Technical developments for online bioprocess monitoring in continuously orbitally shaken microtiter plates

Shaken bioreactors are widespread in the field of biotechnology. Especially, microtiter plates are increasingly used as bioreactors in the early stages of process development. Microtiter plates allow for the investigation of numerous parallel experiments within a short period of time. Furthermore, microtiter plates offer economic advantages compared to larger systems due to reduced material consumption. When used as a bioreactor, microtiter plates are mostly used as bioreactor in combination with shakers with an orbital shaking movement. Due to the orbital shaking movement, the centrifugal force acts on the culture broth and leads to the formation of a rotating liquid sickle. In addition to mixing of the culture broth, the motion of the liquid also increases the gas-liquid exchange surface, which increases the supply of oxygen to microorganisms.However, the full potential of small-scale systems is only guaranteed if (online) monitoring for each bioreactor is available. BioLector technology, an optical monitoring system based on fluorescence and scattered light measurements, allows for the online monitoring of various process parameters such as biomass, NADH and flavin levels during cultivation in orbitally shaken microtiter plates. pH value and dissolved oxygen tension (DOT) are also important biotechnological process parameters. Unfortunately, these parameters cannot be determined directly by fluorescence measurements. Sensitive fluorescent dyes are available for both measurements; these dyes change their fluorescence behavior depending on the respective concentrations. In microtiter plates, these sensitive fluorescent dyes are commonly immobilized at the well bottom. These optical sensor spots are also called optodes. However, previous studies have found interferences between biogenic fluorescence and the fluorescent dyes, leading to erroneous results.In the present work, an alternative measurement system for DOT monitoring in orbitally shaken microtiter plates based on a fluorescent dye was established. The fluorescent dye chosen for DOT determination is excited with red light and emits light in the near-infrared region. With this technique, fluorescence lift time is used as a raw signal. The measurement is conducted outside of the biogenic fluorescence region to prevent interference. Another advantage is that the fluorescent dye is immobilized on nanoparticles, which can easily be added to the culture broth. Accordingly, the system can be used with any type of microtiter plate, as there is no dependence on microtiter plates equipped with optodes. Furthermore, it was shown that dispersed oxygen-sensitive nanoparticles do not influence the growth behavior of Gluconobacter oxydans, Hansenula polymorpha or Escherichia coli.In the present thesis, a BioLector device equipped with the above-mentioned alternative DOT measurement system was combined with the so-called µRAMOS technique. µRAMOS allows for the online determination of the oxygen transfer rate (OTR) in each well of a 48-well-microtiter plate. Because both techniques are based on optical measurement signals, a measuring cycle was developed to obtain interference-free measurements. For the first time, with this new device, multiple data can be collected simultaneously by means of a single experiment. In the past, the combination of elaborated parallel shake flask and microtiter plate experiments were required. By combining both techniques, it was also possible to calculate the volumetric oxygen transfer coefficient (kLa) during E. coli culture. The estimated kLa values are in excellent agreement with an empirical correlation found in the literature.Similar to the DOT, other important process parameters cannot be determined directly via fluorescence. For example, glucose and glycerol are often used as carbon sources in culture media. Unfortunately, these compounds are not fluorescent. No sensitive fluorescent dyes for these substances are commercially available so far. However, multi-wavelength (2D) fluorescence spectroscopy can be applied to determine such process parameters non-invasively. In multi-wavelength (2D) fluorescence spectroscopy, multiple excitation and emission spectra are acquired and combined to an excitation vs. emission matrix (multi-wavelength (2D) fluorescence spectrum). To evaluate the acquired multi-wavelength (2D) fluorescence spectra and great amount of associated data, chemometric methods such as the “Partial Least Squares” (PLS) regression must be applied. So far, this approach has been applied to stirred-tank reactors (with a working volume > 1 L) because the acquisition rates of common fluorescence spectrometers are too slow to monitor each well in microtiter plates. This was overcome by the integration of a “Charge Coupled Device” (CCD) detector-based spectrometer with a BioLector device, which allowed for simultaneous acquisition of the entire emission spectrum. To ensure reliable measurements, a resolution-resolved measurement was established. The usefulness of the new device for online monitoring of glucose, glycerol and acetate concentrations as well as the pH value over time was demonstrated in cultures of H. polymorpha and E. coli.In the present work, different monitoring techniques for small-scale culture systems are described. These techniques provide valuable process parameters that were not easily and reliably accessible beforehand. By means of these novel developed techniques, the user gains significant insight into the investigated cultures. Thus, the clear distinction between secondary screening programs and process development has become noticeably blurred