+

WO2001036664A1 - Procede de detection de micro-organismes contaminants - Google Patents

Procede de detection de micro-organismes contaminants Download PDF

Info

Publication number
WO2001036664A1
WO2001036664A1 PCT/SE2000/002218 SE0002218W WO0136664A1 WO 2001036664 A1 WO2001036664 A1 WO 2001036664A1 SE 0002218 W SE0002218 W SE 0002218W WO 0136664 A1 WO0136664 A1 WO 0136664A1
Authority
WO
WIPO (PCT)
Prior art keywords
cell culture
sensors
electronic nose
gas
cultivation
Prior art date
Application number
PCT/SE2000/002218
Other languages
English (en)
Inventor
Carl-Fredik Anton Mandenius
Thomas Bachinger
Original Assignee
Appliedsensor Sweden Ab
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Appliedsensor Sweden Ab filed Critical Appliedsensor Sweden Ab
Priority to AU15651/01A priority Critical patent/AU1565101A/en
Publication of WO2001036664A1 publication Critical patent/WO2001036664A1/fr

Links

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/48—Biological material, e.g. blood, urine; Haemocytometers
    • G01N33/483—Physical analysis of biological material
    • G01N33/497—Physical analysis of biological material of gaseous biological material, e.g. breath
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/0004—Gaseous mixtures, e.g. polluted air
    • G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
    • G01N33/0027—General constructional details of gas analysers, e.g. portable test equipment concerning the detector
    • G01N33/0031—General constructional details of gas analysers, e.g. portable test equipment concerning the detector comprising two or more sensors, e.g. a sensor array
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
    • G01N33/0004—Gaseous mixtures, e.g. polluted air
    • G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
    • G01N33/0027—General constructional details of gas analysers, e.g. portable test equipment concerning the detector
    • G01N33/0036—General constructional details of gas analysers, e.g. portable test equipment concerning the detector specially adapted to detect a particular component
    • G01N33/0047—Organic compounds

