viability profiles are shown to be in agreement with recently published experimental data from continuous suspension hybridoma cultures. (c) 1992 John Wiley & Sons, Inc.     

Poly(ADP-ribose) synthesis: a useful parameter for identifying apoptotic cells

Cell death by apoptosis was analysed in HeLa cells either treated with the antitumoral drug bleomycin or depleted of growth factors by long-term culture without medium change. The interference of apoptosis with normal cell cycle progression was followed by flow cytometry in cells stained with propidium iodide and with antibody to S-phase-related PCNA protein. Bleomycin-treated cells showed a net accumulation in G₂/M phase paralleled by the appearance of material with a subdiploid DNA content. Cells with a subdiploid DNA content were also present in growth factor-depleted cultures and were shown to derive from all the cell cycle phases. To identify apoptotic features in HeLa cell cultures, we applied a recently developed assay based on the simultaneous analysis in the single cell of three parameters, namely chromatin condensation, DNA degradation and poly(ADP-ribose) synthesis. Apoptotic cells were visualized by sequential reactions: Hoechst staining, terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick-end labelling assay and immunoreaction with anti-poly(ADP-ribose) monoclonal antibody. Positive reactions were obtained for cells at different stages of the apoptotic programme showing condensed nuclei, fragmented chromatin and apoptotic bodies This revised version was published online in November 2006 with corrections to the Cover Date.     

Mammalian cell culture scale-up and fed-batch control using automated flow cytometry

Detailed knowledge of mammalian cell culture proliferation kinetics is important to determine fed-batch strategies for industrial bioreactor operations. In particular, predicting the end of exponential proliferation in batch culture is a critical process parameter during culture scale-up. Using automated flow cytometry we show that an increase in the non-viable sub-population in CHO cell culture can predict the onset of stationary phase by approximately 40 h. This enables a completely automated culture scale-up process as well as a reliable and reproducible control of fed-batch additions during culture expansion. It is shown that the automated scale-up results in a significantly higher total cell count in the reactor than manual scale up initiated in stationary growth phase. During individual, subsequent culture expansions, a significant variation in the proliferation rate was observed despite control of bulk culture parameters. Thus, automated flow cytometry is critical to uncovering useful process parameters that enable new control strategies. Such improved process supervision derived from knowledge-based data analysis is central to the FDA's Process Analytical Technology (PAT) initiative and is expected to result in better and higher quality products.     

Cell cycle analysis of primary sponge cell cultures

Proliferation of sponge cells is generally measured via cell counts or viability assays. However, more insight into the proliferative state of a sponge cell population can be obtained from the distribution of the cells over the different phases of the cell cycle. Cell cycle distribution of sponge cells was measured via flow cytometry after staining the DNA with propidium iodide. The five sponges studied in this paper all showed a large fraction of cells in G1/G0 compared to G2/M and S, indicating that cells were not actively dividing. In addition, some sponges also showed a large apoptotic fraction, indicating cell death. Additional apoptosis measurements, based on caspase activity, showed that harvesting and dissociation of sponge tissue to initiate a primary cell culture was directly correlated with an increase in apoptotic cells. This indicates that for the development of cell cultures, more attention should be given to harvesting, dissociation, and quality of starting material. Finally, cultivation conditions used were ineffective for proliferation, since after 2 d of cultivating Haliclona oculata cells, most cells shifted towards the apoptotic fraction, indicating that cells were dying. For development of in vitro sponge cell cultures, flow cytometric cell cycle analysis is a useful method to assess the proliferative state of a sponge cell culture and can be used to validate improvements in harvesting and dissociation, to select sponges with good proliferative capacities and to study the influence of culture conditions for stimulating cell growth.     

