Letters in Applied Microbiology ISSN -5 ORIGINAL ARTICLE Fermentation in -well plates is an efficient screening platform for filamentous fungi T. Linde, N.B. Hansen,M.L ubeck and P.S. L ubeck Section for Sustainable Biotecnology, Aalborg University, Copenagen SV, Denmark Significance and Impact of te Study: Fermentation studies wit filamentous fungi and especially screening experiments often struggle wit ig inter-vessel variations in metabolite production. Tis study compares two different types of frequently used screening metods namely batc fermentation in Erlenmeyer flasks wit batc fermentation in -well plates. Te results demonstrate tat te variance potentially can be reduced two and a alf-fold using -well plates leading to improved resolution wen testing te impact of varying fermentation parameters on product formation. Keywords bioprocessing, fermentation biotecnology, fungi, optimization. Correspondence Peter Stepensen L ubeck, Section for Sustainable Biotecnology, Aalborg University, A.C. Meyers Vaenge 15, Copenagen SV DK-5, Denmark. E-mail: psl@bio.aau.dk Autors contributed equally to tis work. 13/7: received 1 December 13, revised April 1 and accepted April 1 doi:1.1111/lam.1 Abstract Fermentation by filamentous fungi in Erlenmeyer flasks is a favoured metod for comparing different fermentation conditions. However, significant interflask variation often occurs wen using Erlenmeyer flasks, wic makes te comparison of fermentation product levels less reliable. We ave investigated te use of a -well plate metod for citric acid, etanol and glycerol batc fermentation using te filamentous fungi Aspergillus carbonarius and compared te relative standard deviation (RSD) from sextuplicates obtained using Erlenmeyer flasks and -well plates. Te production levels using te Erlenmeyer flasks sowed a combined average RSD of 9%, wic is two and a alf-fold iger tan wat was measured using te -well plates sowing an average RSD of 1%. We conclude tat fermentation in -well plates is a more reliable screening metod for metabolite production by filamentous fungi and possibly for screening metabolites in general. Introduction Improvement of an organism-mediated process often involves te identification and optimization of several parameters, wic may include temperature, media composition, saking system and oter factors (Sancez- Marroquin et al. 197; Xu et al. 199), as for example growt or morpology of te production organism, or genetic canges (de Jong and Nielsen ). Altoug eac parameter modification migt cause subtle improvements, tey can be difficult to identify because of te variation associated wit biological systems. Te biological variation can be accounted for by setting up experiments in replica; owever, te number of experiments needed to identify successful improvements is often ig, tereby leading to an increase in resources allocated to eac experiment. To overcome tis, a ig-trougput screening process is essential for te identification of multiple successful parameters (Nagy et al. 7). Furtermore, te screening metod as to be reliable, stable and able to generate reproducible results. Traditionally, researc on citric acid production by filamentous fungi as been performed by batc fermentation experiments. Te basic strategy is as follows: an Erlenmeyer flask tat contains a specific kind of media is inoculated wit a filamentous fungus. Te inoculated Erlenmeyer flask is ten incubated at a specific temperature at saking conditions for defined amount of time intervals at wic a sample is witdrawn, measured and used to represent te production capacity of te filamentous fungi under tese conditions (Su and Jonson 19). Letters in Applied Microbiology 59, --3 1 Te Society for Applied Microbiology
T. Linde et al. Screening platform for filamentous fungi Conventional batc screening in flasks or fermenters of newly developed strains is often time-consuming and laborious just because of te large number of strains tat require testing. One approac to ease te workload is to scale down te fermentation process. Since 1951, multiwell plates ave been used for downscaling fermentation processes wit different organisms (Manns 1999). However, tere are only few reports on successful experiments wit filamentous organisms and very few on filamentous fungi (Bills et al. ; Alberto et al. 9; Sooni et al. 1). Screening using -well plate (multiwell culture plates) batc fermentation metods for filamentous fungi would be beneficiary in experiments were a large number of replicas are