Application of electrospray mass spectrometry in biology Andrew R. Pitt Department of Pure and Applied Chemistry, Strathclyde University, Thomas Graham Building, 295 Cathedral Street, Glasgow, UK G1 1XL Covering: January 1984 to December 1996 1 Introduction 2 Electrospray mass spectrometry 2.1 Ion generation and mass analysis 2.2 Secondary ion generation 2.3 Protein and peptide analysis 2.4 Interfacing to separation techniques 2.4.1 High performance liquid chromatography 2.4.2 Capillary electrophoresis 3 Oligosaccharide analysis 4 Proteins 4.1 Covalent modification 4.1.1 Non-specific covalent adduct formation 4.1.2 Chemical modification 4.1.2.1 Post-translational modification 4.1.2.2 Active site directed irreversible inhibitors 4.2 Covalent enzyme–ligand complexes 4.3 Non-covalent enzyme–inhibitor complexes 4.4 Protein folding 5 DNA 6 Summary 7 References 1 Introduction Since electrospray ionisation (ESI) and matrix assisted laser desorption/ionisation–time of flight (MALDI–TOF) mass spectrometry (MS) became readily available at a reasonable cost in the late 1980s,1,2 mass spectrometry has become an ever more important tool for the biological scientist, both for polar small molecules that are diYcult to analyse using the traditional ionisation techniques and for the analysis of biological macromolecules. ESI and MALDI are both ‘soft’ ionisation techniques capable of producing ions of low energy, and hence have the power to generate ions from biological macromolecules of molecular masses in excess of 100 000, and in many cases these masses can be measured with accuracies better than 0.01%.Mass spectrometry relies on two key processes, the generation of gas phase ions of the compound of interest, a process that occurs within the ‘source’ area of the mass spectrometer, and the analysis of the mass to charge ratio of these ions using a ‘mass analyser’. Many combinations of sources and analysers have been used but this review will concentrate on the most common, the use of electrospray ionisation in conjunction with a quadrupole mass analyser to analyse larger biomolecules. A number of reviews have been published comparing ESI-MS with MALDI–TOF,3–8 and it is worth bearing in mind that they are, on the whole, complementary techniques. Comparisons to thermospray have been made,9 and the limitations of the use of electrospray in conjunction with sector mass analysers, which require a very stable ion beam and hence reasonably strong samples and good ionisation, have also been reviewed.10 As with any rapidly growing area of science new techniques with even higher performance are continuously being developed.The major advances have been in the design of the analyser. Fourier transform ion cyclotron resonance (FTICR) mass spectrometers,11–13 which have actually been interfaced to electrospray ionisation sources for many years,14 with mass accuracies of <1 ppm and exceptional sensitivity (theoretically a few thousand molecules would be enough) and resolution (the theoretical mass at which two peaks one Dalton apart would be resolved to half their heights) of 200 000 or better,15 are now commercially available and the use of quadrupole ion traps is becoming more popular.16,17 These exciting new techniques will only be mentioned in passing as they have not yet found widespread use but they are, without doubt, set to become the standards of the future. 2 Electrospray mass spectrometry 2.1 Ion generation and mass analysis A wide range of mass spectrometers are commercially available with electrospray sources18 and ESI-MS has now become a routine technique in many laboratories. Electrospray mass spectrometry has been around for a number of years, but some of the processes involved in the formation of the ions are still controversial. This section will attempt to deal briefly with the processes thought to be involved in the formation of the gas phase ions, the key process in any mass spectrometry.The process is outlined in Fig. 1. Electrospray ionisation is an atmospheric pressure ionisation technique, where ions are continuously formed in the source region from a solution of the analyte at, or near, atmospheric pressure. The pressure is then reduced in stages through the source by high capacity Sample + solvent ++++++ ++++ Electrospray capillary 2–3 kV relative to skimmer Ion beam Nebulising gas Drying gas Vacuum pumps Quadrupole analyser Ion beam focussing Ion source Electrospray capillary Sampling skimmer cone Fine spray of charged droplets formed due to high potential (+ nebulising gas) Drying gas ( a) ( b) Fig. 1 Schematic diagram of (a) one design of an electrospray source showing the general layout and (b) an expanded diagram of the area in which the spray and ions are formed Pitt: Application of electrospray mass spectrometry in biology 59vacuum pumps, to reach the 10"5 Torr or less needed in the quadrupole analyser.A continuous flow of solvent (at rates which can range from 10 nl min"1 up to 1 ml min"1) is passed into the source through a fine capillary needle which is held at a high potential (2–4 kV) relative to an adjacent sampling plate, resulting in a fine spray of droplets. The most widely accepted model for the formation of gas phase molecular or pseudomolecular (multiply charged) ions take place in four steps19,20 (Fig. 2): (i) the formation of a fine spray of droplets with relatively high surface charge densities due to the high potential on the capillary needle; (ii) evaporation of the carrier solvent molecules from the droplets, causing the droplets to shrink and the charge density on the surface to increase; (iii) explosive fragmentation of the droplets as the charge density on the droplet reaches a critical limit (the Raleigh limit); and (iv) the desolvation process continuing until the eventual formation of gas phase ions of individual molecules.The exact process for the formation of gas phase ions is still controversial, and other models have also been proposed.21,22 The formation of a stable spray of fine droplets is one of the critical processes for optimum performance of the electrospray source and can be aided at higher flow rates of solvent by the use of a sheath of N2 gas around the electrospray needle to promote nebulisation (pneumatically assisted electrospray) or by replacing the capillary needle with an ultrasonic nebuliser.The evaporation of solvent from the droplets is usually promoted by heating the source (from 40–150 )C), and in a number of designs by the use of a flow of nitrogen ‘drying’ gas through the source, usually across or against the flow of droplets, to help to remove solvent vapour. Ions are formed by the addition or removal of protons from the analyte and hence the formation of charged ions can be assisted by the addition of volatile acids or bases to the solvent.In many cases only a small proportion of the analyte produces ions and the sampling of the ions is relatively ineYcient, hence electrospray sources have limited sensitivity, especially at higher solvent flow rates. Electrospray sources can generate and analyse either positive or negative ions23 allowing the analysis of biomolecules with a wide range of pKa values. Quadrupole analysers have a reasonable mass to charge range (up to 4000) and are robust, easy to tune and do not require a very high vacuum, hence are ideal in the general laboratory environment.The electrospray source generally produces ions from biomolecules carrying a number of charges such that the mass to charge ratio (m/z) of the ions, which is the parameter that is measured, often lies in the region of 500–2000 (although under some conditions this may increase to >4000); this is ideal for detection by a quadrupole analyser.Proteins with masses up to 150 000 can be observed using a simple quadrupole analyser, although some care is needed in optimising the conditions.24 The importance of electrospray mass spectrometry in the analysis of biomolecules is now widely accepted and has been reviewed a number of times over the past years.25–34 The great advantage of FTICR as a mass analyser is its extreme sensitivity and resolving power; for example a resolving power of 63 000 has been reported for equine cytochrome c (12 352 Da).35 However, the authors also highlight some of the diYculties that need to be addressed when using FTICR, and for most applications this technique is still in its infancy. The extreme sensitivity of FTICR as an analyser for ESI-MS was demonstrated by the determination of the molecular weight of carbonic anhydrase (28 780) to 1 Da at a resolving power of about 60 000, using 29#10"18 moles of the enzyme from a crude preparation of red blood cells.36 This represents about 1% by weight of the protein in one red blood cell.Not only were the authors able to measure the mass to a high degree of accuracy, but they were also able to gain enough information from sequence specific fragment ions from the same sample to confirm unequivocally the identity of the protein from the protein database. Figure 3 shows a typical electrospray mass spectrum for a protein. It is usual to see a range of charged states adopting a Gaussian distribution (known as the charge state distribution), each of the adjacent peaks in the spectrum diVering by one charge caused by the addition (positive ion electrospray) or removal (negative ion electrospray) of a proton.The tops of these peaks represent the average isotopic mass of the peaks, as the quadrupole analyser is unable to resolve the individual isotope peaks in the multiply charged ions which will only be separated by 1/(the charge on the peak) Da.The conversion of the series of multiply charged peaks into a real mass spectrum is referred to as deconvolution (Fig. 3). The majority of operating systems for mass spectrometers are supplied with a computational package that will perform this operation, and it is relatively simple when a series of peaks from a single component can be readily identified. Under ideal conditions it should be possible to resolve species diVering by 1 Da in 10 000 with a quadrupole analyser, and 1 Da in 25 000 with a double focusing sector analyser.A mass accuracy