Preparative scale mass spectrometry: A brief history of the calutron.

Calutrons were developed in the laboratory of E. O. Lawrence at the University of California at Berkeley. They were a modification of the cyclotrons he had invented and used in his Nobel Prize-winning investigations of the atomic nucleus. At the time their construction was undertaken, calutrons represented the only certain means of preparing enriched uranium isotopes for the construction of a fission bomb. The effort was successful enough that every atom of the 42 kg of 235 U used in the first uranium bomb had passed through at least one stage of calutron separation. At peak production, the first stage separators, α tanks, yielded an aggregate 258-g/day 235 U enriched to about 10 at. % from its natural abundance level of 0.72 at. %. The second stage separators, ß tanks, used the 10 at. % material as feedstock and produced a total 204-g/day 235 U enriched to at least 80 at. %. The latter, weapons grade, material was used in fission bombs. Under typical operating conditions, each α tank operated at a uranium beam intensity at the collectors of approximately 20 mA and each ß tank at a beam intensity of approximately 215 mA at the collectors. Bulk separation of isotopes for bomb production ceased in 1945. Since that time calutrons have been used to separate stable isotopes, but on a more limited scale than wartime weapons production. Stable isotope separations since 1960 have taken place using one modified ß tank.

Direct vertical electroelution of protein from a PhastSystem band for mass spectrometric identification at the level of a few picomoles.

An electroelution apparatus prototype of a new design was constructed. In that design, the electric field passes vertically through the protein band located on a horizontal (PhastSystem) minigel polymerized on a net of Gel-Fix (Serva). A simple, home-made apparatus allows for electroelution of protein bands at the level of a few picomoles and their identification, after concentration, by matrix-assisted laser desorption/ionization-time of flight mass spectrometry. The technique is applicable to one-dimensional (1-D) or two-dimensional (2-D) gels of any size, but has been exemplified only by application to 1-D minigels to demonstrate the lower limits of protein load of the method. When in the course of further development of the prototype it will be combined with a modification to two dimensions of the electroelution mechanism under computer control of the high-performance gel electrophoresis apparatus (formerly of LabIntelligence), the new design appears uniquely qualified for an automated spot elution from 2-D gels under avoidance of gel sectioning.

Calcium fractional absorption and metabolism assessed using stable isotopes differ between postpartum and never pregnant women.

Determining the fractional absorption (FA) of calcium using the incorporation into urine of stable isotopes given intravenously (IV) and orally has become a routine procedure. We investigated the FA of calcium in two groups of (2-3 mo) postpartum women lactating (LACT) (n = 6) and nonlactating (PPNL) (n = 6), and in never pregnant (NP) women (n = 7). The women consumed a controlled diet containing 30-33 mmol/d calcium (Ca) for 21 d. On d 7 of the controlled diet, the women received 0.05 mmol of 42Ca IV and 0.25 mmol 44Ca orally in milk. Urine samples (24-h) were collected for the next 14 d and morning blood samples were collected from fasting subjects before dosing and at 24 and 48 h after receiving the isotopes. Milk samples from the LACT women were collected from each feeding beginning 24 h before to 72 h after dosing. There were no significant differences in the FA of calcium as measured by stable isotope incorporation into urine (23.8 +/- 2.9%), serum (24.0 +/- 3.4%) or milk (23.6 +/- 3.6%) of LACT women. The fractional calcium absorption measured in urine of the postpartum women (LACT and PPNL, 23.8 +/- 2.9% and 25.0 +/- 3.3%, respectively) did not differ but was greater (P < 0.028) than that of the NP women (17.3 +/- 1.3%). The postpartum LACT and PPNL women had a reduced urinary excretion of calcium (P < 0.01) compared with the NP women. There was a significantly greater incorporation (P < 0.001) by LACT women of the oral isotope dose into milk than into urine. Calcium FA can be determined from incorporation of stable isotopes into breast milk and serum as well as urine.

Sequential electroelution and mass spectroscopic identification of intact sodium dodecyl sulfate-proteins labeled with 5(6)-carboxyfluorescein-N-hydroxysuccinimide ester.