needed to identify small yet significant canges. Fermentation in -well plates also offers reduced time and labour wen te number of experiments is ig. Also, te volume of medium used for eac experiment is decreased, tus reducing resources allocated to eac experiment. To our understanding, it is not directly stated in any paper tat conventional Erlenmeyer flask batc fermentation suffers from large variations. However, one study on citric acid batc fermentation reports a cut-off value for results yielding <75% of teir maximum citric acid production result, to make sure tat only ig-yielding citric acid production results were taken into consideration (Peksel et al. ). Furtermore, te commercial citric acid producing company Cemtotal states Variations in results often occur even under apparently similar operating conditions. All plants prefer to rely upon teir own experience rater tan te so-called proven facts (Cemtotal 1). Te mere fact tat standard deviations are not reported in many batc production experiments empasizes tis issue (Su and Jonson 197; Tomlinson et al. 195; Kitos et al. 1953; Noguci and Jonson 191; Kubicek-Pranz et al. 199; Dillon et al. 11). Te only way to account for te variation is by setting up experiments in a ig number of replicas. Aspergillus niger can produce ig titres of citric acid (Su and Jonson 19). We ave identified a strain of te closely related species A. carbonarius, wic we ave selected for different projects, involving optimization of growt conditions and patway engineering. Aspergillus carbonarius as sown to be a very good producer of organic acids, especially citric acid and gluconic acid (Yang 1). Te aim of tis study was to apply a -well plate metod as a screening tool for batc production of citric acid wit A. carbonarius and to study weter we could minimize te variation between repeated samples compared wit te citric acid batc production in Erlenmeyer flasks. Hereby, we could attain more representative values for expressing te citric acid productivity of A. carbonarius. Terefore, our ypotesis was tat fermentation in -well plates enables better omogeneity regarding aeration, water evaporation, temperature and surface structure in eac well, tan tat obtained in Erlenmeyer flasks. Also included in te study were te production results of etanol and glycerol in te two systems, to give a more profound picture of te variation problem. Results and discussion In several studies, researcers ave sown tat te microplate metod is at least comparable to Erlenmeyer flask metods as batc fermentation in terms of reproducibility and variation between samples (Sooni et al. 1). In tis study, we successfully scaled down te conventional batc fermentation from Erlenmeyer flasks to -well plates for A. carbonarius metabolite production wile decreasing variability. Sample-to-sample variation for citric acid production In te initial experiment, citric acid titres were measured to assess te variation in fermentation. To test te difference in variation between Erlenmeyer flasks and wells of -well plates, te two set-ups were grown simultaneously and samples taken after 9. To test weter citric acid titres were affected by preinoculating conditions, two individual spore suspensions were prepared from two separate Petri dises containing sporulating A. carbonarius. Eac spore suspension was aliquoted in wells (one plate) and six 5-ml Erlenmeyer flasks. Te outcome of te citric acid fermentation is presented in Fig. 1. Citric acid g l 1 1 1 1 1 Well system Erlenmeyer system Figure 1 Comparison of te citric acid production after 9. Two citric acid batc productions were conducted in -well plates and Erlenmeyer flasks. Two different spore preparations were used to ensure tat citric acid titre was unaffected by preinoculating conditions. Te error bars represent te sample standard deviation obtained from six observations on one -well plate. Eac Erlenmeyer flask result is obtained from six Erlenmeyer flasks. Te ligt grey bars represent spore suspension 1; dark grey bars represent spore suspension. Letters in Applied Microbiology 59, --3 1 Te Society for Applied Microbiology 5
