Development of cell-based immunoassays to measure type I collagen in cultured fibroblasts

The International Journal of Biochemistry &Cell Biology 42 (2010) 1808–1815Contents lists available at ScienceDirectThe International Journal of Biochemistry&CellBiologyj o u r n a l h o m e p a g e :w w w.e l s e v i e r.c o m /l o c a t e /b i o c elDevelopment of cell-based immunoassays to measure type I collagen in cultured fibroblastsBrian Jones,Christine Bucks,Patrick Wilkinson,Michael Pratta,Francis Farrell,Pitchumani Sivakumar ∗Immunology Research,Centocor Research &Development Inc.,145,King of Prussia Road,Radnor,PA 19087,USAa r t i c l e i n f o Article history:Received 10February 2010Received in revised form 4June 2010Accepted 16July 2010Available online 23 July 2010Keywords:CollagenExtracellular matrix TGF ␤Fibroblasts Fibrosisa b s t r a c tExcessive deposition of type I collagen by activated fibroblasts is a hallmark of scarring and fibrotic pathologies.Quantitation of collagen I at the protein level is paramount to measure functionally relevant changes during pathological remodeling of the extracellular matrix.We describe two new cell-based assays to directly quantify the amount of collagen I incorporated into the extracellular matrix of primary human lung fibroblasts.Utilizing a monoclonal antibody specific to native human collagen I,we optimized conditions and parameters including incubation time,specificity and cell density to demonstrate dose-dependent induction of collagen I by transforming growth factor beta,as measured by in-cell enzyme linked immunosorbent assay.The results obtained by this assay were mimicked by an “In situ Quantitative Western Blot”on cultured cells using the same antibody.Results from these assays were comparable to those obtained with a commercial assay for collagen I N-propeptide,which is an index of collagen formation.These assays have been optimized for a 96-well format and provide a novel and useful approach for screening of anti-fibrotic agents in vitro.The assays described here also offer a significant improvement in throughput and specificity over conventional methods that primarily measure soluble collagen.© 2010 Elsevier Ltd. All rights reserved.1.IntroductionType I collagen (COL1)is the major fibrillar protein of the extracellular matrix (ECM)of connective tissues (Ricard-Blum and Ruggiero,2005).Although the primary function of COL1is to pro-vide a structural scaffold to maintain tissue integrity (Nimni,1983;Ottani et al.,2001),recent studies identified that the collagenous ECM could be dynamic rather than static (Kadler,2004),and regu-late multiple aspects of cell behavior including adhesion,migration and signal transduction (Heino,2007;Vogel,2001).Mutations in COL1genes,and misregulated assembly and deposition of COL1lead to several debilitating disorders of ECM metabolism,such as fibrosis (Biagini and Ballardini,1989;Raghow,1994),scar-ring (Shoshan,1979;Zhang et al.,1995),osteogenesis imperfecta (Prockop et al.,1993),and Ehlers Danlos Syndrome (de Paepe,1996).To date,most collagen assay methods have measured changes in gene expression of collagen and/or transcriptional regulation of collagen through COL1promoter-reporter assays (Rishikof etAbbreviations:ELISA,enzyme linked immunosorbent assay;NHLF,normal human lung fibroblasts;NHDF,normal human dermal fibroblasts;COL1,collagen type I;TGF ␤,transforming growth factor beta;ECM,extracellular matrix;ICW,in-cell-Western.∗Corresponding author.Tel.:+16102405735;fax:+16108894525.E-mail address:psivakum@ (P.Sivakumar).al.,2005;Yata et al.,2003).The biosynthetic pathway of COL1is highly complex and involves several post-and co-translational modifications such as hydroxylation of proline and lysine residues,glycosylation of hydroxylysine residues and crosslinking of colla-gen monomers resulting in the deposition of collagen fiber bundles in the ECM (Davidson and Berg,1981;Last et al.,1990;Myllyharju,2003).Although steady state mRNA levels are a useful predictor of COL1alterations,there is a significant lag between the tran-scription of COL1and the production of functional COL1protein,suggesting that measurement of gene expression may not corre-late with protein production.Quantitation