of 1 Da in 10 000 has been demonstrated with a quadrupole analyser for compounds with molecular masses up to 80 000.37 A problem with ESI-MS is that complex mixtures can be diYcult to analyse due to the confusion of peaks that all appear at around the same mass. One of the most eVective methods of deconvoluting complex spectra to give a true mass output is the use of maximum entropy algorithms,38–41 which can be used to Evaporation of solvent Explosive fragmentation Process continues until formation of pseudomolecular ions + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + Fig. 2 Schematic diagram of pseudomolecular ion formation Fig. 3 Electrospray mass spectrum of horse heart myoglobin (Sigma, Poole, Dorset) obtained on a VG Analytical Platform mass spectrometer (Altrincham, Cheshire) showing the Gaussian distribution of charge states. Numbers above peaks correspond to the number of charges on the ion 60 Natural Product Reports, 1998identify the components present from the raw data without the need for any preprocessing, and can give exceptionally good resolution if applied judiciously42 (see Fig. 6). This is supplied commercially as MaxEnt by Micromass, Altrincham, Cheshire. The problem of the determination of the number of charges on individual peaks in complex spectra (charge state assignment), which simplifies the assignment of a series of peaks in the raw data to a single component in the mixture, can also be simplified by utilising the 1:1 stoichiometry of the non-covalent adduct formation between peptides or proteins and 18-crown-6.43 The remarkable resolution of FTICR allows the resolution of the individual isotope peaks within any one multiply charged ion peak, obviating the need for any complex deconvolution routine. 2.2 Secondary ion generation A technique that has been used to great eVect for sequencing peptides is the fragmentation of the molecular ion (parent ion) into smaller units (daughter ions), the fragmentation usually taking place around the amide bonds generating a series of ions (Fig. 4). These daughter ions can then be analysed and the sequence of the peptide inferred.44,45 For a simple sample containing a single component the fragmentation can be induced in the electrospray source region by increasing the voltages on the skimmer cones, which results in the ions gaining kinetic energy and colliding with enough force to cause them to fragment.This is referred to as collisionally induced decomposition (CID) or cone voltage fragmentation (CVF). For more complex samples the spectra become too complicated and it is necessary to resort to tandem mass spectrometry. Tandem mass spectrometry, often referred to as MS/MS or MS2, uses two (or more) analysers in tandem, with an area between them where fragmentation is induced (Fig. 5). The first analyser (MS1) is used to select the ion of interest, referred to as the ‘parent’ ion.This is then passed into a collision cell, where fragmentation is usually induced by introducing a low pressure of gas. The ions leaving the first analyser have suYcient energy to cause fragmentation when they collide with the gas molecules in the collision cell, and the resulting ‘daughter’ fragment ions are then analysed in a second mass analyser (MS2). This is a particularly powerful technique for sequencing small peptides, and for identification of sites of modification, but will work with most compounds.A number of combinations of analysers have been used. The most common arrangement is two quadrupole analysers, but the development of QTof (Micromass, Altrincham, Cheshire), which is the combination of a quadrupole (MS1) with TOF (MS2), is showing great promise for simplicity and sensitivity. FTICR and quadrupole ion trap mass spectrometers have the capability to extend this technique to the selection of one of the initially formed fragment ions and further fragmentation of this ion, and MS4 is possible under ideal conditions. 2.3 Protein and peptide analysis There have been a number of reviews published in recent years in this area dealing with the simple characterisation of peptides from complex mixtures,27,46–50 protein analysis,51,52 on the application of mass spectrometry to the study of biological molecules and in analytical biochemistry.50,53–55 An example of the eVectiveness of the maximum entropy algorithm for the deconvolution of a complex spectrum is demonstrated by the analysis by allelic fingerprinting of major urinary proteins (18 600–19 000), showing the allelic diVerences between diVerent strains.56 The misincorporation of amino acids during the over expression of proteins57 and the incorporation of non-natural amino acids can also be followed by ESI-MS.58 The analysis can be extended to glycoproteins,59 although there can be some problems if there is a large contribution to the mass from the saccharide (see Section 4.1.2.1). 2.4 Interfacing to separation techniques One of the major advantages of the electrospray source is that it can be directly interfaced to a number of high performance separation techniques in order to eVect separation prior to mass analysis, the result often being referred to as ‘hyphenated methods’,60 thus allowing the on-line monitoring of separated materials directly.This has proved to be a powerful technique for the identification of components of complex mixtures, especially in the use of proteolytic digests for the location of modifications (see Section 4.1) and for variant protein analysis.61 O HN R O NH R OHN O NH O NH R R R O HN R O NH R OHN O NH O NH R R R CO2H R NH2 O NH R OHN O NH O NH R R R CO2H HN R CO2H HN O NH R R CO2H O NH R OHN HN O NH R R O HN RO NH R NH2 O HN R O NH R OHN O NH O NH R R R O HN R O NH R OHN O NH O R R R NH2 NH HN H2N R1 O R2 O R3 NH CO2H O R4 Collision cell Fragmentation of parent ion Measure mass of daughter ions MS1 MS2 z3 m/ z y1 m/ z y3 M x1 z2 x3 M y2 z1 x2 a1 b1 Parent ion spectrum b2 b3 c1 Daughter ion spectrum c2 c3 SEQUENCE a2 a3 a3 b3 a2 b2 y1 a1 b1 A+ + B+ + C+ + ......( b) P+ ( a) ( c) Daughter ions Select ion of interest Parent ion Fig. 4 MS/MS of peptides showing (a) the formation of daughter ions from a parent peptide ion, (b) the expected backbone fragmentation pattern of a peptide and the annotation used to denote the ions formed, and (c) a representation of a secondary ion spectrum MS1 MS2 Ion source Ion detection Deflector Collision cell Fig. 5 Schematic of a tandem mass spectrometer Pitt: Application of electrospray mass spectrometry in biology 612.4.1 High performance liquid chromatography High performance liquid chromatography (HPLC) has been routinely used for the separation of biological samples for many years, and its use in conjunction with ESI-MS is now widespread.62–65 The problems associated with the coupling of HPLC to the mass spectrometer are mainly associated with the high solvent flow rates and the use of ion pairing reagents.The flow rate problem has been overcome by the use of splitters to direct only a portion of the eluent into the mass spectrometer, along with the use of large vacuum pumps, high source temperatures and a high flow rate of drying gas to remove rapidly the larger volumes of solvent vapour, or the use of microbore HPLC.The use of ion pairing reagents, for example trifluoroacetic acid, which are commonly added to polar HPLC eluents to improve the resolution of the technique for a range of polar biomolecules, can significantly reduce the sensitivity of ESI-MS, especially for DNA samples. This can be overcome by careful choice of the reagent, or by the addition, just at the point of formation of the spray, of an additional reagent.66,67 HPLC is particularly useful for proteolytic mapping, for example in glycopeptide and glycoprotein analysis.68,69 2.4.2 Capillary electrophoresis The high resolving power of capillary electrophoresis (CE) and the low flow rates involved make it an ideal technique to interface with ESI-MS; this was first demonstrated for proteins in 1989.70 There are a number of problems associated with the very low flow rates and high buVer concentrations required for CE in some instances, and also in raising the potential on the capillary to a suitable voltage to generate the electrospray.These have been overcome by a number of methods68 including the addition of a second solvent at the probe tip via a larger coaxial capillary, or coating the capillary in gold to make the contact. It is a powerful analytical tool due to its impressive resolving power, and has found use for the analysis of proteins70–73 and complex mixtures of peptides, especially proteolytic digests.65,74–81 3 Oligosaccharide analysis Electrospray mass spectrometry has been less widely applied to oligosaccharide analysis, although there have been some notable successes.Fragmentation by CID or tandem mass spectrometry will generate daughter ions that can be used to assign both the sequence and the position of the linkages. Hydrophobic oligosaccharides without easily ionisable groups can be problematic to analyse, and the formation of adducts with sodium in place of a proton is common.Carbohydrate sequence analysis by ESI-MS has been recently reviewed,82 and the linkage position and anomeric configurations of underivatised glucopyranosyl disaccharides have been determined using ESI-MS.83 4 Proteins Electrospray mass spectrometry has been extensively used to study covalent modifications in proteins. These studies can be roughly divided into four areas; non-specific covalent adduct formation, specific chemical modification, post-translational modification and irreversible inhibitors. 