- Email: [email protected]
Simultaneous racemization and isomerization at specific aspartic acid residues in αB-crystallin from the aged human lens
ELSEVIER
Biochimica et Biophysica Acta 1204 (1994) 157-163
etBiochi~ic~a BiophysicaA~ta
Simultaneous racemization and isomerization at specific aspartic acid residues in aB-crystallin from the aged human lens Noriko Fujii a,,, Yoshihiro Ishibashi a, Kenshi Satoh b, Masahiko Fujino a, Kaoru Harada c a Discovery Research Division, Takeda Chemical Industries, Ltd., Wadai-lO, Tsukuba-shi, Ibaraki 300-42, Japan b Department of Applied Biological Science, Science University of Tokyo, Yamazaki 2641, Noda 278, Japan c Shoin Women's College, Shinohara, Obanoyamamachi 1-2-1, Nada-ku, Kobe 657, Japan (Received 6 July 1993)
Abstract
We provide evidence that the racemization and isomerization of aspartyl(Asp) residues occur simultaneously in the aB-crystallin in the lens of aged (mean age: 80 years) and young (age: 11 months) humans. We purified aB-crystallin and subjected it to tryptic digestion. The resulting peptides were separated by reverse-phase high-performance chromatography (RP-HPLC) and were characterized by amino-acid composition, sequence analysis and mass spectrometry. Two specific sites, Asp-36 (D/L of Asp: 0.92) and Asp-62(D/L of Asp: 0.57), among 13 Asp/asparginyl (Asn) residues in aged a B-crystailin, were found to be highly racemized and isomerized to form fl-Asp residues. The fl-Asp-containing peptides were clearly distinguished from normal Asp-containing (a-Asp) peptides by RP-HPLC. The racemization and isomerization of Asp residues in aged aB-crystallin may occur via a succinimide intermediate. In young aB-crystallin, we observed neither racemization nor isomerization. We also found that Met-68 was oxidized to form Met sulfoxide to a greater extent in aged aB-crystaUin than in young aB-crystallin. We concluded that racemization, isomerization, and oxidation of aB-crystallin occur spontaneously in the aging process. Key words: D-Aspartic acid; Lens; Crystallin; Racemization; Isomerization; (Human)
I. Introduction
The longevity of the a-crystallin protein of the vertebrate eye lens, makes it an excellent subject for the study of aging, a-Crystallin is a major structural protein of the lens, and is mainly composed of two types of protein, a A and aB. The amino-acid sequences of these proteins show appprox. 55% homology [1,2]. a A is 173 amino-acid residues in length and a B is 175 residues in length. It has previously been reported that the age-related racemization of L-aspartic acid ( L - A s p ) , leads to the accumulation of enantiomeric D-aspartic acid (D-Asp) in the lens protein [3,4]. We have also found that aA-crystallin is one of the easiest proteins to racemize ( O / L ratio of total Asp: 0.19) in the aged human lens [3]. We determined the D//L ratio of individual Asp and aspargine (Asn) residues in aA-crystal-
* Corresponding author. Fax: + 81 298 645000. 0167-4838/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0 1 6 7 - 4 8 3 8 ( 9 3 ) E 0 1 8 8 - 8
lin by analyzing their tryptic peptides. The A s p / A s n residues in the aA-crystallin racemized in a site-dependent manner. Two specific sites, Asp-58 ( D / L ratio of Asp: 1.4) and Asp-151 ( D / L ratio of Asp: 3.0), among 14 of these residues were the most highly racemized [3]. aB-Crystallin, another major protein of the lens, was recently reported to be expressed in various nonlenticular tissues, including heart, skeletal muscle, kidney, and nervous tissue [5-7]. It has recently been reported that o~B-crystallin is a small heat-shock protein [8] and a-crystallin can function as a molecular chaperone, which suppress aggregation of proteins [9]. Therefore, it is relevant to understand the post-translational modification of aB-crystallin itself. The D / L ratio of total Asp in aB-crystallin of the aged human lens is reported to be 0.11 [3]. If racemization also occurs in aB-crystallin in a site-specific manner, we would expect to find a high D / L ratio in some of the 13 A s p / A s n residues. Therefore, we searched
158
N. Fujiiet aL/ Biochimica et BiophysicaActa 1204 (1994) 157-163
for evidence of racemization at specific sites in the Asp and Asn residues of aB-crystallin in the aged human lens. The present report focuses on the racemization and isomerization of Asp and the oxidation of methionine in the aB-crystallin of the aging human lens.
