Reagentless optical nsing of glutamine using a
dual-emitting glutamine-binding protein
南汇嘴观海公园
Leah Tolosa,Xudong Ge,and Govind Rao *
Department of Chemical and Biochemical Engineering,University of Maryland,1000Hilltop Circle,Baltimore County,Baltimore,MD 21250,USA
Received 3June 2002
Abstract
Glutamine is a major source of nitrogen and carbon in cell culture media.Thus,glutamine monitoring is important in bioprocess control.Here we report a reagentless fluorescence nsing for glutamine bad on the Escherichia coli glutamine-binding protein (GlnBP)that is nsitive in the submicromolar ranges.The S179C variant of GlnBP was labeled at the –SH and N-terminal positions with acrylodan and ruthenium bis-(2,20-bipyridyl)-1,10-phenanthroline-9-isothiocyanate,respectively.The acrylodan emission is quenched in the prence of glutamine while the ruthenium acts as a nonresponsive long-lived reference.The apparent binding谋的成语
constant,K 0
d ,of 0:72l M was calculated from th
e ratio o
f emission intensities of acrylodan and ruthenium ðI 515=I 610Þ.The prence of the long-lived ruthenium allowed for modulation nsin
g at lower frequencies (1–10MHz)approaching an accuracy of Æ0:02l M glutamine.Dual-frequency ratiometric nsing was also demonstrated.Finally,the extraordinary nsitivity of GlnBP allows for dilution of the sample,thereby eliminating the effects of background fluorescence from the culture media.Ó2003Elvier Science (USA).All rights rerved.
ABC transporters are a superfamily of proteins re-sponsible for the active transport of various biochem-ical substances such as ions,amino acids,or sugars in archaea,prokaryotes,and eukaryotes [1].In gram-negative bacteria the types of transport systems are known as binding protein-dependent permeas.The systems consist of a soluble binding protein in the periplasmic space that generally has a micromolar or submicromolar binding affinity for its substrate and two or more pr
oteins as membrane-bound receptors [2].The binding protein shuttles the substrate to the membrane-bound receptors that then internalize the substrate using ATP as the energy source.The soluble binding proteins in the systems have been utilized as potential nsors for various analytes including gluco [3–5],malto [6,7],phosphate [8],and glutamine [9].The main advantage of the binding proteins as nsors is that unlike enzymes,they do not require additional reagents [10].The key event that accompanies molec-ular recognition between protein and substrate is a conformational change.Fig.1provides a schematic
reprentation of this conformational change and how it is exploited in the design of an optical nsor.This figure also shows the principle of analysis for the ratiometric fluorescence-nsing technique described below.
In this paper,we report our efforts to develop a glu-tamine bionsor from an S179C variant of the Esc-herichia coli glutamine-binding protein (GlnBP).1Previous studies have shown that labeling the cysteine in position 179with a polarity-nsitive probe such as acrylodan or anilino-naphthalene sulfonate (ANS)re-sults in changes in the fluorescence properties of the probes in respon to glutamine [9].Although functional in this form,the methodology is far from optimal for a practical and low-cost glutamine nsor.The short life-times of the probes in the nanocond range require higher
frequencies ($100MHz)and more sophisticated instrumentation to detect the glutamine-induced
双系统怎么装lifetime
Analytical Biochemistry 314(2003)199–205
/locate/yabio
ANALYTICAL BIOCHEMISTRY
*
Corresponding author.
E-mail address:grao@umbc.edu (G.Rao).
1
Abbreviations ud:GlnBP,glutamine-binding protein;ANS,anilino-naphthalene sulfonate;NIR,near-infrared;SDS–PAGE,so-dium dedecyl sulfate–polyacrylamide gel electrophoresis;DMEM,Dulbecco Õs modified Eagle medium;PBS,phosphate-buffered saline;FIA,flow injection analysis.
0003-2697/03/$-e front matter Ó2003Elvier Science (USA).All rights rerved.doi:10.1016/S0003-2697(02)00586-9谦辞
changes.To circumvent the difficulties a method of nsing was devid where an external reference —a long-lived metal-ligand complex —was added to the solution or applied to the walls of the cuvette.The combined emission of the labeled protein and the metal-ligand complex allowed for the detection of modulation chan-ges at lower frequencies [5,11].We describe here an improvement on this technique by covalently linking the metal-ligand complex directly to the protein while maintaining its glutamine responsiveness (Fig.1).The resulting dual-emitting protein can be ud not only for low-frequency modulation nsing ð<10MHz Þbut also for low-cost ratiometric nsing [12,13].