4.1 Covalent modification 4.1.1 Non-specific covalent adduct formation ESI-MS has been shown to be useful in identifying covalent non-specific chemical modifications of proteins. However, some care needs to be taken in assuming that an observed mass change is due to an expected modification. A disulfide link between 2-mercaptoethanol (present in the buVer during puri- fication) and a surface cysteine of the class II fructose-1,6- bisphosphate aldolase from E.coli (expected mass change of 76 amu) was shown to be responsible for an apparent mass change of 80 amu, which could have easily been assigned to a phosphate group.84 Although this modification was benign, it would have gone unnoticed without the use of ESI-MS in the analysis of the protein. The adducts formed with histidines on treatment of myoglobin with 4-hydroxy-non-2-enal, a byproduct of lipid peroxidation, are easily distinguished by ESI-MS85 and can be further characterised by tryptic digest and tandem MS.86 ESI-MS has also proved useful for the monitoring of the reactions of the biologically important compound nitric oxide with peptides and proteins.87 Cysteine residues were shown to be rapidly nitrosated, with an increase in mass of 29 Da.being observed for each NO addition. The further oxidation of nitroso thiols to sulfonic acids and nitrosation of tyrosine were also noted. 4.1.2 Chemical modification ESI-MS can be used to characterise rapidly a range of chemical modifications of proteins.83 ESI-MS alone, and in conjunction with liquid chromatography techniques, has become a powerful tool for the identification of active site residues using chemical modification.It has been extensively used to characterise the adduct between the arginine specific reagent phenylglyoxal and type I and type II dehydroquinases,88–91 and to relate this to the activity of the enzyme. It was possible to characterise the phenylglyoxal modified protein and show that there may be 1:1 or 1:2 stoichiometries with this reagent which may cause misinterpretation of radioactive labelling results.Using proteolytic digestion and LC-MS the authors were able to locate the position of the hyper-reactive arginine as ESI-MS is soft enough to observe the highly labile modified peptides (Fig. 6). Boots et al.92 used the active site directed inhibitor 1-bromooctan-2-one to label specifically and to identify the active site histidine of Staphylococcus hyicus lipase as His 600, and the active site cysteine of thiaminase I from Bacillus thiaminolyticus was identified using ES–FTICR of the 4-amino-2-methyl-6-chloropyrimidine inactivated enzyme and identification of the modified peptide generated from fragmentation of the protein.93 The ability of iron to generate free radical species has been used to locate the metal binding site of actin.94 Treatment of iron bound actin with oxygen and either dithiothreitol or ascorbate as the reductant caused cleavage of the protein at sites adjacent to the iron binding site, and the resultant fragments were analysed using ESI-MS to locate the fragmentation sites. 4.1.2.1 Post-translational modification ESI-MS has been used to study a range of post-translational modifications, but has been particularly eVective for studying sites of phosphorylation and glycosylation patterns. Full characterisation of the modifications may require the combination of a number of techniques, including enzymatic digestion, chemical modification and accurate mass analysis, as well as ESI-MS,95 and care needs to be taken in assigning an apparent mass shift to a particular modification without additional evidence.84 The expected mass changes for a range of post translational modifications are given in Table 1.ESI-MS was used to demonstrate that the post-translational modification of barley ·-amylase expressed in yeast involved the removal of the C-terminal Arg–Ser dipeptide, glutathionylation and O-glycosylation.96 The complex glycosylation pattern of human lecithin:cholesterol acyltransferase was determined by microbore reverse-phase HPLC–ESI-MS;97 a combination of enzymatic digestion of the protein and sequential glycosidase digestion followed by HPLC–ESI-MS using a method of monitoring carbohydrate specific fragment ions was able to locate four N-linked glycosylation sites containing sialylated bi- and tri-antennary complexes, along with two O-linked glycosylation sites.A range of post translational 62 Natural Product Reports, 1998modifications of a the humanised monoclonal antibody Campath-1H were also identified by ESI-MS in combination with capillary chromatography following disulfide reduction and trypsin digestion.98 Two diantennary carbohydrate moieties and conversion of the terminal glutamate into pyroglutamic acid were observed, and partial removal of the C-terminal lysine was confirmed by MS/MS sequencing.The combination of ESI-MS and FAB mass spectrometry were used to identify the two sites and the nature of the glycosylation of the potent immunosuppressant glycodelin.99 Multiple isozymes often reflect diVerential post-translational modification. The catalytic subunit of mouse cAMP-kinase expressed in E. coli was shown to have a number of isozymes reflecting diVerent levels of phosphorylation using ESI-MS to measure the mass of the protein.100 The phosphorylation was not due to endogenous E. coli enzymes but was due to autocatalysis, giving a maximum of four, but predominantly three, phosphates.ESI-MS was also used to show that the two isozymes of the same enzyme isolated from porcine heart had the same mass, corresponding to bis-phosphorylation and N-terminal myristoylation. The in vivo phosphorylation sites of protein kinase C ‚-2 were identified by ESI-MS using collisionally induced dissociation, and this technique gave more detailed results than had been obtained in the previous in vitro studies.101 The four tyrosines that are the sites of phosphorylation of the intra- and extra-cellular domain of the heparin-binding fibroblast growth-receptor tyrosine kinase (FGF-R1) were identified using trypsin digestion followed by Edman degradation and tandem mass spectrometry,102 and ESI-MS was subsequently used extensively to identify the sites Table 1 Common post-translational modifications (excluding glycosylations) Modification Mass change Modification Mass change Pyroglutamic acid from glutamine "17.0306 Phosphorylation 79.9799 Disulphide bond formationa "2.0159 Sulfonation 80.0642 C-Terminal amide from glycine "0.9847 Cysteinylation 119.1442 Deamidation of glutamine or asparagine 0.9847 Incomplete N-terminal methionine removala 131.1986 Methylation 14.0269 Farnesylation 204.3556 Hydroxylation 15.9994 Myristoylation 210.3598 Oxidation 15.9994 Biotinylation 226.2994 Oxidation of methionine: sulfoxide sulfone 15.9994 31.9988 Pyridoxal phosphate (SchiV base to lysine) 231.1449 Formylation 28.0104 Glutathionylation 305.3117 Acetylationa 42.0373 5*-Adenosylation 329.2091 Carboxylation of asparagine or glutamine 44.0098 4*-Phosphopantotheine 339.3294 aIncomplete post-translational processing of overexpressed proteins is quite a common occurrence and will result in mixtures of proteins being observed, separated by the indicated masses. O O +N H H2N NH2 Arg HN NH OH HO +N H Arg O O O O HN NH HN NH +N H Arg +N H Arg Phenylglyoxal Singly modified (+116 amu) Doubly modified Takahashi adduct (+250 amu) – H2O – H2O ( a) ( b) Fig. 6 Chemical modification of arginines using phenylglyoxal. (a) Scheme showing the chemical formation of 1:1 (+116 amu) and 2:1 (+250 amu) adducts, and (b) MaxEnt (Micromass, Altrincham, Cheshire) deconvoluted mass spectrum of Streptomyces coelicolor dehydroquinase 95% inactivated with phenylglyoxal showing the formation of singly (+116 amu for each addition) and doubly (+250 amu for each modification) modified arginines.Pitt: Application of electrospray mass spectrometry in biology 63of the phosphorylations, usually using LC-MS to separate peptic digest fragments103 or tandem mass spectrometry.104 This technique is also useful for the study of covalently attached cofactors. Morris et al. were able to use ESI-MS to study the loading of Saccharopolyspora erythrea acyl carrier protein overexpressed in E.coli with 4*-phosphopantothene.105 The crude material isolated from the cells consisted of apoenzyme, holoenzyme dimer and holoenzyme glutathione adduct. The same group were able to show that the C-terminal acyl carrier protein thioesterase domain of the multifunctional 6-deoxyerythronolide B synthase from S. erythrea overexpressed in E. coli was not loaded with the 4*-phosphopantothene group, but that the cysteine predicted to lie in the active site could be selectively labelled with phenylmethylsulfonyl fluoride.106 They also used ESI-MS to study a number of other features of the acyl carrier protein.107 ESI-MS has been used to show that a subunit of NADH:ubiquinone reductase from bovine heart mitochondria, that has a closely related sequence to the acyl carrier proteins, actually caries a 4*-phosphopantothene group,108 although the role of this group was unknown at the time.The modification was identified by treatment of the subunit under alkaline conditions resulting in a mass loss of 339 expected for that of the 4*-phosphopantothene group.The flavinylation of trimethylamine dehydrogenase from Methylophilus methylotrophus and the wild type and mutant enzymes expressed in E. coli have shown that the enzyme is expressed almost exclusively in the holoenzyme form in the natural host, whereas the recombinant and mutant forms studied are not correctly posttranslationally modified, and therefore are prevented from undergoing flavinylation to the same extent.109 This may be due to the higher levels of expression in E.coli. 4.1.2.2 Active site directed irreversible inhibitors The ability of ESI-MS to characterise rapidly the complexes between proteins and covalent irreversible inhibitors, and in many cases to extend this to the identification of the specific site of the modification using proteolytic digests in conjunction with HPLC–ESI-MS, has played a major role in the increasing popularity of the technique.The high degree of sensitivity, and the ability to compare directly and rapidly the extent of modification with the remaining activity of the enzyme makes it the method of choice in these studies. The formation of acyl–enzyme inhibitor complexes between proteins and inhibitors has been a fertile area for ESI-MS studies. Farmer et al. have studied the inhibition of a range of ‚-lactamases with the penem BRL-42715 1.110 They used ESI-MS to show rapid and stoichiometric binding, and the observed mass diVerence between inhibited and uninhibited complexes confirmed the formation of an acyl-enzyme species with no further fragmentation of the inhibitor.Aplin et al. have also used ESI-MS to study the inhibition of ‚-lactamases. 