of COL1at the protein level is paramount to measure functionally relevant changes during pathological remodeling of the ECM.Conventionally,the colorimetric determination of hydroxypro-line has been used as an index to measure COL1protein (McAnulty et al.,1991).Additionally,in vitro biosynthesis of collagen can be quantitated by measuring the incorporation of tritiated pro-line into cells (Nacher et al.,1999).Yet,other studies have used a Sircol dye binding assay to quantitate soluble collagens in culture supernatants.However,all of these methods measure only solu-ble collagen.Moreover,these assays are time-consuming,offer less sensitivity and specificity and may require the use of radioactivity.Therefore,there is a need to develop alternate methods to quantify COL1changes at the protein level.Here,we describe two new cell-based assays to measure rel-ative changes in COL1in cultured fibroblasts.These assays were centered on the hypothesis that an antibody specific to native COL11357-2725/$–see front matter © 2010 Elsevier Ltd. All rights reserved.doi:10.1016/j.biocel.2010.07.011B.Jones et al./The International Journal of Biochemistry&Cell Biology42 (2010) 1808–18151809will bind to COL1in the ECM of cultured cells,and such binding can be quantitated by an immunohistochemical method.We identified a commercially available antibody that bound to ECM-associated COL1and specifically detected increased COL1production induced by TGF␤,through an in-cell ELISA method.Specificity and selec-tivity of this antibody were confirmed through solid phase binding assays and immunohistochemistry.Secondly,a quantitative“in-cell-Western”detection method was utilized to confirm the ability of this antibody to measure the induction of COL1protein with TGF␤.These cell-based quantitative assays,validated and opti-mized for a semi high throughput96-well plate format,provide a novel method to screen and identify agents that modulate COL1 assembly in vitro.2.Materials and methods2.1.Proteins and antibodiesMonoclonal(mab6308)and polyclonal(pab292and pab34710) anti-collagen I antibodies and mouse IgG1isotype control were obtained from Abcam(Cambridge,MA).Rabbit IgG isotype con-trol was obtained from Jackson Immunoresearch(West Grove,PA). Fibronectin and rat tail type I collagen were obtained from BD Biosciences(Bedford,MA).Human collagen types I,III and IV and porcine gelatin were purchased from Sigma Aldrich(St.Louis,MO). Collagen I was denatured by heating native collagen for3min,at 65◦C.TGF␤and anti-TGF␤antibody were purchased from R&D Sys-tems(Minneapolis,MN).All cytokines were frozen at−20◦C at a working stock concentration of10␮g/ml;fresh aliquots were used for each experiment.2.2.Fibroblast culturesNormal human lungfibroblasts(NHLFs)and normal human dermalfibroblasts(NHDFs)were obtained from Lonza(CC-2512-NHLF,CC-2511-NHDFs)(Walkersville,MD).Cells were maintained in Fibroblast Growth Media(FGM-2Bullet kit)(Lonza)per man-ufacturer’s instructions,and were used between passages P3and P6.2.3.Cell-based ELISA protocolNHLF and NHDF cells were cultured in inner wells of a 96-well plate,in either FGM-2media or Dulbecco’s Modified Eagle’s Medium containing5%FBS and100units/ml penicillin and 100␮g/ml streptomycin(5%DMEM).Cells were stimulated with TGF␤for24-h,and then re-stimulated with TGF␤in fresh media containing20␮g/ml ascorbic acid for an additional24h to promote collagen synthesis(Booth and Uitto,1981).After stimulation,cells were washed three times in phosphate buffered saline(PBS),fixed in95%ethanol for10min at room temperature,washed twice in PBS,and blocked in PBS containing1%bovine serum albumin(BSA) for2h at room temperature.To keep the cell monolayer intact, plates were inverted and gently tapped dry,followed by manual washing with a multi-channel pipette.Primary antibodies were added in the range1:500to1:2000(in blocking buffer)and incu-bated for2h at room temperature.Cells were then washed four times with PBS/0.05%Tween20.Secondary antibody(peroxidase conjugated anti-mouse IgG,1:5000in blocking buffer)(Jackson ImmunoResearch)was added and incubated for1h at room tem-perature.The plate was washed as before,developed for20min, in the dark,with TMB one solution(BD Biosciences).The