2. Materials and methods
2.1. Purification of aB-crystallin from human lens We isolated aB-crystallin from the water soluble fraction of aged (mean age: 80 years) and young (age: 11 months) human lenses using ion-exchange chromatography and reverse-phase high-performance liquid chromatography (RP-HPLC), as described previously [3]. 2.2. Enzymatic digestions and isolation of peptides aB-Crystallin was digested with trypsin for 6 h at 37°C in 0.1 M Tris-HCl buffer (pH 7.6), at an enzyme to substrate ratio of 1:50 (mol/mol). The resulting peptides were separated by RP-HPLC (LC-9A, Shimadzu, Kyoto, Japan) using a C18 column (TSK gelODS-80 TM, 6 x 250 mm, Tosoh, Tokyo, Japan) with a linear gradient of 0-40% acetonitrile, 0.1% trifluoroacetic acid, at a flow rate of 1 ml/min, with monitoring at 230 nm. The tryptic peptides were further applied to RP-HPLC with different gradient conditions and the resulting peptides were recovered as a single peptide. The T16 and T3a peptides from the aged aB-crystallin were digested with thermolysin and Staphylococcus aureus V8 proteinase, respectively. Thermolytic digestion of T16 peptide was performed at 40°C for 3 h in 10 mM Hepes, 10 mM CaC12 (pH 7.2), at an enzyme to substrate ratio of 1:20 (mol/mol). The digestion of T3a with V8 proteinase was done to cleave Glu-X in 0.1 M ammonium acetate (pH 4.0), at 37°C for 24 h at an enzyme-to-substrate ratio of 1:20 (mol/mol). The peptide mapping of the second digestions of the tryptic peptides was performed by RP-HPLC under the same conditions as for the aB-tryptic peptides, but with monitoring at 215 nm. 2.3. Amino-acid analysis The peptides (100-200 pmol) were hydrolyzed with gas-phase 6 N HC1 in vacuo at 108°C for 24 h (PicoTag Work Stations, Waters, Tokyo, Japan). The hydrolysates were dried under reduced pressure and were dissolved in amino-acid analyzer sample buffer. The amino-acid analysis was performed on a Hitachi L-8500 amino-acid analyzer (Hitachi, Tokyo, Japan), in the form ofo-phthalaldehyde (OPA) derivatives and was monitored with a Hitachi F-1100 fluorescence spectro-
photometer (Hitachi, Tokyo, Japan), with the excitation and emission wave lengths sets at 350 and 400 nm, respectively. 2.4. Amino-acid sequence analysis The amino-acid sequences were determined using Edman degradation on a pulsed-liquid protein sequencer, equipped with an on-line phenylthiohydantoin (PTH) amino-acid analyzer (Applied Biosystems 477A/120A, Foster City, CA, USA). 2.5. Edman degradation Manual Edman degradation was performed as described elsewhere [10]. The T3a peptide (600 pmol) of aged aB-crystallin was placed in a glass tube and dried in vacuo. The peptide was dissolved in 50% aqueous pyridine (50/~l). Phenylisothiocyanate (2 /zl) was then added, nitrogen was introduced into the tubes, and the mixture was incubated for 20 rain at 55°C. When the coupling step was complete, excess reagent, solvents, and by-products were removed with n-butyl acetate and were dried completely in vacuo. Anhydrous trifluoroacetic acid (40/~l) was added to the residue, nitrogen was blown into the tube, and the mixture was incubated at 55°C for 5 min. The trifluoroacetic acid was removed, water (20 /zl) and n-butyl acetate (150 /xl) were added to the residue, and the contents were mixed thoroughly by vortexing. The two phases were separated by centrifugation, and the upper layer, which contained the anilinothiazolinone (ATZ) derivative, was extracted with n-butyl acetate and was then dried. The peptide-containing aqueous layer was dried and the procedure was repeated for four cycles. The ATZamino acids were hydrolyzed to free amino acids in 6 N HC1 at 150°C for 4 h, and were applied to the aminoacid analyzer. The remaining peptide was applied to the protein sequencer and an aliquot of the peptide was analyzed for the D / L ratio of Asp, as described below. 2.6. Determination of o / r ratio of amino acids Amino-acid contamination was controlled by baking all glassware at 500°C for 4 h. The peptide samples were lyophilized in tubes and were hydrolyzed with gas-phase 6 N HCI for 7 h at 108°C. After hydrolysis, the samples were dried again in vacuo prior to derivatization. The hydrolyzed samples were dissolved in 0.13 M borate buffer (pH 10.4) and were incubated briefly with o-phthalaldehyde (OPA) and n-tert-butyloxycarbonyl-L-cysteine (Boc-L-Cys) to form diastereoisomers [11]. The O / L ratio of the amino acids was determined using RP-HPLC (Shimadzu LC-9A) with a Nova-Pak ODS column (3.9 mm × 300 mm, Waters,
159
N. Fujii et al. / Biochimica et Biophysica A c t a 1204 (1994) 1 5 7 - 1 6 3
T2O
E ¢0 cO
0 e~
Vi ®
,5
/l
1"11
T8
~o
go
go
'(
(~
go go Retention Time (rain)
T14-16
?'o
go
g'G
Fig. 1. An elution profile of tryptic peptides of aged human aB-crystallin. Column, TOSOH ODS 80TM. Eluent: (A) 0.1% trifluoroacetic acid in water; (B) 0.1% trifluoroacetic acid in acetonitrile. Gradient: 0% B to 40% B in 90 min. Flow rate, 1 ml/min. Detection at 230 nm. Each peak was identified by amino-acid composition, sequence analysis, and FAB-MS. The tryptic peptides containing A s p / A s n residues are circled.
