General: • Due to their fast growth and high specific oxygen consumption rates, for bacterial cultures, the desired oxygen-transfer rate (OTR) is a major criterium when choosing the right combination of microplate, culture volume, and shaking conditions (see our list of oxygen transfer rates). Generally, for screenings for improved mutants/constructs or growth media, researchers aim to have OTRs in the wells of the microplate similar to the OTR in their large-scale systems. Typical OTRs aimed at are in the range of 30-50 O 2 l -1 h -1 (corresponding to k L a values of 150-250 h -1 ). • A second criterium is the acceptable statistical variation (standard deviation) between the analytical results (e.g. an enzyme level, or a metabolite concentration) of independent duplicates. In general, the statistical variation gets smaller with larger culture volumes. With 2.5 ml cultures in square 24 deepwell plates, statistical variations of 3% or less are achievable (after optimization of all aspects), while with 96-well plates, it is often difficult (though very much project-dependent) to achieve statistical variations of less than 5%. Further below on this page, you will find a general protocol for the optimization of reproducibility. Below this protocol, you will find a general protocol for the performance of a screening with E. coli or Bacillus. • A third criterium for the choice of the appropriate microplate and culture volume is the amount of cells or supernatant needed for analysis. However, for most projects, this criterium is easily met, and the second criterium (the standard deviation required) is much more important. • Cultivation times for bacterial cultures are generally 3 days or less, so our standard sandwich covers (with relatively large aeration holes above the center of each well) can be generally applied, while the evaporation losses still being acceptable. These standard covers assure a sufficient degree of headspace refreshment rates to keep the O 2 concetrations in the headspace above 18%, also during the growth phase when the specific oxygen consumption rate by the bacterial cells are maximal.

Screening of bacterial collections (mutant libraries or wild-type strains) frozen in 96-well MTPs The origin of our cultivation system lays in precisely this application; in the second half of the 1990s at the ETH Zurich a collection of 1850 wild type bacterial strains was screened routinely for the presence of certain desired enzyme activities, according to the procedure depicted at the right side of this page Suspensions of the wild type strains (containing 15 % v/v glycerol) were frozen in the individual wells of 96-well plates. These so called “master-plates” were frozen at -80 oC, and - for each new screening project - were sampled using our cryo-replicator. The sampled cells were revived on agar medium dispensed in a microtiter plate. After a few days, the cell mass formed was used to inoculate a liquid culture, that was subsequently (after one day of growth during shaking, with a sandwich cover on op) incubated with the educt for the desired bioconversion. After centrifugation, the supernatant was analyzed by LC-MS. For mutant or construct libraries, the above procedure can also be used. To reduce the amount of biological variation often a second - suspended - preculture is used, in order to better “synchronize” the cultures, and so reduce the biological variation in the results (see second protocol below, and also the section ”Procedures for inoculation and synchronization of cultures” at the end of this page).

Protocol for the preparation of a screening: optimization of the reproducibility: 1. Decide what standard deviation between independent duplicates is acceptable. This will a.o depend on the expected percentage-wise improvement of the best mutants or culture condition. 2. Start with optimizing the reproducibility of the analytical procedure. If possible, apply HPLC-UV (in combination with relatively large injection volumes), infrared- based methods or colorimetric/spectroscopic methods, since they generally give rise to the lowest standard deviations (1-3% once optimized). Methods dependent on internal standards (e.g. GC and LC-MS) generally result in relatively poor reproducibilities. 3. Test the standard deviations between duplicate cultures inoculated with the same (overnight) suspended culture (e.g. a shake flask culture) using a range of different MTPs and culture volumes, each in 3-8 fold. If applicable, also vary the cultivation/incubation times. On the basis of the resulting standard deviations, choose a MTP-culture volume combination that will allow you to detect the sort of mutants you are looking for (as defined in step 1). 4. Test the standard deviation between independent duplicates (inoculated with colonies from the same agar plate) using the MTP-culture volume combination selected in step 3. In order to achieve a standard deviation close to the standard deviation achieved in step 3, it is often necessary to synchronise the cultures as described in protocol 2. It may be important to optimize the culture lenghts of the primary and secondary cultures as well as the amounts of inoculation. 5. Now that the general framework of a suitable screening protocol has been established, it may take another few manmonths to further reduce the individual sources of error that contribute to the overall standard deviation. Focus in this stage generally lies on pipetting methods (pipetting/robotic stations versus manual pipetting; manual pipets versus electronic pipets) and the minimization of biological variation by optimizing the inoculation procedures.

