ANALYSIS OF PKA CATALYZED REACTIONS VIA ESI-MS

Darryl L. Davis and Catherine M. Bentzley

University of the Sciences in Philadelphia, Department of Chemistry and Biochemistry 600 S. 43^rd St. Philadelphia, PA 19104

 

ABSTRACT

 

In this work an electrospray ionization-mass spectrometry (ESI-MS) method is presented to track the activity of protein kinases and modulation of the kinase activity. To demonstrate the effectiveness of using ESI-MS as a kinase assay, one of the most well-characterized kinases, cyclic adenosine monophosphate (cAMP)-dependent protein kinase (PKA), is analyzed.  This enzyme catalyzes the phosphorylation of substrates in response to cAMP, an intracellular second messenger of hormone action. The substrates used in this study are kemptide and malantide with one and three potential phosphorylation sites, respectively. The method developed yields quantification of the amount of phosphorylation of known kinase substrates under varying conditions of enzyme, substrate, phosphoryl donor, and activator.  It also allows differentiation of the singly phosphorylated and doubly phosphorylated malantide. The ability to detect the inhibition of PKA by several known inhibitors is also demonstrated.

 

 

Protein kinases, enzymes that catalyze the phosphorylation of various endogenous proteins, are key enzymes for cell regulation, growth, proliferation, and differentiation. These enzymes are not only essential for the regulation of cellular growth and metabolism,1 but also, because of their central role in cell homeostasis, the unregulated activation of protein kinases has been implicated in disease states such as cancer.2  Because these enzymes are critical intermediates in many biochemical pathways, there is an increasing demand for faster and more accurate techniques with which to test for kinase activity. 

            At present, no single technique exists that is capable of achieving the high throughput, specificity, and selectivity that is needed to accurately characterize these enzymes.  To date, the limited techniques used to investigate protein kinase ligand binding include gel electrophoresis, gel filtration, immunoprecipitation, and UV spectroscopy.^3-6 However, each of these techniques has distinct disadvantages. For example, the most common assay for kinase activity utilizes ³²P labeling to track the phosphorylation of proteins by kinases.  In this method, there are two distinct steps: transfer of labeled phosphoryl group and isolation of product.⁷ Although this is a sensitive method (picomole range), it lacks specificity because of the inability to differentiate between the phosphorylation of endogenous proteins and the substrate of interest.  Other problems arise with this technique when the incomplete rinsing of unutilized labeled ATP leads to high background noise.  Finally, the isolation of the product generally involves binding to a solid support such as phosphocellulose paper or nitrocellulose membrane.  One inherent problem with this method is the potential loss of substrate during the binding/washing of the solid supports. Two-dimensional phosphopeptide mapping using gel electrophoresis is time intensive, not amenable to automation, and may not yield quantifiable results. Immunoprecipitation and immunoblot techniques are the most sensitive (femtomole range) methods currently in use, but they can suffer from high background noise (due to non-specific binding) and require the use of an antibody for the analyte of interest.  Methods based on UV spectroscopy provide qualitative information including kinetic properties but often require interfacing to other analytical methods, such as those previously mentioned, for direct confirmation of a species identity.

Several researchers have already used mass spectrometry to confirm phosphorylation of peptides and proteins post-reaction.  For example, Matsumoto et al used matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF) to quantify the phosphopeptide produced in a calcium/calmodulin-dependent protein kinase II (CaMK II) reaction.⁸ In this work, the authors were interested in using MALDI-TOF as an assay of phosphopeptide production by CaMK II.  Although their MALDI method required smaller volumes than the standard HPLC method (0.5 μL  vs. 100 μL), a systematic correction of 5% in the degree of phosphorylation of the peptides was unfortunately necessary.  In other works, Newton et al used fast atom bombardment mass spectrometry (FAB-MS) to assay the activity of cyclic adenosine monophosphate-dependent protein kinase.⁹ They discovered that the FAB-MS method produced results comparable with the traditional radioactive method while enabling the monitoring of multiple species.  Unfortunately, direct confirmation of phosphorylated peptides was not achieved.  Graham et al used electrospray ionization mass spectrometry (ESI-MS) to determine the extent of phosphorylation and the number of phosphorylation sites associated with cyclic adenosine monophosphate-dependent protein kinase (PKA) phosphorylation of tyrosine hydroxylase.  However, they found that HPLC was still necessary for cleaning up the sample prior to MS.¹⁰

This work focuses on the development of an ESI-MS method to track the activity of protein kinases and modulation of the kinase activity. To demonstrate the effectiveness of ESI-MS for this process, a well-characterized kinase, cyclic adenosine monophosphate(cAMP)-dependent protein kinase (PKA), is studied.  PKA has become a model of protein kinase action because it is easily aquired, and the crystal structure of the catalytic subunit with various bound ligands has been solved.11-20  This enzyme catalyzes the phosphorylation of substrates in response to cAMP, an intracellular second messenger of hormone action.21-23 The PKA holoenzyme is an inactive tetrameric enzyme consisting of 2 regulatory (R subunit) and 2 catalytic subunits (C subunit).  The binding of cAMP to the regulatory subunits (2cAMP molecules per R subunit) releases the C subunits, which are catalytically active.24 The typical substrates for in vitro reactions are kemptide, casein, histone, and protamine. In this work, we will be using ESI-MS to monitor the phosphorylation of kemptide and malantide after activation of PKA by cAMP.  This method is ideal because it allows for direct confirmation, high throughput, specificity, and selectivity.