Tokyo, Japan), using fluorescence detection (344 nm of excitation wavelength and 433 nm of emission wavelength). Elution was carried out with a linear gradient of 7-47% acetonitrile, 3% tetrahydrofuran in 0.1 M acetate buffer (pH 6.0) in 120 min at a flow rate of 0.8 ml/min, at 30°C [12]. 2. 7. Fast atom bombardment mass spectrometry
Fast atom bombardment (FAB) mass spectra were recorded on a double-focussing mass spectrometer equipped with a cesium ion gun (JMS-HXllOHF, JEOL, Tokyo, Japan), as previously described [13]. 2.8. Nomenclature of peptides
The peptides are indicated by prefixes corresponding to the type of cleavage by which they are produced: T, tryptic digestion; Th, thermolytic digestion; V8, S. aureus V8 proteinase digestion; Ed, Edman degradation.
3. Results
tryptic peptides of aB-crystallin shown in Fig. 2. The tryptic peptides containing A s p / A s n residues are indicated by circles. The peptides T1 and Tla were searched for by FAB-MS (Fig. 3), since they are known to be NHz-terminally blocked by an acetyl group [14]. The D / L ratio of the respective aspartyl residues in aged aB-crystallin indicated that each aspartyl and asparginyl residue was racemized to a different extent (Table 1). Two tryptic peptides, T3a(1) (D/L ratio of Asp: 0.46) and T4'(1) (D/L ratio of Asp: 0.57) were extremely racemized. All other aspartyl residues were not racemized to a great extent (Table 1). Peptide T4 was separated into 3 peaks, as shown in Fig. 1. FAB-MS revealed that the mass of peptides T4'(1) and T4'(2) were the same (M + H += 1512.7), and that they were 16 mass units heavier than T4
I0 20 30 ac-M-D-I-A-I-H-H-P-W-I-R-~R-P-F-F-P-F~H-S-P-S-R-~L--F-D--(~-F-F-G-E-H-L-L,,I 1"1a ~', ~ I ] ~3aEd-1-~.-+I3W+d-2-40 50 60 E-S-D-L-F-P-T-S-T-S-L-S-P-F~Y-L-R-p-p-s-F-L-R4A-P-S-W-F-D-T-G-L--S-E-M-
i
3.1. Analysis of aB-crystallin from the aged human lens
In order to determine the D / L ratio of the A s p / A s n residues in the otB-crystallin, the purified aB-crystallin was subjected to tryptic digestion and analyzed by RP-HPLC. Fig. 1 shows a typical separation of the tryptic peptides of aB-crystallin. To confirm their purity, each peak was again subjected to RP-HPLC, under different gradient conditions. The peaks were assigned by amino-acid composition, sequence analysis, and FAB mass spectrometry (MS). The symbols on the respective peaks in Fig. 1 correspond to the numbers of
"~',+-2
~[
I
70 80 90 100 R+L-E-K~-D-R~-F-S-V-N-L--D-V-K~H-F-S-P-E-E-L--K~V-K~V-L-G-D-V- I-E-V-H-G-
~
110
120
130
E-E-R.~Q-D-E-H-G-F-I-S-R.[E-F-H-R4K-y-R~I-p-A-D--V-D-P-L-T-I.:T-S-S-L_ 1"11-~.-~-------~2-----------~-~T13-~T14-15 ]~ L_ T 1 6 ~
S_SID40_G_V_L_T_V_N_G_P_R IIS_~v_s _G_p_E_RIT_1_1_64~_T_RIE_E_K_P_A_V_~_O_A_
Fig. 2. The primary structure of human aB-crystallin ([2],[14]). a B crystallin was digested with trypsin (T). The tryptic peptides, T3a and T16 were further digested with S. aureus V8 proteinase (V8) and thermolysin (Th), respectively. T3a peptide was also characterized by Edman degradation (Ed).
160
N. Fujii et al. / Biochimica et Biophysica Acta 1204 (1994) 157-163
Table 2 Characterization of the T4 peptide of aB-crystallin from the aged human lens
100.
1430.8
Ca)
80.
58
40
v
Tryptic aB
Sequences
D/L of Asp
Linkage
M+H+
T4'(1) T4'(2) T4
APSWFD62TGLSEM(O)R APSWFD62 TGLSEM(O)R APSWFD62 TGLSEMR
0.57 0.04 0.09
/3 a a
1512.7 1512.7 1496.7
28
The linkage a and /3 correspond to the a-aspartyl and /3-aspartyl residues, respectively. M(O); methionine sulfoxide.