A gel electrophoresis apparatus capable of scanning the migration path fluorometrically and of computer-directed electroelution of bands was applied to the mass spectrometric identification of sequentially electroeluted 5(6)-carboxyfluorescein-N-hydrosuccinimide ester (FLUOS)-labeled sodium dodedyl sulfate (SDS)-proteins. The masses of four electroeluted SDS-proteins under study determined by matrix assisted laser desorption/ionization-time of flight (MALDI-TOF) spectrometry are changed by 1% due to their reaction with FLUOS in a 1:5 molar ratio of protein:label, allowing for the identification of the labeled intact proteins on the basis of mass. More importantly, the partial (10 or 50%) derivatization of proteins with FLUOS does not preclude their tryptic hydrolysis, and identification of the protein on the basis of the mass spectrometric analysis of its tryptic peptides. Potentially, the procedure allows for the automated mass spectrometric identification of SDS-proteins globally labeled with FLUOS and electrophoretically separated, without need for any gel sectioning.

Stacking of unlabeled sodium dodecyl sulfate-proteins within a fluorimetrically detected moving boundary, electroelution and mass spectrometric identification.

The previously reported fluorimetric detection of sodium dodecyl sulfate (SDS)-protein in the presence of cascade blue in agarose gel electrophoresis using barbital buffer was found to be equally feasible in the absence of the fluorescent marker and using Tris-Tricinate buffer, provided that SDS was loaded with the sample but not contained in the catholyte. That fluorescent detection is thought to be due to the formation of a moving boundary between leading SDS and trailing barbital, or Tricinate buffer. This hypothesis is supported by the following evidence: (i) The fluorometrically detected band disappears with addition of SDS to the catholyte; (ii) band area is proportional to protein and/or SDS load; (iii) mobility of SDS-proteins differing in mass is the same at agarose concentrations up to 3%; (iv) lowering of protein mobility by increase in gel concentration and/or increase in the size of the SDS-protein leads to band disappearance. Fluorescent detection of the band is like to be nonspecific and due to the light scattering properties of a stack comprising moving boundaries of any analytes with net mobilities intermediate between SDS (or micellar SDS) and the trailing buffer constituent at their regulated very high concentrations. The steady-state stack of SDS-proteins in the size range of 14.4-45.0 kDa, and the transient stack of an SDS-protein of 66.2 kDa have lent themselves to electroelution and characterization by mass of the proteins after removal of SDS and buffer exchange using matrix assisted laser desorption/ionization-time of flight (MALDI-TOF)-mass spectrometry. The possibility to form a stack of protein between leading SDS and trailing buffer anions under conditions of weak molecular sieving (open-pore gel and small-sized protein) contributes to the understanding of moving boundaries in gel electrophoresis, but in view of the narrowly defined conditions, under which this stack forms, is of limited practical significance for the gel electrophoresis of SDS-proteins.

Electroelution of proteins from bands in gel electrophoresis without gel sectioning for the purpose of protein transfer into mass spectrometry: elements of a new procedure.

Electroelution of protein bands from a gel has advantages over the competitive common technique requiring gel sectioning with respect to yield, speed and the potential for computer-controlled application to multicomponent two-dimensional (2-D) gels. The electroelution design for the commercial high-performance gel electrophoresis (HPGE) apparatus represented the most advanced technique to date until the recent discontinuation of its production. The present report serves to summarize the necessary design elements for the purpose of renewing and further developing the electroelution technique. A rudimentary technique is presented by which the electroeluate is collected in a glass tube superimposed on a reversibly stained gel band and connected to an anolyte reservoir. Although the stain used is insufficiently sensitive, the technique allowed for the qualitative verification of its usefulness in the transfer of the electroeluate into mass spectrometry.

Specific lectin binding to beta1 integrin and fibronectin on the apical membrane of madin-darby canine kidney cells.

Although lectins have previously been used to identify specific cell types in the kidney and various other tissues, the proteins labeled were not identified. We hypothesized that fluorescently labeled lectins could provide a useful tool for direct labeling of membrane-associated glycoproteins. Protein fractions from Madin-Darby canine kidney (MDCK) cells were exposed to a panel of 16 fluorescently labeled lectins to identify suitable lectin-protein pairs. Peanut agglutinin (PNA) selectively bound a 220-240 kDa O-linked glycoprotein with a slightly acidic isoelectric point, while Sambucus nigra agglutinin (SNA) labeled a 130 kDa glycoprotein with a highly acidic isoelectric point. Both proteins were readily labeled by lectins applied to the apical surface of living confluent cells. The proteins were isolated by lectin affinity columns and identified by mass spectrometry. Peptides from the PNA-binding protein shared molecular weight and amino acid composition with fibronectin. Fragments of the SNA-binding protein showed amino-acid identity with peptides from beta1 integrin. The identities of these proteins were validated by Western blotting. Binding of PNA to a 220 kDa protein was inhibited by an anti-fibronectin antibody, and binding of a 130 kDa protein by SNA was diminished by an anti-beta1 integrin antibody. We conclude that PNA and SNA can be used as specific markers for fibronectin and beta1 integrin, respectively, in MDCK cells.