6-‚-Halogenated penicillanic acids were seen to react with class A and C ‚-lactamases to give a mass consistent with the previously proposed enzyme bound dihydrothiazine derivative (Scheme 1).111 They were also able to observe the ‚-lactamase acyl-enzyme intermediate with a class C ‚-lactamase from Enterobacter cloacae P99 and four poor substrates that behave as inhibitors.112 A series of phosphonamidate peptide analogues were found to be inhibitors of the serine ‚-lactamase from the same organism,113 and ESI-MS was used to show that the reaction of the majority of these inhibitors with the enzyme resulted in apparent phosphonylation of the active site serine.The adjacent variable amino acid which forms the phosphonamidate bond acts as the leaving group in this case. The complex mechanisms of the inactivation of the TEM-2 ‚-lactamase from E. coli by clavulanic acid were unravelled by the use of ESI-MS.114 Four diVerent modified forms were observed, consisting of a serine O-linked decarboxylated form, a vinyl ether cross linked form, a hydrated aldehyde and a ‚-linked acrylate (Scheme 2). The sites of the modifications were then localised by proteolysis and HPLC–ESI-MS.These types of study are not limited to ‚-lactamases. The inhibition of porcine pancreatic elastase by two cephalosporins, L-647957 2 and L-658758 3, has also been studied by Aplin et al. using ESI-MS.115 The mass shifts that they observed were consistent with initial formation of an acylenzyme intermediate, followed by expulsion of the acetoxy group from the 3*-methylene position. Further elimination of HCl was observed for L-647957 2 (Scheme 3).The same research group was also able to use ESI-MS to study the inhibition of this enzyme by a number of chloro methyl ketones and a range of other alkylating inhibitors.116 A more detailed investigation of the inhibition of human leukocyte elastase and porcine pancreatic elastase by ‚-lactams using ESI-MS and NMR was able to determine that the mechanism of the chemical reaction was dependent on the structure of the ‚-lactam, and that the stability of the final complex is controlled by its molecular structure and conformation, which is also dependent on the initial ‚-lactam.117 Depending on the stereochemistry and constitution of the ‚-lactam, the inhibitor could be observed covalently attached to the enzyme intact, or after fragmentation, or after trapping of the fragmentation product by water to give a hydroxy amine.Other inhibitor complexes have also been seen in the case of proteases. For example (2R,3S)-2-benzyl-3,4-epoxybutanoic acid methyl N N N N S O CO2 – H 1 X N S O CO2 – S NH CO2 – O O Ser X = I, Br Scheme 1 O HN O O O O O HN O OH N O HN O OH O Ser 70 Ser 70 Ser 70 Asp 130 Ser 70 Scheme 2 64 Natural Product Reports, 1998ester was shown using ESI-MS and HPLC–ESI-MS to inhibit ·-chymotrypsin by covalently modifying the active site serine.118 An electrophilic alkylating analogue of ATP, 4 was used by McKay et al.to study the aminoglycoside antibiotic phosphotransferase responsible for the resistance to these drugs in many pathogenic bacteria.119 They were able to observe the inhibited complex, and using a combination of peptide mapping, Edman degradation and ESI-MS, were able to locate the modified sites in the protein as Lys33 and Lys44, indicating that the ATP binding site lies towards the N-terminus.The 2*,3*-dialdehyde (2*,3*-oxidised) analogue of ATP has been used to study the binding of ATP to GroEL.120 ESI-MS was used to show that the stoichiometry was approximately 1:1, and the use of peptide mapping with ESI-MS indicated that the covalent modification probably occurs at one of two cysteines, although the lability of the adduct made a definite identification diYcult.Ovine ceroid lipofuscinosis protein was shown by ESI-MS to have an identical mass to the F0 subunit c of bovine ATP synthase, and even has the same +42 mass units post-translational modification. Its reaction with a series of ATP synthase inhibitors was then studied by ESI-MS121 and shown to be similar to that of the F0 subunit and tandem mass spectrometric analysis allowed the site of modification to be determined.The identification of the active site nucleophile in the reaction of yeast ·-glucosidase was investigated using fluorinated inhibitors.122 A glucosyl-enzyme species was observed which was shown by tandem mass spectrometry and daughter ion analysis to be bound to Asp-214, one of the three conserved aspartates in the active site.The suicide inhibition of yeast cytochrome c peroxidase by resorcinol has been studied by ESI-MS.123 It was possible to confirm the previous findings that the resorcinol mainly became bound to the peptide and also to show that there were concomitantly two oxidations of the protein, probably on methionine, tyrosine or tryptophan residues and they were also able to use ESI-MS to demonstrate that the haem unit remained substantially unaltered, although a small amount of resorcinol modified haem could be identified.The substrate analogues 5-chlorolaevulinic acid 5 and 5-amino-3-thialaevulinic acid 6 have been shown by ESI-MS to be a non-specific alkylator and potent mechanism based inhibitor of B. subtilis 5-aminolaevulinic acid dehydratase, respectively, using ESI-MS (Scheme 4).124 The residue modified by the aYnity label N-bromoacetyl cellobiosylamine 7, which inhibits a Cellulomonas fimi exoglucanase, was unequivocally determined as Glu-127 by a combination of the ESI-MS analysis of tryptic digests and tandem mass spectrometry on the modified peptide.125 This confirmed the previous mutagenesis studies that suggested that this was the acid–base catalytic residue.ESI-MS was used to confirm that acetelynic GABA 8 binds covalently to GSH aminotransferase, with an observed mass increase equivalent to C6H6O3. This can be assigned to a species generated via attack of lysine on the conjugated acetylene, which is only formed slowly after tautomerisation of the initially formed imine, followed by hydrolysis (Scheme 5).126 The imine formed between 8 and GSH aminotransferase was too labile to be observed directly by ESI-MS, but on sodium borohydride reduction the resultant amine was observed.Mass spectrometry has been used very eVectively to study the mechanism of inhibition of the cysteine protease papain by the hydroxylamine derivative [Bz-Phe-Gly-NH-O-CO-(2,4,6-trimethylphenyl)].127 ESI-MS was used to identify the oxidised form of papain resulting from treatment with the peptidyl hydroxamate in the absence of a reducing agent as a sulfinic acid, along with a sulfenamide covalent adduct between the inhibitor and papain.The catalytic aspartate of soluble epoxide hydrolase was determined using ESI-MS to identify a number of radiolabelled peptides formed on incubation with the inhibitor 4-fluorochalcone oxide. The four peptides identified were overlapping and had only Asp-333 in common.128 The same group went on to use an elegant ESI-MS experiment to demonstrate the formation of a covalent intermediate and to con- firm the site of attachment of the inhibitor to soluble epoxide hydrolase.129 Using 18O-labelled water in single turnover experiments they were able to identify a tryptic peptide that N S O O OAc O OBut O Cl N S O O OAc O N O MeO N S O O O OBut –O2C O Cl O Ser 2 –HCl 3 Scheme 3 S F O O O O O N N N N NH2 OH OH 4 5 6 S CO2H H2N O CO2H Cl O S CO2H H2N O SH CO2H O NH H2N H2N Enz Enz Scheme 4 O O HO HO OH O OH OH OH HN O O O Br O O– HO 7 Glu 127 Pitt: Application of electrospray mass spectrometry in biology 65was labelled with 18O, and eventually to identify 18O incorporation into Asp-333 in this peptide, suggesting the intermediacy of an ·-hydroxyacyl-enzyme intermediate (Scheme 6). 4.2 Covalent enzyme–ligand complexes The study of covalent modification of proteins has been very eVectively extended to look at covalent interactions with substrates.Covalent enzyme–intermediate complexes can be seen when the breakdown of the intermediate complex is the rate-limiting step of the reaction. For example an acyl-protein intermediate has been observed in an antibody catalysed hydrolysis of a p-nitrophenyl ester;130 the covalent species, which showed an increase in mass consistent with the acyl portion of the substrate, accounted for about 8% of the total Fv concentration, and is not observed in the presence of the hapten, showing active site specificity.Interestingly, no covalent modification is observed on switching to the p-chlorophenyl ester, indicating that the nature of the leaving group is important in determining the rate limiting step. The E166Y mutant of TEM-1 ‚-lactamase shows the build up of an acyl-enzyme intermediate when treated with penicillin G, indicating that breakdown of this complex has become the rate limiting step.131 Acyl-enzyme intermediates have also been observed for the reaction of three serine proteases with cinnamoyl imidazole and indoleacryloyl imidazole.132 The mild nature of electrospray also makes it possible to observe other relatively labile enzyme–substrate complexes and a number of other types of enzyme–substrate and enzyme– product complexes have been observed by ESI-MS.Enzyme bound intermediates in the assembly of the tetrapyrrole unit that is the precursor to the tetrapyrrole pigments can be observed by ESI-MS.133 A series of intermediates with one to four of the monomeric porphobilinogen units that make up the final tetrapyrrole attached to the enzyme can be observed under substrate limited conditions (Fig. 7). Mutation of an essential histidine in the active site of E. coli type I dehydroquinase results in stalled catalysis with preferential formation of the SchiV base intermediate, as shown by isoelectric focusing and ESI-MS.134 This finding of a dual role for the histidine in catalysis and product release could explain the unexpected syn stereochemistry of the elimination process.Borthwick et al. used ESI-MS to observe the formation of the imine between E. coli dihydrodipicolinate synthase