The direct confirmation of the phosphorylation of substrates from the mass spectra generated is indicated by an increase of 80 daltons, which corresponds to the addition of one phosphate group.  There is no non-specific background arising from unutilized radiolabled ATP and/or endogenous protein phosphorylation.   In ESI –MS, selectivity can be achieved through standard data acquisition protocol by varying the m/z range and ionization mode.  One can also view negatively charged ions versus positively charged ions.  Another benefit of an ESI-MS based assay is the potential to track the modification of substrates over time using time-resolved ESI-MS.  The time-resolved data allows us to track progressive stages in the phosphorylation of substrates over time and to perform on-line changes to analyte concentrations. 

The method presented in this work is the first time ESI-MS has been used to monitor an active PKA catalyzed reaction.  The limitations of this method include the high buffer and salt concentrations required for biochemical reactions.  In addition, the ESI acquisition of mixtures of various components can lead to ion suppression, thereby making the quantification of analytes in a reaction mixture difficult.

 

EXPERIMENTAL SECTION

  Chemicals.  Kemptide, malantide, PKA, adenosine triphosphate, and cAMP were acquired from Sigma Chemical Company (St. Louis, MO)

The PKA catalytic subunit was acquired from Roche Molecular Biochemicals (Indianapolis, IN)

Sample Preparation.  The PKA reaction mixtures were prepared to a final volume of 0.5 mL as follows: 1) stock solutions of ATP, cAMP, kemptide, and malantide were made in water, and O.5 μM PKA was prepared in 4.0 mM-K₂HPO₄ aqueous buffer 2) each component was added to its final concentration; 200 mM cAMP, 200 mM ATP, 0.5 mM PKA, and 200 mM of substrate.  For the inhibited reactions, 1 μM inhibitor was added to the reaction mixture.

Data Aquisition. 

The reaction mixture was infused into the mass spectrometer using a direct infusion syringe pump (Harvard Apparatus) at a rate of 2 mL/min.

All spectra were acquired on a Micromass Trio 2000 ESI source region.  The scanning rate was 20 scans/minute.

RESULTS AND DISCUSSION

Quantitation of Analytes      

The most beneficial aspect of using ESI-MS as a kinase assay is the direct confirmation of reactants (ATP, cAMP, kemptide, and malantide) and products (ADP, phosphkemptide, and phosphomalantide).  The ability to quantitate several analytes from the same mixture without prior chromatographic separation facilitates the rapid analysis of many samples.

            Quantitation was accomplished by mixing all reactants except the substrate, adding a quencher (acetic acid) followed by substrate addition.  The reaction mixture was infused for 1 minute, and the summation of all the spectra were used.  The spectra were deconvoluted using an in-house program, and the areas of the peaks were calculated using Origin version 3.5.  Alternatively, each analyte was infused separately without the other components of the reaction.  Both sample preparation methods produced similar results in regards to detector response. As expected, no background phosphorylation was present. The standard curves of the two analytes of interest, kemptide and malantide, were found to be linear within the range of interest, 1 mM to 200 mM.  Figures 1 and 2 are representative spectra of 200 mM solutions of malantide and kemptide infused at a rate of 2 mL per minute, respectively. 

            Both peptides produced a strong signal in the positive ESI mode.  Detection was possible for less than 1 mM solutions; however, the detector response was no longer linear below 1 mM. 

Phosphorylation of Substrates.

PKA phosphorylates the serine and threonine residues of peptides and proteins through the reaction scheme shown in Figure A. The known consensus sequence, the sequence of amino acid residues on the substrate that is needed for phosphorylation to occur, of PKA is RRXSX; however, it is not all inclusive.  There are examples of substrates lacking this sequence that are phosphorylated by PKA.  PKA also phosphorylates threonine residues.  The lack of complete data regarding sequence specificty arises from the time necessary, using traditional methods such as radiolabeling, to determine phosphorylated substrates and the site of phosphorylation.  ESI-MS excels at this.  The peptide substrate malantide (Figure 3) contains three residues, two serines and one threonine, which represent possible phosphorylation sites.  However, only one of the residues, serine 5, lies within an   experimentally known consensus sequence.  This peptide, however, undergoes up to two phosphoryl transfers on the same peptide.  By examining the region of the triply charged peptide, this fact becomes evident.  The peak at m/z 572 represents a singly phosphorylated peptide, and the peak at 600 m/z  represents the doubly phosphorylated peptide.  The addition of an HPO₃ to the doubly charged peptide is evident at peak 857 m/z.  Trypsinolysis of phosphorylated malantide prior to MS reveals that serine 5 is indeed the predominant site of phosphorylation, followed by threonine 2 with a very minor amount of both serines being phosphorylated.