0 ¢-
t~
"e0.
d
.Q
< 180
1
(b)
1161.6
88
60
|
tt
1G~0
1708
, 1608
, l~Hfl
I 20Bfl
M/Z
Fig. 3. Assignments of peptides T1 and Tla. (a) Partial FAB mass spectrum of peak eluting at 67.1 rain, showing an intense signal at m / z 1430.8, which corresponded to peptide T1 (theoretical value 1430.7). (b) Partial FAB mass spectrum of peak eluting at 68.4 rain, showing an intense signal at m / z 1161.6, which corresponded to peptide T la (theoretical value 1161.6).
(M + H ÷= 1496.7), indicating that methionyl residues of T4'(1) and T4'(2) were oxidized to methionine sulfoxide. It was clear that the separation of T4 and T4' on the RP-HPLC was due to the oxidation of methionine. Sulfoxide is a chiral moiety with respect to the sulfur atom. Under this condition, however, the diastereoisomers, constituted from the S- and R-forms of Table 1 D/L ratio of aspartyl residues in tryptic peptides of ~B-crystallin from the aged human lens Peptide
Sequences
D//L
of
Asp T1 Tla T3a(1) T3a(2) T4 T4'(1) T4'(2) T6 T7 T10 T12 T16
acMD 2 IAIHHPWlR acMD2IAIHHPW LFD 25QFFGEHLLESD 36LFPTSTSLSPF LFD 25QFFGEHLLESD 36LFPTSTSLSPF APSWFD 62TGLSEMR APSWFD62TGLSEM(O)R APSWFD62 TGLSEM(O)R D7aR FSVNTSLDSOVK VLGD96VIEVHGK QDt°gEHGFISR IPAD 127VD 129PLTITSSLSSD 14°GVLTV NI46GPR
0.00 0.00 0.46 0.02 0.09 0.57 0.04 0.00 0.00 0.00 0.07 0.01
Methionyl residues of T4'(1) and T4'(2) were oxidized to methionine sulfoxide M(O). The detailed characterization of the T4, T4', T3a and T16 peptides are shown in Tables 2, 3 and 4, respectively.
L-methionine sulfoxide, could not be separated by RPHPLC. The ratio of T4' to T4 was estimated to be appprox. 1.0. The masses and amino-acid compositions of T4'(1) and T4'(2), were the same, but the D / L ratio of Asp was different. The D / L value (D/L ratio of Asp: 0.57) of the T4'(1) peptide, indicated that the separation of T4' into T4'(1) and T4'(2) was not caused by the diastereomeric effect. Had this not been the case, the L-Asp-containing peptide (L-Asp content: 100%) would need to be distinguished from the D-Asp-containing peptide (D-Asp content: 100%), chromatographically. Therefore, the linkage types of the Asp residues in T4'(1) and T4'(2), were analyzed by a protein sequencer. The analysis of T4'(2) detected phenylthiohydantoin(PTH)-Asp-62, while that of T4'(1) was stopped at Phe-61, which indicates a- and/3-type linkages, respectively. The separation of T4'(1) and T4'(2) on the RP-HPLC (Fig. 1) was almost certainly due to this difference. Table 2 summarizes the modifications, such as oxidation, racemization, and isomerization that were observed in the T4 peptide. Truscott et al. [15] have reported oxidation of methionine in the senile cataract human lens. The present study also demonstrated the oxidation of Met-68 to methionine sulfoxide in aB-crystallin from the aged human lens. T3a was also separated into two peaks (Fig. 1): highly racemized T3a(1) (D/L ratio of Asp: 0.46) and non-racemized T3a(2) ( D / L ratio of Asp: 0.02). Peptide T3a contained two Asp residues, at positions 25 and 36 (Fig. 2). In order to determine which Asp residue in T3a(1) was susceptible to racemization, and to clarify the linkage type of Asp in T3a(1) and T3a(2), both peptides were further digested with S. a u r e u s V8 proteinase, and separated into two peptides, T3aV8-1 (containing Asp-25) and T3aV8-2 (containing Asp-36). The non-racemized T3a(2) was completely digested with V8 proteinase and the T3a(2)V8-1 and T3a(2)V8-2 peptides were isolated by RP-HPLC (data not shown). These peptides were applied to the protein sequencer, and it was confirmed that the linkages of Asp-25 and Asp-36 were both a-type. After the hydrolysis of these peptides, no racemization was observed in either Asp residue (Table 3). The racemized T3a(1) was barely
N. Fujii et aL / Biochimica et Biophysica Acta 1204 (1994) 157-163
161
Table 3 Characterization of the T3a peptide of aB-crystallin from the aged human lens
Table 5 D//L ratio of aspartyl residues in the tryptic peptides of aB-crystallin from the young human lens
Tryptic aB
Sequences
D//L Linkage of ASp
Peptide
T3a(1)Ed-1 T3a(1)Ed-2 T3a(2)V8-1 T3a(2)V8-2
L/F/D25/Q F27FGEHLLESD36LFPTSTSLSPF LFD 25QFFGEHLLE SD36LFPTSTSLSPF
0.00 0.92 0.02 0.00
T1 T3a T4 T4' T6-7 T7 T10 T12 T16
a /3 a a
T3a(2) peptides were further digested with Staphylococcus aureus V8 proteinase, and separated into two peptides, T3a(2)V8-1 (containing D 25) and T3a(2)V8-2 (containing D36). T3a(1) was not digested with the enzyme. The separation of D 2s and D 36 of T3a(1) was performed using Edman degradation. The linkage a and /3 correspond to the a-aspartyl and /3-aspartyl residues, respectively.