Relationship of intestinal calcium absorption to 1,25-dihydroxyvitamin D [1,25(OH)2D] levels in young versus elderly women: evidence for age-related intestinal resistance to 1,25(OH)2D action.

Intestinal calcium absorption decreases with aging, but it is unclear whether this is attributable to an age-related intestinal resistance to 1,25-dihydroxyvitamin D [1,25(OH)2D] action. Thus, we assessed the in vivo dose response of active intestinal calcium absorption to a broad range of circulating 1,25(OH)2D levels in elderly [age (mean +/- SD), 72.5+/-3.0 yr] vs. young women (age, 28.7+/-5.3 yr; n = 20 per group), who were stratified into 5 subgroups: group 1 was given a high calcium intake of 75 mmol/day, suppressing 1,25(OH)2D levels; group 2 was given a normal calcium diet of 15-30 mmol/day, representing basal 1,25(OH)2D levels; group 3 was given a low-calcium diet of 5 mmol/day to stimulate endogenous 1,25(OH)2D production; group 4 was given the low-calcium diet plus 1 microg/day 1,25(OH)2D; and group 5 was given a low-calcium diet plus 2 microg/day 1,25(OH)2D. After 7 days of diet and/or 1,25(OH)2D treatment, fasting fractional calcium absorption (FCA) was assessed by a double-tracer method using stable calcium isotopes. Serum 1,25(OH)2D and vitamin D-binding protein levels were measured concurrently, and the free 1,25(OH)2D index [molar ratio of 1,25(OH)2D to DBP] was calculated. FCA was significantly correlated with the free 1,25(OH)2D index in the young (R = 0.63, P = 0.003) but not in the elderly women (R = 0.27, P = 0.25). Moreover, the slope of the relationship between FCA and free 1,25(OH)2D index (representing intestinal sensitivity to 1,25(OH)2D) was significantly greater in the young (compared with the elderly) women [mean +/- SEM, 0.15+/-0.04 (young) vs. 0.03+/-0.02, elderly, P = 0.03]. Thus, using an experimental design that allowed us to assess FCA over a wide range of 1,25(OH)2D levels, we demonstrate that elderly women have a resistance to 1,25(OH)2D action that may contribute to their negative calcium balance, secondary hyperparathyroidism, and bone loss.

Recovery of sodium dodecyl sulfate-proteins from gel electrophoretic bands in a single electroelution step for mass spectrometric analysis.

Mass spectrometric analysis of proteins derived from bands in gel electrophoresis is incompatible with the covalent fluorescent labeling of the protein. Thus, if one wishes to take advantage of the capacity for computer-directed electroelution of electrophoresis apparatus with intermittent fluorescent scanning of the migration path, the protein must be labeled fluorescently in a noncovalent, reversible fashion. This was recently achieved by staining of SDS-proteins with Cascade blue and electrophoresis in barbital buffer. However, the method was not a practical one for the purpose of isolating proteins from gel electrophoretic bands and their transfer into the mass spectrometer for three reasons: (i) Ten consecutive electroelution steps were required to obviate pH changes in the electroelution chamber; (ii) electroeluates from six gel electrophoretic lanes needed to be pooled; (iii) excessive protein loads ranging from 7 to 33 microg/pool were required. The present study reports the solution to those three problems. Mass spectrometric (MALDI-TOF) characterization of five proteins was demonstrated (i) after a single electroelution step; (ii) using electroelution from a single gel of 0.3-cm(2) cross-sectional area; and (iii) using a protein load of 2 (in one case 4) microg. However, the migration rates of the Cascade blue-SDS-protein-barbital complexes derived from proteins with widely varying molecular weights proved to be the same. Thus, despite the three advances made, the method to date remains restricted to samples of single proteins.

Calcium kinetics in children with osteogenesis imperfecta type III and IV: pre- and post-growth hormone therapy.