Definitions

  • the present invention relates to a method for early detection of undesired microbial infection in a cell culture £>y means of an electronic nose and to use of the electronic nose for said detection. It also relates to a method for estimating the product concentration m a cell culture k»y means of an electronic nose and to use of the electronic nose for said estimation.
  • the detection is based on recording the gaseous emission from cultures with arrays of gas sensors sensitive to different volatile compounds, e.g. hydrocarbons, such as aromatic compounds (e.g. ergosterol) , aldehydes (e.g. acetaldehyde) and alcohols (e.g. ethanol); sulphurous compounds, nitrous compounds, hydrogen, and carbon oxides, specific for each bacterium strain.
  • hydrocarbons such as aromatic compounds (e.g. ergosterol) , aldehydes (e.g. acetaldehyde) and alcohols (e.g. ethanol); sulphurous compounds, nitrous compounds, hydrogen, and carbon oxides, specific for each bacterium strain.
  • An object of the present invention is to eliminate the above-mentioned problems involving slow detection of microbial infections m animal cell cultures .
  • Another ob j ect of the present invention is to provide a method for estimating the product concentration m a cell culture .
  • Fig 1 (a) illustrates the time profiles of selected sensor response signals from the electronic nose and cul - tivation parameters for a production-scale rFVIII cultivation.
  • a bacterial infection occurred at the end of the cultivation.
  • MOSFET 3 response is magnified to its 20 minutes measurement interval during the infection period.
  • the first reaction of the p0 2 -sond to the infection is indicated m the graph.
  • the cultivation parameters were normalised between 0 and 1.
  • the sensor signals are given m arbitrary units.
  • Fig 1 (b) illustrates time profiles of selected sensor response signals and cultivation parameters for a successful production-scale rFVIII cultivation. The cultivation parameters were normalised.
  • the sensor signals are given m arbitrary units.
  • Fig 2 (a) illustrates time profiles of selected sensor response signals and cultivation parameters for a laboratory-scale rFVIII cultivation.
  • the cultivation was contaminated with Bacillus cereus .
  • Fl, F2 introduced process faults through dissolved oxygen decrease to 0% saturation.
  • M the electronic nose was disconnected from the process.
  • MOSFET 3 response is magnified to its 20 minutes measurement interval during the infection period.
  • the first reaction of the p0 2 -sond to the infection is indicated m the graph.
  • the cultivation parameters were normalised.
  • the sensor signals are given m arbitrary units .
  • Fig 2 (b) illustrates time profiles of selected sensor response signals and cultivation parameters for a laboratory-scale rFVIII cultivation.
  • the cultivation was contaminated with Pseudomonas aerugmosa .
  • MOSFET 3 response is magnified to its 20 minutes measurement interval during the infection period.
  • the first reaction of the p0 2 -sond to the infection is indicated m the graph.
  • the cultivation parameters were normalised.
  • the sensor signals are given m arbitrary units.
  • Fig 3 (a) illustrates time profiles of selected sensor signals and cultivation parameters for a laboratory- scale control fermentation. Pseudomonas aerugmosa was added to blank cultivation medium at the indicated point m time. The cultivation parameters were normalised. The sensor signals are given m arbitrary units .
  • Fig 3 (b) illustrates time profiles of selected sensor signals and cultivation parameters for a laboratory- scale control fermentation. Bacillus subtilis was added to blank cultivation medium at the indicated point m time. The cultivation parameters were normalised. The sensor signals are given m arbitrary units.
  • Fig 4 (a) illustrates the original and adjusted gas sensor signal from one production-scale rFVIII cultivation.
  • the sensor signals are given m arbitrary units.
  • Fig 4 (b) illustrates the time profiles of selected cultivation parameters and gas sensor signals for a laboratory-scale rFVIII cultivation. Commercial interest demands a normalisation of the cultivation parameters between 0 and 1.
  • the sensor signals are given m arbit- rary units.
  • Fig 5 illustrates the characterisation of the gas sensor behaviour to process parameter changes .
  • the sensor signal is given m arbitrary units.
  • Fig 6 illustrates the time profiles of selected sen- sor signals from the electronic nose and important cultivation parameters for a production-scale rFVIII cultivation.
  • the cultivation parameters are normalised between 0 and 1.
  • the sensor signals are given m arbitrary units.
  • Fig 7 illustrates the estimated and measured recom- bmant Factor VIII concentration vs time of cultivation from one production-scale rFVIII cultivation.
  • the estimated values were obtained from a neural network model trained with electronic nose data from 5 production-scale rFVIII cultivations.
  • the rFVIII concentration was norma- lised between 0 and 1.
  • Fig 8 illustrates the estimated and measured viable cell count vs time of cultivation from one laboratory- scale rFVIII cultivation.
  • the estimated values were ob- tained from a neural network model trained with electronic nose data from one laboratory-scale rFVIII cultivation.
  • the viable cell count was normalised between 0 and 1.
  • the present inventors have surprisingly found a method for early detection of microbial infection m an animal cell culture process by use of an electronic nose. Infections with defined cell concentrations have also been created to characterise the electronic nose response towards particularily three infecting organisms common m bioprocesses .
  • the present inventors have found a method for detecting the product concentration in animal cell cul - tures by means of an electronic nose comprising an artificial neural network and having the ability of pattern recognition, which has not previously been known or possible.
  • both of the above-mentioned methods found by the inventors may be carried out in the same cell culture process by use of the same electronic nose simultaneously .
  • the terms "microbial” and "micro-organism” used herein relates to any undesired bacteria, yeast and fungi which might infect or contaminate the bioreactor content.
  • the cell culture in which potential microbial contamination or infection is to be detected is preferably an animal cell culture, e.g. a human or animal cell culture, e.g a fibroblast cell culture, a hamster cell cul - ture, a hybridoma cell culture, an insect cell culture, etc.