Procedures for inoculation and "synchronization" of cultures: When screening libraries of mutants, the inoculation procedure is crucial for reproducible results. In many cases, the mutant library will be initially available as single colonies on an agar plate. A robotic colony picker may be used to inoculate liquid medium dispensed in the wells of a 96-well MTP. Alternatively for small libraries, one may use toothpicks for this initial inoculation. After an sandwich cover has been put on, this primary MTP may be incubated on an orbital shaker for 1-2 days. In the case of a qualitative screening, e.g. to screen for the presence or absence of a certain gene, product or enzyme, the resulting cell suspension can be used immediately for analysis or bioassay. For quantitative screenings, e.g. when searching for high-activity mutants, it is often advisable to use this primary MTP to inoculate a second MTP, by transferring 1-5 µl, with a multichannel pipette. The rest of the primary MTP can be stored at minus 80 o C, after addition of glycerol, for later use, if desired. The cultures in this secondary MTP will be more or less "synchronized", i.e. the cells grow in parallel, and reach the stationary phase at the same time. The latter is especially relevant if the product is unstable or is prominently formed in the stationary phase. Non-synchronized cultures (such as from direct inoculation starting from colonies) often give rise to large numbers of false positives, and possibly, false negatives. In practice, synchronizing cultures becomes increasingly difficult at smaller culture volumes, most notably as a result of a larger variation in the size of the inoculum. When starting with a library in 96-MTP format (cells frozen in the presence of 15% glycerol, v/v), one can use a 96-pin replicator with fixed pins to sample the library after melting the master-plate. Alternatively, one can press a sterile spring-loaded replicator onto the frozen cultures (no melting required, so there is no viability loss of the remaining frozen cultures) to get a small film of cell suspension on the tip of the pins. In either case, one can subsequently transfer the sampled cells either directly into a liquid culture, or onto an agar surface in a rectangular Petri dish. The choice depends on the viability of the host strain in liquid medium after the cells have been frozen at -80 o C. Many strains grow much better in liquid medium after having been pre-cultured on an agar-based medium.

Eaxmple protocol for the screening of a mutant/construct library in E. coli or Bacillus: 1. Fill all 96 wells of a sterile half deepwell plate CR1496c square wells with a total volume of approx. 1.1 ml) with 0.25 ml of an appropriate sterile rich medium (e.g. LB medium) 2. Inoculate each well with a single colony from an agar plate. Use either sterile toothpicks, or - alternatively - a colony picking robotic system. 3. Cover the inoculated MTP with a sandwich cover. 4. Incubate the MTP+cover in a clamp system mounted on an orbital shaker, at 30-37 o C, for 16-24 hours. Shaker conditions: 250 rpm, 50 mm shaking amplitude. 5. Prepare a second MTP (as in step 1) in case synchronised cultures are desired (see also text below). 6. Inoculate this second MTP by transfer of 5 ul from each well of the first MTP. For this purpose, use either a 12 channel multipipette, a 96 channel pipetting machine, or a pipetting robot. 7. Repeat step 3 and 4. 8. Optional: use the first MTP to prepare a frozen master-plate for possible re-screening at a later time: add 250 µl of a 30% (v/v) glycerol in water solution to teach well. Use a 12-channel multipipette (move up and down) for thorough mixing. Use wide-orifice tips in case cultures are viscous, or are not fully homogeneous. Put on a polystyrene lid. Freeze at -80 o C, and store in a cryo-box. 9. Harvest the cells and/or supernatant by centrifugation of the second MTP. Optionally: add 250 µl sterile water or buffer prior to centrifugation: this will make it easier to take off the supernatant after centrifugation 10. Perform the assay on either the supernatant (in case of an extracellular product or enzyme), or lysed cells (in case of a cytoplasmatic enzyme product).

Glucose feeding systems: Many large scale bioprocesses are fed-batch systems: when the initially present glucose (maximally 100-200 mM because of toxicity reasons) is fully consumed, the continuous feeding of a highly concentrated glucose solution is started. Mutant or medium screenings aimed at an increased productivity of such fed-batch systems are ideally also done in a similar fed-batch mode. This may be a realistic option for 24 or 96 cultures using peristaltic pumps or syringe-based feeding systems, but such external dosing systems are practically not feasable for thousands of cultures. For large numbers of cultures in MTPs, two internal glucose delivery systems are applicable. Firstly, small silicone elastomer disks containing glucose crystals may be added to each well (available via Kuhner AG). The glucose will slowly diffuse from the disks into the medium. A second option is the addition of a combination of glucosidases and starch or cyclodextrin to each well. By varying the amount of glucosidase in the medium, the glucose supply rate can be adjusted to a level that mimics the large scale bioprocess. This second strategy was patented by Green and Rheinwald from the MIT in 1975, and recently further elaborated upon by Panula-Perälä et al. In the latter paper, the starch is added into an agar gel on the bottom of each well, while glucoamylase is added to the growth medium itself. A distinct advantage of the silicone elastomer disks is that also other compounds than glucose can be used, and that it can also be applied for strains that produce proteases (that would destruct the glucosidase used in the enzymatic method). A practical and logistic advantage of the enzymatic method is that the necessary components can be added as liquids. The substantially higher cell densities that can be reached with these systems also makes it more challenging to keep the pH within certain limits in the absence of an active pH control system. A relatively strong buffer (e.g. 0.15-0.2 M phosphate buffer) is recommended. Also, ammonia is preferably not used as a nitrogen source since with the consumption of each molecule of NH 4 + one proton is released, and the medium may thus acidify rapidly. Use of an acid carbon and energy source (e.g. succinic acid or acetic acid) leads to a pH rise when consumed, and may thus counteract a pH drop due to the consumption of NH4+. If a sugar is used as a carbon and energy source, applying high OTR shaking conditions will keep the fermentative formation of acids to a minimum for many microbial strains.