Glutamine nsing is very important in small and large-scale bioprocess involving eukaryotic cell cul-ture.Glutamine is a major nitrogen and carbon source in tissue culture media,and is considered t
ogether with gluco as a limiting factor in cell growth and product yield [14–16].Additionally,unfavorable levels of gluta-mine can lead to the deleterious production of ammonia,which is toxic to cell cultures [17–19].Monitoring of glutamine concentrations is therefore an esntial aspect of process control.
Currently available glutamine bionsors rely on en-zymes such as glutamina (EC 3.5.1.2)in combination with glutamate oxida (EC 1.4.3.11).Glutamate oxi-da is required to suppress the interference of glutamic acid to the measurements [20–22].In another assay,glutamine reacts with three different enzymes to produce NADH,which is then determined spectrophotometri-cally [23].High-pressure liquid chromatograpy [24]and LC-MS-MS [25]have been ud,but the techniques both require expensive instrumentation.Near-infrared (NIR)spectroscopy allows for noninvasive quantifica-tion of glutamine but requires the generation of an elaborate calibration model [26–28].
风娃娃教学反思The glutamine binding protein that was ud in this report has none of the disadvantages of currently available glutamine bionsors and nsing techniques.GlnBP is not an enzyme.Thus,it does not consume glutamine or require other ‘‘reagents’’for its activity.
尼古丁英文
The primary signal transduction mechanism involves a change in conformation (Fig.1)from an ‘‘open’’to a ‘‘clod’’structure when glutamine is questered in the binding site [29,30].This change in conformation is easily detected with polarity-nsitive fluorescent probes.Additionally,it is highly lective for glutamine over other amino acids and has a submicromolar bind-ing affinity ðK d ¼0:2l M Þ[9].Materials and methods GlnBP expression and isolation
Dr.Joph Lakowicz generously provided the plas-mid containing the S179C variant of GlnBP.Trans-formation and expression of the protein was carried out li strain HB101.To relea the periplasmic GlnBP,cells from 150-ml overnight cultures were first pelleted by spinning for 10min at 6000g and resus-pended in 3ml of deionized water.Three ml of chlo-roform was added to the cells,which were vortexed briefly and incubated for 10min at room temperature.Then 12ml of 20mM phosphate buffer,pH 7.5,was added.Samples were vortexed briefly and spun for 20min at 6000g .The supernatant containing the peri-plasmic proteins was then decanted or pipetted into a clean sterile plastic tube.The amount of S179C GlnBP was estimated to be >80%of the total periplasmic extract after sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE)and staining with bril-liant blue G (Sigma–Aldrich,St.Louis,MO).Reaction of the crude extract with 6-acryloyl-2-dimethylamino-naphthalene or acrylodan (Molecular Probes,Eugene,OR)followed by SDS–PAGE re
vealed clo to 100%labeling of the GlnBP with almost no detectable la-beling of the other proteins.This was determined by illuminating the unstained gel on a UV box to show the fluorescent band corresponding to the dye-labeled protein.A prestained standard protein ladder allowed for the estimation of the molecular weight.Thus,
no
Fig.1.Schematic reprentation of the principle of analysis.The acrylodan is covalently attached to a cysteine mutation on the opposite side of the glutamine-binding site.In the ‘‘open’’or glutamine-free conformation the protein shields the acrylodan while in the ‘‘clod’’or glutamine-bound conformation it is expod to the solvent.This is obrved as a decrea in the acrylodan intensity.In both cas,the ruthenium is unaffected by the conformational changes thereby rving as a long-lived ref
erence.
200L.Tolosa et al./Analytical Biochemistry 314(2003)199–205
further purification of the periplasmic extract was necessary.
