`
`www.elsevier.com:locate:ijpharm
`
`The degradation pathways of glucagon in acidic solutions
`
`Anjali B. Joshia, Elena Rusb, Lee E. Kirscha,*
`a Di6ision of Pharmaceutics, College of Pharmacy, The Uni6ersity of Iowa, Iowa City, IA 52242, USA
`b Molecular Analysis Facility, The Uni6ersity of Iowa, Iowa City, IA 52242, USA
`
`Received 2 March 2000; received in revised form 28 April 2000; accepted 1 May 2000
`
`Abstract
`
`Objecti6e: Glucagon is a 29 amino acid peptide hormone that exhibits degradation via both chemical and physical
`pathways. The objective of the studies reported herein was to identify the degradation products and scheme for
`glucagon hydrolysis in acidic solutions. Methods: Solutions of glucagon in 0.01 N HCl (pH 2.5) were degraded at
`60°C for 70 h. One isocratic and two gradient RP-HPLC methods were developed to separate the degradation
`products. Structure elucidation of the separated peaks was achieved using amino acid sequencing, amino acid
`analysis, and mass spectrometry. Degradation was carried out in the pH range 1.5–5 to check for changes in
`degradation scheme with pH. Authentic samples of degradation products were degraded under similar acidic
`conditions to confirm precursor successor relationships in the degradation scheme. Results: Sixteen major degradation
`products were isolated and identified. The major pathways of degradation were found to be aspartic acid cleavage at
`positions 9, 15, and 21 and glutaminyl deamidation at positions 3, 20, and 24. Cleavage occurred on both sides of
`Asp-15 but only on the C-terminal side of Asp-9 and Asp-21. Deamidation of the Asn residue at position 28 was not
`detected. © 2000 Elsevier Science B.V. All rights reserved.
`
`Keywords: Degradation pathways; Deamidation; Glucagon; Glutaminyl; Hydrolysis; Peptide cleavage
`
`1. Introduction
`
`Glucagon is a polypeptide hormone that is used
`for the emergency treatment of insulin induced,
`sulphonylurea induced, and spontaneous hypo-
`glycemia. It has the ability to raise blood glucose
`concentrations by increasing hepatic glycogenoly-
`sis through activation of liver phosphorylase and
`to stimulate insulin secretion by direct action on
`pancreatic b-cells. Recovery of consciousness is
`
`* Corresponding author. Fax: (cid:27)1-319-3359349.
`E-mail address: lee-kirsch@uiowa.edu (L.E. Kirsch).
`
`usually achieved within 15–30 min of the injec-
`tion whereupon oral glucose may be used to
`continue treatment (Marks, 1983). Glucagon also
`has the ability to reduce smooth muscle tone and
`motility. It is used as a diagnostic aid in radio-
`logic procedures of the gastrointestinal tract that
`require a diminished tone and motility of the
`organ under study (Diamant and Picazo, 1983).
`Glucagon for injection is a mixture of glucagon
`hydrochloride with one or more suitable, dry dilu-
`ents supplied in single-dose or multiple-dose con-
`tainers. The pH of the reconstituted solution is
`between 2.5 and 3.0 (US Pharmacopeia 24, 1999).
`
`0378-5173:00:$ - see front matter © 2000 Elsevier Science B.V. All rights reserved.
`PII: S 0 3 7 8 - 5 1 7 3 ( 0 0 ) 0 0 4 3 8 - 5
`
`Page 1
`
`NPS EX. 2048
`CFAD v. NPS
`IPR2015-01093
`
`
`
`116
`
`A.B. Joshi et al. :International Journal of Pharmaceutics 203 (2000) 115–125
`
`Thus the stability of glucagon in acidic aqueous
`solutions is relevant to drug product quality.