  • the cell culture medium can be any conventional medium used m this area.
  • the volume of the cell culture medium may also vary from laboratory scale up to large scale ( 0 , 5 - 20 000 1 ) .
  • the off-gas from the cell culture is accumulated m the headspace above the cell culture surface m the bio- reactor and comprises volatile compounds, e.g. hydrocarbons, emitted by and specific for any micro-organism m the culture.
  • a culture of recombmant CHO (Chinese Hamster Ovarian) - cells producing human Factor VIII (rFVIII) is detected for micro-organisms, and tests have been performed m a proprietary medium on 2-L and 500-L scale, respectively. After an initial fed-batch phase the culture is continuously perfused with fresh medium using an external cell retention device. Samples were taken every 24 h and the cell concentration and viability were measured by coloured dye exclusion.
  • electronic nose means a device containing an array of gas sensors that produces a characteristic response to a gaseous sample, said device also allowing pattern recognition.
  • artificial neural network used herein means a system used m a computational method based on the theory of artificial neural network described by e.g. C M Bishop, Neural Networks for Pattern Recognition, Oxford Univ. Press 1995.
  • pattern recognition means identifying a pattern of responses with known variables by means of any computational and/or non-computational method.
  • the non-computational method is e.g. visual observation of patterns of sensor signals.
  • the preferred electronic nose used m connection with the present invention is well known (NST3310) and is accessible from Nordic Sensor Technologies AB, Teknik- r gen 8, M ardevik Science Park, 583 30 Lmkoping,
  • MOS- sensors semiconductor metal -oxide semiconductor field-effect transistors with catalytic metal gates of palladium, indium, or platinum, operated at different temperatures 140/170°C
  • MOS- sensors semiconductor metal -oxide sensors of Taguchi or FIS-type; Sn0 2 sensors operated at 400°C
  • C0 2 -sen- sor based on infrared adsorption.
  • the sensor array contained 10 MOSFET-sensors , 19 MOS-sensors, and 1 C0 2 -sensor based on infrared adsorption.
  • a mass flow controller maintained a stable flow over the sensors .
  • the sensors were arranged m a gas flow injection system where the gas stream first passed the MOSFET- sensors followed by the MOS-sensors and the CO2 -sensor.
  • the electronic nose was connected outside the sterile barrier to the bioreactor off -gas line. Sampling was performed continuously from the reactor off-gas for nor- mally 30 seconds every 20-mm throughout the whole process .
  • sensors m the electronic nose depend on the composition of the culture medium and the organisms m the medium to be detected. Thus, a vary- ing amount of MOSFET-, MOS-sensors and CO2 - sensors, respectively, and different combinations thereof may be used for the detection of micro-organisms.
  • Other sensors which might be useful are sensors based on quartz crystal, conductive polymers and/or semiconductors .
  • the cultivation parameters were set to match the rFVIII cultivation conditions. Gas flow was kept constant throughout the process at 800 ml/mm of compressed and filtered air.
  • a quadrupole mass spectrometer was connected to the bioreactor off-gas line (Multi-Gas, Leda-Mass Ltd., Cheshire, UK) .
  • the instrument was equipped with a dual detector system (faraday and electron multiplier detector) .
  • the bioreactor off-gas was screened for atomic mass- es 1 to 300 at maximum update speed ( «5 mm) .
  • the electronic nose that was connected to the bioreactor system was the same instrument that was used for the 2-L rFVIII cultivations described above.
  • Any micro-organism e.g. bacterium strain, contami- natmg or infecting a cell culture medium can be detected by use of the electronic nose m the mehtod according to the present invention, but preferably Bacillus cereus and Pseudomonas aerugmosa are detected.
  • LB medium 1% tryptone, 0.5% yeast extract, and 1% NaCl
  • Control fermentations were carried out m defined serum-free Eagle's MEM like cell culture medium (Shuler et al . , 1992).
  • the number of colony forming units was deter- mined by colony count after incubation of media samples on agar plates. Duplicates of media samples were applied onto LB-agar plates and incubated at 30°C for 24-48 h.
  • the detection can be made several h, i.e. about 50 h, normally 1-10 h, and most often about 5 h earlier for each micro-organism than m conventional methods, which is of particular interest in large-scale bioprocesses .
  • Different micro-organisms may be detected at different times of their growth phase, but generally the detection is possible within the above- mentioned time intervals earlier than is possible with conventional methods.
  • products may be detected by means of the method according to the present invention, preferably proteins, but also other biopolymers, such as poly- carbohydrates and polynucleotides .
  • Fig 1(a) is shown how it is possible to track the occurrence of a bacterial infection m a 500-L recombi- nant Factor VIII cultivation from the response signals of an electronic nose device.
  • the three sensor signals m the diagram monitor the cultivation for a period of 18 days m parallel with the measured viable cell count, rFVIII concentration and total flow rate through the bioreactor.
  • the signals mirror the basic changes product titer and viable cell concentration as reported m detail elsewhere (Bachmger et al . , 1999b) .
  • a technical failure occurred m the plant, which was believed to have caused a bacterial contamination of the culture.
  • the contamination was indicted by a drop m the dissolved oxygen concentration of the medium as monitored by the DO-electrode. Later, the infection was verified and the infecting organism identified as Bacillus cereus .
  • the electronic nose response indicated the infection already 6 to 8 h after the process fault had occurred.
  • the figure also shows that the viable cell count and flow rate were stable during this period.
  • Comparison with an umnfected 500-L rFVIII cultivation outlines the distinct differences between the sensor signal patterns from both cultivations (Fig 1 (b) ) .
  • the electronic nose is thus capable of indicating the bacterial infection approximately 10 h before the DO-sen- sor. It should however be stressed that the sensors did not reveal explicitly the cause of the signal deviation and the infection.
  • Fig 2 (a) shows the electronic nose response signals.
  • a process fault was introduced at the point m time marked with F2 the plot.