Sequences
D//L of Asp
acMD 2IAIHHPWIR LFD 25QFFGEHLLESD 36LFPTSTSLSPF APSWFD62TGLSEMR APSWFD62TGLSEM(O)R D73RFSVN78LD8°VK FSVN7SLDS°VK VLGDq6VIEVHGK QDI°gEHGFISR IPAD127VD 129PLTITSSLSSD 14°GVLTV N146GPR
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
The methionyl residue of T4' was oxidized to methionine sulfoxide
cleaved by the V8 proteinase under any of the reaction conditions. The first four amino-acid residues (Leu 23Phe24-Asp25-Gln26)from the N-terminus of the T3a(1) peptide could be analyzed by a protein sequencer and showed an a-type linkage of Asp-25. Therefore, the PTH-Asp-25 obtained from protein sequencer, was collected and hydrolyzed with 6 N HCI for 17 h at 108°C to free Asp-25 (this condition prevents racemization). The Asp-25 thus obtained was converted to BOC-LC y s / O P A derivative and RP-HPLC analysis confirmed that the Asp-25 was not racemized (Table 3). We then attempted to determine the D / L ratio of Asp-36 in the remaining peptide obtained by releasing the first four residues in T3a(1) using manual Edman degradation. When the degradation of T3a(1) had proceeded to the first four residues, the resulting ATZ-amino acids were hydrolyzed and the progress of the reaction was confirmed using the amino-acid analyzer. The sequence analysis of the remaining peptide proceeded from Phe27 to Ser-35, with a repetitive yield of 92%, and little PTH-Asp-36 was detected (below 10% of the theoretical value). The results indicated that the configuration of Asp-36 was /3-type. An aliquot from the remaining peptide, was hydrolyzed and the D / L value of Asp-36 was estimated to be 0.92 (Table 3). The racemization of Asp-36 was accompanied by isomerization, as in the case of Asp-62. Non-digestion with V8 proteinase may be caused by the presence of D-Asp or/3-linkage aspartic acid residue (/3-Asp) in the T3a(1) peptide. Since T16 contained three Asp residues and one Asn residue, the peptide was digested with thermolysin Table 4 D//L ratios and linkage types of the ASp/ASn residues in thermolytic peptides of the T16 peptide of aB-crystallin from aged human lens Peptide
Sequences
D//L of Asp
Linkage
T16Thl T16Th4 T16Th5
IPADI27VD129p LSSD14°GVLT VN146GPR
0.01 0.01 0.02
a a ot
T16 was further digested with thermolysin (Th). The linkage a and /3 correspond to the ct-aspartyl and /3-aspartyl residues, respectively.
M(O). in order to determine whether site-specific racemization and isomerization of A s p / A s n had occurred. The resulting peptides were isolated by RP-HPLC and applied to the protein sequencer to assign the peptide and to determine the linkage type of Asp/Asn. The D / L value of A s p / A s n was determined following hydrolysis of the peptides and no racemization was found (Table 4). The results are consistent with the ones recently reported [16].
3.2. Analysis of aB-crystallin from the young human lens We also analyzed the tryptic peptides of aB-crystallin obtained from the young human lens. None of the Asp residues were racemized (Table 5). Only a small amount (7%) of the T4' peptide (methionine sulfoxide-containing peptide) was detected. In contrast to the peptides of the aged lens, the T4' and T3a peptides of young aB-crystallin were all single peptides. We also confirmed that the linkages of Asp of the T4, T4' and T3a peptides of young aB-crystallin were all atype.
4. Discussion
We found that the racemization of Asp residues occurs in both aA-and aB-crystallin from the aged human lens, and we found site-specific racemization of Asp residues in aA-crystallin [3]. Site-specific racemization was also found in the aB-crystallin of the aged human lens. Two specific positions at Asp-36 (D/L ratio of Asp: 0.92) and Asp-62 (D/L ratio of Asp: 0.57) were extremely racemized and the racemization of Asp at both sites was accompanied by isomerization (Tables 2 and 3). This is the first study to demonstrate the formation of fl-Asp and D-Asp residues by spontaneous isomerization and racemization of a-aspartyl residues in aB-crystallin during the aging.