            The peptide substrate kemptide (Figure 4) contains only one potential phosphorylation site.  The peaks at m/z 853 and 427 correspond to the single phosphorylation of the singly and doubly charged ion, respectively.

            As mentioned earlier, one of the areas of kinase research that is lacking is the determination of sequence specificty (as evidenced by gaps in known consensus sequences).  Because ESI-MS can rapidly determine the extent of phosphorylation of a substrate and the number of phosphorylation sites, the ESI-MS technique would be better suited than current techniques in these types of study.

PKA Inhibition

As stated earlier, one of the major areas of interest in kinase research is the search for and screening of possible kinase inhibitors. This area of research demands that the method of detection is capable of high throughput to screen the large numbers (>2000) typically produced by combinatorial techniques used to produce inhibitor candidates.  To test whether ESI-MS could be used to monitor the inhibition of an active reaction, the inhibition of the PKA catalyzed reaction by two known protein kinase inhibitors was monitored.  Figures 5 and 6 demonstrate the inhibition of the PKA catalyzed phosphorylation of malantide (200 mM) by 1mM concentrations of N-(2-guanidnoethyl)-5-isoquinoline sulfonamide (a strong, ATP competitive PKA inhibitor) and 1-(5-chloronapthalene-1-sulfonyl)-1H-hexahydro-1,4diazepine (a weak, ATP competitive PKA inhibitor), respectively.  In comparison with the uninhibited reaction demonstrated in Figure 3, the phosphorylation of malantide is reduced by greater than 50% by the strong inhibitor although the reaction mixture containing the weak inhibitor shows little reduction in malantide phopshorylation.  This is expected because the weak inhibitor is specific for protein kinase C not PKA.

Time-Resolved ESI-MS

Measurement of changes in several analytes simultaneously over time using the same reaction mixture and conditions is one of the unique aspects of ESI-MS.  From Figure 7, it can readily be seen that the second phosphorylation site on malantide is phosphorylated much more slowly than the first site.  From this data, we can conclude that the substrate is phosphorylated, released from the enzyme and then is phosphorylated a second time.  Comparing Figure 8, which shows the time-resolved phosphorylation of kemptide, with Figure 7, which shows the time-resolved phosphorylation of malantide, the phosphorylation of kemptide occurs at a slower rate (10 minutes for 50% phosphorylation) than that of malantide (7 minutes).  The remaining graphs demonstrate that the presence of a strong inhibitor (Figure 9), while slowing the reaction rate, does not significantly change the rate at which the first and second sites are phosphorylated in comparison with Figure 7.  Suprisingly, the time-resolved data for the weak inhibitor (Figure 10), while increasing the time until 50% phosphorylation of malantide by only 1 minute, shows that it is the second site of phosphorylation that is mainly affected.

CONCLUSIONS

        The method established by this initial work is capable of quantitating the amount of phosphorylation of known kinase substrates under varying conditions of enzyme, substrate, phosphoryl donor, and activator. In addition, the method allows for the differentiation between singly phosphorylated and doubly phosphorylated malantide.  The ability to actively monitor the PKA catalyzed reaction over time is a further benefit.  The method is suitable for rapid determination of PKA inhibition and sequence specificity. 

            Several extensions to this method are being pursued.  We are automating the method utilizing an auto injector and software macros. The screening of peptide libraries with this method would also be advantageous and straightforward.  Both developments would increase the throughput of kinase activity screening allowing one to rapidly test for kinetic parameters, suitable inhibitors, and sequence specificity.

 

 

Literature Cited

(1) Lehninger, A.; Nelson, D.; Cox, M. (1993) Principles of Biochemistry.  2nd Ed.

     p.777. 

(2) Cho-Chung, Y.; Pepe, S.; Clair, T.; Budillon, A.; Nesterova, M. (1995) Crit. Rev. in

     Oncology/Hematology 21, 33-61.

(3) Bullock, B.; Habener, J. (1998) Biochemistry 37, 3795-3809.

(4) Herberg, F.; Doyle, M.; Cox, S.;Taylor, S. (1999) Biochemistry 38, 6352-6360.

