Articles & News Stories

Design and scaleup of downstream processing of monoclonal antibodies for cancer therapy: from research to clinical proof of principle

Design and scaleup of downstream processing of monoclonal antibodies for cancer therapy: from research to clinical proof of principle
of 14
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Design and scaleup of downstream processing of monoclonalantibodies for cancer therapy: from research to clinical proof of principle Alberto L. Horenstein*, Federico Crivellin, Ada Funaro,Marcela Said, Fabio Malavasi  Laboratory of Immunogenetics, Department of Genetics, Biology and Biochemistry, University of Torino Medical School, Turin, ItalyCenter for Experimental Research and Medical Studies (CeRMS), San Giovanni Battista Hospital, Turin, Italy Received 12 September 2002; received in revised form 18 December 2002; accepted 19 December 2002 Abstract Murine monoclonal antibodies (mAb) from cell culture supernatants have been purified in order to acquire clinical grade for in vivo cancer treatment. The starting material was purified by high performance liquid chromatography (HPLC) systemsranging from the analytical scale process to a scaleup to 1 g per batch. Three columns (Protein A affinity chromatography withsingle-step elution, hydroxyapatite (HA) chromatography followed by linear gradient elution and endotoxin removing-gelchromatography), exploiting different properties of the mAb were applied. The final batches of antibody were subjected to alarge panel of tests for the purpose of evaluating the efficacy of the downstream processing.The resulting data have allowed us to determine the maximum number of times the column can be used and to precisely andthoroughly characterize antibody integrity, specificity, and potency according to in-house reference standards. The optimized bioprocessing is rapid, efficient, and reproducible. Not less importantly, all the techniques applied are characterized by costswhich are affordable to medium-sized laboratories. They represent the basis for implementing immunotherapeutic protocolstransferable to clinical medicine. D  2003 Elsevier Science B.V. All rights reserved.  Keywords:  Purified monoclonal antibody; Clinical grade; Immunotherapy0022-1759/03/$ - see front matter   D  2003 Elsevier Science B.V. All rights reserved.doi:10.1016/S0022-1759(03)00006-1  Abbreviations:  BSA, bovine serum albumin; CEA, carcinoembryonic antigen; CIP, cleaning and sanitization in place; ELISA, enzyme-linkedimmunosorbent assay; EU, endotoxin unit; FPLC, fast protein liquid chromatography; IgG, immunoglobulin G; LAL,  Limulus  amoebocytelysate; LC, liquid chromatography; mAb, monoclonal antibodies; HA, hydroxyapatite; HPLC, high performance liquid chromatography; IEF,isoelectric focusing; IIF, indirect immunofluorescence; FACS, fluorescent activated cell sorter; FDA, US Food and Drug Administration; PBS, phosphate-buffered saline; rProtA, recombinant Protein A; SDS-PAGE, sodium dodecylsulfate polyacrylamide gel electrophoresis; SEC-HPLC,size-exclusion HPLC.* Corresponding author. Laboratory of Immunogenetics, Department of Genetics, Biology and Biochemistry, via Santena 19, 10126 Turin,Italy. Tel.: +39-011-696-1734; fax: +39-011-696-6155.  E-mail address: (A.L. Horenstein) of Immunological Methods 275 (2003) 99–112  1. Introduction Monoclonals are antibodies which are mass-pro-duced in the laboratory to recognize a s pecific molec-ular target  (Ko¨hler and Milstein, 1975). After decadesof disappointing results, monoclonal antibodies (mAb)are finally emerging as viable drugs, and the US Foodand Drug Administration (FDA) has approved 11 of them, to date, to treat cancer and transplant rejectionand to combat autoimmune diseases (Reichert, 2001).At least 400 other mAb are known to be in clinical and preclinical trials (Gura, 2002).Recent developments in the fields of genomics and proteomics have increased the likelihood of discover-ing many antibodies with desirable therapeutic func-tions, but so far, the processing of biotherapeuticsantibodies to meet clinical demand is a major bottle-neck in the biotechnology