Induction of apoptosis in normal cultured rat hepatocytes and in Hep3B, a human hepatoma cell line

The in vitro occurrence of apoptosis in hepatic cells has not been well characterized because it depends on apoptosis inducing-agents and culture conditions. Furthermore, for a given hepatic cell and the same agent, discrepant results have been reported depending on the technique used to evaluate the proportion of apoptotic cells. In this study, we compared the effects of several apoptosis-inducing agents – transforming growth factor β₁ (TGF-β₁), retinoic acid (RA), okadaic acid (OA), and cycloheximide (CY) – on two types of hepatic cells, the human hepatoma cell line Hep3B and normal rat hepatocytes, maintained either plated for 24 to 48 h or in suspension for 20 h. Chromatin condensation and/or nucleus fragmentation were investigated morphologically by DAPI staining. DNA fragmentation was investigated biochemically by agarose gel electrophoresis and poly(ADP-ribose) polymerase (PARP) cleavage was studied by western blot. Apoptotic cells were quantified either by counting cells on UV microscopy after DAPI staining or by flow cytometry. Nuclear changes, the ladder pattern on DNA electrophoresis and PARP cleavage were observed in plated cells, hepatoma cells and normal rat hepatocytes, with all inducers but especially with OA. Semiquantification confirmed that OA was a strong inducer in plated cells under the present conditions, since about 14% and 30% of Hep3B cells (with DAPI staining and flow cytometry, respectively) were apoptotic after 48 h treatment, while, with the other inducers, apoptosis was weaker and discrepancies were also observed between the two counting methods (TGF-β₁; 4% and 12%; RA, 7% and 12%; CY, 4% and 16%, with DAPI staining and flow cytometry, respectively). OA induced a moderate apoptosis in cultured hepatocytes (13% with DAPI staining), while TGF-β₁, RA and CY were found to be weakly apoptotic (respectively 4% for the first two and 6% for the last ) after 48 h. In contrast, in suspension cells, apoptosis was observed neither in Hep3B cells nor in normal hepatocytes, whatever the apoptotic inducer and whatever the techniques used to detect apoptosis. In conclusion, our results show that induction of apoptosis in hepatic cells depends not only on the apoptosis-inducing agent but also on the culture conditions. This revised version was published online in July 2006 with corrections to the Cover Date.     

Cellular characteristics of head and neck cancer stem cells in type IV collagen-coated adherent cultures

Although head and neck squamous carcinoma cancer stem cells (HNSC-CSCs) can be enriched in serum-free suspension cultures, it is difficult to stably expand HNSC-CSC lines in suspension due to spontaneous apoptosis and differentiation. Here, we investigated whether HNSC-CSCs can be expanded without loss of stem cell properties by adherent culture methods. Cell culture plates were coated with type IV collagen, laminin, or fibronectin. We examined cancer stem cell traits of adherent HNSC-CSCs grown on these plates using immunocytochemistry for stem cell marker expression and analyses of chemo-resistance and xenograft tumorigenicity. We also assessed the growth rate, apoptosis rate, and gene transduction efficiency of adherent and suspended HNSC-CSCs. HNSC-CSCs grew much faster on type IV collagen-coated plates than in suspension. Adherent HNSC-CSCs expressed putative stem cell markers (OCT4 and CD44) and were chemo-resistant to various cytotoxic drugs (cisplatin, fluorouracil, paclitaxel, and docetaxel). Adherent HNSC-CSCs at the limiting dilution (1000 cells) produced tumors in nude mice. Adherent HNSC-CSCs also showed less spontaneous apoptotic cell death and were more competent to lentiviral transduction than suspended HNSC-CSCs. In conclusion, compared to suspension cultures, adherence on type IV collagen-coated culture plates provides better experimental conditions for HNSC-CSC expansion, which should facilitate various refined cellular studies.     