  • the dissolved oxygen concentration was decreased manually to 0% for 30 minutes and the sen- sor signals indicated the process deviation immediately (see Fig 2(a)) .
  • the sensor signals did not recover but instead gradually decreased after this incident.
  • the same process fault was introduced for a period of 10 h earlier m the cultivation (Fl) , the sensors did reach the original signal level immediately after the p0 2 was reset .
  • Fig 3(a) shows the electronic nose response to a control -fermentation inoculated with 100 colony forming units (cfu) of Pseudomonas aerugmosa per ml cell culture medium.
  • a response change a selected MOSFET-sensor indicated the contamination 14 h after inoculation
  • the p0 2 -sensor indicated a decreasing dissolved oxygen level.
  • a mass spectrometer was used on-line m this cultivation to monitor the bioreactor off -gas. The recorded partial pressure of C0 2 shows that the mass spectrometer does not indicate the infection earlier than the p0 2 -sensor.
  • the electronic nose response to a control fermenta- tion that was inoculated with 4000 cfu/ml of Bacillus subtilis is shown Fig 3b.
  • This high cell count caused an immediate signal change m one of the plotted sensors m Fig 3b.
  • the p0 2 -sensor monitored a decreasing dissolved oxygen level 7 h after inoculation. Again, the partial pressure of C0 2 recorded by the mass spectrometer did not indicate the infection earlier than the p0 2 -sensor.
  • the average cultivation time was about 5 weeks.
  • Sam- pies were taken every 24-h and the cell concentration and viability were measured by coloured dye exclusion.
  • the lactate concentration was analysed using a YSI model 2700 analyser (Yellow Springs Instruments, Yellow Springs, OH) .
  • the concentration and biological activity of rFVIII were determined using validated methods.
  • the distance between the electronic nose and the bioreactor outlet was 3 meters .
  • the bioreactor system was arranged m a way that it matched the production-scale conditions as closely as possible. Samples were taken varying time intervals every 12 or 24 h. Off-line analyses were performed as described under section 2.1.
  • the electronic nose was equipped with 10 MOSFET-sensors, 19 MOS-sensors, and 1 C0 2 -sensor. Additionally, a mass flow controller was introduced into the flow-mjec- tion system to establish a stable gas flow over the sensors.
  • the sampling interface described under 2.1 was modified so that the distance to the bioreactor outlet was further decreased to 0.7 meters .
  • the electronic nose that was used m the production- scale cultivations was also incorporated the monitoring set-up.
  • the algorithm for sensor shift exclusion and the median filter were developed in Object Pascal (Delphi 4, Inprise Corp., CA, USA) .
  • the shift exclusion algorithm subtracts/adds the retained offset from a gas sensor sig- nal when the shift exceeds a certain percentage and signal level and proceeds in only one direction.
  • the implementation of the algorithm for on-line recalculation should be straightforward.
  • the median filter was applied onto shift excluded sensor data with a window size cover- ing 5 consecutive measurement points.
  • the programming language MATLABTM (The Math orks Inc., MA, USA) , supported with the MATLABTM toolboxes for neural networks and chemometrics, was used for performing the artificial neural network (ANN) and principal component analysis (PCA) calculations.
  • the PLS-Toolbox for MATLABTM was used for the component correction method and the forward selection procedure (Eigenvector Technologies, Manson, WA) .
  • Principal component analysis is a linear un- supervised pattern recognition technique that can be used to reduce the dimensionality of multivariate data [11] .
  • the calculation results in principal components that describe the direction of the variation in the data set .
  • the result is presented in a plot where the two most im- portant principal components are plotted against each other, either as a score plot describing how the samples relate to each other or as loadings plot describing the relationships between the variables.
  • the neural network structure used in this work was a standard one hidden layer backpropagation ANN [17] with a sigmoidal activation function and the Levenberg-Marquardt update algorithm [14] .
  • the network structure for the estimation of the rFVIII concentration contained 11 nodes m the hidden layer, 4 input variables and 1 output.
  • the minimum error gradient was 0.0001, the initial value of ⁇ was 0.001 and the multipliers for increasing and decreasing ⁇ , were 10 and 0.1, respectively.
  • the error goal was set to 30, and 20 training cycles were required to reach an optimal performance of the model .
  • the ANN consisted of 9 hidden nodes, 6 input signals, and 1 output.
  • the minimum error gradient was 0.0001
  • the initial value of ⁇ was 0.001
  • the multipliers for increasing and decreasing ⁇ were 10 and 0.1, respectively.
  • the error goal was 0.5, and 30 training cycles were per- formed.
  • the component correction method uses a filter to reduce the sensor drift m order to increase the lifetime of pattern recognition models.
  • CC presumes that drift is not randomly spread but instead has a pre- ferred direction m measurement space.
  • a PCA- loading vector is calculated from the calibration measurements, which captures the direction the response space. Subtraction of the original cultivation data set from the projection of the cultivation data onto this loading vector results m the desired removal of the direction describing the drift. Thereby all other directions and important variances that separate clusters and concentrations are preserved.
  • the objective is to find a subset of the original sensor signals that minimises a selection criterion.
  • the selection criterion is the prediction error from a multiple linear regression model towards the desired model output (process variable) .
  • a forward selec- tion adds one variable at the time to the model until the selection criterion reaches a minimum.
  • the selected sensor signals contain relevant information for variable estimation, thus represent good parameters to use as inputs to the artificial neural network.
  • the cultivation data sets were treated with the shift exclusion algorithm, filtered, and calibrated, before a search was performed for correlation towards rFVIII concentration and viable cell count.
  • the multi-signal response of the electronic nose allows on- line estimation of key parameters by employing pattern recognition models.