`Structurally, glucagon contains 29 amino acids
`(Fig. 1). In the crystalline state, a completely
`helical conformation was proposed for glucagon
`(Sasaki et al., 1975). However, optical rotatory
`dispersion (Gratzer et al., 1968) and circular
`dichroism studies (Srere and Brooks, 1969) of
`glucagon in dilute aqueous solutions revealed a
`predominantly random coil conformation with at
`most 15% a-helix at the C-terminal end. The
`molecule has an isoelectric point of (cid:2) 7 and
`exhibits very low solubility (B0.1 mg:ml) in the
`approximate pH range of 4–8. It is readily soluble
`(\10 mg:ml) at pH values less than 3 or greater
`than 9 (Bromer, 1983). It is known to self-associ-
`ate at high concentrations and forms aggregates
`and gels at mild temperatures in acidic and basic
`
`Fig. 1. Sequences of glucagon and its major degradation
`products in acidic aqueous solutions. The sequence of each
`degradation fragment is denoted using the one-letter abbrevia-
`tions for the amino acids and by the numbers of the first and
`last amino acid residues. Deamidated fragments are denoted
`by parentheses followed by ‘dn’ where ‘n’ is the number of the
`glutamine residue which deamidates.
`
`solutions. The aggregation is promoted by salt,
`increased pH (within the acid range of solubility),
`agitation, and increased temperature up to 30°C
`(Beaven et al., 1969). Aggregation is accompanied
`by an increase in secondary structure in the form
`of anti-parallel b-sheets (Gratzer et al., 1968).
`The degradation pathways of glucagon have
`not been reported. In one study, glucagon degra-
`dation in 0.03 N HCl at 105°C was found to
`release three aspartic acid residues
`in 13 h
`(Schultz, 1967). The possibility of iso-aspartyl for-
`mation at the position of Asp-9 at neutral pH has
`also been suggested (Ota et al., 1987). The objec-
`tive of the studies presented herein was to identify
`the degradation products
`and scheme
`for
`glucagon hydrolysis in acidic solutions.
`
`2. Materials and methods
`
`2.1. Materials
`
`Purified porcine glucagon was obtained from
`Lilly Research Laboratories (Indianapolis, IN).
`Sodium dihydrogen phosphate,
`o-phosphoric
`acid, acetic acid, formic acid, hydrochloric acid,
`trifluoroacetic acid, potassium chloride, sodium
`formate, and sodium acetate were from Fisher
`(Springfield, NJ). All chemicals were of reagent
`grade and used as received. Solvents used for
`chromatography were HPLC grade. Cellulose es-
`ter dialysis membranes were from Spectrum
`(Houston, TX). Degradation product fragments
`were synthesized manually by SynPep Corpora-
`tion (Dublin, CA).
`
`2.2. Degradation of glucagon and glucagon
`fragments in acidic pH range
`
`(0.1–1 mg:ml)
`Aqueous glucagon solutions
`were degraded at 60°C in the pH range 1.5–5
`using dilute HCl, phosphate, formate, or acetate
`buffers. Degradation was allowed to proceed to
`two to three half-lives. Aliquots were removed
`from reaction mixtures at various time intervals
`and stored at 4°C until analyzed. Authentic
`degradation products were subject to hydrolysis at
`pH 1.5 and 2.5 under conditions identical to those
`used for glucagon.
`
`Page 2
`
`
`
`A.B. Joshi et al. :International Journal of Pharmaceutics 203 (2000) 115–125
`
`117
`
`2.3. Separation of degradation products
`
`chromatography
`liquid
`performance
`High
`(HPLC) analyses were performed using a Shi-
`madzu RP-HPLC system consisting of an SCL-
`10AVP system controller, LC-10ATVP pumps,
`SIL-10ADVP auto-injector, SPD-10AVP UV-VIS
`detector, and a CTO-10ASVP column oven.
`Chromatograms were integrated and data stored
`using Class VP Chromatography Data System
`software (Version 4.2).
`products was
`Separation
`of
`degradation
`achieved on a Lichrospher RP-18, 4.6(cid:29)250 mm,
`5m-column using isocratic elution (Method A).
`The mobile phase contained 29:71 acetonitrile:
`buffer
`(buffer(cid:30)0.1 M sodium phosphate
`monobasic and 0.002 M cysteine free base, ad-
`justed to pH 2.6 with 85% phosphoric acid). Flow
`rate was 1.0 ml:min, detection was at 214 nm, and
`column temperature was 35°C. Degradation prod-
`ucts
`that were unresolved with the isocratic
`method were separated by gradient elution using a
`Zorbax 300SB C8, 4.6(cid:29)250 mm, 5m column. The
`mobile phase contained methanol and a 0.024%
`solution of trifluoroacetic acid in water. The gra-
`dient used was 5–35% methanol
`in 30 min
`(Method B). Flow rate was 1.0 ml:min, detection
`was at 214 nm, and column temperature was
`25°C. Another gradient method (Method C) was
`developed in an attempt to separate additional
`degradation peaks using the same column and
`mobile phase as the previous gradient method.