Screening platform for filamentous fungi T. Linde et al. Te citric acid titre levels obtained after 9 of fermentation in -well plates were 5 g l 1 1% (mean relative standard deviation (RSD)) and 59 g l 1 %. In comparison, te titres obtained from te Erlenmeyer flasks were g l 1 5% and 7 g l 1 5%. Te very similar results from te two -well plates indicated tat te preinoculation conditions did not ave any significant effect on te outcome of te citric acid titre and variation. Citric acid fermentation in Erlenmeyer flasks To gain insigt into te causes of te ig RSD observed in Erlenmeyer flasks, a time-course experiment was carried out were A. carbonarius spores were added to six Erlenmeyer flasks in a final concentration of 1 5 spores per ml. Samples from te six replicates were taken for ig-performance liquid cromatograpy (HPLC) analysis at, 9, 1 and 19 after inoculation (Fig. ). Te average RSD was very ig in te flask experiment (%). At, te citric acid production was almost identical in te different flasks. However, already at 9 and beyond, te RSD increased in all analysed parameters. An exception to te variation observed was te pelleted morpology (Fig. ) and also te amount of biomass accumulated at day eigt (3 gl 1 ). Comparison of metabolite production in -well plates and Erlenmeyer flasks To make a clear comparison, fermentation was conducted in -well plates wit te same set-up; te only differences were te vessel type and te incubator type. Samples were Citric acid g l 1 1 1 1 1 1 7 9 1 1 1 19 Figure Citric acid fermentation in Erlenmeyer flasks. Six citric acid fermentations were performed in six Erlenmeyer flasks. Observations were made at, 9, 1 and 19 after inoculation to determine citric acid titres. Te figure sows te individual observation obtained from eac Erlenmeyer flask. ( ) Flask 1, ( ) Flask, (D) Flask 3, ( ) Flask, (9) Flask 5 and (+) Flask. taken at, 9, 1 and 19. From te Erlenmeyer setup, one ml was sampled from six individual flasks and te flasks reincubated. In contrast, six wells were emptied wen sampling from te -well plate fermentation, tereby sampling in a noncontinuous manner. Te resulting citric acid, glycerol and etanol fermentation profiles over time are sown in Fig. 3. Te RSD was lower for all of te -well plate experiments compared to te RSD of te Erlenmeyer flask experiments except for citric acid production at. Te mean RSD for Erlenmeyer production from, 9, 1 and 19 in te citric acid, etanol and glycerol production experiment was, 5 and 1%, respectively. In contrast, in te production in te -well plates, te mean RSD from all samples taken from to 19 was 11, and %, respectively. Te mean RSD for all te Erlenmeyer samples combined was 9%, wile te mean RSD for all te well samples combined was 1% (Table 1). Our studies wit metabolite production, using A. carbonarius in batc fermentation in Erlenmeyer flasks wit cotton stoppers, sowed tat tis metod in fact as large variations in production levels. Fermentation in -well plates largely reduced te RSD associated wit citric acid production by A. carbonarius, making it possible to attain more representative citric acid productivity values. Te igest citric acid titre level at 35 gl 1 was obtained after 19 in -well plates, wic also is te igest amount of citric acid concentration reported from A. carbonarius. From a statistical point of view, te percentage needed to obtain a significant cange from te mean value was determined. If for example a two-tailed t test was employed in eac of te fermentation metods, te percentage needed to obtain a significant cange from te mean value would need to be 9 1% (Student/Welc statistics) to sow a significant cange (P < 5). In contrast, te -well plate set-up sowed an RSD at 1%, and only a 17% difference (independent on Student/ Welc statistics) would sow a significant cange (P < 5). Results in -well plates more effectively resolve te effects of ow different parameters affect te experimental outcomes. As wit conventional batc fermentation, fermentation in microwell plates as numerous parameters tat ave to be adjusted to obtain useful and reliable screening results (Sooni et al. 1). One very important parameter in aerobic fermentation and especially in citric acid fermentation is to ensure a sufficient oxygen transfer rate (OTR) for te organism to proliferate (Kubicek et al. 195). One side effect of ig OTRs is increased evaporation. It is very important to minimize evaporation from te production media, because te evaporation increases te concentrations of components in te media and tereby Letters in Applied Microbiology 59, --3 1 Te Society for Applied Microbiology