Children with osteogenesis imperfecta (OI) type III and type IV were studied using a (42)Ca stable isotope technique. Serum dilution kinetics of (42)Ca were studied pre- and post-growth hormone (GH) treatment in 9 OI III (age range 5-9 years) and 8 OI IV patients (age range 5-12 years). Each subject was studied twice: at baseline and following GH therapy (range 1-1.5 years). Isotopic enrichments of (42)Ca were followed over 7 days using thermal ionization mass spectrometry. A binding site model, which describes reversible and irreversible binding of calcium (Ca) ions to postulated short- and long-term binding sites in bone, was used to analyze the kinetic data. In type III patients, GH treatment (1) increased the fraction of short-term binding sites, theta (0.777 +/- 0.112 versus 0.877 +/- 0.05, respectively; P = 0.034); (2) increased the apparent half-life of a Ca ion attached to the long-term binding site by 76% (P = 0. 009); (3) although not statistically significant (P = 0.098), a trend toward an increased growth rate was observed with increasing change in theta (Deltatheta); (4) patients experienced a 75% increase in growth rate during the first 6 months of treatment. In type IV patients, GH treatment increased the apparent half-life of a Ca ion attached to the long-term binding site by 83% (P = 0.048), however, no trend toward an increased growth rate was observed with increasing Deltatheta in these patients. These significant changes in Ca binding to bone may influence growth in type III patients.

Tracking pathology with proteomics: identification of in vivo degradation products of alphaB-crystallin.

Soemmerring's ring is one form of "after cataract" that can occur following cataract surgery. The ring structure is formed by adherence of the anterior lens capsule to the posterior lens capsule. Epithelial cells remaining after surgery differentiate into lens fiber cells but the resulting tissue mass does not remain transparent. The protein in normal lens and in Soemmerring's rings from four individuals was analyzed using two-dimensional (2-D) gel electrophoresis, matrix assisted laser desorption/ionization-time of-flight-mass spectrometry (MALDI-TOF-MS) and image analysis with Phoretix software. The 2-D protein patterns of the Soemmerring's rings were generally similar to that of cortical fiber cells of normal human lens with some notable exceptions. Several post-translationally modified forms of alphaB-crystallin(1-175) were identified. Two degradation products, alphaB-crystallin(1-170) and alphaB-crystallin(1-174), each make up 9.5-27% of the total alphaB-crystallin in the Soemmerring's rings and less than 1% in the normal lenses. Other modified forms of alphaB-crystallin are aberrant in the fiber cells of the Soemmerring rings relative to normal lens.

Insulin-like growth factor I and growth hormone (GH) treatment in GH-deficient humans: differential effects on protein, glucose, lipid, and calcium metabolism.

We examined the effects of recombinant human (rh) insulin-like growth factor I (IGF-I) vs. rhGH in a variety of metabolic paths in a group of eight severely GH-deficient young adults using an array of contemporary tools. Protein, glucose, and calcium metabolism were studied using stable labeled tracer infusions of L-[1-13C]leucine, [6,6-2H2]glucose, and 42Ca and 44Ca; substrate oxidation rates were assessed using indirect calorimetry; muscle strength was determined by isokinetic and isometric dynamometry of the anterior quadriceps, as well as growth factors, hormones, glucose, and lipid concentrations in plasma before and after 8 weeks of rhIGF-I (60 microg/kg, sc, twice daily), followed by 4 weeks of washout, then 8 weeks ofrhGH (12.5 microg/kg-day, sc); the treatment order was randomized. In the doses administered, rhIGF-I and rhGH both increased fat-free mass and decreased the percent fat mass, with a more robust decrease in the percent fat mass after rhGH; both were associated with an increase in whole body protein synthesis rates and a decrease in protein oxidation. Neither hormone affected isokinetic or isometric measures of skeletal muscle strength. However, rhGH was more potent than rhIGF-I at increasing lipid oxidation rates and improving plasma lipid profiles. Both hormones increased hepatic glucose output, but rhGH treatment was also associated with decreased carbohydrate oxidation and increased glucose and insulin concentrations, indicating subtle insulin resistance. Neither hormone significantly affected bone calcium fluxes, supporting the concept that these hormones, by themselves, are not pivotal in bone calcium metabolism. In conclusion, rhIGF-I and rhGH share common effects on protein, muscle, and calcium metabolism, yet have divergent effects on lipid and carbohydrate metabolism in the GH-deficient state. These differences may allow for better selection of treatment modalities depending on the choice of desired effects in hypopituitarism.