162
N. Fujii et al. / Biochimica et Biophysica Acta 1204 (1994) 157-163
RP-HPLC clearly separated the peptides containing /3- Asp and a-Asp, and the peptide containing /3-Asp was eluted faster than that containing a-Asp in both the T4 and T3a peptides (Fig. 1). The tendency is consistent with the separation of synthetic peptides containing a-Asp and /3-Asp residues [17]. On the other hand, the resolution of diastereomeric peptides (O- and L-Asp-containing-peptides, and the S- and Rforms of L-methionine sulfoxide-containing-peptides) could not be achieved under these conditions. The racemization of the Asp residue occurs to a greater extent in/3-Asp than in a-Asp in both T3a and T4 (Tables 2 and 3). It has been known that/3-Asp and D-Asp residues are derived from L-Asn/Asp residues under physiological conditions through the formation of a succinimide intermediate [17-21] in small peptides. The intermediate may be hydrolyzed at either side of its two carbonyl groups, yielding both /3- and a-Asp residues in small peptides. The succinimide can racemize far more rapidly than open chain structures. It is well known that o-succinimide is prefer opened to /3-D-Asp, tO a-D-Asp [19]. The rate of succinimide ring formation was found to be influenced by the neighboring amino-acid residue. The cyclization reaction to form succinimide takes place most easily in the Asn-Gly sequence in small peptides because there is less steric hindrance of the Gly residue [19-21]. In the case of aB-crystaUin, however, the isomerized and racemized Asp residues originated from L-Asp adjoined to bulky Leu and Thr, respectively, which may cause difficulty in forming a succinimide intermediate. Yet, despite the fact aB-crystallin contains both Asp140-Gly and Asn146-Gly sequences, none of the Asp and Asn residues in the T16 peptide were racemized or isomerized (Table 4). Additionally, it is known that /3-Asp residues are derived from the L-Asp residue in glucagon (Asp-Tyr) [22] and calmodulin (Asp-Gin, Asp-Thr) [23]. These observations indicate that the succinimide formation in the protein must depend on the higher order structure rather than on the primary structure. The molar ratio of /3-Asp to a-Asp residues in synthetic peptides has been reported to be 3 to 1 [19]. However, this was not the case for either T3a or T4, which may indicate that, under physiological conditions, the racemization and isomerization reactions do not reach equilibrium among the/3-Asp, a-Asp, and succinimidyl forms in the aged lens. We predicted the secondary structure of aB-crystallin using Robson algorithms (data not shown). The result suggests that Asp-36 and Asp-62 may exist in a /3-turn structure and that the following sequences are in a random coil structure. These sites may be located in the flexible surface region of the aB-crystallin protein. In particular, the fact that methionine was oxidized in the T4 peptide indicates that the surrounding region of Asp-62 may be exposed outside of the pro-
tein. Shapira et al. recently suggested that the racemized aspartic-acid residue of myelin basic protein in brain may exist in a/3-turn region [24]. The racemization and isomerization of Asp residues in aB-crystallin is likely to take place through a succinimide intermediate, which produces the D-Asp and /3-Asp residues observed in this study. Regardless of the mechanism of racemization and isomerization, this posttranslational modification of the amino-acid residues of aB-crystallin is likely to occur spontaneously in the aging process. This is likely because similar alternations were not observed in young aB-crystallin and such modifications would be expected to change properties such as solubility and conformation of the protein. Since aggregated, insoluble a-crystallin protein is reported to be more racemized than soluble a-crystallin [25], it is possible that the insolubilization of crystallins is closely related to the configuration of a-crystallin in the aging lens. If acrystallin have a protective role for some lens proteins from aggregation in transparent eye lens [9], the disruption of their structure by the modification of aBcrystaUin itself, such as the racemization and isomerization of Asp residues, and the oxidation of Met residue, could have wider effects on lens supramolecular organization.
References [1] De Jong, W.W., Terwindt, E.C. and Bloemendal, H. (1975) FEBS Lett. 58, 310-313. [2] Dubin, R.A., Ally, A.h., Chung, S. and Piatigorsky J. (1990) Genomics, 7, 594-601. [3] Fujii, N., Muraoka, S., Satoh, K., Hori, H. and Harada, K. (1991) Biomed. Res. 12, 315-321. [4] Groenen, P.J.T.A., Van den IJssel, P.R.L.A., Voorter, C.E.M., Bloemendal, H. and De Jong, W.W. (1990) FEBS Lett. 269, 109-112, [5] Bhat, S.P. and Nagineni, C.N. (1989) Biochem. Biophys. Res. Commun. 158, 319-325. [6] Dubin, R.A., Wawrousek, E.F. and Piatigorsky, J. (1989) Mol. Cell. Biol. 9, 1083-1091. [7] Kato, K., Shinohara, H., Kurobe, N., tnaguma, Y., Shimizu, K. and Oshima, K. (1991) Biochim. Biophys. Acta. 1074, 201-208. [8] Klemenz, R., Frohli, E., Steiger, R.H., Schafer, R. and Aoyama, A. (1991) Proc. Natl. Acad. Sci. USA 88, 3652-3656. [9] Horwitz, J. (1992) Proc. Natl. Acad. Sci. USA 89, 10449-10453. [10] Allen, G. (1981) in Sequencing of Proteins and Peptides (Work T.S. and Burdon R.H., eds.), pp. 174-221, North-Holland, Amsterdam. [11] Bruckner, H. and Wittner, R. (1989) J. Chromatogr. 476, 73-82. [12] Hashimoto, A., Nishikawa, T., Oka, T., Takahashi, K. and Hayashi, T. (1992) J. Chromatogr. 582, 41-48. [13] Ishibashi, Y., Kikuchi, T., Wakimasu, M., Mizuta, E. and Fujino, M. (1991) Biol. Mass. Spectrom. 20, 703-708. [14] Kramps, J.A., De Man, B.M. and De Jong, W.W. (1977) FEBS Lett. 74, 82-84. [15] Truscott, R.J.W. and Augusteyn, R.C. (1977) Biochim. Biophys. Acta. 492, 43-52.