T. Linde et al. Screening platform for filamentous fungi (a) 5 Table 1 Te relative standard deviation (RSD) obtained in tis study 35 Experiment Time () -well RSD (%) Erlenmeyer RSD (%) Citric acid g l 1 (b) Etanol g l 1 (c) Glycerol g l 1 3 5 15 1 5 9 7 5 3 1 1 1 1 1 1 9 1 19 9 1 19 9 1 19 Citric acid production 13 9 1 1 3 57 19 1 5 Etanol production 9 19 1 7 1 19 3 35 Glycerol production 5 9 9 9 1 15 19 1 Average RSD 1 9 From RSDs obtained in te metabolite production study, an average RSD was deduced from te -well plate metod as well as te Erlenmeyer metod. No RSD value could be obtained from etanol production at because of lack of etanol production. gradually canges te foundation of te fermentation in an uncontrolled manner (Duetz and Witolt 1). Te OTR in batc experiments is directly related to wat type of flask or microplate is used, wat speed and diameter te saking device is using and wat amount of air is available troug different lid types (Maier and B ucs 1; Zang et al. ). In te present study, te results suggest tat te -well plates facilitate better omogeneity regarding OTR, water evaporation, temperature and surface structure compared to wat can be acieved in Erlenmeyer flasks. Compared to Erlenmeyer flasks, -well plates are a better coice for screening of fungal strains for te production of at least citric acid because of ig reproducibility. Moreover, te format facilitates screening large numbers of strains. Finally, Erlenmeyer flask could be omitted in te process of going from laboratory-scale production in -well plates to fermentation in 1 1-l fermentation vessels, because te production in -well plates is at least as good as in Erlenmeyer flasks and te reproducibility is muc better. Figure 3 Fermentation profile for -well plates and Erlenmeyer flasks. Over a 19- period, samples were taken and analysed. Eac data entry represents te average titre for six samples. Eac -well plate data entry represents te termination of six wells from te -well plate, wereas te Erlenmeyer flask fermentation is sampled continuously. Te ligt grey bars represent te Erlenmeyer flasks; te dark grey bars represent te wells. Te tree figures a, b and c sow te results gatered for tree different metabolites in te same experiment. (a) Citric acid production. Te flask experiment values ave been multiplied by two for easier visual comparison of te relative standard deviation. (b) Etanol production. Te well experiment values ave been multiplied by two. (c) Glycerol production. Morpology and biomass production Te mycelia morpology during citric acid batc production wit Erlenmeyer flasks and -well plates was similar. During bot fermentation set-ups, A. carbonarius growt consists of pellets wit a diameter varying between 5 and lm. However, during te Erlenmeyer flask production, considerable wall growt occurred near te upper waterfront (Fig. ). Altoug te morpology of te micro-organism plays a role in controlling and augmenting citric acid formation Letters in Applied Microbiology 59, --3 1 Te Society for Applied Microbiology 7
Screening platform for filamentous fungi T. Linde et al. -wells 1 3 5 Erlenmeyer flasks Figure Te morpology of Aspergillus carbonarius during citric acid batc production after 9 of citric acid batc production in -well plates and Erlenmeyer flasks. First row sows te growt in -well plates. Second and 3rd rows sow te growt in Erlenmeyer flasks. (Max et al. 1), biomass and pellet form did not differ significantly in our experiments and tus could not explain te variation in citric acid production. Materials and metods Fungal culture Te fungal strain A. carbonarius ITEM 51 was originally isolated from grapes in Italy (ttp://server.ispa.cnr. it/item/collection/). It was kindly supplied from Jens Frisvad, Danis Tecnical University, Lyngby (Department of Systems Biology, Center for Microbial Biotecnology, Fungal Cemodiversity). Te culture was stored in spore suspension at C in 5% glycerol and was revived on potato dextrose agar (PDA) medium (Scarlau Cemie S.A. Barcelona, Spain). Inoculum preparation Ten microlitre spores from te culture stock was spread onto PDA plates and incubated 5 days at 3 C. Spores were arvested by adding 3 ml milliq water and scraping te spores wit a spatula. Te concentrated spore suspension was defined by counting te spores in a sample under 9 magnification (Olympus CX1, Olympus Europa SE & Co. KG, Hamburg, Germany) and a counting camber (Fucs Rosental, Hausser Scientific Company, Piladelpia, PA). Te concentration of spores in te spore suspension was 1 7 1 spores per ml. Te spore suspension was used directly as inoculum for te fermentation. Te fermentation media was prepared, and te spore suspension was added to a final spore concentration of 1 5 spores per ml. Fermentation media and spores were prepared in one single flask for an experiment. Te flask was ten stirred using a magnetic stirrer to obtain a uniform solution, and afterwards, te replicate samples were distributed uniformly into te experiment vessel type. Sample analysis Eigt undred microlitre fermentation brot was filtered troug 5-lm HPLC-grade regenerated cellulose membrane filters. One undred and fifty microlitre of te filtrates was analysed for te content of cellobiose, glucose, xylose, glycerol, citric acid, etanol, arabinose, lactate and acetate by HPLC on a Dionex Ultimate Letters in Applied Microbiology 59, --3 1 Te Society for Applied Microbiology