industry. While severalcompanies have resorted to establishing larger produc-tion capabilities to cope with the production needs,others are opting for contracts with manufacturingorganizations rather than taking the risk of establishingtheir own manufacturing facilities. This paper dealswith our experience as a Research Department involved in setting up a laboratory environment that is fully dedicated to large-scale purification of mAb totumor antigens for diagnostic and therapeutic clinicalapplications (Horenstein et al., 1992; Funaro et al.,1998; Malavasi et al., 1999; Funaro et al., 2000).ThediversityofmAbapplications ismatched bythevariety of purification methods available. From aca-demic laboratories to biotech manufacturing plants,small- and large-scale purification of mAb for clinicalapplications from mouse ascites or cell culture super-natants(Stankeretal.,1985;Darbyetal.,1993)impliesthe use of a combination of various separation techni-ques, the most popular and efficient of which is liquidchromatography (LC). The various retention modes of LC including bioaffinity (Protein A/G, specific antigenormimeticligand),mixed mode(hydroxyapatite (HA),dye ligands), and physicochemical chromatographicmechanisms (size exclusion, ion exchange and hydro- phobic interactions) have been in use for decades, but only recently has the development of new resinsimproved those methods exploiting the physical andchemical properties of the antibodies. In addition tothese properties, an appropriate combination of chro-matographic materials in at least two orthogonal (i.e. based on two different mechanisms) steps, as per FDAspecifications,areimportantfactorsinobtaininghighly pure antibodies on a preparative scale.Purification of mAb by LC generally takes place inthree phases: (i) a capture step, in which the mAb present in hybridoma culture supernatant is separatedfrom other sample components; (ii) an intermediatestep, in which the mAb is isolated from contaminantssimilar in size or other biochemical properties; and (iii)a cleaning step, for the complete removal of tracecontaminants. The procedure should yield a goodamount of product in a reasonable turn-around time,without the need for specialized equipment or harsh purification procedures which may damage or modifyan antibody’s properties. Adhering to these criteria,selected murine hybridoma cell lines were grown insmall-scale bioreactors to acquire a clinical grade product for the therapeutical treatment of tumors. Thestartingmaterial waspurifiedbyathree-stepchromato-graphic procedure combining the resolving power of LC with the precision and speed of high performanceliquid chromatography (HPLC), after establishing theappropriate conditions on a small scale. The basic stepsfor the recovery and purification of the mAb or down-stream processing were: (i) clarification by removal of cells and cell debris by centrifugation and ultracentri-fugation; (ii) concentration by using a tangential flowdevice; (iii) preparative sequential HPLC on immobi-lized recombinant Protein A (rProtA) and hydroxyapa-tite (HA) chromatography; (iv) endotoxin-removinggelchromatographyandvirusclearancebyhydrophilicmembranes; and (v) quality control tests.The clinical applications of the processed mAbhave confirmed the efficacy of the optimized purifi-cation protocol. Indeed, tumor regression devoid of side effects has been observed so far in patients withmalignant glioma who have participated in immuno-therapeutic trials with an anti-tenascin mAb (Grana et al., 2002). 2. Materials and methods 2.1. Monoclonal antibodies All antibodies used in this study were murine mAb produced by conventional procedures (Horenstein et al., 1986, 1987; Malavasi and Albertini, 1992) and  A.L. Horenstein et al. / Journal of Immunological Methods 275 (2003) 99–112 100  harvested from supernatants of hybridomas culturedon a miniPERM bioreactor (Heraeus Instruments,South Plainfield, NJ) according to Falkenberg et al.