Fluorophore coupling
The lone cysteine in S179C GlnBP was labeled with acrylodan as described in [9].The labeled protein was parated from free dye by gel-permeation chromatog-raphy on a Sephadex G-25column eluted with phos-phate-buffered saline,pH 7.5.The N-terminal of the acrylodan-labeled GlnBP was lectively labeled with ruthenium bis-(2,20-bipyridyl)-1,10-phenanthro-line-9-isothiocyanate by maintaining the pH at 7.5.Ruthenium bis-(2,20-bipyridyl)-1,10-phenanthroline-9-isothiocyanate was prepared as previously reported in [31].The dual-labeled protein (Ru-GlnBP-Acr)was collected by elution from a Sephadex G-25column.Fluorescence measurements
Steady-state emission spectra were recorded on a Varian Cary Eclip spectrofluorimeter (Varian Instru-ments,Walnut Creek,CA).Time-resolved luminescence decays were measured on a frequency domain fluorim-eter (ISS-Koala,Champaign,IL)with the following modifications.Blue LED LNG992CF
BW (Panasonic,Secaucus,NJ)driven by a current source was ud as the excitation source.The modulation voltage was applied through bias T.The standard radiofrequency amplifier for the photomultiplier tubes was replaced with a ZHL-6A (Mini-circuits,Brooklyn,NY)to enhance the low-frequency performance.The excitation light was filtered by 500-,550-,and 650FL07short-wave pass filters (Andover,Salem,NH).The emission light was filtered by a 500FH90long-wave pass filter (Andover).Lumi-nescence decay data were analyzed by nonlinear least-squares methods.
Theory of modulation nsing
The theory of modulation nsing for combinations of short and long-lived luminophores is described in detail in [11].For a mixture of fluorophores,the pha and modulation can be calculated using the sine and cosine transforms of the intensity decays,N x and D x ,respectively,at a given frequency x
N x ¼X f i m i sin /i ;ð1ÞD x ¼
X
f i m i cos /i ;
ð2Þ
where f i is the fractional steady-state intensity,/i is the pha,and m i is the modulation.The modulation at frequency x is given by
m ¼N 2ÀþD 2Á1=2
:ð3Þ
In the ca of Ru-GlnBP-Acr,the difference in life-time between Ru and Acr is large.It is,therefore,rea-sonable to assume that a frequency x can be identified
where the modulation of Acr is clo to 1.0while the modulation of Ru is clo to 0.0.At this frequency,N ¼f Acr sin /Acr ;ð4ÞD ¼f Acr cos /Acr :
ð5Þ
Using Eq.(3)we arrive at m ¼f Acr :
ð6Þ
This final conclusion proves that the modulation of a probe emitting both short and long-lived compo
nents is the fractional intensity of the short-lived component.The implication is that signal transduction need not be accompanied by lifetime changes but can be limited to intensity changes of the short-lived component.Addi-tionally,the modulation changes are obrved at fre-quencies lower than tho required if the only emitting species is the short-lived dye as reported in [11].The lower frequencies allow for the design of simpler,low-cost instrumentation.
Results
The absorbance spectra of Ru-GlnBP-Acr and GlnBP-Acr are shown in Fig.2.Total protein concen-tration for both samples determined by the brilliant blue G-perchloric acid colorimetric assay (Sigma
Diagnostics,
L.Tolosa et al./Analytical Biochemistry 314(2003)199–205201
St.Louis,MO)is 9:0l M.The calculated amount of protein-bound acrylodan bad on an extinction coeffi-cient of 20,000cm À1M À1[32]is 7.8and 7:5l M in GlnBP-Acr and Ru-GlnBP-Acr,respectively,or about 86%labeling efficiency.This is consistent with the esti-mated >80%GlnBP in the periplasmic extract after SDS–PAGE as described in the experimental ction.The abnce of cysteine-containing proteins in the periplasm is an advantage in this ca becau further purification of the periplasmic fluid is not absolutely necessary.The calculated amount of protein-bound ruthenium bad on an extinction coefficient of 15,000cm À1M À1[31]is 12:0l M.This amount is more than the expected quantity of bound ruthenium if only the N-terminal of GLnBP was available at the reaction pH of 7.5.However,the prence of other proteins in the periplasmic extract may explain this larger value.Nonetheless,the prence of the ruthenium-labeled proteins does not interfere with the glutamine mea-surements becau the role of ruthenium is as a nonre-sponsive reference.A possible drawback is that the acrylodan:ruthenium ratio may change from batch to batch and may require an initial calibration step before measurement.Further purification of the crude peri-plasmic extract will rectify this but may spell the dif-ference between a low-cost and a more expensive nsor.