assay was stopped by addition of1N sulfuric acid,and the optical density(OD) was read at450nm using a Spectramax Plus plate reader.2.4.In-cell-Western assayNHLF cells were cultured and stimulated as described for the ELISA protocol.After stimulation,cells were washed three times in PBS andfixed in95%ethanol for20min at room temperature. All washing and staining steps were performed using gentle agi-tation on plate rotator.Cells were permeabilized by washing5 times(5min each time)with PBS/0.1%Triton X-100.To keep the cell monolayer intact,plates were inverted and gently tapped dry, followed by manual washing with multi-channel pipette.Plates were blocked with100␮l per well Odyssey Blocking Buffer(LI-COR biosciences,Lincoln,Nebraska),for1.5h at room temperature or overnight at4◦C.Block buffer was removed,and primary antibody (mab6308)or IgG1isotype control(used in the range25–2.5␮g/ml, in blocking buffer)was added for2.5h at room temperature.Plates were washed5-times(5-min per wash)with PBS-0.1%Tween20. Secondary antibody(goat anti-mouse-IR800,LI-COR biosciences) was added for1h at room temperature.To discriminate between live and dead cells,DRAQ5(LI-COR)was added during the sec-ondary incubation period.The plate was washed as before,and then 100␮l PBS was added to each well prior to scanning the plate.Sig-nal intensity of stained cells was acquired using the Odyssey Imager and Odyssey2.1software.2.5.Immunofluorescence microscopyNHLFs were seeded at20,000cells per well on4-well chamber slides and allowed to attach overnight.Cells were stimulated for24-h with TGF␤and then re-stimulated with TGF␤in the presence of 20␮g/ml ascorbic acid for an additional24h.After stimulation,cells were washed three times in PBS,fixed in95%ethanol,and blocked for2h in PBS containing1%donkey serum and0.05%sodium azide. Cells were incubated with mab6308(primary antibody,1:1000 in blocking buffer)for2h at room temperature,Cells were then washed three times in PBS,and incubated for1h at room tempera-ture with Cy3conjugated donkey anti-mouse IgG–antibody(1:200 in blocking buffer)(Jackson ImmunoResearch,West Grove,PA). Washing was repeated as before,the chamber wells were removed, and cells were mounted in anti-fade media(9:1glycerol:PBS,with 1%n-propyl gallate).Images were acquired with the Nikon Eclipse E800Upright Microscope at20×magnification.2.6.Solid phase binding assays96-Well plates were coated with antigens(collagens,gelatin orfibronectin)serially diluted in PBS,for2h at37◦C.Plates were washed twice in PBS and blocked in PBS/0.5%BSA for2h at room temperature with gentle shaking.Blocking buffer was removed and plates were incubated with primary antibody(mab6308,diluted 1:2000in blocking buffer)for2h at room temperature.Plates were washed four times with PBS/0.05%tween-20and incubated with secondary antibody(peroxidase conjugated goat anti-mouse IgG, 1:5000in blocking buffer)for1h at room temperature.Plates were washed four times as before,and TMB one solution was added and incubated for15min in the dark.The reaction was stopped with1N sulfuric acid and plates were read at450nm in a Spectramax plus plate reader.2.7.Radioimmunoassay for procollagen I N-propeptideP1NP levels in NHLF supernatants were measured using a com-mercial Radio Immuno Assay Kit(Immunodiagnostic Systems, Scottsdale,AZ),per the manufacturer’s protocol.1810 B.Jones et al./The International Journal of Biochemistry&Cell Biology42 (2010) 1808–18152.8.Data analysisFor the cell-based ELISA,data was exported from SoftMax Pro 4.2,and imported into EXCEL for calculation of mean OD and normalization.Optical Densities(OD)from secondary only or iso-type controls were used to normalize the data.Data was graphed in GraphPad Prism4and expressed as mean OD±SEM,or fold change over un-stimulated controls.For the in-cell-western,data was exported from Odyssey2.1and imported into EXCEL.Data is reported as integrated intensity(ii800),or normalized for viability as integrated intensity800/integrated intensity700(ii800/ii700). Statistical significance was assessed by one-way ANOVA followed by Student’s–Newman Keuls or Tukey post-test.Differences were considered statistically significant at p≤0.05.3.Results3.1.Identification of a suitable