(1995). Cells were grown in serum-free culturemedium (Hybridoma-SFM, InVitrogen, GIBCO), to prevent risks associated with bovine-derived material,in an atmosphere of 5% CO 2 :95% air at 37  j C at a celldensity range of 3–10  10 6 cell/ml. Once the cells, producing mAb at a concentration of   f 50–100 mgIgG/l, were adapted to those culture conditions, amaster cell bank was generated for each hybridoma.For purification, about 5 l of cell culture supernatant was collected and pooled to obtain a single batch. Analiquot of the batch was submitted to an externalcontractor for a series of controls according to FDAguidelines on requirements for the production andquality control of  mAb of murine origin intendedfor use in humans (FDA, US Department of Healthand Human Services, 1997). The specific targets andthe isotypes of the mAb used to demonstrate thereproducibility and feasibility of downstream process-ing were: (i) anti-tenascin (clone CB-TNX, IgG 1 )mAb (Davico Bonino et al., 1995); (ii) anti-carci- noembryonic antigen (CEA) (clone CB30, IgG 1 )mAb; and (iii) anti-CD38 (clone IB4, IgG 2a  ) mAb(Malavasi et al., 1994). 2.2. Equipment  The chromatographic equipment used was a Beck-man HPLC system (Beckman Instruments, Fullerton,CA), consisting of a Programmable Solvent Module166, a Programmable Detector Module 166 NM set on 280 nm and a solvent delivery system with anautomatic analytical injector and a prep-load pumpaccessory for rapid, large-volume sample loading. Thechromatographic data were collected and processedusing System Gold software (Beckman). For recoverystudies, a Beckman DU 640 spectrophotometer wasemployed to measure UV absorbance. Other mainequipment included a class 2 laminar flow work-station (ICN Biomedicals, Milan, Italy); a floor modelcentrifuge to spin up to 15,000   g   (Beckman J 2-21M/E) with a fixed angle JA-17 and swing-out JS-7.5rotors (Beckman); peristaltic pumps (1–2.5 l/minflow capability and a P-1 from Pharmacia, Uppsala,Sweden). Sodium dodecylsulfate polyacrylamide gelelectrophoresis (SDS-PAGE) and IEF analyses weredone on a Mini-Protean II gel electrophoresis appara-tus (Bio-Rad Laboratories, Hercules, CA) usingPower Supplies Model 3000/300 (Bio-Rad Laborato-ries). 2.3. Reagents Analytical grade Tris base, sodium mono- and di- phosphate, sodium chloride and sodium hydroxidewere provided by Sigma (Sigma-Aldrich, Italy).SDS, TEMED, ammonium persulfate, acrylamideand bisacrylamide were provided by Bio-Rad (Rich-mond, CA, USA). Methanol and ethanol of HPLCgrade were from Sigma-Aldrich. All salts and other chemicals were of analytical grade. Ultra-pure dis-tilled water was obtained using a Milli-Q generatingsystem (Millipore, Bedford, MA, USA). Buffers were prepared with sterile and pyrogen-free distilled water (Bieffe Medital, Grosotto, Italy), filtered through a0.22- A m filter, and degassed prior to use. 2.4. Preparation of the rProtA and HA chromato- graphic columns The rProtA Sepharose Fast Flow resin (IPA-400HC, RepliGen, Needham, MA), designed for lowligand leakage (<5 ng rProtA in 1.0 mg/ml antibody product), in the absence of enterotoxins and residualgram-negative endotoxin (FDA Drug Master Files),was used as bulk support media. It was packed usingan in-house procedure into glass Bio-Scale MT10HPLC columns (Bio-Rad Laboratories, 150-mmheight [h.]  10 mm internal diameter [i.d.]) to givea bed height of 9–10 cm. Briefly, the solution con-taining degassed rProtA was delivered through thecolumn under fast protein LC (FPLC) conditionsfollowed by a packing step by the HPLC pump at room temperature ( f 22  j C).The Macro-Prep ceramic HA BioGel HT (Bio-RadLaboratories) column was prepared by recycling a preparative column (50-mm h.  25-mm i.d., Bio-RadLaboratories) or a Bio-Scale MT10 column (Bio-RadLaboratories) with a solution of HA resuspended in 10mM sodium phosphate buffer of pH 6.7 and packed by the HPLC conditions at a flow rate of 1 ml min  1 for 5 h.The results of the rProtA column packing wereanalyzed by comparison of the retention time of a  A.L. Horenstein et al. / Journal of Immunological Methods 275 (2003) 99–112  101  control antibody and by protein analysis of the non-retained peak, i.e. the flow-through, on the HAcolumn. 