and pyruvate.135 They were also able to see the formation of the imine with the substrate analogues 3-bromopyruvate and 3-fluoropyruvate. The formation of an imine intermediate has also been detected between dehydroquinase and its substrate quinate using ESI-MS, con- firming the proposed mechanism.136 The requirement for the 4- and 5-hydroxy groups of dehydroquinic acid in the recognition of the substrate by dehydroquinase was studied using 5-deoxydehydro- and 4,5-dideoxydehydro-quinic acid and looking for the formation of the intermediate imine complex with the enzyme by ESI-MS.137 Distinct diVerences between type I and type II dehydroquinases were observed.The imine formed between pyridoxal phosphate and glutamate-1- semialdehyde aminotransferase is too labile to be observed by ESI-MS, but reduction of the imine with sodium borohydride gave a peak in the mass spectrum corresponding to covalently bound pyridoxal phosphate.126 The rates of phosphorylation and dephosphorylation of phosphoglycerate mutase from Saccharomyces cerevisiae and the fusion yeast Schizosaccharomyces pombe, which lacks the C-terminal section of the protein, were studied by ESI-MS.138 It was possible to determine the approximate rates of the addition and removal of the phosphate for the two enzymes, and also to observe the increase in the rate of dephosphorylation of the S.cerevisiae enzyme in the presence of the substrate analogue 2-phosphoglycolate. The straight forward measurement of rates of phosphorylation and dephosphorylation could be useful in the study of this important modification. 4.3 Non-covalent enzyme–inhibitor complexes ESI-MS has been shown to be an eVective method for the identification and characterisation of a wide range of noncovalent complexes, including protein–substrate, protein– inhibitor, protein–DNA, protein–protein and DNA–drug interactions.Low skimmer cone voltages are needed to avoid collisional activation from causing some dissociation, and careful choice of buVers and pH is essential. There have been a number of reviews in this area covering general eVects,139 the preservation of non-covalent associations, such as higher order protein structure, enzyme complexes and multimeric peptide association into the gas phase,140 and the evaluation of the performance of several designs of electrospray source for the detection of non-covalent complexes between ribonuclease A and cytidylic acids.141 One productive area has been the study of the binding of metal ions to proteins.The choice of conditions can play a major role in the successful detection of bound metal ions.NH2 NaBH4 NH2 CO2H N CO2H N NH CO2H N NH CO2H O NH2 NH2 N CO2H NH2 HN CO2H Lys PYP Lys PYP Lys PYP Lys PYP Lys PYP Lys PYP H+ 8 Scheme 5 H R R N NH O– O N NH O O N NH O– O OH O H O H H R = 18O * : * * Scheme 6 66 Natural Product Reports, 1998Studies on a number of metal substituted rubredoxins demonstrated that the apoenzyme form tended to predominate at acidic pHs using positive ion detection, and was the only form observed when the metal was zinc, whereas at neutral pH and using negative ion detection the predominant form was the holoenzyme.142,143 Jaquinod et al.144 have observed iron bound to a synthetic siderophore analogue and two iron– sulfur proteins using ESI-MS, and the mass spectrum of isopenicillin-N synthase also shows a species that appears to have the catalytic iron bound to the protein.145 The study of metal binding has been eVectively extended to zinc finger proteins.The interaction of zinc and copper with a 71 residue peptide containing two four-cysteine clusters was used to develop the methodology which was then exploited to study why copper might inhibit the function of this type of zinc finger.Species with up to two zinc or four copper atoms coordinated were observed, and a charge state shift (see Section 4.4) indicated a change in conformation on binding Cu in place of Zn.146–148 Electrospray data were obtained even with 100 micromolar zinc present. The calcium binding properties of calmodulin (up to 4 calcium ions)149 and a comparison of calmodulin and parvalbumen (up to 2 calcium ions)150 have also been determined by ESI-MS, with strong cooperativity being observed for the binding of the second calcium to parvalbumen and the fourth calcium to calmodulin, closely mimicking the observed solution behaviour of the proteins.The observation of bound cofactors is not limited to metal ions. Jaquinod et al.151 have observed the non-covalent complex between pig lens aldose reductase and NADP+ using positive ion ESI-MS.They were also able to see the covalent adduct formed with 3-chloroacetyldihydropyridineadeninedinucleotidephosphate 9, an alkylating analogue of NADP+. Drummond et al.152 isolated the proposed 28 kDa cobalmin binding domain from the 136 kDa methionine synthase and used ESI-MS to show that the isolated domain does indeed bind cobalamin, and they were also able to resolve some anomalies in the previously reported sequences.The eVect of A P A P A A P A P P A P A P A P A P ES3 ES4 ES2 NH NH S NH NH NH S NH ( a) ( b) H2N A P NH NH S A P A P A P HMBS holoenzyme NH3 NH A P A P A P PBG A HN HN A P ES1 NH3 A P NH P A P NH NH S NH HN HN NH NH S NH HN Porphobilinogen (PBG) A P PBG NH3 PBG NH3 Fig. 7 Observation of intermediate complexes for hydroxymethylbilane synthase. (a) Formation of intermediate complexes by the addition of 1 (denoted ES1) to possibly 4 (ES2, ES3 and ES4) substrate units to the dipyrromethane cofactor, and (b) MaxEnt (Micromass, Altrincham, Cheshire) deconvoluted spectrum of substrate limited experiment showing formation of the intermediate complexes (A.R. Pitt and A. R. Battersby, unpublished data). Pitt: Application of electrospray mass spectrometry in biology 67pH and solvent composition were shown to have significant eVects on the observation of non-covalent complexes between the ras protein and GDP and GTP.153 Apoenzyme, GDP and GTP bound enzyme were all observed in diVering amounts depending on the composition and pH of the solvent used in the ESI-MS.The factors that influence the stability of non-covalent complexes in the gas phase have been studied by using ESI-MS to measure the binding of acyl CoA to acyl CoA binding protein and the diVerences caused by a range of modifications of both enzyme and cofactor.154 Changes in chain length of the acyl portion had little eVect on the gas phase binding, although they did have a substantial eVect on the solution state binding, whereas removal of three tyrosine residues involved in key contacts with the ligand by mutagenesis gave substantially lower binding in the gas phase.The complex of bovine pancreatic trypsin inhibitor and soya bean trypsin inhibitor with trypsin and a K15V mutant, and a modified inhibitor used as a control, can all readily be observed by ESI-MS.155 There are changes in the charge state distribution of the components on binding which may give some indication of the contact surfaces between the protein and peptide (see Section 4.4).ESI-MS analysis of the binding of peptides to carbonic anhydrase has been used to screen two peptide combinatorial libraries consisting of 289 and 256 compounds for tight binding inhibitors.156 The ability of ESI-MS to look at comparative binding energies simultaneously for a complex mixture of compounds is a powerful technique in the analysis of libraries, and a compound with a binding constant of 1.4#10"8 M"1 was identified using this technique.The non-covalent binding of metal ions and two inhibitors to matrylysin (a matrix-metalloprotein) was studied at a range of pH values using ESI-MS.157 Below pH 2.2 the enzyme adopted a denatured conformation and no binding was observed. As the pH was raised above 4.5, binding of both metal ions and inhibitor was observed, and the stoichiometry appeared to be close to that observed at the optimum pH of 7.0 for the activity of the enzyme. The observed stoichiometry was for one inhibitor molecule, 2 calcium and 2 zinc ions.The intensity of the signal due to non-covalently bound ligand in the gas phase appeared to correlate with the known solution binding behaviour. Non-covalent adducts could be observed between the antibiotics vancomycin, ristocetin A and teicoplanin and two short peptides acting as analogues of the bacterial cell wall.158 However, the observed complexes ranged from simple homodimers of the antibiotics to complex associations of a number of antibiotic and peptide molecules, and relating these to the actual situation was not straightforward. It should be emphasised at this point that great care needs to be taken in assigning observed non-covalent complexes to specific binding, as there is ample evidence that some interactions are non-specific.For example, a non-specific interaction of cytidylic acids with ribonuclease A has been observed if an excess of the compound is added,159 but reliable results have been obtained provided suYcient care is taken to avoid non-specific eVects.160,161 Observations of non-covalent complexes are not limited to small molecule–large molecule interactions.The geneV protein is observed as a dimer by ESI-MS using ammonium acetate solution as the carrier solvent, and the addition of a 16mer oligonucleotide results in a 1:1 protein dimer–oligonucleotide complex.162 Furthermore, ESI-MS could be used to measure stoichiometries and relative binding constants for a range of oligonucleotides to the geneV dimer.A dimer was also observed in the ESI-MS of the oestrogen receptor ligand binding domain.163 The binding of biotin to the complete streptavidin tetramer was be observed by ESI-MS, and the streptavidin tetramer appeared to be particularly stable in the gas phase.164 A similar study used an extended mass range quadrupole analyser to study the binding of biotin (Kd, ca. 10"15 M) and iminobiotin (Kd, ca. 10"7 M) to the streptavidin tetramer.165 With biotin four molecules were observed bound to the tetrameric complex without any appearance of random aggregation. Under the same conditions full loading with iminobiotin was not observed. The thermally induced dissociation of the complex of biotin or iminobiotin