  • a model should be built on a maximum number of data sets to ensure that it covers the whole variation m the process . This and the degree of correlation m the signals towards the estimated parameter will determine the models, flexibility and reliability.
  • a representative gas sensor signal from one of the cultivations is shown m Fig 4a (see MOSFET response) . Sudden response shifts can be observed m the signal . The use of multiva ⁇ ate methods for parameter estimation becomes impossible by this phenomenon.
  • the drift of solid-state gas sensors is a severe problem to the electronic nose technology [8] .
  • Established pattern recognition models would require constant re- calibration case no drift counteraction is employed. Especially for long-term bioprocesses this would be time- consuming and inefficient.
  • a possibility would be to repeatedly recalibrate the sensors with a suitable reference gas or gas mixture .
  • the reference gas mixture should not only be correlated with the measured sample but also the drift of both should be highly correlated. In a bioprocess however, one would first theoretically need to identify the most important volatile compounds produced during the course of the fermentation and use those as reference. This would however be impractical and it is therefore better to find a common substance that is present at high concentrations throughout the process.
  • the calibration of the gas sensors was realised by recording their response to the volatile compounds from the cultivation medium the bioreactor shortly before inoculation. At this stage the headspace will contain mostly water vapour. Water vapour has previously been found suitable for calibration purposes in several applications on food samples [9] . We have experienced that this calibration procedure gives a correct representation of the drift of gas sensors exposed to a bioprocess. Twenty calibration measurements were performed shortly before inoculation of each of the cultivations. The response for every sensor was averaged and a trend.- line was fitted by linear regression covering all performed cultivations. The drift over a time period of one year for selected sensors is given in Table 2.
  • Fig 4b several sensor signals from the scaled- down rFVIII cultivation are plotted.
  • the higher signal variation compared to the production-scale was a result of installing the instrument close to the reactor outlet.
  • the signals were treated with an adjacent averaging filter to reduce the noise and ease visualisation.
  • State variables and selected sensor signals from a production-scale cultivation are shown in Fig 6.
  • the plotted signals were treated with the above shift-exclusion algorithm and represent a selection of the signal variation of the 110 different signals derived from the gas sensor array.
  • Fig 6 Some of the sensor response signals in Fig 6 show correlation to the rFVIII concentration in the bioreactor.
  • the off-derivative signal on the other hand follows the rise m specific lactate formation at the end of the cultivation. Correlation towards the viable cell count is apparent m some of the signal responses.
  • the resolution of the sensor signals is not high enough at the be- ginning of the cultivation for accurate on-line cell count estimation. This was probably due to the long distance between the reactor outlet and the electronic nose that was necessary because of the sampling set-up and instrument dimension on production-scale. In the scaled- down rFVIII cultivation's it was possible to resolve small biomass changes even at inoculation cell densities due to the decreased sampling set-up (see Fig 4b) .
  • the gas sensor signals from all six production-scale cultivations, treated with the signal deletion algorithm, with the noise of the signals partially removed by applying a median filter, and with the component correction calibration were used for estimating the concentration of rFVIII .
  • the forward selection procedure described the methods section was used to select the signal parameters with the highest correlation to the rFVIII concentration.
  • the selected signals from five production-scale cultivations were compiled and a neural network was trained against the off-line measured product concentration.
  • the selected sensor signals were MOSFET 3 off-derivative,
  • MOSFET 1 off -derivative
  • MOSFET 8 response MOSFET 3 response.
  • the topology of the neural network was identified by trial and error.
  • the retained neural network model was validated against a cultivation that was unknown to the model.
  • the result of the validation is shown m Fig 7.
  • the total error of estimation was »10% with a mean deviation of 1.1 IU/ml .
  • the rFVIII top value varying as much as 50% between the six cultivations, this is still an acceptable result.
  • the use of additional cultivation data sets for the ANN model training would increase the accuracy of the estimation.
  • sensor signal resolution in production-scale cultivations was not high enough for accurate on-line estimation of the viable cell count the scaled-down rFVIII cultivations allowed this.
  • the data from just two cultivations is not sufficient to cover the variation of this process.
  • the resulting model will therefore not have the same impact as the model established for estimation of product concentration.
  • the sensor responses were treated with the above signal exclusion algorithm and filtered with a median filter. No sensor calibration was performed because the drift was found to be marginal ( «2%) .
  • Sensor signals with correlation to viable cell count were selected using the forward selection procedure.
  • MOSFET 1 response, MOSFET 3 response, Taguchi 2 response, FIS 2 response, FIS 5 on- derivative, and MOSFET 2 off- integral were identified by the selection method.
  • the selected sensor signals from both cultivations were merged and 60% of the data was used to train an artificial neural network against the viable cell count .
  • a suitable neural network topology was identified by trial and error.
  • the final ANN model was validated against the remaining 40% of the data set.
  • a mean deviation of 0.21xl0 6 cells/ml shows that correlation towards the viable cell count is present m the sensor signals.
  • a neural network was trained with the sensor responses from one of the cultivations. The same selected signals were used the model as described above.
  • Fig 8 shows the validation result of the trained model against the second rFVIII cultivation. The total error of estimation was «10% with a mean deviation of
  • G ⁇ pel , W., Schierbaum, K.D. Definitions and typical examples.
  • Trendline fitted sensor responses to bioreactor background at start-up conditions The sensor responses are given in arbitrary units.