Electroelution of nonfluorescent stacked proteins detected by fluorescence optics from gel electrophoretic bands for transfer into mass spectrometry.

The extreme accuracy of spectrometrically determined masses of proteins has opened the possibility to identify proteins separated as gel electrophoretic bands in the absence of specific immunologic ways of identification. For the purpose of protein transfer from gel electrophoretic bands to mass spectrometer, electroelution from the intact gel has advantages, in particular when apparatus with capacity for fluorescent scanning allows one to direct the electroelution cell over the band under computer control. To avoid fluorescent labeling of the protein which is incompatible with mass spectrometric identification, it is proposed to selectively stack the unlabeled protein and detect it by comigrating tracking dye prior to electroelution. The feasibility of the approach is exemplified in case of a single protein, but still remains to be demonstrated in conjunction with the selective stacking or unstacking of a single protein from a mixture of several proteins.

Transfer of SDS-proteins from gel electrophoretic zones into mass spectrometry, using electroelution of the band into buffer without sectioning of the gel.

Five SDS-proteins, ranging in molecular weight from 14 to 66 kDa, were detected without covalent fluorescent labeling by the automated gel electrophoresis apparatus with intermittent fluorescence scanning (HPGE apparatus, LabIntelligence) during electrophoresis in barbiturate buffer in the presence of Cascade Blue. The SDS-proteins were electroeluted from the gel into 220 microl of buffer by a modification of the procedure of Gombocz and Cortez. The electroeluate was freed of SDS, ultrafiltered and subjected to MALDI-TOF mass spectrometry. The masses of the five native proteins were found to be maintained after electrophoresis and electroelution in the presence of the potential contaminants SDS, barbituric acid and Cascade Blue. The procedure of protein transfer from SDS-PAGE into mass spectrometry, without excision of bands, gel maceration and protein recovery by diffusion, therefore is shown to be suitable for the identification by mass of intact proteins derived from gel electrophoretic bands.

Effect of growth hormone treatment on calcium kinetics in patients with osteogenesis imperfecta type III and IV.

Using a dual stable isotope technique, the effect of growth hormone (GH) on whole body calcium (Ca) metabolism was studied in children (ages 5-14 years) with type III (n = 9) and IV (n = 8) osteogenesis imperfecta. Each subject was studied twice: at baseline and following a GH (0.1-0.2 U/kg per day) treatment period of 1-1.5 years. Subjects were given 42Ca intravenously and 44Ca orally. The sera and urine 42Ca and 44Ca isotopic enrichments were followed over 7 days using thermal ionization mass spectrometry. The SAAM program was used to fit a three-compartment model to the tracer data. No significant differences were observed between: (1) children with type III and IV disease; or (2) baseline studies of boys and girls within each disease type. However, GH treatment significantly increased: (1) the exchangeable calcium pool (EP) in type III patients (2086 vs. 4422 mg/day, p = 0.02); and (2) the parameter associated with bone calcium accretion in type IV patients (Vo+: 973 vs. 1560 mg/day,p = 0.03) with boys responding with a significantly greater increase than girls (p = 0.008). Although not statistically significant, a trend toward an increase in Vo+ in type III patients and in EP in type IV was observed following treatment. Our observations imply that more Ca was available for bone mineralization following GH treatment in these subjects.

Profound hypogonadism has significant negative effects on calcium balance in males: a calcium kinetic study.