2.5. Cleaning in place (CIP) of chromatographicmedia and systems For sanitizing chromatographic media and systems,sodium hydroxide was used after purification cycles.Solutions of sodium hydroxide for rProtA and HAresins, 1 and 0.1 M, respectively, were used with acontact time of 30 min at a flow rate of 1 ml min  1 .The system was filled with 50% (v/v) methanol untilcommencing the purification of the next batch. Thewhole purification process was carried out under sterile conditions and samples were tested for bacterialcontamination at each purification step.The detergent E-Toxa-clean (Sigma) was appliedfor endotoxin removal from glass bottles and multiusetubes. 2.6. Purification of monoclonal antibodies The harvested supernatants deprived of cells andcell debris were concentrated by employing a tan-gential flow ultrafiltration device (Ultrasette, Filtron,Pall Italia) with a 50-kDa cutoff membrane andmaterial stored at 4  j C until purification. Sampleswere purified at room temperature ( f 22  j C), com- bining a rapid and reliable three-step procedure usingHPLC. 2.6.1. First step The concentrated material was passed through anrProtA column equilibrated with 5 column volumes(CV) of binding buffer (1.5 M glycine/3 M NaCl, pH 8.9). The sample, together with 112.6 g/l glycine,175 g/l NaCl, and 3 g/l NaOH to give a pH of 8.9,was loaded onto the column at a 2 ml min  1 flowrate and then the column washed with binding buffer until the absorbance dropped to baseline. Bound Igswere eluted with 3 CV of 0.1 M sodium citrate buffer, pH 3.5, and immediately neutralized with 1M Tris, pH 8, and concentrated/diafiltrated against 10 mM sodium phosphate buffer, pH 6.7 (HA preparative column starting buffer), using a Centri-con Plus-80, MWCO 30k Dalton, concentrator fromMillipore. 2.6.2. Second step Ceramic HA preparative chromatography wasequilibrated at room temperature with 5 CV of 10mM sodium phosphate buffer, pH 6.7, and loadedwith 50 ml of the sample at a concentration of 2–3mg/ml and at a flow rate of 4 ml min  1 . After beingwashed with 5 CVof the same buffer, at which point the UV baseline was stable, the mAb were eluted witha 90-min linear gradient from 10 to 400 mM sodium phosphate buffer, pH 6.7, at a flow rate of 4 mlmin  1 . The pooled HA peak fractions containingthe mAb were concentrated and extensively dialyzedagainst phosphate-buffered saline (PBS), pH 7.2,using a Centricon Plus-80, MWCO 30k Dalton (Milli- pore), and finally filtered through a 0.22- A m dispos-able hydrophilic Posidyne membrane syringe filter (Pall, Ann Arbor, MI, USA, 25-mm diameter).Throughout the above mentioned procedures, IgGconcentration was monitored and measured byabsorbance at 280 nm. Calculations concerning reten-tion times, peak heights and relative peak areas weredone using Gold Beckman software which integratesthe obtained antibody peak and displays the resultsimmediately after the purification run has finished. 2.6.3. Third step The purification of mAb for in vivo applicationsaddresses the removal of contaminants (endotoxin andviral particles). Thus, the scaleup involved a final stepof endotoxin-removing gel chromatography followed by the clearance of contaminants by hydrophilicPVDF membrane filtration (Brandwein and Aranha-Creado, 2000). Endotoxin removal.  Column chromatogra- phy was prepared by packing a 5 ml-bed volumecolumn using as solid phase chromatographic sup- port the depyrogenated resin polimyxin B-Sepharose(Pharmacia) or Actigel ALD (Acticlean Etox R , Ster-ogene Bioseparations, CA, USA), resuspended inPBS, pH 7.2. The mAb preparation was passedthrough, at least twice, under sterile conditions toincrease endotoxin binding efficiency. Briefly, thecolumn was equilibrated with PBS, pH 7.2, andthe filtered sample applied at a flow rate of 1 mlmin  1 under FPLC conditions. At the end of the procedure the column was regenerated by perfusionwith 20 CV of 1 M sodium hydroxide, allowed to  A.L. Horenstein et al. / Journal of Immunological Methods 275 (2003) 99–112 102  stand at 4  j C overnight, and washed with sterile pyrogen-free water until neutrality. Limulus amebocyte lysate (LAL) test.  