T. Linde et al. Screening platform for filamentous fungi 3-LC system (Dionex Corporation, Sunnyvale, CA) wit an Aminex HPX-7H column coupled to an RIdetector. Te eluent was mmol l 1 H SO wit a flow rate of mlmin 1 at C. All cromatograms were integrated using te Cromeleon software (Dionex Corporation). Fungal biomass measurements Biomass was measured at te end of an experiment. Te remaining fermentation brot was filtered for biomass, and te liquid pase was discarded. Te biomass was dried in an oven at 1 C for and weigted. Media preparation Potato dextrose agar plates for spore production were prepared using PDA mixed wit milliq water according to te manufactures instructions and autoclaved at 11 C for min. Fifteen millilitre of te ot PDA solution was poured into Petri dises wit a diameter of 9 cm and was solidified in a sterile environment. Te recipe for te fermentation media was obtained from (Su and Jonson 19), and no alteration was made to te original recipe. A total of 9 mmol l 1 sucrose, 1 mmol l 1 KH PO, 1 mmol l 1 MgSO 7H O, 31 mmol l 1 NH NO 3 and 1 mol l 1 HCl were used to adjust te ph to 3. Te solution was autoclaved at 11 C for min. After cooling, filter-sterilized trace metals were added to a final concentration of 9 lmol l 1 CuSO, 3 lmol l 1 ZnSO and lmol l 1 FeSO (metal-based weigt). Te fermentation media were prepared and used fres to avoid any precipitation. Twenty-four-well plate and flask types Te strain was fermented in eiter Greiner CELLSTAR multiwell culture plates ( wells) for suspension cultures (Cruinn, Dublin, Ireland) or 5-ml unbaffled Erlenmeyer flasks. Te -well plates are capable of olding 33 ml media per well. However, 15 ml fermentation media including spore inoculum were aliquoted in eac well. Te plates were covered wit plastic lids during te fermentation. Te plastic lids were perforated wit 15 1 mm oles to increase airflow to te set-up. In all Erlenmeyer flask experiments, volume of fermentation media was 5 ml (1% v/v). Te flasks were sealed wit cotton plugs for gas excange and to prevent contamination. Incubator set-up parameters Te incubator type used for te -well plates was an incubator saker from Glas-Col model 99A DPMINC (Glas-Col, Terre Haute, IN). Te incubator was saking at 5 rev min 1, 5 C. Te saking amplitude was 7 cm in diameter. Te umidity in te incubator was controlled by mounting plastic boxes wit open milliq water inside te closed incubator, tereby creating a very umid environment to avoid excess evaporation from te fermentation brot. Te incubator type used for te Erlenmeyer flasks was an incubator rotator from IKA model KS -ic-control (IKA-Werke GmbH & Co. KG, Staufen, Germany). Te incubator was rotating at 7 rev min 1 wit 5 cm amplitude in diameter. Bot incubators were run at 5 C in te same incubator room. All incubations were run wit a comparable flask or plate containing milliq water to account for water loss due to evaporation. Statistical analysis Te mean value, standard deviation based on sample and RSD were used as comparison factors trougout te researc. Mean values are separated from teir sample standard deviation wit te plus minus sign (). Te RSD was calculated as te coefficient of variation. Te sample coefficient of variation is defined as te ratio of te sample standard deviation to te sample mean. All samples consist of six observations to be able to give a iger statistical impact tan te conventional triplicates. Levene s test verified tat bot Student s t test and Welc s t test, bot wit a two-tailed distribution, sould be used to give a ubiquitous and comparable P (P < 5). All calculations and modelling were performed in R-Development (). Acknowledgements Tecnician Kristian Hansen is tanked for excellent tecnical assistance. Tis work was supported by te European Community s Sevent Framework Programme (7 13) [SupraBIO, Grant Number 1], Novozymes A/S and Te Danis Council for Strategic Researc [MycoFuelCem, Grant Number 11-113]. Conflict of Interest No conflict of interest declared. References Alberto, F., Navarro, D., De Vries, R., Aster, M. and Record, E. (9) Tecnical advance in fungal biotecnology: development of a miniaturized culture metod and an automated ig-trougput screening. Lett Appl Microbiol 9, 7. Letters in Applied Microbiology 59, --3 1 Te Society for Applied Microbiology 9