The impact of estrogen deficiency on bone has been extensively studied in the female; however, the effects of androgen deficiency on calcium fluxes in males have been less well characterized. We investigated the effect of short-term, severe androgen deficiency on measures of calcium absorption and kinetics as well as on markers of bone turnover in males. To accomplish this, 11 healthy male volunteers were recruited (mean age 23.3 +/- 0.5 years [SEM], body mass index 25.3 +/- 0.8 kg/m2). They consumed a weight maintenance diet for at least 3 days prior to admission to our Research Unit, with a calcium intake of approximately 1200 mg/day. At baseline (D1), subjects received 42Ca intravenously as well as 44Ca PO mixed with milk or juice. A 29-h urine collection was begun and blood samples collected at frequent intervals for the measurement of the isotopic enrichment of 42Ca and 44Ca using thermal ionization mass spectrometry. Twice daily urine samples were collected for 5 days after the administration of the isotopes. A gonadotropin-releasing hormone agonist (Lupron) was given after D1, again 3 weeks later, and studies repeated identically 4 weeks (D2, n = 6) and 10 weeks from baseline (D3, n = 7) (two subjects completed three studies). Testosterone concentrations were markedly suppressed on both D2 and D3 (-95%, p < 0.006), whereas there were no detectable changes in growth hormone and insulin-like growth factor-1 concentrations. Urinary calcium excretion increased significantly after 4 weeks (43%, p = 0.0007) and 10 weeks (73%, p = 0.003) of sustained hypogonadism. Using a multicompartmental kinetic model, the contribution of oral calcium to the urinary losses was decreased by D3 (-41%, p = 0.01), yet the contribution of bone calcium to urine losses increased by 10 weeks (+11%, p = 0.01). There was a 21% decrease in bone calcium deposition (Vo+) by D3 (p < 0.05) with no significant change in bone resorption rates (Vo-). There was a significant correlation between the decrease in testosterone concentration and the increase in urinary calcium excretion, especially at 10 weeks (R2 = 0.84, p = 0.004). These kinetic changes were accompanied by a decrease in osteocalcin concentrations on D2, with improvements by D3. Urinary N telopeptide, a measure of bone resorption, also increased during the studies. In summary, profound hypogonadism in young males is associated with marked increases in urinary calcium losses, with a greater contribution of bone calcium to those losses and decreased kinetic markers of bone calcium deposition. We conclude that even short-term, severe deficiency in gonadal steroids can have profound negative effects on calcium and bone metabolism in males.

Mass spectrometric analysis of the electroeluates of fluorescent proteins after preparative electrophoresis in the automated HPGE-1000 apparatus.

Bands of green fluorescent protein (GFP) and R-phycoerythrin (PHYCO) in gel electrophoresis on the automated apparatus for gel electrophoresis with periodic fluorescence scanning (HPGE), the HPGE-1000 apparatus, were retrieved from the gel by electroelution. While PHYCO was recovered in a single volume of electroeluate buffer after the predicted migration time, GFP fluorescence was lost under the same conditions and could only be recovered using multiple changes of electroeluate buffer. The multiple volumes of buffer necessitated pooling, concentration, and storage, conditions under which a minor GFP component, GFP-II, formed artifactually. PHYCO after electroelution also exhibits a minor component present in the original preparation. The electroeluate of GFP, transferred into a mass spectrometer after pooling, concentration and storage, is indistinguishable in mass from the original preparation.

Studies towards neoglycoconjugates from the monosaccharide determinant of Vibrio cholerae O:1, serotype Ogawa using the diethyl squarate reagent.

The effect of reaction time, concentration and molar excess of hapten upon the efficiency of the conjugation of carbohydrates to proteins using the diethyl squarate reagent has been studied using chicken serum albumin (CSA) as the carrier protein and a linker-equipped D-glucose derivative as the hapten. A high degree of incorporation of the latter into CSA was achieved with high efficiency, and the use of a large excess of the ligand was not necessary. Conjugation of the immunodominant monosaccharide determinant of Vibrio cholerae O:1, serotype Ogawa, bearing the same spacer, followed a similar pattern, showing that the nature of the carbohydrate does not substantially affect the outcome of the conjugation and that a predicted degree of antigen-loading onto carrier protein is possible to achieve.

Simultaneous assessment of conformation and aggregation of beta-amyloid peptide using electrospray ionization mass spectrometry.

Electrospray ionization mass spectrometry was used to study conformation and aggregation of the synthetic beta-amyloid peptide, residues 1-40 (betaA4), as a function of concentration and sample aging. All mass spectra showed a major envelope of peaks corresponding to charge states of 7-3 of the monomeric form of betaA4. In addition, weaker envelopes of peaks corresponding to charge states of dimeric, trimeric, and tetrameric betaA4 species were seen under gentle ionization conditions. The average charge state of the envelope associated with the monomeric form decreased by ca. 0.5 z as samples were aged, indicating that the relatively open form (likely random coil) of the peptide was modified into the more compact form (likely beta-sheet) as a function of sample aging. The aggregate forms became weaker and ultimately were absent both in the more dilute solutions and in aged aliquots of the concentrated sample. These aggregates were interpreted as assemblies of the random coil form. We interpret our inability to see an ion envelope that can be associated with aggregates of the beta-sheet form to be a consequence of the presumed very compact nature of this form. A model for the formation of betaA4 fibrils is proposed and discussed.