`The gradient was 5–50% methanol
`in 90 min.
`Flow rate was 1.0 ml:min, detection was at 214
`nm, and column temperature was 25°C.
`Degradation product peaks were collected from
`repeated 200 ml reaction mixture samples using a
`FRC 10A fraction collector (Shimadzu). Seven
`peaks were collected with the isocratic method
`(Method A), five more peaks with the first gradi-
`ent method (Method B), and an additional three
`peaks with the second gradient method (Method
`C). The collection procedure was validated by
`re-injecting peak fractions back on the column to
`check peak integrity. Individual peaks were col-
`lected in flat top 1.5 ml polypropylene microcen-
`trifuge tubes (Fisher Scientific) and stored in the
`refrigerator at 4°C until analysis. Since analysis of
`
`the fractions was typically done 4–7 days after
`fraction collection, the stability of the fractions at
`4°C was evaluated. No significant degradation of
`peptide fractions was observed up to 18 days at
`4°C.
`
`2.4. Degradation product identification
`
`Structure elucidation of degradation products
`was done using either amino acid sequencing,
`amino acid analysis, or mass spectrometry. Most
`of the fractions were identified using more than
`one method.
`Amino acid analysis was performed by hy-
`drolyzing an aliquot of each fraction using 100 ml
`of 6 N HCl at 110°C for 24 h. When hydrolysis
`was completed, the vials were placed in a Speed-
`Vac® (Model SC210A) where all the residual liq-
`uid was
`removed by
`evaporation. Sample
`hydrolysates were then re-dissolved in sodium cit-
`rate buffer diluent. A Beckman 6300 high-perfor-
`mance ion-exchange analyzer was used to analyze
`each sample. Separation was done using a 12 cm
`hydrolysate column and a three-step temperature
`program in combination with three sodium citrate
`buffers to separate amino acids in various charged
`states. Standards consisting of a mixture of all
`amino acids plus an internal standard were also
`hydrolyzed. The system used ninhydrin to react
`with the amino acid giving a color reaction. The
`intensity of the color was proportional to the
`concentration of amino acid.
`For sequencing, fraction aliquots were concen-
`trated in a SpeedVac® and passed through a
`Prosorb sample preparation cartridge (Perkin
`Elmer). The adsorbent filter in these cartridges
`draws sample solution through a membrane by
`capillary action. The membrane
`immobilizes
`proteins and peptides while buffer components
`that could potentially interfere with sequencing
`pass through. The process not only desalts but
`also concentrates the sample. The membrane
`holding the peptide was directly loaded on the
`sequencer. Sequencing was done using an Applied
`Biosystems 492 automatic sequencer with an on-
`line PTH analyzer. A standard consisting of a
`mixture of all amino acids was first injected to
`obtain reference retention times for each amino
`
`Page 3
`
`
`
`118
`
`A.B. Joshi et al. :International Journal of Pharmaceutics 203 (2000) 115–125
`
`acid. One fraction was assigned more than one
`sequence (one major and one minor) due to the
`presence of a mixture of peptides.
`Degradation product identity was also verified
`using Fast Atom Bombardment Mass Spectrome-
`try (FABMS). The instrument used was a Hewlett
`Packard 1100 LCMSD. Organic solvents present
`in peak fractions collected with the isocratic
`method (Method A) were removed by evapora-
`tion under nitrogen. Sample peak fractions were
`then desalted using cellulose ester dialysis mem-
`branes (MWCO(cid:30)500). De-ionized water was
`used as the dialysis medium and was changed
`twice daily. The extent of dialysis was monitored
`by measuring the conductivity of the dialysate on
`an Orion Model 160 conductivity meter. Dialysis
`was stopped when the conductivity measurement
`was close to that of pure de-ionized water. The
`dialyzed fractions were then freeze dried (Virtis
`Advantage lyophilizer) and analyzed by FABMS.
`Peak fractions collected using the gradient meth-
`ods (Methods B and C) were simply concentrated
`in a SpeedVac® before FABMS analysis since
`these fractions contained only water and volatile
`solvents.