antibody for cell-based ELISAA cell-based ELISA approach has been previously described for the detection and quantification of cytoskeletal proteins such as alpha smooth muscle actin(Cushing et al.,2008).However,to our knowledge a robust cell-based ELISA for COL1does not exist.Since COL1is an ECM protein and the epitopes on native ECM-assembled COL1could be readily available for antibody binding,we hypothe-sized that an antibody to native COL1would bind and quantitatively detect COL1incorporated in the ECM of cultured cells.To address this hypothesis,we cultured normal human lungfibroblasts in the presence of increasing doses of TGF␤,a known pro-fibroticstim-Fig.1.Identification of a suitable antibody for cell-based ELISA.(A–C)The effect of TGF␤on COLI expression in lungfibroblasts was measured by a cell-based ELISA,using three different antibodies(mab6308,pab292and pab34710).(D)dose-dependent induction of COL1with TGF␤,measured by in-cell ELISA with mab6308under optimized conditions(relative quantification).(E)COL1standard curve was generated using plate bound native COL1and mAb6308antibody for detection.(F)Dose response of TGF␤on total amount of COLI measured through cell-based ELISA with mab6308(absolute quantification)(*p<0.05,***p<0.001versus no treatment.One-way ANOVA and Student’s–Neuman–Keul’s post-test).B.Jones et al./The International Journal of Biochemistry&Cell Biology42 (2010) 1808–18151811ulus(McAnulty et al.,1991;Sheppard,2006).Two stimuli were administered24h apart,with the second one in the presence of ascorbic acid,to promote collagen synthesis and deposition(Booth and Uitto,1981).Cells were then processed via ELISA as described in Section2.Three different antibodies were tested in the ELISA–two rabbit polyclonals(pab292and pab34710)and a monoclonal(mab6308) to type I collagen,together with appropriate isotype controls. pab34710and mab6308were raised against full length native COL1 whereas pab292was raised against pepsin soluble COL1.As seen in Fig.1A–C,although all three antibodies detected COL1,only mab6308detected increased COL1as a function of TGF␤treat-ment.We then optimized the dose range of TGF␤for detection with mab6308,and found a linear dose response in the range of 0.05–1ng/ml of TGF␤(Fig.1D).The effect of TGF␤was saturated at 1ng/ml.In all experiments,a no primary(secondary only)control was included for each treatment group and a background sub-traction was performed.Experiments were performed with lung fibroblasts derived from three different donors yielding identical results with TGF␤treatment(data not shown).To determine if the assay could be extrapolated to measure abso-lute increases in COL1,we modified the protocol to include a COL1 standard curve.In this experiment,cells were cultured in the inner wells of a96-well plate and stimulated with a dose range of TGF␤for48-h.Prior tofixation,the outer empty wells were coated with standard rat tail COL1in the range of0–2␮g,and the plate was incubated at37◦C for a further2h.Subsequently,the ELISA was performed on thefixed cells as well as the collagen standard.After adjusting for the blank(no collagen),a linear standard curve within the range of2–125ng of total collagen was observed(Fig.1E).Using the standard curve as a reference,we observed a robust6-fold induction of collagen at the highest dose of TGF␤tested(Fig.1F).3.2.Mab6308is specific for native COL1We next determined if mab6308could specifically detect native ECM-associated COL1,using solid phase binding assays.96-well plates were coated with increasing concentrations of type I,III and IV collagens orfibronectin,and an ELISA was performed with mab6308.As shown in Fig.2A,mab6308bound strongly to COL1 and very weakly to type III,but not to other ECM molecules such as COLIV orfibronectin.To determine specificity to native COL1, similar solid phase binding ELISAs were performed with gelatin or heat denatured COL1.As expected,mab6308specifically bound to native COL1but not to denatured COL1or gelatin(Fig.2B).In addi-tion,immunofluorescent staining was performed with mab6308 on NHLF cells.Consistent with the ELISAfindings,1.0ng/ml TGF␤treatment showed an increased staining for COL1compared to untreated controls.Importantly,most of