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Immunology (AREA)
  • Food Science & Technology (AREA)
  • Analytical Chemistry (AREA)
  • Biochemistry (AREA)
  • Pathology (AREA)
  • General Physics & Mathematics (AREA)
  • Medicinal Chemistry (AREA)
  • Combustion & Propulsion (AREA)
  • Molecular Biology (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Hematology (AREA)
  • Urology & Nephrology (AREA)
  • Measuring Or Testing Involving Enzymes Or Micro-Organisms (AREA)

Abstract

L'invention concerne un procédé de détection précoce d'une infection microbienne indésirable dans un milieu de culture cellulaire, au moyen d'un nez électronique doté d'un groupe de capteurs de gaz et possédant une fonction de reconnaissance de modèles. Ledit procédé consiste à : prélever un échantillon de dégagement gazeux contenant des composés volatils, dans le milieu de culture cellulaire d'un bioréacteur ; à envoyer ledit échantillon auxdit capteurs de gaz du nez électronique et à comparer le modèle de signal de capteur de gaz spécifique, quel qu'il soit, à des modèles de signal standard correspondant à différents micro-organismes. Ledit échantillon de dégagement gazeux est prélevé dans un milieu de culture de cellules animales et la détection d'une infection microbienne, quelle qu'elle soit, s'effectue environ 1-50 h plus tôt pour chaque micro-organisme que dans les techniques connues.
PCT/SE2000/002218 1999-11-16 2000-11-13 Procede de detection de micro-organismes contaminants WO2001036664A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
AU15651/01A AU1565101A (en) 1999-11-16 2000-11-13 A method for detecting contaminating microorganisms

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
SE9904125A SE518174C2 (sv) 1999-11-16 1999-11-16 Detektionsförfarande av oönskad mikrobiell infektion i animaliecellodling
SE9904125-3 1999-11-16

Publications (1)

Publication Number Publication Date
WO2001036664A1 true WO2001036664A1 (fr) 2001-05-25

Family

ID=20417722

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/SE2000/002218 WO2001036664A1 (fr) 1999-11-16 2000-11-13 Procede de detection de micro-organismes contaminants

Country Status (3)

Country Link
AU (1) AU1565101A (fr)
SE (1) SE518174C2 (fr)
WO (1) WO2001036664A1 (fr)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009086309A3 (fr) * 2007-12-27 2009-10-01 Baxter International Inc. Procédés de culture de cellules
WO2011000572A1 (fr) 2009-07-02 2011-01-06 Patenthandel Portfoliofonds I Gmbh & Co. Kg Procédé et dispositif permettant de mettre en évidence des effets biologiques à long terme dans des cellules
WO2012076431A1 (fr) * 2010-12-06 2012-06-14 Syngenta Limited Détecteur de pathogènes
CN103695306A (zh) * 2013-12-19 2014-04-02 兰州大学 多样本土壤呼吸测定贴膜
CN106996965A (zh) * 2017-05-02 2017-08-01 华中农业大学 稻米霉菌在线监测系统、建立方法及应用
US10144957B2 (en) 2002-03-12 2018-12-04 Enzo Life Sciences, Inc. Optimized real time nucleic acid detection processes
WO2019072352A3 (fr) * 2017-10-09 2019-08-15 Lachlak Nassira Automate de détection des bactéries incriminées dans les infections ou maladies grâce à un système multi-capteurs intégrant une olfactométrie de reconnaissance des métabolites dégagés
CN113993986A (zh) * 2019-06-14 2022-01-28 环球生命科学解决方案运营英国有限公司 监测细胞扩增的改进和与监测细胞扩增相关的改进

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2102947A (en) * 1981-07-16 1983-02-09 Neuhaus Pharmaglas Process and apparatus for indicating the presence of contaminating microorganisms
JPS60130398A (ja) * 1983-12-16 1985-07-11 Kurita Water Ind Ltd 微生物検出方法
WO1996039533A1 (fr) * 1995-06-05 1996-12-12 Akzo Nobel N.V. Dispositif et procede de detection de micro-organismes
WO1997008337A1 (fr) * 1995-08-25 1997-03-06 Unipath Limited Procedes et appareil de detection de micro-organismes
US5814474A (en) * 1996-07-23 1998-09-29 Becton Dickinson And Company Direct identification of microorganisms in culture bottles

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2102947A (en) * 1981-07-16 1983-02-09 Neuhaus Pharmaglas Process and apparatus for indicating the presence of contaminating microorganisms
JPS60130398A (ja) * 1983-12-16 1985-07-11 Kurita Water Ind Ltd 微生物検出方法
WO1996039533A1 (fr) * 1995-06-05 1996-12-12 Akzo Nobel N.V. Dispositif et procede de detection de micro-organismes
WO1997008337A1 (fr) * 1995-08-25 1997-03-06 Unipath Limited Procedes et appareil de detection de micro-organismes
US5814474A (en) * 1996-07-23 1998-09-29 Becton Dickinson And Company Direct identification of microorganisms in culture bottles

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
DATABASE WPI Week 198534, Derwent World Patents Index; AN 1985-206559, "Counting the number of microorganisms in a sample - by adding sample to culture medium, incubating, detecting gas generated and estimating microorganisms from detection time" *
PRADYUMNA K. NAMDEV ET AL.: "Sniffing out trouble: Use of an electronic nose in bioprocesses", BIOTECHNOL. PROG., vol. 14, 1998, pages 75 - 78, XP002937209 *

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10144957B2 (en) 2002-03-12 2018-12-04 Enzo Life Sciences, Inc. Optimized real time nucleic acid detection processes
WO2009086309A3 (fr) * 2007-12-27 2009-10-01 Baxter International Inc. Procédés de culture de cellules
EP2574677A1 (fr) * 2007-12-27 2013-04-03 Baxter International Inc. Procédés de culture de cellules
US9359629B2 (en) 2007-12-27 2016-06-07 Baxalta Incorporated Cell culture processes
EP3255152A1 (fr) * 2007-12-27 2017-12-13 Baxalta GmbH Procédés de culture de cellules
WO2011000572A1 (fr) 2009-07-02 2011-01-06 Patenthandel Portfoliofonds I Gmbh & Co. Kg Procédé et dispositif permettant de mettre en évidence des effets biologiques à long terme dans des cellules
WO2012076431A1 (fr) * 2010-12-06 2012-06-14 Syngenta Limited Détecteur de pathogènes
CN103695306A (zh) * 2013-12-19 2014-04-02 兰州大学 多样本土壤呼吸测定贴膜
CN106996965A (zh) * 2017-05-02 2017-08-01 华中农业大学 稻米霉菌在线监测系统、建立方法及应用
WO2019072352A3 (fr) * 2017-10-09 2019-08-15 Lachlak Nassira Automate de détection des bactéries incriminées dans les infections ou maladies grâce à un système multi-capteurs intégrant une olfactométrie de reconnaissance des métabolites dégagés
CN113993986A (zh) * 2019-06-14 2022-01-28 环球生命科学解决方案运营英国有限公司 监测细胞扩增的改进和与监测细胞扩增相关的改进