`To verify the sequencing results, two peaks
`obtained by isocratic elution (Method A) were
`analyzed by FABMS. The molecular ion masses
`obtained corresponded to the fragments obtained
`by sequencing. Hence, additional FABMS analy-
`ses were not performed on fractions that were
`sequenced. However, FABMS was used to
`confirm every sequence obtained with amino acid
`analysis alone.
`
`3. Results
`
`3.1. Chromatographic separation of degradation
`products
`
`degraded
`of
`separation
`Chromatographic
`glucagon using isocratic elution (Method A) gave
`seven major degradation peaks (Peaks I-VII, Fig.
`2). Sequencing of these products revealed that
`they were peptide fragments containing the C-ter-
`minus of glucagon. A gradient elution method
`(Method B) was developed to separate the corre-
`
`sponding fragments containing the more polar
`amino acids of the N-terminus end. An additional
`five peaks were obtained (Peaks VIII–XII, Fig.
`3). A second gradient method (Method C) was
`developed to check for additional degradation
`products. This method involved a slow gradient
`(90 min) from 5 to 50% methanol and yielded
`several peaks. The peak at retention time of 5.5
`min was found to be the same as peak XII
`obtained with Method B (Fig. 3). The peaks
`eluting at 35.3, 59.3 min, and 69.3 min were found
`to be the same as peak I, peak III, and peak V
`respectively, obtained by Method A (Fig. 2).
`These peak assignments were made by co-inject-
`ing peak fractions or authentic samples of peaks
`XII, I, III, and V and comparing retention times.
`The peak eluting at 72.1 min was glucagon. Thus,
`three new peaks, at 15.3, 20.2, and 26.6 min were
`obtained by Method C (peaks XIII, XIV, and XV
`in Fig. 4).
`An overlay of sample chromatograms obtained
`from degradation reactions conducted over the
`pH range 2.5–5 (Fig. 5) demonstrated that the
`retention times of the major degradation product
`peaks were similar at pH 2.5 and 3.5. At higher
`pH values (4.5 and 5), the chromatograms were
`somewhat similar but separation quality was com-
`promised by glucagon aggregation.
`
`3.2. Degradation product identities
`
`Of the seven peaks obtained by Method A (Fig.
`2), peaks I, III, and V were composed of peptide
`fragments 22–29, 16–29, and 10–29 respectively,
`indicating that peptide cleavage had occurred on
`the C-terminal side of Asp-21, Asp-15, and Asp-9.
`Peaks II, IV, and VI eluted immediately after
`peaks I, III, and V (Fig. 2) and were found to be
`the deamidated forms of fragments 22–29 (peak
`I), 16–29 (peak III), and 10–29 (peak V) wherein
`Gln-24 converted to Glu-24. Peak VII was the
`deamidated form of glucagon with Glu at position
`24. A small peptide fragment starting at Tyr-10
`co-eluted with deamidated 16–29 in peak IV.
`Both these peptides could be detected through
`separate PTH-AA signals. However, both frag-
`ments have Asp as their sixth residue, hence the
`chromatogram for the sixth cycle of sequencing
`
`Page 4
`
`
`
`A.B. Joshi et al. :International Journal of Pharmaceutics 203 (2000) 115–125
`
`119
`
`Fig. 2. Separation of degradation products of glucagon by isocratic elution using RP-HPLC Method A (column was a Lichrospher
`RP-18, mobile phase was 29:71 acetonitrile:buffer, buffer(cid:30)0.1 sodium phosphate monobasic and 0.002 M cysteine free base,
`adjusted to the pH 2.6 with 85% phosphoric acid).
`
`Page 5
`
`
`
`120
`
`A.B. Joshi et al. :International Journal of Pharmaceutics 203 (2000) 115–125
`
`Fig. 3. Separation of degradation products containing the more polar amino acid residues of the N-terminal end using RP-HPLC
`Method B (column was a Zorbax SB300 C8, mobile phase was methanol and 0.024% TFA in water, gradient was 5–35% methanol
`in 30 min).
`
`Page 6
`
`
`
`A.B. Joshi et al. :International Journal of Pharmaceutics 203 (2000) 115–125
`
`121
`
`Fig. 4. Separation of additional degradation products containing the more polar amino acid residues of the N-terminal end using
`RP-HPLC Method C (column was a Zorbax SB300 C8, mobile phase was methanol and 0.024% TFA in water, gradient was 5–50%
`methanol in 90 min).