the staining appeared to be fibrillar suggesting that mab6308detects ECM-associatedfibrillar collagen(Fig.2C).3.3.Optimization of media conditions,time course and cellseeding densitiesAfter validating the specificity and reproducibility of COL1 detection with mab6308,we next performed experiments to opti-mize media and cell seeding conditions for the use of this assay in a96-well semi high throughput format.To determine whether mab6308detection of COL1could be influenced by media conditions,NHLF cells were cultured in one of two commonfibroblast media–DMEM/5%FBS or FGM-2media. Although increase in COL1with TGF␤was detected in both media conditions,the signal and the fold increase was greater in DMEM/5% FBS media as compared to FGM-2(Fig.3A and B).Sincefibrob-lasts accumulate more collagen with increased time in culture,we Fig.2.mab6308specifically detects native COL1.(A)Solid phase binding assays for mab6308against collagens I,III,IV andfibronectin.Plates were coated with indicated proteins,blocked and incubated with mab6308followed by detection with HRP-conjugated anti-mouse IgG.mab6308specifically binds to native COLI and shows either no or negligible binding to other matrix proteins.(B)Binding of mab6308 to native and denatured collagens measured through solid phase assay(note strong binding to only native collagen).(C)Immunofluorescent staining for COL1in human lungfibroblasts using mab6308shows the ability to detectfibrillar ECM-associated collagen(magnification20×,bar=100␮m).determined if lengthening the incubation times with TGF␤could improve the detection window of COL1.However,there was no discernable difference in total COL1in cells stimulated for72-h as compared to48-h(Fig.3C and D).Finally,we determined if the seeding density of thefibroblasts could have an effect on COL1pro-duction and detection.As shown in Fig.4A,the relative amount of COL1produced increased with increasing cell densities.How-ever,the greatest window of detection of TGF␤induced COL1was obtained at a seeding density of5–10×103cells per well(Fig.4B).From these experiments,NHLFs seeded at a density of 5–10×103cells per well in a96-well plate,and stimulated for 48h post-attachment in DMEM/5%FBS media provided the opti-mal conditions to detect an increase in COL1production induced by TGF␤.1812 B.Jones et al./The International Journal of Biochemistry &Cell Biology42 (2010) 1808–1815Fig.3.Effect of culture time and media conditions on the measurement of TGF ␤induced COLI production in human lung fibroblasts using mab6308.NHLF cells were cultured in either DMEM/5%FBS (A)or FGM-2(B)media to determine the effect of growth media on collagen production.Cells were stimulated with TGF ␤for 48h (C)or 72h (D)to determine the impact of culture time on collagen production.(*p <0.05,**p <0.01,***p <0.001versus no treatment.One-way ANOVA and Student’s–Neuman–Keul’s post-test).3.4.TGF ˇinduced COL1production in fibroblasts can be reversed using a neutralizing anti-TGF ˇantibodyIn our experiments,a dose-dependent increase in COL1pro-duction by stimulation with TGF ␤was consistently detected.To confirm the specificity of COL1induction by TGF ␤,ELISA was per-formed after stimulating cells with TGF ␤,or TGF ␤pre-incubated with a neutralizing antibody or an isotype control antibody.As shown in Fig.5,TGF ␤induced increase in COL1could be reversed by the neutralizing TGF ␤antibody dose-dependently.The isotype control antibody did not have an effect on the COL1response.3.5.In-cell-Western (ICW)–an alternate method for COL1detection in fibroblastsTo confirm the utility of mab6308to detect changes in COL1production and validate the cell-based ELISA,we utilized an in-cell-Western (ICW)detection assay.The ICW method is similar to an ELISA,but is potentially more sensitive due to the use of near-infrared conjugated secondary dyes for detection.Additionally,by using two detection dyes with non-overlapping spectral emissions,we can simultaneously determine cell viability and COL1produc-tion and/or the relative quantity of two ECM components.This represents a significant advantage over a traditional ELISA read-out,since the COL1production can be normalized by cell number or other house-keeper of choice.To determine whether