(5) Cho-Chung, Y.; Pepe, S.; Clair, T.; Budillon, A.; Nesterova, M. (1995 ) Crit. Rev. in

     Oncology/Hematology 21, 37.

(6) Shaffer, J.; Adams, J. A. (1999) Biochemistry 38, 5572-5581.

(7) Glover, C.; Allis C. (1991) Methods in Enzymology 200, 85-90.

(8) Matsumoto, H.; Kahn, E.; Komori, N. (1998) Anal. Biochem. 260, 188-194.

(9) Graham, M.; Dickson, P.; Dunkley, P.; Von Nagy-Felsobuki, E. (1998) Rapid   

      Commun. Mass Spectrom. 12, 746-748.

 

(10) Goodlett, D.; Aebersold, R.; Watts, J. (2000) Rapid Commun. Mass Spectrom. 14,

        344-348.

(11) Narayana, N.; Diller, T.; Koide, K.; Brunton, L.; Xuong, N.; Ten Eyck, L.; Taylor,

        S. (1999) Biochemistry 38, 2367-2376.

(12) Narayana, N.; Cox, S.; Xuong, N.; Ten Eyck, L.; Taylor, S. (1997) Structure 5, 921.

(13) Knighton, D.; Zheng, J.; Ten Eyck, L.; Ashford, V.; Xuong, N.; Taylor, S.;

       Sowadsji, J. (1991) Science 253, 407.

(14) Zheng, J.; Knighton, D.; Ten Eyck, L.; Karlsson, R.; Xuong, N.; Taylor, S.; 

       Sowadski, J. (1993) Biochemistry 32, 2154-2161.

(15) Karlsson, R.; Zhenj, J.; Xuong, N.; Yaylor, S.; Sowadski, J. (1993) Acta  

       Crystallogr., Sec D, 49, 381-388.

(16) Zheng, J.; Knighton, D.; Xuong, X.; Taylor, S.; Sowadski, J. (1998) Protein Sci. 2, 

       1559-1573.

(17) Engh, R.; Girod, A.; Kinzel, V.; Huber, R.; Bossenmeyer, D. (1996) J. Biol. Chem. 

        271, 26157.

(18) Bossemeyer, D.; Engh, R.; Kinzel, V.; Ponstingl, H.; Huber, R. (1993) EMBO J. 12, 

       849-859.

(19) Prade, L.; Engh, R.; Girod, A.; Kinzel, V.; Huber, R.; Bossenmeyer, D.  (1997)

        Structure 5, 1627.

(20) Zheng, J.; Trafny, E.; Knighton, D.; Xuong, N.; Taylor, S.; Ten Eyck, L.; Sowadski,

       J. (1993) Acta. Crystallogr., Sect D, 49, 362.

(21) Cheng, X.; Shaltiel, S.; Taylor, S. (1998) Biochemistry 37, 14005-14013.

(22) Krebs, E. (1972) Curr. Top. Cell Regul. 5, 99-133.

(23) Shaffer, J.; Adams, J. (1999) Biochemistry 38, 5572-5581.

(24) Murphy, B.; Scott, J. (1998) Trends in Cardio. Med. 8,89-95.

 

Figure Legends

Figure A.

Illustration of the phosphorylation of substrate.

Figure 1.

Spectrum of 200 uM solution of malantide.

Figure 2.

Spectrum of 200 uM solution of kemptide.

Figure 3.

Spectra of PKA(1uM) catalyzed phosphorylation of malantide (200uM) at

0,10,and 40 minutes.  · - singly phosphorylated substrate, ··- doubly

phosphorylated substrate

Figure 4.

Spectra of PKA(1uM) catalyzed phosphorylation of kemptide (200uM) at

0,10,and 20 minutes.  * - singly phosphorylated substrate (+2 ion), **-

singly phosphorylated substrate (+1 ion).

Figure 5.

Spectra (0,10,and 15 minutes) of PKA(1uM) catalyzed phosphorylation of

malantide (200uM) in the presence of N-(2-guanidnoethyl)-5-isoquinoline

sulfonamide (1uM).  · - singly phosphorylated substrate, ··- doubly

phosphorylated substrate.

 

 

 

Figure 6.

Spectra (0,10,and 15 minutes) of PKA(1uM) catalyzed phosphorylation of

malantide (200uM) in the prescence of 1-(5-chloronapthalene-1-sulfonyl)-

1H-hexahydro-1,4 diazepine (1uM).  · - singly phosphorylated substrate,

··- doubly phosphorylated substrate.

Figure 7.

Graph of time monitored malantide phosphorylation.

Figure 8.

Graph of time monitored kemptide phosphorylation.

Figure 9.

Graph of time monitored malantide phosphorylation in the prescence of a

strong inhibitor.

Figure 10.

Graph of time monitored malantide phosphorylation in the prescience of a

weak inhibitor.