Flow cytometric analysis of protein content in Taxus protoplasts and single cells as compared to aggregated suspension cultures

Plant suspension cultures are highly aggregated, preventing the direct application of flow cytometry for the study of population dynamics. The utility of single cells to accurately represent aggregated suspension cultures was tested through the analysis of total protein content. Specifically, protein content of two Taxus cuspidata suspension culture lines was studied using the Bradford assay for aggregated suspension cultures, and flow cytometry with fluorescein isothiocyanate staining for protoplasts and single cells. Taxus protein levels were measured at 75–160 mg per gram dry weight via the Bradford assay. Aggregated suspension cultures, protoplasts, and single cells predicted the same trend of protein content over the culture period (21 days). Normalized protein content of isolated single cells was statistically equivalent to aggregated suspensions for both cell lines. However, normalized protein content of isolated protoplasts showed significant differences from aggregated suspensions for one of the two cell lines. Elicitation with methyl jasmonate (MJ) is commonly utilized to increase paclitaxel accumulation in suspension cultures, and therefore the effect of MJ elicitation on protein content in aggregated suspensions, isolated single cells and protoplasts was assessed. Aggregated suspension cultures, protoplasts, and single cells did not show any change in total protein content following elicitation with MJ at 200 M on day 7. This study illustrates the usefulness of flow cytometry for obtaining culture population information and the value of using intact single cells for the study of plant metabolism.     

Apoptosis: A mammalian cell bioprocessing perspective

Apoptosis is a form of programmed and controlled cell death that accounts for the majority of cellular death in bioprocesses. Cell death affects culture longevity and product quality; it is instigated by several stresses experienced by the cells within a bioreactor. Understanding the factors that cause apoptosis as well as developing strategies that can protect cells is crucial for robust bioprocess development. This review aims to a) address apoptosis from a bioprocess perspective; b) describe the significant apoptotic mechanisms linking them to the most relevant stresses encountered in bioreactors; c) discuss the design of operating conditions in order to avoid cell death; d) focus on industrially relevant cell lines; and e) present anti-apoptosis strategies including cell engineering and model-based optimization of bioprocesses. In addition, the importance of apoptosis in quality-by-design bioprocess development from clone screening to production scale are highlighted.     

A plant cell bioreactor with medium-perfusion for control of somatic embryogenesis in liquid cell suspensions

The Braun Biostat BF2 bioreactor system employs a novel aeration and agitation system, designed to enhance gaseous exchange and reduce shear stresses on submerged cell suspension cultures. The Biostat BF2 bioreactor employs a central pivoting spindle, around which the aeration tubing is wound forming a large paddle-type structure suspended from the top-plate and swung in a circle by a solid-state magnetic stirrer.The aeration tubing is a polypropylene capillary membrane, which has a unique microporous structure and is ideal for aeration, permitting two-way, bubble-free, gaseous exchange of the medium. This tubing can be rendered porous and can be used in the perfusion of aqueous solutions, enabling cell-free media exchange to be conducted. Thin-walled silicone rubber tubing, although gas permeable to a degree, cannot be made porous to aqueous solutions.The bioreactor was inoculated with a suspension culture of Sitka spruce (Picea sitchensis [Bong.] Carr.) known to be embryogenic and capable of maturing to plantlets on solidified medium. The perfusion capability of the bioreactor was employed to replace the inital proliferation medium with maturation medium in order to induce the development of the somatic embryos in submerged cell culture. The size ratio of the somatic embryo heads was monitored over 7 weeks. This cell line was found to mirror just the initial elongation, previously observed in shake-flask culture.Abbreviations BAP 6-benzylaminopurine - 2,4-D 2,4-dichlorophenoxyacetic acid - SSPM Selby Sitka proliferation medium - SSMM Selby Sitka maturation medium The following was presented at the NERC TBLG '95 Meeting as the Bioreactor Workshop     

Plasmodium falciparum: erythrocytic stages die by autophagic-like cell death under drug pressure

It has been reported that an apoptotic cell death process can occur with protozoans, but no consensus on Plasmodium susceptibility to apoptosis was reached till now. Thus, we evaluated if Plasmodium falciparum blood forms undergo apoptosis after in vitro pressure with chloroquine, S-nitroso-N-acetyl-penicillamine (SNAP) or staurosporine. Inhibition of parasite growth and loss of viability were observed in treated cultures by both light microscopy and flow cytometry. When DNA fragmentation was verified, only a small number of TUNEL-positive parasites was detected in treated cultures and pretreatment of parasite with a general caspase inhibitor was not able to prevent parasite death. Considering the lack of apoptotic characteristics and the observation of parasites with cytoplasmatic vacuolization by electron microscopy, we conclude that P. falciparum parasites under chloroquine, SNAP or staurosporine pressures do not die by apoptosis but by a process similar to autophagy. The autophagic pathway could be explored as an alternative target for the development of new antimalarial drugs.     