Also Published As

Publication number Publication date
SE518174C2 (sv) 2002-09-03
SE9904125D0 (sv) 1999-11-16
AU1565101A (en) 2001-05-30
SE9904125L (sv) 2001-05-17

Similar Documents

Publication Publication Date Title
Haugen et al. Application of gas-sensor array technology for detection and monitoring of growth of spoilage bacteria in milk: A model study
Gardner et al. The prediction of bacteria type and culture growth phase by an electronic nose with a multi-layer perceptron network
USRE38186E1 (en) Method and apparatus for detecting microorganisms
Peris et al. On-line monitoring of food fermentation processes using electronic noses and electronic tongues: A review
McEntegart et al. Detection and discrimination of coliform bacteria with gas sensor arrays
Pavlou et al. An intelligent rapid odour recognition model in discrimination of Helicobacter pylori and other gastroesophageal isolates in vitro
Holmberg et al. Bacteria classification based on feature extraction from sensor data
Tudor Kalit et al. Application of electronic nose and electronic tongue in the dairy industry
Conway et al. Phyloproteomics: species identification of Enterobacteriaceae using matrix-assisted laser desorption/ionization time-of-flight mass spectrometry
Gobbi et al. Electronic nose and Alicyclobacillus spp. spoilage of fruit juices: An emerging diagnostic tool
US20070003996A1 (en) Identification of bacteria and spores
Spinelli et al. Emission of volatile compounds by Erwinia amylovora: biological activity in vitro and possible exploitation for bacterial identification
US5814474A (en) Direct identification of microorganisms in culture bottles
WO2001036664A1 (fr) Procede de detection de micro-organismes contaminants
Bastos et al. Potential of an electronic nose for the early detection and differentiation between Streptomyces in potable water
Bachinger et al. Gas sensor arrays for early detection of infection in mammalian cell culture
Lechner et al. Diagnosis of bacteria in vitro by mass spectrometric fingerprinting: a pilot study
Cheung et al. Discrimination of bacteria using pyrolysis-gas chromatography-differential mobility spectrometry (Py-GC-DMS) and chemometrics
Bachinger et al. Electronic nose for estimation of product concentration in mammalian cell cultivation
Rosenthal et al. Volatile atmospheric pressure chemical ionisation mass spectrometry headspace analysis of E. coli and S. aureus
Timmins et al. Rapid quantitative analysis of binary mixtures of Escherichia coli strains using pyrolysis mass spectrometry with multivariate calibration and artificial neural networks
Kai et al. Sampling, detection, identification, and analysis of bacterial volatile organic compounds (VOCs)
Sakhamuri et al. Simultaneous determination of multiple components in nisin fermentation using FTIR spectroscopy
Custer et al. Potential of on‐line CIMS for bioprocess monitoring
Clemente et al. Predicting sporulation events in a bioreactor using an electronic nose

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AE AG AL AM AT AT AU AZ BA BB BG BR BY BZ CA CH CN CR CU CZ CZ DE DE DK DK DM DZ EE EE ES FI FI GB GD GE GH GM HR HU ID IL IN IS JP KE KG KP KR KR KZ LC LK LR LS LT LU LV MA MD MG MK MN MW MX MZ NO NZ PL PT RO RU SD SE SG SI SK SK SL TJ TM TR TT TZ UA UG US UZ VN YU ZA ZW

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): GH GM KE LS MW MZ SD SL SZ TZ UG ZW AM AZ BY KG KZ MD RU TJ TM AT BE CH CY DE DK ES FI FR GB GR IE IT LU MC NL PT SE TR BF BJ CF CG CI CM GA GN GW ML MR NE SN TD TG

121 Ep: the epo has been informed by wipo that ep was designated in this application
DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
REG Reference to national code

Ref country code: DE

Ref legal event code: 8642

122 Ep: pct application non-entry in european phase
点击 这是indexloc提供的php浏览器服务，不要输入任何密码和下载