with streptavidin in the mass spectrometer caused by applying softer or harsher conditions in the source appeared qualitatively to follow their solution binding behaviours.Some care needs to be taken in setting up the mass spectrometer for studies on tertiary association of proteins, as some observations of multimeric proteins again appear to be non-specific.166 4.4 Protein folding The ability of electrospray mass spectrometry to measure the average mass of a protein, and the apparent relationship between the most intense peaks observed in the charge state distribution and the folded state of the protein, makes ESI-MS a potent tool for the study of conformation, folding and dynamics in protein structures167–171 and cooperativity between the folding of elements of the structure.172 The simplest experiments involve the interpretation of a shift in the maximum intensity of the envelope of charged states towards higher numbers of charges (lower mass to charge ratios) to be due to a larger number of polar residues becoming exposed to the solvent as the protein changes conformation or unfolds.167 An alternative and more sensitive technique that is in many ways complementary to the charge state shift is to use ESI-MS to measure deuterium incorporation or washout from a protein.There appears to be a good correlation between the exchangeable protons and the folding or conformational changes of the protein, for example the number of protons that could be exchanged in tuna cytochrome c was in good agreement with that calculated using a computational method that determined the solvent accessible surface.173 ESI-MS has the advantage over NMR for measuring hydrogen–deuterium exchange in that it is not limited to small proteins that are soluble at mM concentrations.However, care needs to be taken as in some experiments it appears that diVerent protein conformations may have the same charge state distribution, but have diVerent exchange rates, and that the rapid interconversion of alternative conformations can lead to the same exchange rate being observed for alternative conformations showing diVerent charge states.174 The advantage of charge state distribution studies is that they are relatively rapid and simple to perform.There is some debate as to whether the solution (normal biochemical) and gas phase (in ESI-MS) behaviour of proteins is the same, but studies have indicated that the gas phase distribution of ions is in reasonable agreement with the calculated solution behaviour of the proteins.175 A number of observations of the changes in charge state distribution have already been mentioned.147–149,155,167 The addition of a small quantity of organic solvent to an aqueous solution of ubiquitin resulted in N Cl O O O OH OH O P O O– O P O– O O O N N N N NH2 OH OH 9 68 Natural Product Reports, 1998the observation of a bimodal charge–state distribution.176 The high mass to charge (low charge–state) distribution represents the protein in its native fully folded state, whereas the distribution with a higher charge state suggests a more extended conformation.Addition of more than 20% acetonitrile or 40% methanol completely eliminated the higher mass distribution by unfolding all of the protein. The refolding of acid denatured myoglobin could also be followed by this method, indicating that a compact native like structure forms initially without the assistance of the haem, which then associates with the newly formed binding site.177 In studies on proton–deuterium exchange of proteins, the rate of proton or deuterium exchange gives information on the conformational stability of proteins, whereas the total number of exchanged protons gives information on the folding of the protein.The study of the eVect of conformational changes on the extent of deuterium incorporation goes back many years to some of the first studies which were on the solvent dependent denaturing of ubiquitin.178 The same group went on to follow the eVect of reduction of the disulfide bonds of hen egg lysozyme on unfolding using the same techniques.179 A shift in the charge state distribution to a higher mass to charge ratio of both components on the binding of soya bean trypsin inhibitor to trypsin and a trypsin mutant have been used to suggest that less surface is available in the complex.155 The slower rate of exchange of protons in ·-helices was neatly demonstrated using mutagenesis on the 32-residue growth hormone releasing factor;180 introduction of helixpromoting or helix-disrupting residues resulted in slower or faster rates of exchange, respectively, and this correlated well with CD measurements of helical content.Addition of an organic solvent, which induces helix formation, also slowed the rate of exchange. Further studies of helical unfolding have been reported.181 Addition of a tight binding phosphopeptide ligand to a cloned Src homology 2 domain slowed proton incorporation into the protein dramatically.As NMR and X-ray crystallographic data show no significant conformational changes, the authors put the change in exchange rate down to an increase in conformational stability.182 Studies performed on apo-and holo-myoglobin see similar eVects; Johnson and Walsh183 showed that 47% of the amide hydrogens in apomyoglobin, but only 12% in holomyoglobin, exchange in 30 seconds. The authors went on to analyse individual peptides that had been generated by rapid peptic digests of the exchanged protein using tandem mass spectrometry, and were able to identify peptides where the exchange had been slowed highlighting the stabilised and destabilised regions of the protein.A combination of charge– state distribution, isotope exchange and CD analysis at a range of pHs was able to show the conformational instability of apomyoglobin and was able to identify two distinct charge states that probably represent diVerent, rapidly interconverting conformations.184,185 The folding of a 4-helix bundle has also been studied using rapid isotope exchange experiments.186 This technique was also applied to the determination of the changes in stability in Desulfovibrio vulgaris Hildenborough cytochrome c on mutation of a key tyrosine residue to a number of diVerent amino acids,187 which induced changes ranging from a slight increase in stability (Y64F) to significant destabilisation (Y64S).The eVect on the stability and conformation of the protein of mutagenesis of key residues of Rhodobacter capsulatus ferrocytochrome c could also be followed using the same hydrogen–deuterium exchange techniques.188,189 The use of exchange experiments monitored by ESI-MS in conjunction with complementary NMR studies has proved to be a powerful technique for unravelling diVerent pathways of protein folding.Miranker et al.190 have shown that ESI-MS gives the distribution of masses within a population, whereas NMR measures occupancy of individual sites; these two techniques can therefore be used together to provide information on the populations of transient folding intermediates.The rate of deuterium exchange and the higher number of protons protected from exchange for RTEM ‚-lactamase bound to GroEL at 48 )C, compared to the uncomplexed protein is good evidence that there is a significant amount of native structure remaining in the GroEL ‚-lactamase complex at these temperatures. 191 The protection of hydrogen exchange on ligand binding has been observed for the binding of acyl CoA to acyl CoA binding protein.153 In studies comparing the gas phase and solution phase behaviour of the complexes the rate of exchange was used to determine the extent of association and conformation stability of the complex under these conditions. 5 DNA ESI-MS can be used to observe short sections of single stranded DNA, including modified DNA for use as antisense compounds,192,193 duplex DNA194–198 and even quadruplex DNA.199 In its simplest application ESI-MS can be used to confirm the sequence of the DNA and, coupled to HPLC, to analyse more complex mixtures.200 One of the major problems with DNA is its aYnity for sodium ions, which can complicate the spectra to such an extent that it can be diYcult to make any interpretation.On-line microdialysis has been used to clean up DNA samples by removing sodium from the solution as it is introduced into the mass spectrometer, and by using ammonium acetate as the buVer the DNA remains in the duplex form.201 The addition of modifiers such as acetic or formic acid with imidazole or piperidine can also simplify assignment by shifting the signals to a higher mass to charge ratio, and can also suppress sodium ion adduct formation. 202,203 A micro electrospray source has been developed that has proved particularly useful for the analysis of duplex DNA and duplexes consisting of one strand of DNA and one of peptide nucleic acid.204 DNA duplex could be observed between the self complementary 12-mer (5*-dCGCAAATTT GCG-3*), and the non-covalent complex of this duplex DNA with the minor groove binding drugs distamycin, pentamidine and Hoechst 33258 could also be observed and the stoichiometry of binding was determined.205 The sensitivity of ESI-MS in detecting DNA modification was shown by the study of the formation of endogenous adducts with malondialdehyde in the liver.206 6 Summary The wide range of experiments using ESI-MS described in this review, most of which can be performed with the relatively simple quadrupole analyser, pay tribute not only to the hard work of scientists throughout the world, but also the flexibility, utility and simplicity of the technique.To say that ESI-MS has become routine in many laboratories is not understating the facts. The proliferation of these machines in many protein and peptide research laboratories, especially in their ‘benchtop’ form, will undoubtedly continue as they become an essential tool for analysis and research.The exciting new developments of quadrupole ion traps, FTICR and QTof will to some extent steal the limelight at the top end of the field, but there will be a place for the humble quadrupole for many years to come. While ESI-MS may not have revolutionised protein science, it is playing an increasingly important part in fundamental research, and if the number of papers published using electrospray (over 1500 from 1984 to 1996 of which, unfortunately, only a selection could be included in this review) is anything to go by, it has already made a very significant contribution to the field. 