N. Fujii et al. / Biochimica et Biophysica Acta 1204 (1994) 157-163
[16] Groenen, P.J.T.A., van Dongen, M.J.P., Voorter, C.E.M., Bloemendal, H. and De Jong, W.W. (1993) FEBS Lett. 322, 69-72. [17] Murray, E.D., Jr. and Clarke, S. (1984) J. Biol. Chem. 259, 10722-10732. [18] McFadden, P.N. and Clarke, S. (1982) Proc. Natl. Acad. Sci. USA 79, 2460-2464. [19] Geiger, T. and Clarke, S. (1987) J. Biol. Chem. 262, 785-794. [20] Stephanson, R.C. and Clarke, S. (1989) J. Biol. Chem. 264, 6164-6170.
163
[21] Cross, R.T. and Schirch, V. (1991) J. Biol. Chem. 266, 2254922556. [22] Ota, I.M., Ding, L. and Clarke, S. (1987) J. Biol. Chem. 262, 8522-8531. [23] Ota, I.M. and Clarke, S. (1989) J. Biol. Chem. 264, 54-60. [24] Shapira, R., Wilkinson, K.D. and Shapira, G. (1988) J. Neurochem. 50, 649-654. [25] Muraoka, S., Fujii, N., Ueda, Y., Mitsui, Y., Satoh, K. and Harada, K. (1991) Biomed. Res. 12, 61-64.
Biochimica et Biophysica Acta 1204 (1994) 157-163
etBiochi~ic~a BiophysicaA~ta
Simultaneous racemization and isomerization at specific aspartic acid residues in aB-crystallin from the aged human lens Noriko Fujii a,,, Yoshihiro Ishibashi a, Kenshi Satoh b, Masahiko Fujino a, Kaoru Harada c a Discovery Research Division, Takeda Chemical Industries, Ltd., Wadai-lO, Tsukuba-shi, Ibaraki 300-42, Japan b Department of Applied Biological Science, Science University of Tokyo, Yamazaki 2641, Noda 278, Japan c Shoin Women's College, Shinohara, Obanoyamamachi 1-2-1, Nada-ku, Kobe 657, Japan (Received 6 July 1993)
Abstract
We provide evidence that the racemization and isomerization of aspartyl(Asp) residues occur simultaneously in the aB-crystallin in the lens of aged (mean age: 80 years) and young (age: 11 months) humans. We purified aB-crystallin and subjected it to tryptic digestion. The resulting peptides were separated by reverse-phase high-performance chromatography (RP-HPLC) and were characterized by amino-acid composition, sequence analysis and mass spectrometry. Two specific sites, Asp-36 (D/L of Asp: 0.92) and Asp-62(D/L of Asp: 0.57), among 13 Asp/asparginyl (Asn) residues in aged a B-crystailin, were found to be highly racemized and isomerized to form fl-Asp residues. The fl-Asp-containing peptides were clearly distinguished from normal Asp-containing (a-Asp) peptides by RP-HPLC. The racemization and isomerization of Asp residues in aged aB-crystallin may occur via a succinimide intermediate. In young aB-crystallin, we observed neither racemization nor isomerization. We also found that Met-68 was oxidized to form Met sulfoxide to a greater extent in aged aB-crystaUin than in young aB-crystallin. We concluded that racemization, isomerization, and oxidation of aB-crystallin occur spontaneously in the aging process. Key words: D-Aspartic acid; Lens; Crystallin; Racemization; Isomerization; (Human)
I. Introduction
The longevity of the a-crystallin protein of the vertebrate eye lens, makes it an excellent subject for the study of aging, a-Crystallin is a major structural protein of the lens, and is mainly composed of two types of protein, a A and aB. The amino-acid sequences of these proteins show appprox. 55% homology [1,2]. a A is 173 amino-acid residues in length and a B is 175 residues in length. It has previously been reported that the age-related racemization of L-aspartic acid ( L - A s p ) , leads to the accumulation of enantiomeric D-aspartic acid (D-Asp) in the lens protein [3,4]. We have also found that aA-crystallin is one of the easiest proteins to racemize ( O / L ratio of total Asp: 0.19) in the aged human lens [3]. We determined the D//L ratio of individual Asp and aspargine (Asn) residues in aA-crystal-
* Corresponding author. Fax: + 81 298 645000. 0167-4838/94/$07.00 © 1994 Elsevier Science B.V. All rights reserved SSDI 0 1 6 7 - 4 8 3 8 ( 9 3 ) E 0 1 8 8 - 8