Screening platform for filamentous fungi T. Linde et al. Bills, G.F., Platas, G., Fillola, A., Jimenez, M.R., Collado, J., Vicente, F., Martın, J., Gonzalez, A. et al. () Enancement of antibiotic and secondary metabolite detection from filamentous fungi by growt on nutritional arrays. J Appl Microbiol 1, 1 15. Cemtotal (1) Profile Citric Acid. [Online] Available from: ttp://www.cemtotal.com/citric-acid-a-profile.tml. [Accessed: June 19, 13]. Dillon, G.S., Brar, S.K., Verma, M. and Tyagi, R.D. (11) Apple pomace ultrafiltration sludge: a novel substrate for fungal bioproduction of citric acid: optimisation studies. Food Cem 1, 71. Duetz, W.A. and Witolt, B. (1) Effectiveness of orbital saking for te aeration of suspended bacterial cultures in square-deepwell microtiter plates. Biocem Eng J 7, 113 115. de Jong, W.A. and Nielsen, J. () Enanced citrate production troug gene insertion in Aspergillus niger. Metab Eng 1, 7 9. Kitos, P.A., Campbell, J.J.R. and Tomlinson, N. (1953) Influence of temperature on te trace element requirements for citric acid production by Aspergillus niger. Appl Environ Microbiol 1, 15 159. Kubicek, C.P., R or, M. and Rem, H. (195) Citric acid fermentation. Crit Rev Biotecnol 3, 331 373. Kubicek-Pranz, E.M., Mozelt, M., Ror, M. and Kubicek, C.P. (199) Canges in te concentration of fructose, - bispospate in Aspergillus niger during stimulation of acidogenesis by elevated sucrose concentration. Biocim Biopys Acta 133, 5 55. Maier, U. and B ucs, J. (1) Caracterisation of te gasliquid mass transfer in saking bioreactors. Biocem Eng J 7, 99 1. Manns, R. (1999) Microplate History. [Online] Available from: ttp://www.plastiquarian.com/index.pp?id=13&subid=179. [Accessed: May 1, 13]. Max, B., Salgado, J.M., Rodrıguez, N., Cortes, S., Converti, A. and Domınguez, J.M. (1) Biotecnological production of citric acid. Braz J Microbiol 1, 75. Nagy, V., Seidl, V., Szakacs, G., Komon-Zelazowska, M., Kubicek, C.P. and Druzinina, I.S. (7) Application of DNA bar codes for screening of industrially important fungi: te aplotype of Tricoderma arzianum sensu stricto indicates superior citinase formation. Appl Environ Microbiol 73, 7 75. Noguci, Y. and Jonson, M.J. (191) Citric acid fermentation of sugars purified wit celating resin. J Bacteriol, 53 51. Peksel, A., Torres, N., Liu, J., Juneau, G. and Kubicek, C.P. () 13C-NMR analysis of glucose metabolism during citric acid production by Aspergillus niger. Appl Microbiol Biotecnol 5, 157 13. R-Development (). R: A Language and Environment for Statistical Computing. [Open source] ISBN 3-951-7-. Vienna, Austria: R Foundation for Statistical Computing. Sancez-Marroquin, A., Carreno, R. and Ledezma, M. (197) Effect of trace elements on citric acid fermentation by Aspergillus niger. Appl Environ Microbiol, 9. Su, P. and Jonson, M.J. (197) Effect of te composition of te sporulation medium on citric acid production by Aspergillus niger in submerged culture. J Bacteriol 5, 11 17. Su, P. and Jonson, M.J. (19) Citric acid. Ind Eng Cem, 1 15. Sooni, S.V., Bapat, P.M., and Lantz, A.E. (1) Robust, small-scale cultivation platform for Streptomyces coelicolor. Microb Cell Fact doi: 1.11/175-59-11-9. Tomlinson, N., Campbell, J.J.R. and Trussell, P.C. (195) Te influence of zinc, iron, copper, and manganese on te production of citric acid by Aspergillus niger. J Bacteriol 59, 17. Xu, D.B., Madrid, C.P., R or, M. and Kubicek, C.P. (199) Te influence of type and concentration of te carbon source on production of citric acid by Aspergillus niger. Appl Microbiol Biotecnol 3, 553 55. Yang, L. (1) Metabolic Engineering of Aspergillus carbonarius for Organic Acids. A Tesis Submitted in partial fulfilment of te Requirements of Aalborg University for te Degree of Master of Science. Copenagen, Denmark: Aalborg University Denmark. Zang, H., Lamping, S.R., Pickering, S.C., Lye, G.J. and Samlou, P.A. () Engineering caracterisation of a single well from -well and 9-well microtitre plates. Biocem Eng J, 13 19. 3 Letters in Applied Microbiology 59, --3 1 Te Society for Applied Microbiology