The presence or absence of endotoxins, pyrogenic lipopo-lysaccharides derived from Gram-negative bacteria,was determined in the final product, and buffers usedin the purification procedures with t he  Limulus  ame- bocyte lysate (LAL) gelation test  (Pearson, 1985)according to the manufacturer’s recommendations(Pyrogent Kit, BioWhittaker, Walkersville, MD). Thedetection limit of the kit was 0.125 endotoxin unit (EU)/ml. Murine DNA control.  The detection of murine DNA in purified samples was performed usinga dot-blot apparatus as described elsewhere (Marianiet al., 1989). The sensitivity of the assay allows thedetection of 5 pg total murine DNA/mg protein. Detection of rProtA in the presence of  monoclonal antibodies.  rProtA was detected by anenzyme-linked immunosorbent assay (ELISA) accord-ing to manufacturer’s (RepliGen) instructions. Briefly,mAb samples at a protein concentration  V 0.5 mg/mlwere twofold serially diluted with 0.5 M acetate, 0.1 M NaCl, pH 3.5, containing 0.1% Tween-20, and incu- bated in microtiter plates coated with chicken anti-rProtein A k  antibodies at room temperature for 30min. At the end of the incubation, the plates werewashed with PBS–Tween-20 and the captured rProtAwere detected by the addition of a biotinylated rabbit anti-ProtA probe. After 30-min incubation and further washing, the biotin probe was reacted with streptavidin peroxidase followed by a colorimetric reaction usingtetramethylbenzidine substrate. The developed color was measured with an ELISA plate reader (Tosoh) at 450 nm. The rProtA contamination was determinedusing a standard curve and the content expressed at thelevel of part per million (ppm) obtained by the ratio between mean rProtA concentration (ng/ml) and theconcentration of the undiluted mAb sample (mg/ml). 2.7. Characterization of purified monoclonal anti-bodies Characterization typically included identification by sodium dodecylsulfate polyacrylamide gel electro- phoresis (SDS-PAGE) of the migration pattern andisoelectrofocusing (IEF) profile, analysis of the purityand integrity by size-exclusion HPLC (SEC-HPLC),evaluation of protein concentration and of the bindingto the target molecule. 2.7.1. Gel electrophoresis Chromatographic fractions were analyzed bySDS-PAGE performed under denaturing conditions(Laemmli, 1970) in the presence and absence of areducing agent using a minigel system (Bio-Rad Lab-oratories) with 12.5% polyacrylamide gels asdescribed elsewhere (Horenstein et al., 1998). Proteins on SDS-PAGE gel were detected by silver nitrate (Bio-Rad Laboratories) or Coomassie brilliant blue R-250staining. 2.7.2. Isoelectrofocusing  The isoelectric point (p  I  ) values of the purifiedmAb were determined by running an isoelectrofocus-ing (IEF) slab gel with a pH range of 3–10 using avertical gel apparatus (Bio-Rad Laboratories). Briefly,the running conditions involved a run time of 1 h at 500 and 750 V for 15 min. The IEF gel was fixed witha solution of 12% (w/v) trichloroacetic acid in deion-ized water for 30 min before staining with Coomassie brilliant blue R-250. Proteins with known p  I   valueswere used as standards to calculate the p  I   of the mAb. 2.7.3. Size exclusion chromatography (SEC) A SEC-HPLC was applied (i) to analyze the purityof the antibodies and (ii) to monitor production andconcentration of antibodies in the loading materialwhich cannot be measured using ultraviolet (UV)absorbance because impurities obscure UV detectionof antibodies. The SEC provided a quantitative meas-urement of the mAb concentration once a standardcurve was produced (Gadowski and Abdul-Wajid,1995). The assay used an UltraSpherogel SEC 3000HPLC column (Beckman, 300 mm [h.]  7.5 mm[i.d.]) equilibrated with 0.1 M potassium phosphatecontaining 0.1 M sodium sulfate, pH 7. The columnwas standardized using gel-filtration standards, whichconsisted of thyroglobulin (670 kDa), gamma glob-ulin (158 kDa), bovine serum albumin (BSA, 66kDa), ovalbumin (44 kDa), ribonuclease A (13.7kDa), and uracil (2 kDa). The flow rate was 1.0 mlmin  1 , detection was absorbance at 280 nm, and the  A.L. Horenstein et al. / Journal of Immunological Methods 275 (2003) 99–112  103
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks