Article pubs.acs.org/ac
Measurement of the Activity of Individual Subunits of Single Molecules of the Tetrameric Enzyme β-Galactosidase Douglas B. Craig,*,† Thomas T. Morris,† and Coleen Marie Q. Ong-Justiniano‡ †
Chemistry Department, University of Winnipeg, Winnipeg, Manitoba, Canada Department of Biochemistry and Medical Genetics, University of Manitoba, Winnipeg, Manitoba, Canada
‡
ABSTRACT: Escherichia coli β-galactosidase was incubated in the presence of the slow-release inhibitor D-galactal for 30 min at a concentration of 70 times its Ki. The sample was then diluted 20000-fold into buffer containing the fluorogenic substrate 9H-(1,3-dichloro-9,9-dimethylacridin-2-on-7-yl) β-D-galactoside, reducing the inhibitor concentration to Ki/280. The sample was subjected to a capillary electrophoresis continuous flow single enzyme molecule assay. As the inhibitor dissociated while the enzyme traveled the length of the capillary, a fraction of molecules showed stepwise increases in activity. This was due to the activation of individual subunits within single molecules. The changes in activity can be largely explained in terms of each molecule containing subunits of indistinguishable activity. rates of 2.7 × 102 M−1 s−1 and 4.6 × 10−3 s−1, respectively. The Ki of D-galactal has been reported to be 14 μM.11 This inhibitor was used by Walt’s group in the first report of the inhibition of single molecules of β-galactosidase.12 In this study, β-galactosidase was incubated in the presence of 1 mM D-galactal, which is approximately 70 times its Ki. After 30 min, the sample was diluted 20000-fold into assay buffer, reducing the inhibitor concentration to Ki/280. Under these conditions inhibitor dissociation occurs with little probability of reassociation for any given enzyme molecule. The sample was immediately assayed at the single-molecule level. We report for the first time the activity of individual subunits on single enzyme molecules, which were observed as the individual subunits became active due to inhibitor dissociation.
S
ingle-molecule assays of different enzymes have demonstrated that individual molecules of a given enzyme have different properties.1−4 Heterogeneous properties include Km,5 Vmax, activation energy of catalysis,2 and electrophoretic mobility.6 Furthermore, the catalytic rate has been found to vary over time for individual molecules, likely due to fluctuations in conformation.7 Escherichia coli β-galactosidase (EC 3.2.1.23) is a 464 kDa tetramer consisting of four identical 1023 amino acid residue subunits and contains four identical active sites. The enzyme is functional only as a tetramer.8 The continuous flow capillary electrophoresis assay has been used to determine the activity of individual molecules of βgalactosidase.9 In this study, this assay was used to provide a continuous read of the catalytic rate of single β-galactosidase molecules over the course of approximately 10 min. Buffer containing the fluorogenic substrate 9H-(1,3-dichloro-9,9dimethylacridin-2-on-7-yl) β-D-galactoside (DDAO-gal) and very dilute enzyme was continuously mobilized through a narrow bore capillary. The enzyme concentration and capillary volume were such that approximately five β-galactosidase molecules were present within the 40 cm long capillary at a time. Since the enzyme and product have different net mobilities, in the buffer used here the former migrating faster than the latter, as the enzyme traversed the capillary it moved away from the product it formed. This resulted in the formation of a smear of fluorescent product. As this smear passed the ultrasensitive detector, a box-shaped peak appeared in the resulting electropherogram. The height of the box-shaped peak at its leading edge represents the activity of the enzyme as it exited the capillary and that at the trailing edge the activity as it entered. The height at any point along the length of the peak represents the activity at the corresponding time between entering and exiting. D-Galactal is a nonclassical competitive inhibitor of βgalactosidase.10 This inhibitor shows slow binding and release © 2012 American Chemical Society
■
METHODS Wild-type E. coli strain 35321 (American Type Culture Collection) was grown for 18 h at 37 °C in 12.29 g/L M9 minimal salts containing 1% (w/v) glycerol, 0.2% (w/v) hydrolyzed casein, 1 mM MgCl2, 50 μM Fe2SO4, and 50 μM ZnCl2. To induce the production of β-galactosidase, stock 25 mM isopropyl β-D-thiogalactoside (IPTG) was added to a final concentration of 250 μM and the cultures further incubated for 90 min. Cells were harvested by centrifugation at 10000g at 4 °C for 10 min. The pellet was resuspended in approximately 0.5 mL of 100 mM N-(2-hydroxyethyl)piperazine-N′-ethanesulfonic acid (HEPES; pH 7.3) containing 2 mM MgCl2 and 2% (v/v) Sigma protease inhibitor cocktail (no metal chelators) and frozen in N2(l). To lyse the cells, the pellet was ground, briefly thawed, and then refrozen. After five freeze/grind/thaw cycles the cold slurry was diluted with 3 mL of cold 100 mM Received: March 20, 2012 Accepted: April 13, 2012 Published: April 13, 2012 4598
dx.doi.org/10.1021/ac300777u | Anal. Chem. 2012, 84, 4598−4602
Analytical Chemistry
Article
Figure 1. Resultant electropherogram from a single-molecule assay of β-galactosidase as the slow-release inhibitor D-galactal dissociates. The initial shift in baseline is due to the start of the elution of the substrate DDAO-gal, and the second is due to the start of the elution of the product DDAO. Subsequent to this five peaks are observed. Each represents the pool of product formed by a single β-galactosidase molecule as it traveled the length of the capillary.
HEPES (pH 7.3) containing 2 mM MgCl2 and 2% (v/v) protease inhibitor cocktail and then centrifuged for 10 min at 10000g to remove cellular debris. The supernatant was passed through a sterile 0.45 μm filter, and the filtrate was diluted with an equal volume of sterile glycerol and stored at −20 °C.13 Assays were performed using an in-laboratory-constructed capillary electrophoresis instrument which utilizes postcolumn laser-induced fluorescence detection within a sheath flow cuvette. The instrument design has been previously published.14 The injection end of a 40 cm long, 5 μm internal diameter, 145 μm external diameter uncoated fused silica capillary (Polymicro Technologies) and a 0.5 mm diameter platinum wire connected to a high-voltage power supply (Spellman model CZE 2000) were placed into a buffercontaining vessel in the injection carousel and held at a positive potential. The detection end of the capillary, from which approximately 1 mm of the external polyimide coating was removed by flame, was inserted into a quartz sheath flow cuvette containing a 250 × 250 μm inner bore (Hellma). The capillary was grounded through the sheath flow buffer within the cuvette. The 10 mW output at 633 nm of a HeNe laser (Melles Griot) was focused using a 6.3×, NA (numerical aperture) = 0.2 microscope objective (Melles Griot) approximately 10 μm below the detection end of the capillary. Emission was collected at 90° using a 60×, NA = 0.7 microscope objective (Universe Kogaku) and passed through a 670DF40 optical filter (Omega Optical) and a slit and onto a photomultiplier tube (Hamamatsu model 1477). The analog signal was collected and digitized using a Pentium 4 computer through a PCI-MIO-16XE I/O board utilizing LabView software (National Instruments) at 10 Hz. The same board was used to control the electrophoresis voltage and photomultiplier tube (PMT) bias. Prior to each assay, 1.7 μL of 24 mM DDAO-gal (Invitrogen/Molecular Probes) was added to 198.3 μL of sample buffer and washed three times with an equal volume of toluene to remove the 7-hydroxy-9H-(1,3-dichloro-9,9-dime-
thylacridin-2-one) (DDAO) that was present as an impurity and would otherwise result in a high background signal. A 125 μL volume of washed substrate was added to 372.5 μL of sample buffer. Stock enzyme was diluted 100-fold into the sample buffer containing a 1 mM concentration of the inhibitor D-galactal. After a 30 min incubation period the enzyme was diluted 100-fold into sample buffer, 2.5 μL was added to the substrate solution, and the sample was immediately assayed. The final sample contained buffer, 50 μM substrate, 50 nM inhibitor, and approximately 1 fM enzyme. The sample was continuously injected into the capillary at 16 kV (injection end positive).