`
`Page 7
`
`
`
`122
`
`A.B. Joshi et al. :International Journal of Pharmaceutics 203 (2000) 115–125
`
`Fig. 5. Overlaid chromatograms of glucagon degraded in the pH range 2.5–5.0 using Method C.
`
`Page 8
`
`
`
`A.B. Joshi et al. :International Journal of Pharmaceutics 203 (2000) 115–125
`
`123
`
`gave only one peak for Asp. This made it difficult
`to determine whether the signal came just from
`the major fragment (deamidated 16–29) or from
`both fragments. A comparison of the combined
`picomole yields for the two fragments (as the
`sequencing progressed) indicated that the Asp pi-
`comole yield was not large enough to include two
`signals. Hence the sequence for the co-eluting
`peak was assigned as 10–14. This suggested the
`possibility that the Asp-15 residue cleaved on
`both its C-terminal and N-terminal sides.
`Peaks VIII through XII (Fig. 3) were obtained
`by gradient elution (Method B) and contained the
`peptide
`fragments
`containing the N-terminal
`residues that were not recovered by isocratic elu-
`tion. Peaks VIII and X were fragments 1–9 and
`10–21, respectively. This confirmed the result that
`Asp-9 and Asp-21 did not cleave on the N-termi-
`nal side. Peak XII was fragment 10–15, which
`further confirmed cleavage on both sides of the
`Asp-15 residue. Peaks IX and XI were the deami-
`dated products of peaks VIII and X respectively,
`which indicated that the glutamine residues in
`positions 3 and 21 were susceptible to deamida-
`tion as well as Gln-24 (peaks II, IV, VI, and VII).
`The three peaks obtained with Method C (Fig.
`4) were analyzed using amino acid analysis and
`confirmed by FABMS. Peak XIII was found to be
`fragment 1–21, peak XIV was fragment 1–15,
`and peak XV was fragment 1–14. This further
`confirmed that peptide cleavage occurred on both
`sides of Asp-15 and only on the C-terminal side of
`Asp-21. The peaks that
`immediately followed
`peaks XIII, XIV, and XV in Fig. 4 were not
`analyzed. These are expected to be the deami-
`dated forms of the fragments 1–21 (peak XIII),
`1–15 (peak XIV), and 1–14 (peak XV), respec-
`tively. The results of structure elucidation of the
`degradation products are summarized in Fig. 1.
`
`3.3. Verification of precursor-successor
`relationships
`
`The synthesized peptide fragments 22–29 and
`16–29 corresponding to peaks I and III in Fig. 2
`were subjected to degradation under acidic condi-
`tions. The resultant degradation chromatograms
`indicated that these fragments were subject to the
`
`expected peptide cleavage and deamidation path-
`ways. Degradation of fragment 22–29 resulted in
`the appearance of a peak with retention time
`corresponding to deamidated 22–29. Degradation
`of the peptide fragment 16–29 gave rise to HPLC
`peaks corresponding to 22–29 and the deami-
`dated forms of 16–29 and 22–29. The peaks
`corresponding to fragment 10–29 (peak V, Fig. 2)
`and fragment 1–15 (peak XIV, Fig. 3) were col-
`lected and also degraded under acidic conditions.
`Fragment 10–29 gave peaks corresponding to
`22–29 and 16–29, and the deamidated forms of
`10–29. Fragment 1–15 gave peaks corresponding
`to 1–14 and 1–9.
`
`4. Discussion
`
`Sixteen degradation products from thermally
`stressed, acidic solutions of glucagon were iso-
`lated and identified. The major pathways of
`glucagon degradation were glutaminyl deamida-
`tion and aspartyl peptide cleavage. The former
`was demonstrated to occur at residues 3, 20, and
`24 and the latter at residues 9, 15, and 21. Cleav-
`age at Asp-9 and Asp-21 occurred on the C-termi-
`nal side of the amino acid whereas cleavage at
`Asp-15 occurred on both N- and C-terminal sides.
`The proposed degradation scheme is illustrated in
`Fig. 6.
`Deamidation is a common protein degradation
`pathway that involves the loss of ammonia from
`the side chain amides of asparagine or glutamine
`to form the corresponding side chain carboxylic
`acid residues: aspartic or glutamic acid (Manning
`et al., 1989). Numerous examples of asparaginyl
`and, to a lesser extent, glutaminyl deamidation
`have been reported (Robinson and Rudd, 1974;
`Geiger and Clarke, 1987; Patel and Borchardt,
`1990a; Windisch et al., 1997). Some of the effects
`of primary, secondary, tertiary, and quaternary
`structure on deamidation kinetics have been de-
`scribed (Kossiakoff, 1988; Wearne and Creighton,
`1989; Patel and Borchardt, 1990b; Wright, 1991;
`Darrington and Anderson,
`1994; Xie
`and
`Schowen, 1999). In general, the presence of neigh-
`boring glycine (specifically at the N(cid:27)1 residue),
`serine
`or
`threonine
`facilitates
`deamidation
`
`Page 9
`
`
`
`124
`
`A.B. Joshi et al. :International Journal of Pharmaceutics 203 (2000) 115–125
`
`(Wright, 1991) whereas the presence of significant
`secondary structure
`(Wearne and Creighton,
`1989) or the lack of solvent access to the potential
`deamidating residue due to tertiary structure may
`tend to slow reaction rates. Moreover, asparaginyl
`deamidation is reported to be more facile than
`glutaminyl deamidation (Robinson and Rudd,
`1974). It has been suggested that this kinetic
`difference may be due to the greater length of the
`glutamine side chain which precludes stabilization
`of a side chain carbonyl oxyanion transition state
`by hydrogen bonding with the neighboring pep-
`tide nitrogen (Wright and Robinson, 1982).
`In glucagon, the lack of asparaginyl deamida-
`tion at residue 28 in the presence of glutaminyl
`deamidation at residues 3, 20 and 24 was a some-
`what unexpected observation. Although the sole
`asparagine residue in glucagon is located in a
`region of the peptide associated with some helical
`
`Fig. 6. Proposed degradation scheme of glucagon in acidic
`aqueous solutions. Deamidation and aspartyl cleavage path-
`ways are represented by dashed and solid arrows respectively.
`Only observed, but not hypothesized, degradation products
`are depicted. * Starting materials; ** has not been chromato-
`graphically separated.
`
`structure, this structure has been shown to be lost
`as the temperature is raised in the range 22–50°C
`(Yi et al., 1992). Since the current study was
`conducted at 60°C, secondary structure effects
`were unlikely.
`Most of the studies that support the greater
`susceptibility of asparaginyl residues to deamida-
`tion were conducted at pH conditions approach-
`ing neutrality (Robinson and Rudd, 1974; Geiger
`and Clarke, 1987; Patel and Borchardt, 1990a).
`The relative reactivity of glutamine and as-
`paragine and the effects of primary, secondary,
`and tertiary structure on glutaminyl deamidation
`in acidic conditions have not been reported and
`are the focus of our ongoing studies.
`Preferential peptide cleavage at aspartic acid
`residues in the presence of acid is a well-known
`hydrolysis pathway that was historically used to
`fragment proteins
`for
`compositional analysis
`(Light, 1967). Cleavage can occur between aspar-
`tic acid and either the N(cid:28)1 or N(cid:27)1 neighboring
`residue. The mechanism is believed to involve
`intramolecular catalysis by the side chain car-
`boxylic acid and formation of a six- or five-mem-
`bered ring intermediate (Inglis, 1983). Peptide
`cleavage was observed at all three of the aspartic
`acid residues in glucagon. Cleavage on both sides
`of Asp-15 was observed. The relative rates of
`peptide cleavage at the three residues are expected
`to vary based on sequence effects. Future work
`will evaluate the relative reactivity of the aspartic
`acid residues.
`
`References
`
`Beaven, G.H., Gratzer, W.B., Davies, H.G., 1969. Formation
`and structure of gels and fibrils from glucagon. Eur. J.
`Biochem. 11, 37–42.
`Bromer, W.W., 1983. Chemical characteristics of glucagon. In:
`Lefe`bvre, P.J. (Ed.), Handbook of Experimental Pharma-
`cology, Vol 66:1. Springer-Verlag, Berlin, pp. 1–22.
`Darrington, R.T., Anderson, B.D., 1994. The role of in-
`tramolecular nucleophilic catalysis and the effects of self-
`association on the deamidation of human insulin at low
`pH. Pharm. Res. 11, 784–793.
`Diamant, B., Picazo, J., 1983. Spasmolytic action and clinical
`use of glucagon. In: Lefe`bvre, P.J. (Ed.), Handb. Exp.