mab6308could be used to detect COL1in the ICW assay,NHLFs were cultured for 48h in the pres-ence of an increasing concentration of TGF ␤.After stimulation,the cells were processed as described in Section 2.Analogous to our findings with the cell-based ELISA,mab6308detected TGF ␤induced increase in COL1in NHLFs (Fig.6A and B).Impor-tantly,isotype control antibody did not yield a significant signal above the background stain of secondary IR800dye alone,indi-cating that mab6308specifically detected COL1in the ICW assay.Moreover,the TGF ␤stimulation did not affect cell proliferation,as the dose-dependent induction of COL1was still maintained after normalization to the live cell stain (Fig.6C).The three-fold increase in COL1after TGF ␤stimulation,detected by the ICW assay,was consistent with that observed in the cell-based ELISA,suggesting that either of these methods can be used to detect native COL1,increased as a function of a pro-fibrotic stimu-lus.3.6.Detection of COL1in dermal fibroblasts with the cell-based ELISA approachTo extend the utility of our method to measure COL1in other cells,we performed the cell-based ELISA on dermal fibroblasts using a similar TGF ␤stimulation as with lung fibroblasts.As shown in Fig.7,TGF ␤stimulation caused a dose-dependent increase in COL1production in dermal fibroblasts from the two donors tested.These data suggest that this method can be effectively applied to detect COL1protein changes in fibroblast cells from different organs.B.Jones et al./The International Journal of Biochemistry &Cell Biology 42 (2010) 1808–18151813Fig.4.Effect of initial cell seeding density on the measurement of TGF ␤induced COLI production in NHLFs using mab6308.(A)Dose response to TGF ␤at different seeding densities and (B)fold change in collagen production with TGF ␤at dif-ferent seeding densities (***p <0.001versus no treatment.One-way ANOVA and Student’s–Neuman–Keul’spost-test).Fig.5.Effect of anti-TGF ␤1neutralizing antibody on TGF ␤induced COLI production in NHLFs measured using cell-based ELISA with mab6308.NHLFs were stimulated with TGF ␤or TGF ␤and neutralizing antibody for 48h (***p <0.001versus no treat-ment.One-way ANOVA and Student’s–Neuman–Keul’spost-test).Fig.6.mab6308detects collagen production in stimulated NHLFs in an ICW assay.NHLFs were seeded and stimulated for 48-h as previously described.Mab6308or IgG1isotype control were used for detection of cell associated,native COL1in the ICW assay.(A)Scanned image of the 96-well plate after staining with mab6308or isotype control,shows specificity of COL1detection.(B)COL1production increases after stimulation with TGF ␤(shown as signal intensity).(C)Normalization by cell viability indicates a true increase in COL1on a per-cell basis (*p <0.001,One-way ANOVA and Tukey post-test).parison of collagen ELISA with measurement of procollagen N-propeptideTo determine how our current method compares to other exist-ing methods of collagen measurement,we performed our assay in parallel with a commercial assay that measures soluble propep-tide of COL1(P1NP)in fibroblast culture supernatants.Collagen I propeptides generated upon cleavage of procollagen molecules are a reliable marker of collagen metabolism under normal and patho-1814 B.Jones et al./The International Journal of Biochemistry &Cell Biology42 (2010) 1808–1815Fig.7.mab6308detects collagen production and modulation in dermal fibroblasts.NHDF cells were cultured and stimulated with increasing amounts of TGF ␤for 48h,as described for lung fibroblasts.Changes in COLI production were detected using mab6308(**p <0.01,***p <0.001versus no treatment for each donor.One-way ANOVA and Student’s–Neuman–Keul’s post-test).logical conditions (Orum et al.,1996;Risteli et al.,1995).NHLF cells were seeded and stimulated with a dose range of TGF ␤as described.The cells were fixed and processed for ELISA and the supernatants were collected and assayed for levels of P1NP.As shown in Fig.8,the TGF ␤dose response was comparable between bothmethods.parison of collagen ELISA with P1NP measurement.NHLF cells were cul-tured and stimulated with increasing amounts of TGF ␤for 48h.Cells were processed for ELISA as described and P1NP levels were measured in the supernatants.(A)Detec-tion of ECM COLI with cell-based ELISA and (B)P1NP levels in supernatant (**p <0.01,***p <0.001versus no treatment.One-way ANOVA and Student’s–Neuman–Keul’s post-test).4.DiscussionWe have described two new cell-based immunoassays for the relative and/or absolute quantification of type I collagen in cultured fipared to existing methods,the assays described here are relatively simple,have increased specificity and through-put,and will provide a useful method for screening pro-fibrotic agents in vitro.The methods are based on the hypothesis that an antibody to native COL1molecule would bind to ECM-bound COL1in cultured cells,and that such binding can be quantitated using an immunological method.To this end,we tested several com-mercially available antibodies and identified a mouse monoclonal antibody (mab6308)that bound to COL1in human fibroblasts.Although other antibodies bound COL1,binding of mab6308was selective and sensitive to detect increases in COL1induced by TGF ␤.Since fibroblasts produce a large amount of COL1,high basal levels can significantly hamper the ability to detect measurable dif-ferences with stimuli,more significantly in post-confluent cultures (Grinnell et al.,1989).Therefore it is important to establish optimal conditions under which the basal levels of COL1can be kept at a low but detectable level.We controlled these conditions by identifying the optimal seeding density,limiting the total time in culture,and adding ascorbic acid (a key cofactor for collagen biosynthesis)to post-confluent cells for 24h.Accordingly,we determined that a total stimulus period of 48h with cells seeded at 5–10×103/well in a 96-well plate was an optimal ing these conditions,a three-fold induction of COL1was consistently achieved at the high-est dose of TGF ␤tested.Although initially set up to measure relative amounts of collagen,we extended this method to measure absolute amounts of collagen by including a standard collagen curve.An additional complexity in the quantitation of COL1is the high degree of homology between the different species of collagen that would affect specificity.While over 25different types of collagen have been characterized to date,many of these collagens play sup-portive roles as minor components of the ECM (Kadler et al.,2007).Of the major collagens,type II collagen is exclusively present in car-tilage.Type III collagen is mostly confined to fetal tissues (Kadler et al.,2007)but is known to be upregulated in a variety of fibrotic pathologies (Chesnutt et al.,1997;Yamada et al.,1993;Zhang et al.,1995).In vivo,lung fibroblasts reside in an ECM environment con-sisting predominantly of COL1,together with basement-membrane localized type IV collagen.Further,cultured adult fibroblasts are known to express type I and IV collagen (Lam et al.,2004;Olsen et al.,1988;Shimonishi et al.,2005).Our solid phase assays clearly demonstrate that mab6308binds strongly to type I collagen but not type III or IV –a significant improvement over conventional methods such as hydroxyproline measurement that cannot distin-guish collagen phenotypes.Another advantage of our method is the ability of the antibody to selectively bind to native COL1,but not denatured COL1,as confirmed by solid phase binding assays against heat denatured collagen or gelatin.In summary,our method pro-vides a means to specifically quantify functionally relevant changes in COL1in the ECM.The validity of our findings with the cell-based ELISA approach was confirmed by using an alternate method based on a similar ing mab6308in an ICW assay,we quantified the TGF ␤dose-dependent induction of COL1in lung fibroblasts.The ICW assay supports the data generated with the cell-based ELISA,and also has the additional advantage of assessing cell viability.The utility of the ICW assay has previously been shown for detection of tissue factor in human mononuclear cells,another cell associated protein (Egorina et al.,2005;Egorina et al.,2006).In our studies we normalized COL1production to cell viability by staining lung fibroblasts with both mab6308and the DNA specific dye,DRAQ5.Based on this dual staining,we determined that TGF ␤induced increased COL1production was not simply due to increased cell。