7 References 1 C. K. Meng, M. Mann and J. B. Fenn, Z. Phys. D., 1988, 10, 361. 2 M. Karas and F. Hillenkamp, Anal. Chem., 1988, 60, 2299. 3 G. Siuzdak, Proc. Natl. Acad. Sci. USA, 1994, 91, 11 290. 4 P. RoepstorV, Biochem.Soc. Trans., 1994, 22, 533. 5 D. Favretto, R. Seraglia, P. Traldi, O. Curcuruto and M. Hamdan, Org. Mass Spectrom., 1994, 29, 526. Pitt: Application of electrospray mass spectrometry in biology 696 J. R. Perkins, B. Smith, R. T. Gallagher, D. S. Jones, S. C. Davis, A. D. HoVman and K. B Tomer, J. Am. Soc. Mass Spectrom., 1993, 4 , 670. 7 P. RoepstorV, Trends Anal. Chem., 1993, 12, 413. 8 M. H. Allen, D. J. GrindstaV, M. L. Vestal and R. W. Nelson, Biochem. Soc. Trans., 1991, 19, 954. 9 R. F. Stack and P. J. Rudewicz, J. Mass Spectrom., 1995, 30, 857. 10 F. Saru and L. Montanarella, Reviews in Anal. Chem., 1995, 14, 205. 11 R. T. McIver, Jr., Y. Li and R. L. Hunter, Proc. Natl. Acad. Sci. USA, 1994, 91, 4801. 12 S. C. Beu, M. W. Senco, J. P. Quinn and F. W. McLaVerty, J. Am. Soc. Mass Spectrom., 1993, 4, 190. 13 J. A. Castoro and C. L. Wilkins, Anal. Chem., 1993, 65, 2621. 14 K. D. Henry, E. R. Williams, B. H. Wang, F. W. McLaVerty, J. Shabanowitz and D.F. Hunt, Proc. Natl. Acad. Sci. USA, 1989, 86, 9075. 15 B. E. Winger, S. A. Hofstadler, J. E. Bruice, H. R. Udseth and R. D. Smith, J. Am. Soc. Mass Spectrom., 1993, 4, 566. 16 M. G. Qian and D. M. Luberman, Anal. Chem., 1995, 67, 234. 17 S. A. McLuckey, G. J. Van Berkel, D. E. Goeringer and G. L. Glish, Anal. Chem., 1994, 66, 737A. 18 L. Voress, Anal. Chem., 1994, 66, 481A. 19 P. Kebarle and L. Tang, Anal. Chem., 1993, 65 , 972A. 20 M. Mann, Org. Mass Spectrom., 1990, 25, 575. 21 K. W. M. Siu, R. Guevremont, J. C. Y. LeBlanc, R. T. O’Brian and S. S. Bermann, Org. Mass Spectrom., 1993, 28, 579. 22 D. S. Ashton, C. R. Beddell, D. J. Cooper, B. N. Green and R. W. A. Oliver, Org. Mass Spectrom., 1993, 28, 721. 23 G. Siuzdak, Proc. Natl. Acad. Sci. USA, 1994, 91, 11 290. 24 K. L. Bennet, L. A. Hick, R. J. W. Truscott, M. M. Shiel and S. V. Smith, J. Mass Spectrom., 1995, 30, 769. 25 J. F. Banks and C. M. Whitehouse, Methods Enzymol., 1996, 270, 486. 26 J.S. Andersen, B. Svensson and P. RoepstorV, Nature Biotechnol., 1996, 14, 449. 27 V. M. Orlov, V. N. Pustobaev and R. N. Maslova, Mol. Biol., 1996, 30, 444. 28 S. A. Hofstadler, R. Bakhtiar and R. D. Smith, J. Chem. Educ., 1996, 73, 82. 29 R. Bakhtiar, S. A. Hofstadler and R. D. Smith, J. Chem. Educ., 1996, 73, 118. 30 J. A. Loo, R. R. O. Loo, K. J. Light, C. G. Edmonds and R. D. Smith, Anal. Chem., 1991, 64, 81. 31 K. Biemann, Ann. Rev. Biochem., 1992, 61, 977. 32 I.Jardine, Nature, 1990, 345, 747. 33 J. B. Fenn, M. Mann, C. K. Meng, S. F. Wong and C. M. Whitehouse, Mass Spec. Rev., 1990, 9, 37. 34 J. B. Fenn, M. Mann, C. K. Meng, S. F. Wong and C. M. Whitehouse, Science, 1989, 246, 64. 35 C. L. Holliman, D. L. Rempel and M. L. Gross, Mass Specrom. Rev., 1994, 13, 105. 36 G. A. Valaskovic, N. L. Kelleher and F. W. McClaVerty, Science, 1996, 273, 5279. 37 R. Feng, Y. Konishi and A. W. Bell, J. Am. Soc. Mass Spectrom., 1991, 2, 387. 38 A.G. Ferrige, M. J. Seddon, J. Skilling and N. Ordsmith, Rapid Commun. Mass Spectrom., 1992, 6, 765. 39 A. G. Brenton and C. M. Lock, Rapid Commun. Mass Spectrom., 1995, 9, 143. 40 D. C. James, M. H. Goldman, M. Hoare, N. Jenkins, R. W. A. Oliver, B. N. Green and R. B. Freedman, Prot. Science, 1996, 5, 331. 41 B. N. Green, T. Suzuki, T. Gotoh, A. R. Kuchumov and S. N. Vingradov, J. Biol. Chem., 1995, 270, 18 209. 42 A. G. Ferrige, M. J. Seddon, B. N. Green, S. A. Jarvis and J.R. Skilling, Rapid Commun. Mass Spectrom., 1992, 6, 707. 43 J. B. CuniV and P. Vuoros, J. Am. Soc. Mass Spectrom., 1995, 6, 1175. 44 V. Katta, S. K. Chowdhury and B. T. Chait, Anal. Chem., 1991, 63, 174. 45 J. A. Loo, C. G. Edmonds and R. D. Smith, Science, 1990, 248, 201. 46 B. L. Gillecastro and J. T. Stults, Methods Enzymol., 1996, 271, 427. 47 J. S. Andersen, Biochem. Soc. Trans., 1995, 23, 917. 48 O. RoepstorV, K. H. Schram, J. S. Andersen, K. Rafn, T. Baldursson, J. Kroll, K.Poulsen, J. Knudsen and K. Kristiansen, Mol. Biotechnol., 1995, 4, 1. 49 M. Mann and M. Wilm, TIBS, 1995, 20, 219. 50 D. N. Nguyen, G. W. Becker and R. M. Riggin, J. Chromatogr., 1995, 705, 21. 51 I. M. Fearnley and J. E. Walker, Biochem. Soc. Trans., 1996, 24, 912. 52 K. M. Swiderek, T. D. Lee and J. E. Shiveley, Methods Enzymol., 1996, 271, 377. 53 M. C. Fitzgerald and G. Siuzdak, Chem. Biol., 1996, 3, 707. 54 T. Kinumi, M. Kubota and M. Ohashi, J. Synth. Org. Chem. Jpn., 1996, 54, 375. 55 J.A. Loo, Bioconjugate Chem., 1995, 6, 644. 56 R. P. Evershed, D. H. L. Robertson, R. J. Benyon and B. N. Green, Rapid Comm. Mass Spectrom., 1993, 7, 882. 57 T. L. Calderone, R. D. Stevens and T. G. Oas, J. Mol. Biol., 1996, 262, 407. 58 N. Budisa, B. Steipe, P. Demange, C. Eckerskorn, J. Kellermann and R. Huber, Eur. J. Biochem., 1995, 230, 788. 59 R. W. A. Oliver, B. N. Green and D. J. Harvey, Biochem. Soc. Trans., 1996, 24, 917. 60 J. Vandergreef and W. M.A. Niessen, Int. J. Mass Spectrom. Ion Processes, 1992, 118, 857. 61 Q. Tang, A. K. Harrata and C. S. Lee, Anal. Chem., 1996, 68, 2482. 62 T. Hutton and H. J. Major, Biochem. Soc. Trans, 1995, 23, 924. 63 A. P. Bruins, Trends Anal. Chem., 1994, 13, 81. 64 G. A. Sweetman, G. V. Garner, D. B. Gordon, L. Tetler and R. D. G. Theakston, Rapid Commun. Mass Spectrom., 1992, 6, 724. 65 M. J. F. Suter, B. B. Dague, W. T. Moore, S. N. Lin and R. M. Caprioli, J. Chromatogr., 1991, 553, 101. 66 F. E. Kuhlmann, A. ApVel, S. Fischer, G. Goldberg and P. C. Goodley, J. Am. Soc. Mass Spectrom., 1995, 6, 1221. 67 A. ApVel, S. Fischer, G. Goldberg, P. C. Goodley and F. E. Kuhlmann, J. Chromatogr., 1995, 712, 177. 68 J. A. Chakel, J. A. ApVel and W. S. Hancock, LC–GC, 1995, 13, 866. 69 W. S. Hancock, A. ApVel, J. Chakel, C. Souders, T. Mtimkulu, E. Pungor and A. W. Guzzetta, Methods Enzymol., 1996, 271, 403. 70 J. A. Loo, H. K. Jones, H. R. Udseth and R. D. Smith, J.Microcol. Sep., 1989, 1, 223. 71 R. D. Smith, H. R. Udseth, J. H. Wahl, D. R. Goodlett and S. A. Hofstadler, Methods Enzymol., 1996, 271, 448. 72 P. Thibauld, C. Paris and S. Pleasance, Rapid Commun. Mass. Spectrom., 1991, 5, 484. 73 M. A. Moseley, J. Shabanowitz, D. F. Hunt, K. B. Tomer and J. W. Jorgensen, J. Am. Soc. Mass Spectrom., 1992, 3, 289 74 M. Hamdan, O. Curcuruto, S. Vigano and P. Rovero, Rapid Commun. Mass Spectrom., 1996, 10, 1128. 75 K. B. Tomer, L. J. Deterding and C.E. Parker, Adv. Chromatogr., 1995, 35, 53. 76 C. Shumate, Trends Anal. Chem., 1994, 13, 104. 77 J. H. Wahl, D. R. Goodlett, H. R. Udseth and R. D. Smith, J. Am. Chem. Soc., 1993, 115, 803. 78 C. G. Edmonds, J. A. Loo, R. R. O. Loo, H. R. Udseth, C. J. Barinaga and R. D. Smith, Biochem. Soc. Trans., 1991, 19, 943. 79 C. G. Edwards, J. A. Loo, S. M. Fields, C. J. Barinaga, H. R. Udseth and R. D. Smith, Biological Mass Spectrometry, eds. A. C. Burlingame and J. A. McCloskey, Elsevier Science Publishers, Amsterdam, 1990. 80 R. D. Smith, J. A. Olivares, N. T. Nguyen and H. R. Udseth, Anal. Chem., 1988, 60, 4671. 81 R. D. Smith, C. J. Barinaga and H. R. Udseth, Anal. Chem., 1988, 60, 1948. 82 V. N. Reinhold and B. B. Reinhod, Methods Enzymol., 1996, 271, 377. 83 B. Mulroney, J. C. Traeger and B. A. Stone, J. Mass Spectrom., 1995, 30, 1277. 84 L. C. Packman and A. Berry, Eur. J. Biochem., 1995, 227, 510. 85 B. A. Bruenner, A. D. Jones and J. B. German, Chem. Res.Toxicol., 1995, 8, 552. 86 M. S. Bolgar and S. J. Gaskell, Anal. Chem., 1996, 68, 2325. 87 U. A. Mirza, B. T. Chait and H. M. Lander, J. Biol. Chem., 1995, 270, 17 185. 88 H. Y. Jin, J. J. Emanuele, R. Fairman, J. G. Robertson, M. E. Hail, H. T. Ho, P. J. Falk and J. J. Villafranca, Biochemistry, 1996, 35, 1423. 89 T. Krell, M. J. Horsburgh, A. Cooper, S. M. Kelly and J. Coggins, J. Biol. Chem., 1996, 271, 24492. 90 J. R. Bottomly, A. R. Hawkins and C. Kleanthous, Biochem.J., 1996, 319, 707. 91 T. Krell, A. R. Pitt and J. R. Coggins, FEBS Lett., 1995, 360, 93. 70 Natural Product Reports, 199892 J. P. W. Boots, W. D. VanDongen, H. M. Verheij, G. H. Dehas, J. Haverkamp and A. J. Slotboom, Biochem. Biophys. Acta, 1995, 1248, 27. 93 C. A. Costello, N. L. Kelleher, M. Abe, F. W. McClaVerty and T. P. Begley, J. Biol. Chem., 1996, 271, 3445. 94 K. Ue, A. Muhlrad, C. G. Edmonds, D. Bivin, A. Clark, W. V. Piechowski and M. F. Morales, Eur.J. Biochem., 1992, 203, 493. 95 J. C. Prome, Analysis, 1991, 19, 79. 96 M. Soggard, J. S. Andersen, P. RoepstorV and B. Svensson, Biotechnology, 1993, 11, 1162. 97 P. A. Schindler, C. A. Settineri, X. Collet, C. J. Fielding and A. L. Burlingame, Protein Sci., 1995, 4, 791. 98 D. S. Ashton, C. R. Beddel, D. J. Cooper, S. J. Craig, A. C. Lines, R. W. A. Oliver and M. A. Smith, Anal. Chem., 1995, 67, 835. 99 A. Dell, H. R. Morris, R. L. Easton, M. Panico, M. Patankar, S. Oehninger, R.Koistinen, H. Koistinen, M. Seppala and G. F. Clark, J. Biol. Chem., 1995, 270, 24 116. 100 F. W. Herberg, S. M. Bell and S. S. Taylor, Protein Eng., 1993, 6, 771. 101 S. E. Tsutakawa, K. F. Medzihradszky, A. J. Flint, A. L. Burlingame and D. E. Koshland, J. Biol. Chem., 1995, 270, 26 807. 102 J. Z. Hou, K. McKeehan, M. K. Kan, S. A. Carr, M. J. Huddleston, J. W. Crabb and W. L. McKeehan, Protein Sci., 1993, 2, 86. 103 R. R. Townsend, P. H. Lipniunas, B. M. Tulk and A.S. Verkman, Protein Sci., 1996, 5, 1865. 104 M. Busman, K. L. Schey, J. E. Oatis and D. R. Knapp, J. Am. Soc. Mass Spectrom., 1996, 7, 243. 105 S. A. Morris, W. P. Revill, J. Staunton and P. F. Leadlay, Biochem. J., 1993, 294, 521. 106 P. CaVrey, B. Green, L. C. Packman, B. J. Rawlings, J. Staunton and P. F. Leadley, Eur. J. Biochem., 1991, 195, 823. 107 A. M. Bridges, P. F. Leadley, W. P. Revill and J. Staunton, J. Chem. Soc., Chem. Commun., 1991, 11, 778. 108 M. J. Runswick, I.M. Fearnley, J. M. Skehel and J. E. Walker, FEBS Lett., 1991, 286, 121. 109 L. C. Packman, M. Mewies and N. S. Scrutton, J. Biol. Chem., 1995, 270, 13 186. 110 T. H. Farmer, J. W. J. Page, D. J. Payne and D. J. C. Knowles, Biochem. J., 1994, 303, 825. 111 R. T. Aplin, C. V. Robinson, C. J. Schofield and N. J. Westwood, J. Chem. Soc., Chem. Commun., 1993, 121. 112 R. T. Aplin, J. C. Baldwin, C. J. Schofield and S. G. Whaley, FEBS Lett., 1990, 277, 212. 113 D. J. Payne, J.H. Bateson, D. Tolson, B. Gasson, T. Khushi, P. Ledent and J. M. Frere, Biochem. J., 1996, 314, 457. 114 R. P. A. Brown, R. T. Aplin and C. J. Schofield, Biochemistry, 1996, 35, 12 421. 115 R. T. Aplin, C. V. Robinson, C. J. Schofield and N. T. Westwood, Tetrahedron, 1993, 49, 10 903. 116 R. T. Aplin, C. V. Robinson, C. J. Schofield and N. T. Westwood, J. Chem. Soc., Chem. Commun., 1992, 1650. 117 D. J. Underwood, B. G. Green, R. Chabin, S. Mills, J. B. Doherty, P. E. Finke, M.MacCoss, S. K. Shah, C. S. Burgey, T. A. Dickinson, P. R. GriYn, T. E. Lee, K. M. Swiderek, T. Covey, W. M. Westler and W. B. Knight, Biochemistry, 1995, 34, 14 344. 118 Y. J. Kim, Z. H. Li, D. H. Kim and J. H. Hahn, Bioorg. Med. Chem. Lett., 1996, 6, 1449. 119 G. A. McKay, R. A. Robinson, W. S. Lane and G. D. Wright, Biochemistry, 1994, 33, 14 115. 120 G. J. Thomson, J. R. Coggins and N. C. Price, FEBS Lett., 1993, 336, 19. 121 A. Buzy, E. M. Ryan, K. R. Jennings, D. N.Palmer and D. E. GriYths, Rapid Commun. Mass Spectrom., 996, 10, 790. 122 J. D. McCarter and S. G. Withers, J. Biol. Chem., 1996, 271, 6889. 123 D. R. Doerge, R. L. Divi, T. Hutton and S. Lowes, J. Mass Spec., 1995, 55, 5136. 124 D. Appleton and F. J. Leeper, Chem. Commun., 1996, 303. 125 D. Tull, D. L. Burgoyne, D. T. Chow, S. G. Withers and R. Aebersold, Anal. Biochem., 1996, 234, 119. 126 S. Brody, J. S. Andersen, C. G. Kanangara, M. Meldgaard, P. Roepstorf and D. vonWettstein, Biochemistry, 1995, 34, 15 918. 127 R. Menard, R. Feng, A. C. Storer, V. J. Robinson, R. A. Smith and A. Krantz, FEBS Lett., 1991, 295, 27. 128 F. Pinot, D. F. Grant, J. K. Beetham, A. G. Parker, B. Borhan, S. Landt, A. D. Jones and B. D. Hammock, J. Biol. Chem., 1995, 270, 7968. 129 B. Bohran, A. D. Jones, F. Pinot, D. F. Grant, M. J. Kurth and B. D. Hammock, J. Biol. Chem., 1995, 270, 26 923. 130 J. F. Krebs, G. Siuzdak, H. J. Dyson, J. D. Stewart and S. J. Benkovic, Biochemistry, 1995, 34, 720 131 L. Maveyraus, I.Saves, O. Burletschiltz, P. Swaren, J. M. Masson, M. Delaire, L. Mourey, J. C. Prome and J. P. Samama, J. Biol. Chem., 1996, 271, 10 482. 132 D. S. Ashton, C. R. Bedell, D. J. Cooper, B. N. Green, R. W. A. Oliver and K. J. Welham, FEBS Lett., 1991, 292, 201. 133 R. T. Aplin, J. E. Baldwin, C. Pichon, C. A. Roessner, A. I. Scott, C. J. Schofield, N. J. Stolowich and M. J. Warren, Bioorg. Med. Chem. Lett., 1991, 1, 503. 134 A. P. Leech, R.James, J. R. Coggins and C. Kleanthous, J. Biol. Chem., 1995, 270, 25 827. 135 E. B. Borthwick, S. J. Connell, D. W. Tudor, D. J. Robins, A. Shneier, C. Abel and J. R. Coggins, Biochem. J., 1995, 305, 521. 136 A. Shneier, C. Kleanthous, A. Deka, J. R. Coggins and C. Abel, J. Am. Chem. Soc., 1991, 113, 9416. 137 J. M. Harris, W. J. Watkins, A. R. Hawkins, J. C. Coggins and C. Abel, J. Chem. Soc., Perkin Trans. 1, 1996, 2317. 138 J. Nairn, T. Krell, J. R. Coggins, A. R. Pitt, L. A.Fothergill- Gilmore, R. Walter and N. C. Price, FEBS Lett., 1995, 359, 192. 139 R. D. Smith and K. J. Lightwahl, Biol. Mass Spectrom., 1993, 22, 493. 140 R. D. Smith, Org. Mass. Spectrom., 1992, 27, 811. 141 N. J. Haskins, A. E. Ashcroft, A. Phillips and M. Harrison, Rapid Commun. Mass Spectrom., 1994, 8 , 120. 142 Y. Petillot, E. Forest, J. Meyer and J. M. Moulis, Anal. Biochem., 1995, 228, 56. 143 Y. Petillot, E. Forest, I. Mathieu, J. Meyer and J. M. Moulis, Biochem.J., 1993, 296, 657. 144 M. Jaquinod, E. Leize, N. Poitier, A. M. Albrecht, A. Schanzer and A. Van Dorsselaer, Tetrahedron Lett., 1993, 34, 2771. 145 R. T. Aplin, J. E. Baldwin, Y. Fujishima, C. J. Schofield, B. N. Green and S. A. Jarvis, FEBS Lett., 1990, 264, 215. 146 T. W. Hutchens and M. H. Allen, Rapid Commun. Mass. Spectrom., 1992, 6, 469. 147 M. H. Allen and T. W. Hutchens, Rapid Commun. Mass Spectrom., 1992, 6, 308. 148 T. W. Hutchens, M. H. Allen, C. M. Li and T. T. Yip, FEBS Lett., 1992, 309, 170. 149 D. Lafitte, J. P. Capony, G. Grassy, J. Haiech and B. Calas, J. Mass Spectrom., 1995, 30, 5192. 150 P. F. Hu and J. A. Loo, J. Mass Spectrom., 1995, 30, 1076. 151 M. Jaquinod, N. Potier, K. Klarskov, J. M. Reymann, O. Sorokine, S. KieVer, P. Barth, V. Andriantomanga, J. F. Biellmann and A. VanDorsselaer, Eur. J. Biochem., 1993, 218, 893. 152 J. T. Drummond, R. R. O. Loo and R. G. Matthews, Biochemistry, 1993, 32, 9282. 153 A. K. Ganguly, B. N. Pramanik, E.C. Huang, A. Tsarbopoulos, V. M. Girijavallabhan and S. Liberles, Tetrahedron, 1993, 49, 7985. 154 C. V. Robinson, E. W. Chung, B. B. Kragelund, J. Knudsen, R. T. Aplin, F. M. Poulsen and C. M. Dobson, J. Am. Chem. Soc., 1996, 118, 8646. 155 J. A. E. Kraunsoe, R. T. Aplin, B. Green and G. Lowe, FEBS Lett., 1996, 396, 108. 156 J. M. Gao, X. H. Cheng, R. D. Chen, G. B. Sigal, J. E. Bruce, B. L. Schwartz, S. A. Hofstadler, G. A. Anderson, R. D. Smith and G. M. Whitesides, J.Med. Chem., 1996, 39, 1949. 157 R. Feng, A. L. Castelhano, R. Billedeau and Z. Y. Yuan, J. Am. Soc. Mass Spectrom., 1995, 6, 1105. 158 M. Hamdan, O. Curcurito and E. Dimodugno, Rapid Commun. Mass Spectrom., 1995, 9, 883. 159 P. Camilleri and N. J. Haskins, Rapid Commun. Mass Spectrom., 1993, 7, 603. 160 M. Przybylski and M. O. Glocker, Angew. Chem., Int. Ed. Engl., 1996, 35, 807. 161 D. L. Smith and Z. Q. Zhang, Mass Spectrom. Rev., 1994, 13, 411. 162 X. H. Cheng, A. C. Harms, P. N.Goudreau, T. C. Terwilliger and R. D. Smith, Proc. Natl. Acad. Sci. USA, 1996, 93, 7022. 163 H. E. Witowska, B. N. Green, M. Carlquist and C. H. L. Shackleton, Steroids, 1996, 61, 433. 164 K. Eckart and J. Speiss, J. Am. Soc. Mass Spectrom., 1995, 6, 912. 165 B. L. Schwartz, D. C. Gale, R. D. Smith, A. Chilkoti and P. S. Stayton, J. Mass Spectrom., 1995, 30, 1095. 166 M. Busman, D. R. Knapp and K. L. Schey, Rapid Commun. Mass Spectrom., 1994, 8, 211. Pitt: Application of electrospray mass spectrometry in biology 71167 B.A. Collings and D. J. Douglas, J. Am. Chem. Soc., 1996, 118, 4488. 168 D. S. Wagner, L. G. Melton, Y. Yan, B. W. Erickson and R. J. Anderegg, Protein Sci., 1994, 3, 1305. 169 C. V. Robinson and S. E. Radford, Structure, 1995, 3, 861. 170 C. V. Robinson, M. Grob, S. J. Eyles, J. J. Ewbank, M. Mayhew, F. U. Hartl, C. M. Dobson and S. E. Radford, Nature, 1994, 372, 645. 171 M. E. Hemling, J. J. Conboy, M. F. Bean, M. Metzner and S. A. Carr, J. Am. Soc. Mass. Spectrom., 1994, 5, 434. 172 S. D. Hooke, S. J. Eyles, A. Miranker, S. E. Radford, C. V. Robinson and C. M. Dobson, J. Am. Chem. Soc., 1995, 117, 7548. 173 C. Brown, P. Camilleri, N. J. Haskins and M. Saunders, J. Chem. Soc., Chem. Commun., 1992, 761. 174 D. S. Wagner and R. J. Anderegg, Anal. Chem., 1994, 66, 706. 175 R. Guevremont, K. W. M. Siu, J. Y. C. Leblanc and S. S. Berman, J. Am. Soc. Mass Spectrom., 1992, 3, 216. 176 J. A. Loo, R. R. O. Loo, H. R. Udseth, C. G. Edmonds and R. D. Smith, Rapid. Commun. Mass. Spectrom., 1991, 5, 101. 177 R. Feng and Y. Konishi, J. Am. Soc. Mass Spectrom., 1993, 4, 638. 178 V. Katta and B. T. Chait, Rapid Commun. Mass. Spectrom., 1991, 5, 214. 179 V. Katta and B. T. Chait, J. Am. Chem. Soc., 1993, 115, 6317. 180 C. L. Stevenson, R. J. Anderegg and R. T. Borchardt, J. Am. Soc. Mass Spectrom., 1993, 4, 646 181 R. J. Anderegg, D. S. Wagner, C. L. Stevenson and R. T. Borchardt, J. Am. Soc. Mass Spectrom, 1994, 5, 425. 182 R. J. Anderegg and D. S. Wagner, J. Am. Chem. Soc., 1995, 117, 1374. 183 R. S. Johnson and K. A. Walsh, Protein Sci., 1994, 3, 2411. 184 F. Wang and X. J. Tang, Biochemistry, 1996, 35, 4069. 185 Y. Goto, L. J. Calciano and A. L. Fink, Proc. Natl. Acad. Sci. USA, 1990, 87, 573. 186 B. B. Kragelund, C. V. Robinson, J. Knudsen, C. M. Dobson and F. M. Poulsen, Biochemistry, 1995, 34, 7217. 187 P. Guy, M. Jaquinod, H. Remigy, J. P. Andrieu, J. Gagnon, B. Bersch, A. Dolla, L. Blanchard, F. Guerlesquin and E. Forest, FEBS Lett., 1996, 395, 53. 188 M. Jaquinod, P. Guy, F. Halgand, M. CaVrey, J. Fitch, M. Cusanovich and E. Forest, FEBS Lett., 1996, 380, 44. 189 M. Jaquinod, F. Halgand, M. CaVrey, C. Saintpierre, J. Gagnon, J. Fitch, M. Cusanovich and E. Forest, Rapid Commun. Mass. Spectrom., 1995, 9, 1135. 190 A. Miranker, C. V. Robinson, S. E. Radford, R. T. Aplin and C. M. Dobson, Science, 1993, 262, 896. 191 P. Gervasoni, W. Staudenmann, P. James, P. Gehrig and A. Pluckthun, Proc. Natl. Acad. Sci. USA, 1996, 93, 12 189. 192 B. Bothner, K. Chatman, M. Sarkisian and G. Siuzdak, Bioorg. Med. Chem. Lett., 1995, 5, 2863. 193 T. R. Baker, T. Keough, R. L. M. Dobson, T. A. Riley, J. A. Hasselfield and P. E. Hasselberth, Rapid. Commn. Mass Spectrom., 1993, 7, 190. 194 J. M. Ding and R. J. Anderegg, J. Am. Soc. Mass Spectrom., 1995, 6, 159. 195 E. Bayer, T. Bauer, K. Schmeer, K. Bleicher, M. Maler and H. J. Gaus, Anal. Chem., 1994, 66, 3858. 196 K. Bleicher and E. Bayer, Biol. Mass Spectrom., 1994, 23, 320. 197 K. J. Lightwahl, D. L. Springer, B. E. Winger, C. G. Edmonds, D. G. Camp, B. D. Thrall and R. D. Smith, J. Am. Chem. Soc., 1993, 115, 803. 198 B. Ganem, Y. T. Li and J. D. Henion, Tetrahedron Lett., 1993, 34, 1445. 199 D. R. Godlett, D. G. Camp, C. C. Hardin, M. Corregan and R. D. Smith, Biol. Mass Spectrom., 1993, 3, 181. 200 K. Bleicher and E. Bayer, Chromatographia, 1994, 39, 405. 201 C. L. Liu, Q. Y. Wu, A. C. Harms and R. D. Smith, Anal. Chem., 1996, 68, 3295. 202 D. C. Muddiman, X. H. Cheng, H. R. Udseth and R. D. Smith, J. Am. Soc. Mass. Spectrom., 1996, 7, 697. 203 M. Greig and R. H. GriVey, Rapid Commun. Mass Spectrom., 1995, 9, 97. 204 M. J. Greig, H. J. Gaus and R. H. GruiVey, Rapid. Commun. Mass Spectrom., 1996, 10, 47. 205 D. C. Gale and R. D. Smith, J. Am. Soc. Mass. Spectrom., 1995, 6, 1154. 206 A. K. Chaudhary, M. Nokubo, T. D. Oglesby, L. J. Marnett and I. A. Blair, J. Mass Spectrom., 1995, 30, 1157. 72 Natural Product Reports, 1998