`Pharm, vol. 66:2. Springer-Verlag, Berlin, pp. 611–643.
`
`Page 10
`
`
`
`A.B. Joshi et al. :International Journal of Pharmaceutics 203 (2000) 115–125
`
`125
`
`Geiger, T., Clarke, S., 1987. Deamidation, isomerization, and
`racemization at asparaginyl and aspartyl residues in pep-
`tides. Succinimide-linked reactions
`that contribute to
`protein degradation. J. Biol. Chem. 262, 785–794.
`Gratzer, W.B., Beaven, G.H., Rattle, H.W.E., Bradbury,
`E.M., 1968. A conformational study of glucagon. Eur. J.
`Biochem. 3, 276–283.
`Inglis, A.S., 1983. Cleavage at aspartic acid. Methods Enzy-
`mol. 91, 324–332.
`Kossiakoff, A.A., 1988. Tertiary structure is a principal deter-
`minant to protein deamidation. Science 240, 191–194.
`Light, A., 1967. Partial acid hydrolysis. Methods Enzymol. 11,
`417–420.
`Manning, M.C., Patel, K., Borchardt, R.T., 1989. Stability of
`protein pharmaceuticals. Pharm. Res. 6, 903–918.
`Marks, V., 1983. Glucagon in the diagnosis and treatment of
`hypoglycemia. In: Lefe`bvre, P.J. (Ed.), Handbook of Ex-
`perimental Pharmacology, vol. 66:2. Springer-Verlag,
`Berlin, pp. 645–672.
`Ota, I.M., Ding, L., Clarke, S., 1987. Methylation at specific
`altered aspartyl and asparaginyl residues in glucagon by
`the erythrocyte protein carboxyl methyltransferase. J. Biol.
`Chem. 262, 8522–8531.
`Patel, K.P., Borchardt, R.T., 1990. Chemical pathways of
`peptide degradation. II Kinetics of deamidation of an
`asparaginyl residue in a model hexapeptide. Pharm. Res. 7,
`703–711.
`Patel, K., Borchardt, R.T., 1990. Chemical pathways of pep-
`tide degradation. III Effect of primary sequence on the
`pathways of deamidation of asparaginyl
`residues
`in
`hexapeptides. Pharm. Res. 7, 787–793.
`Robinson, A.B., Rudd, C.J., 1974. Deamidation of glutaminyl
`
`and asparaginyl residues in peptides and proteins. Curr.
`Top. Cell Regul. 8, 247–295.
`Sasaki, K., Dockerill, S., Adamiak, D.A., Tickle, I.J., Blun-
`dell, T., 1975. X-ray analysis of glucagon and its relation-
`ship to receptor binding. Nature 257 (5529), 751–757.
`Schultz, J., 1967. Cleavage at aspartic acid. Methods Enzymol.
`11, 255–263.
`Srere, P.A., Brooks, G.C., 1969. The circular dichroism of
`glucagon solutions. Arch. Biochem. Biophys. 220, 708–
`710.
`US Pharmacopeia 24, 1999. US Pharmacopeial Convention,
`Rockville, MD, pp. 777.
`Wearne, S.J., Creighton, T.E., 1989. Effect of protein confor-
`mation on rate of deamidation: ribonuclease-A. Proteins 5,
`8–12.
`Windisch, V., DeLuccia, F., Duhau, L., Herman, F., Mencel,
`J.J., Tang, S.Y., Vuilhorgne, M., 1997. Degradation path-
`ways of salmon calcitonin in aqueous solution. J. Pharm.
`Sci. 86, 359–364.
`Wright, H.T., 1991. Sequence and structure determinants of
`the non-enzymatic deamidation of asparagine and glu-
`tamine in proteins. Protein Eng. 4, 283–291.
`Wright, H.T., Robinson, A.B., 1982. Cryptic amidase active
`sites catalyze deamidation in proteins. In: Kaplan, N.O.,
`Robinson, A.B. (Eds.), From Cyclotrons to Cytochromes.
`Academic Press, New York, pp. 727–743.
`Xie, M., Schowen, R., 1999. Secondary structure and protein
`deamidation. J. Pharm. Sci. 88, 8–13.
`Yi, G.S., Choi, B.S., Kim, H., 1992. The conformation of
`glucagon in dilute aqueous solutions as studied by 1H
`NMR. Biochem. Int. 28, 519–524.
`
`.
`
`Page 11