Comparative characterization of cell death between Sf9 insect cells and hybridoma cultures

Physiological cell death (PCD) in Sf9 insect cell batch cultures was comprehensively characterized using simultaneous determinations of qualitative and quantitative assays, including agarose gel electrophoresis, confocal, epifluorescence, and transmission electron microscopy, and DNA content by flow cytometry. Results were compared to hybridoma cultures where abundant information of apoptosis exists. Both cultures shared some typical apoptosis features, including cell shrinkage, loss of sphericity, swollen endoplasmic reticulum and Golgi apparatus, chromatin condensation, and specific DNA degradation. However, distinctive morphological and kinetic differences between both cultures revealed that Sf9 cells died by an atypical PCD process characterized by absence of nuclear fragmentation, scarce association of condensed chromatin to the nuclear envelope, swollen mitochondria, and high nonspecific DNA degradation. These features, distinctive of necrosis, were not observed in the normal apoptotic process of hybridomas. Glucose depletion marked the appearance of apoptotic Sf9 cells, which there up on increased gradually, whereas apoptotic hybridomas rapidly increased upon glutamine depletion. Furthermore, active phagocytosis was found in Sf9 viable cells, a characteristic phenomenon during in vivo apoptosis but uncommon for in vitro cultures. Sf9 cells contained unusually high numbers of phagosomes, particularly after glucose depletion. Additionally, few apoptotic bodies accumulated in culture, suggesting their elimination by phagocytosis. Other distinctive characteristics of Sf9 cells were the presence of a polynucleated hypertrophic population fraction, polyploidy, cell cycle arrest in G2/M phase, and more necrosis compared to hybridomas. Such phenomena prevented a reliable quantification of apoptosis from determination of the sub-G1 peak. Nonetheless, emergence of a bimodal Sf9 cell size distribution coincided with the increase in the sub-G1 population and onset of death. The fraction of particles in the smaller peak (6-11 microm diameter) closely correlated with the fractions of apoptotic bodies, late apoptotic, and secondary necrotic cells. Accordingly, Sf9 cell size was shown to be an effective, rapid, and simple parameter for quantifying death. Altogether, the results of this study provide new insights into PCD and other phenomena in insect cell culture important for biotechnological applications of Sf9 cells.     

Unstructured models for suspension cultures of Taxus media cells in a bioreactor under substrate-sufficient conditions

For a better understanding of the simulation, optimization, and process control in cell cultures, good kinetic models are necessary for large scale plant cell culture. In this paper, the systematic kinetics of taxol production by Taxus media cell suspension cultures in a stirred 15-L bioreactor under substrate-sufficient conditions and the absence of inducer intervention were studied. A kinetic model of cell growth was established by logistic equation, and kinetic unstructured models of substrate consumption, product synthesis and rheological behavior were constituted, which incorporated energy spilling. These models were verified by comparing the simulation results with those obtained experimentally. These results showed that energy spilling was a key factor that must be considered in constructing unstructured kinetic models of Taxus media cell suspension cultures in a stirred bioreactor under substrate-sufficient conditions. Besides, an optimized operation measure of decreasing energy spilling was proposed. An increase of 17.64% in cell biomass and 14.88% in taxol concentration were obtained when the strategy of limiting added carbon several times was experimentally implemented in a 15-L bioreactor. Results demonstrated that these established models should be helpful in the process prediction and operation optimization to guide the production and amplification of Taxus media cell suspension cultures in a bioreactor.