Fig.3shows the emission spectra of Ru-GlnBP-Acr in the prence of glutamine.This figure eks to illus-trate two things.The fluorescence intensity of acrylodan ðk max ¼515nm Þdecreas with increasing concentration of glutamine as obrved previously [9].Concurrently,the luminescence intensity of Ru remains constant at k max ¼610nm.There are very few examples of small-molecule fluorescent dyes that show ratiometric emis-sion in the prence of an analyte [33–35].In fact,the design and synthesis of ratiometric probes are major challenges in probe chemistry.Ratiometric dyes are desirable becau factors such as the concentration of the dye,the intensity of the light source,the path length,and sample positioning are internally corrected.Here we show that this is possible with a dual-emitting protein nsor.Secondly,the standard solutions of glutamine (0.1to 6.4mM)were prepared in glutamine-free Dul-becco Õs modified Eagle media (DMEM).DMEM is the most commonly ud media for tissue culture.The re-sulting solutions were then diluted 1000Âwith PBS and ud in the assay.The submicromolar nsitivity of the protein for glutamine remains virtually unchanged
from
202L.Tolosa et al./Analytical Biochemistry 314(2003)199–205
that obrved with the acrylodan label alone [9].This proves that the Ru label in the N-terminal has no effect on the binding activity of the protein.More impor-tantly,the other components prent in the complex DMEM media did not interfere with the binding of glutamine.The int in Fig.3is the ratio of the inten-sities of acrylodan and Ru ðI 515=I 610Þas a function of glutamine concentration.The apparent binding constant K 0d for a single binding site was calculated from the data to be 0:72Æ0:10l M glutamine.This is clo to the reported 100–300nM dissociation constant for the wild type [10].应急救援学
Frequency-domain intensity decay data are shown in Fig.4.The data were fit by the least-squares method to a biexponential decay to obtain the approximate lifetimes and fractional intensities of acrylodan and ruthenium as listed in Table 1.As expected,the lifetime of acrylodan decread from 2.4to 1.7ns with the binding of gluta-mine while the lifetime of Ru remained practically the same at about 685ns.The ratio of the fractional inten-sities of acrylodan to Ru decread from 1:1to 1:2.The modulation at frequencies 0.1, 2.5, 4.0,and 10MHz is plotted as a function of glutamine concen-trations in Fig.5.Modulation measurements are easily accurate to Æ0:01[11],which leads to an accuracy of Æ0:02l M glutamine or approximately 1.4parts per 10billion.Although this is quite impres
sive,in real-life situations,modulation measurements have the disad-vantage of requiring careful shielding from ambient light.This can be remedied by calculating the ratio of the modulation at two frequencies:(1)the frequency where no modulation change is detected in the prence of analyte and (2)the frequency where the bionsor emission is modulated.In the data prented here,0.1MHz is practically constant at all glutamine con-centrations,while the data at 2.5,4.0,and 10MHz are responsive to glutamine levels.The int in Fig.5shows the plot for the modulation ratios.
Discussion
Increasing demand for the production of important biological products by eukaryotic cell cultures has in-tensified efforts in the development of nsing devices for monitoring nutrient levels,available oxygen,and cell density in bioreactors.Glutamine together with gluco is an esntial nutrient that needs to be con-trolled in order to maximize product formation.Here we showed an extremely nsitive glutamine nsor bad on the glutamine-binding protein li豆浆的英文
.
Table 1
Calculated lifetimes,s ,and fractional intensities,f ,for acrylodan and ruthenium in Ru-GlnBP-Acr whe
n the data in Fig.3are fit to a biexponential decay Glutamine,l M s Ru ,ns f Ru s Acr ,ns f Acr 0.06760.490 2.360.5100.16830.539 2.250.4760.26870.585 2.110.4150.46820.626 1.900.3740.86930.653 1.780.3471.6
688
0.669
1.66
0.331
L.Tolosa et al./Analytical Biochemistry 314(2003)199–205203