■
RESULTS In the presence of an inhibitor, enzyme molecules are often in either an active or an inactive state. However, with oligomeric enzymes, there remains the possibility of intermediate states due to the binding of inhibitor to one or more monomers and the remainder being free of inhibitor. If such were the case, a stepwise decrease or increase in activity of an individual enzyme molecule might be predicted to occur as inhibitor either binds or is released. Figure 1 shows an electropherogram resulting from the continuous injection of buffer containing the substrate DDAOgal and β-galactosidase as the inhibitor D-galactal slowly dissociated. At 337 s there is a shift in the baseline due to the weakly fluorescent substrate that has started to elute. This is followed by a second baseline shift at 644 s due to the elution of the product DDAO. This DDAO is largely present due to the nonenzymatic hydrolysis of the substrate. Beyond this second shift, box-shaped peaks are observed. These peaks represent pools of product formed as individual enzyme molecules traversed the capillary. The width of the peaks reflects the difference in the electrophoretic mobility of the DDAO product and the enzyme. Differences in these widths are attributed to differences in the electrophoretic mobilities of the individual enzyme molecules. Peak areas represent the total 4599
dx.doi.org/10.1021/ac300777u | Anal. Chem. 2012, 84, 4598−4602
Analytical Chemistry
Article
in Figure 2. A constant activity was observed for a period of time after the enzyme molecule entered the capillary (trailing edge of the peak, see the Discussion), and this was followed by a sudden change to a higher activity which thereafter remained constant until the enzyme molecule exited (leading edge of the peak). This is also observed with the peak eluting at 1250 s in Figure 1. Such peaks are consistent with partially active enzyme molecules releasing one or more inhibitor molecules, resulting in one or more additional active sites becoming active. Whether this shift in activity occurred shortly after the enzyme entered the capillary or shortly before it exited or some time in between appeared to be random. Conversion from a more active to a less active rate was not observed. In the middle trace of Figure 2, two peaks are observed. In the first peak two shifts in activity are observed. This is consistent with two separate releases of inhibitor occurring. Such a peak was observed three times. Since β-galactosidase is a tetramer, a maximum of three shifts from a lower activity to a higher one are possible. However, no such peak was observed. On occasion, two enzyme molecules in relatively close proximity migrate down the capillary, resulting in partially overlapping peaks. The second peak seen in the upper trace of Figure 2 depicts such an instance. In the case of overlapping peaks, the central portion of the peak is elevated, and this height is equal to the sum of the heights of the plateaus at the leading and trailing ends. The peaks referred to in the lower and middle traces cannot be explained in terms of two or more overlapping peaks. In three instances short peaks which were less than half the width of the next narrowest peaks were observed. These may represent molecules which were completely inhibited for the majority of the time they traveled the length of the capillary and only lost one or more of their associated D-galactal molecules shortly before exiting. Assays were also performed in the presence of 1 mM inhibitor. Under these conditions no peaks were observed.
amount of product formed by each enzyme molecule. Differences in these areas represent differences in the catalytic rates of the enzyme molecules.9 In this study the large majority of peak areas fell within a range of 1 order of magnitude. Since DDAO-gal is neutral, the initial shift in the baseline can be used to calculate the electroosmotic flow. The time of the second shift can be used to determine the electrophoretic mobility of DDAO. From these values and the width of each peak, the electrophoretic mobility of each enzyme molecule can be calculated. This can then be used to determine the travel time of each enzyme molecule.9 The average time for a βgalactosidase molecule to travel the length of the capillary in this study was 557 s. Figure 2 shows a portion of the resultant electropherograms from the continuous flow assay of individual molecules of β-
■
Figure 2. Portions of the resultant electropherograms from singlemolecule assays of β-galactosidase as the slow-release inhibitor Dgalactal dissociates. The leading edge of each box-shaped peak represents the activity of the enzyme molecule as it exited the capillary and the trailing edge the activity as it entered. The height of the peak at any point along its length represents the activity at the corresponding time between entering and exiting. In the lower trace, box-shaped peaks for three individual molecules are observed. In the middle peak, there is a shift in the height of the peak due to the activation of putatively a single subunit. In the middle trace, two peaks are observed. In the earlier eluting peak, there are two stepwise changes in peak height putatively due to the sequential activation of two individual subunits. In the upper trace, the later eluting peak represents the overlapping of peaks from two different enzyme molecules which each show a constant activity.
DISCUSSION In Walt’s study,12 β-galactosidase was incubated with D-galactal at a concentration of approximately 5 times its Ki. The sample was then diluted 1000-fold, reducing the D-galactal concentration to Ki/200, and assayed. In the assay individual molecules of β-galactosidase were housed in femtoliter-sized vessels containing the substrate resorufin β-D-galactoside (res-gal). The fluorescent signal emanating from each vessel was monitored over time as the slowly released inhibitor dissociated. It was observed that there was an initial lag phase of up to 1000 s followed by a linear increase in signal over time. These results were explained in terms of stochastic inhibitor dissociation whereupon all inhibitor molecules were simultaneously released from a given molecule. Despite having four distinct active sites, each molecule was found to be either fully active or fully inhibited with no states in between. We employed a buffer system similar to that of Walt,12 who used 1.5 mM KH2PO4, 8.1 mM Na2HPO4, 2.7 mM KCl, 136 mM NaCl, and 1 mM MgCl2 (pH 7.3). In this study the concentrations of KCl and NaCl were reduced by half to reduce the Joule heating that would otherwise occur during electrophoresis. Additionally, the phosphate was replaced with 10 mM HEPES due to the poor solubility of magnesium phosphate. In our initial studies we utilized buffer containing 1 mM citrate, being present due to the washing of the substrate res-gal to remove the resorufin present as an impurity.15 Upon
galactosidase as the inhibitor D-galactal dissociated. In the lower trace, three box-shaped peaks are observed. In the case of the first and last peaks, the variation in signal at the top of the peak is not dissimilar to that of the background, indicating that the catalytic rate remained approximately constant during the travel time through the capillary. This was observed for the majority of the enzyme molecules. This was consistent with the major finding of Walt’s study; all inhibitor molecules dissociate from a given molecule at once, yielding a fully active molecule with a constant activity.12 However, in a minority of cases we observe a shift in activity which was not reported by Walt. Such an occurrence is represented by the middle peak of the lower trace 4600
dx.doi.org/10.1021/ac300777u | Anal. Chem. 2012, 84, 4598−4602
Analytical Chemistry
Article
reflect the activity of a single subunit. In two of the peaks there was a ratio of shifts in signal between the baseline and successive plateaus of approximately 1:1:1 and 2:1:1. These may be explained as a molecule with a single active subunit undergoing the sequential activation of one and then another subunit of similar activity and a molecule with two active subunits having a third and then a fourth subunit of similar activity becoming active. In the former case, this may represent a molecule which had insufficient time to become fully active. The peak shown in the middle trace of Figure 2 shows neither a clear 1:1:1 nor a 2:1:1 pattern and may possibly reflect subunits which have somewhat different activities. On 16 occasions a peak was observed where there was a single shift in activity, as is seen in the middle peak from the bottom trace in Figure 2. On 15 of these occasions the ratios of the shift in height from the baseline to the lower plateau and the shift in height from the lower to higher plateau were approximately 1:1, 2:1, and 3:1. These may reflect molecules with a single active subunit having a second subunit become active or possibly two subunits simultaneously becoming active in a molecule which had two subunits initially active, a molecule with two active subunits having a third become active, and finally a molecule with three active subunits having its fourth and final subunit activate. Molecules which did not become fully active may be represented by peaks showing the 1:1 or 2:1 ratios of shifts in plateau height. These observed relative changes in activity can be explained without invoking the notion of subunits of a given molecule having different activities. However, in one instance there was a relative shift in signal from approximately 5 to 6, which is not consistent with a molecule consisting of four subunits of indistinguishable activity. Upon dissociation of D-galactal, the majority of molecules observed in this study behaved as that reported by Walt’s group,12 who found that all active sites on a given molecule become active simultaneously or at least within the time resolution of the assay. In this study stepwise changes in activity were also observed but only for a minority of the enzyme molecules. The observation that a fraction of the enzyme molecules appeared to show an all-or-none activity with the remainder showing a stepwise change in activity may be due to a population heterogeneity with respect to inhibitor binding. One subpopulation appears to be displaying cooperative inhibitor release and the second stepwise release. In all but three of these instances where stepwise release was observed, a single change in activity was observed. For the large majority of the peaks where a shift in activity was observed, the shift can most simply be explained in terms of subunits within a given tetramer having indistinguishable activities. However, the range of activity for different individual molecules has been reported to be in excess of 10-fold.15 Furthermore, a range of evidence suggests that conformational differences may underlie these differences in catalytic rate.16 That the subunits of a given molecule appear to have similar activities, despite the different molecules having widely different activities, may reflect an adherence to the core premise of the symmetry model for multisubunit proteins.17 That is, for a given molecule, all subunits may need to retain a similar conformation. Differences between more and less active molecules may reflect differences in the conformation that all four subunits adopt.
switching to the substrate DDAO-gal in later studies, the citrate was maintained in the buffer.9 However, in this study the buffer contained no citrate, and this lead to the unexpected finding that the presence of citrate substantially alters the electrophoretic mobility of β-galactosidase, presumably due to its binding to the enzyme. In static incubation assays,1,2 upon being filled with substrate and very dilute enzyme, a capillary is incubated without mobilization of the contents. During this time, and since the enzyme molecules are on average several centimeters apart, product formed by one enzyme molecule does not have time to diffuse to and mix with that from a different enzyme molecule. Rather distinct pools of product accumulate about each enzyme molecule. If the incubation is followed by a brief period of electrophoresis, the enzyme is moved away from its product pool and into fresh substrate. A second incubation can then be performed. Following this, and upon mobilization past the detector, two peaks for each enzyme molecule are observed in the electropherogram, one for each incubation. The peak from the first incubation, having more time for diffusion to occur, is wider and shorter than that from the second. In static incubation β-galactosidase assays using an uncoated capillary with DDAO-gal as substrate and buffer containing 1 mM citrate, the peak from the second incubation exits after that from the first incubation.5 This indicates that, during the electrophoresis period between incubations, the first product pools moved ahead of the enzyme molecule. This demonstrates that the enzyme molecule must have had a more negative electrophortic mobility than the product in the presence of the citrate. In the absence of citrate, the orders of elution of the peaks from the first and second incubations are reversed (data not shown), indicating that under these conditions the electrophoretic mobility of the enzyme is less negative than that of DDAO. In the continuous flow assay reported previously, the presence of citrate led to a box-shaped peak where the earlier eluting edge of the peak corresponded to the activity of the enzyme molecule as it entered the capillary and the later eluting edge that as it exited.9 In the absence of citrate this is reversed. It is for this reason that the box-shaped peaks shown in this study are mirror images of that expected on the basis of studies where citrate was present.13 Walt’s group found that inhibitor dissociation happened within 1000 s of the start of the assay.12 In this study an enzyme molecule entering the capillary exited on average 557 s later. Not withstanding that a different buffer was used which may affect the dissociation rate, one might not predict peaks with shifts in height to be observed in the electropherograms at times past approximately 1600 s. This was the case with the exception of one peak which eluted at approximately 1700− 1750 s. In the earlier eluting peak from the middle trace in Figure 2, the changes in average height from the baseline to the first plateau and between successive plateaus (from longer retention time toward shorter) were 0.078, 0.049, and 0.038. For the other two such peaks observed in this study, the height differences were 0.036, 0.041, and 0.035, as well as 0.097, 0.045, and 0.046. Each shift may reflect a change in activity due to the activation of one subunit. Another possibility is that, for a given molecule, one of the shifts represents the activity of two subunits and the remaining two each the activity of a single subunit. Regardless, and since β-galactosidase is a tetramer, for all three of these peaks, at least two of the shifts in height need 4601
dx.doi.org/10.1021/ac300777u | Anal. Chem. 2012, 84, 4598−4602
Analytical Chemistry
■
Article
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS This project was supported by a grant from the Natural Sciences and Engineering Research Council (NSERC). REFERENCES
(1) Xue, Q.; Yeung, E. S. Nature 1995, 373, 681−683. (2) Craig, D. B.; Arriaga, E. A.; Wong, J. C. Y.; Lu, H.; Dovichi, N. J. J. Am. Chem. Soc. 1996, 118, 5245−5253. (3) Lu, H. P.; Xun, L.; Xie, S. Science 1999, 282, 1877−1882. (4) Edman, L.; Foldes-Papp, Z.; Wennmalm, S.; Rigler, R. Chem. Phys. 1999, 247, 11−22. (5) Craig, D. B.; Haslam, A. M.; Coombs, J. M. L.; Nichols, E. R. Biochem. Cell Biol. 2010, 88, 451−458. (6) Nichols, E. R.; Craig, D. B. Electrophoresis 2008, 29, 4257−4269. (7) Tan, W.; Yeung, E. S. Anal. Chem. 1997, 69, 4242−4248. (8) Jacobson, R. H.; Zhang, X.-J.; DuBose, R. F.; Matthews, B. W. Nature 1994, 369, 761−766. (9) Craig, D. B.; Nichols, E. R. Electrophoresis 2008, 29, 4298−4303. (10) Lee, Y. C. Biochem. Biophys. Res. Commun. 1969, 35, 161−167. (11) Wentworth, D. F.; Wolfenden, R. Biochemistry 1974, 13, 4715− 4720. (12) Gorris, H.-H.; Rissen, D. M.; Walt, D. R. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 17680−17685. (13) Craig, D. B.; Chase, L. N. Anal. Chem. 2012, 84, 2044−2047. (14) Zhao, J.-Y.; Chen, D.-Y.; Dovichi, N. J. J. Chromatogr. 1992, 608, 117−120. (15) Shoemaker, G. K.; Juers, D. H.; Coombs, J. M. L.; Matthews, B. W.; Craig, D. B. Biochemistry 2003, 42, 1707−1710. (16) Craig, D. B.; Schwab, T.; Sterner, R. Biochem. Cell Biol. 2012, in press. (17) Monod, J.; Wyman, J.; Changeux, J.-P. J. Mol. Biol. 1965, 12, 88−118.
4602
dx.doi.org/10.1021/ac300777u | Anal. Chem. 2012, 84, 4598−4602