lin by analyzing their tryptic peptides. The A s p / A s n residues in the aA-crystallin racemized in a site-dependent manner. Two specific sites, Asp-58 ( D / L ratio of Asp: 1.4) and Asp-151 ( D / L ratio of Asp: 3.0), among 14 of these residues were the most highly racemized [3]. aB-Crystallin, another major protein of the lens, was recently reported to be expressed in various nonlenticular tissues, including heart, skeletal muscle, kidney, and nervous tissue [5-7]. It has recently been reported that o~B-crystallin is a small heat-shock protein [8] and a-crystallin can function as a molecular chaperone, which suppress aggregation of proteins [9]. Therefore, it is relevant to understand the post-translational modification of aB-crystallin itself. The D / L ratio of total Asp in aB-crystallin of the aged human lens is reported to be 0.11 [3]. If racemization also occurs in aB-crystallin in a site-specific manner, we would expect to find a high D / L ratio in some of the 13 A s p / A s n residues. Therefore, we searched
158
N. Fujiiet aL/ Biochimica et BiophysicaActa 1204 (1994) 157-163
for evidence of racemization at specific sites in the Asp and Asn residues of aB-crystallin in the aged human lens. The present report focuses on the racemization and isomerization of Asp and the oxidation of methionine in the aB-crystallin of the aging human lens.
2. Materials and methods
2.1. Purification of aB-crystallin from human lens We isolated aB-crystallin from the water soluble fraction of aged (mean age: 80 years) and young (age: 11 months) human lenses using ion-exchange chromatography and reverse-phase high-performance liquid chromatography (RP-HPLC), as described previously [3]. 2.2. Enzymatic digestions and isolation of peptides aB-Crystallin was digested with trypsin for 6 h at 37°C in 0.1 M Tris-HCl buffer (pH 7.6), at an enzyme to substrate ratio of 1:50 (mol/mol). The resulting peptides were separated by RP-HPLC (LC-9A, Shimadzu, Kyoto, Japan) using a C18 column (TSK gelODS-80 TM, 6 x 250 mm, Tosoh, Tokyo, Japan) with a linear gradient of 0-40% acetonitrile, 0.1% trifluoroacetic acid, at a flow rate of 1 ml/min, with monitoring at 230 nm. The tryptic peptides were further applied to RP-HPLC with different gradient conditions and the resulting peptides were recovered as a single peptide. The T16 and T3a peptides from the aged aB-crystallin were digested with thermolysin and Staphylococcus aureus V8 proteinase, respectively. Thermolytic digestion of T16 peptide was performed at 40°C for 3 h in 10 mM Hepes, 10 mM CaC12 (pH 7.2), at an enzyme to substrate ratio of 1:20 (mol/mol). The digestion of T3a with V8 proteinase was done to cleave Glu-X in 0.1 M ammonium acetate (pH 4.0), at 37°C for 24 h at an enzyme-to-substrate ratio of 1:20 (mol/mol). The peptide mapping of the second digestions of the tryptic peptides was performed by RP-HPLC under the same conditions as for the aB-tryptic peptides, but with monitoring at 215 nm. 2.3. Amino-acid analysis The peptides (100-200 pmol) were hydrolyzed with gas-phase 6 N HC1 in vacuo at 108°C for 24 h (PicoTag Work Stations, Waters, Tokyo, Japan). The hydrolysates were dried under reduced pressure and were dissolved in amino-acid analyzer sample buffer. The amino-acid analysis was performed on a Hitachi L-8500 amino-acid analyzer (Hitachi, Tokyo, Japan), in the form ofo-phthalaldehyde (OPA) derivatives and was monitored with a Hitachi F-1100 fluorescence spectro-
photometer (Hitachi, Tokyo, Japan), with the excitation and emission wave lengths sets at 350 and 400 nm, respectively. 2.4. Amino-acid sequence analysis The amino-acid sequences were determined using Edman degradation on a pulsed-liquid protein sequencer, equipped with an on-line phenylthiohydantoin (PTH) amino-acid analyzer (Applied Biosystems 477A/120A, Foster City, CA, USA). 2.5. Edman degradation Manual Edman degradation was performed as described elsewhere [10]. The T3a peptide (600 pmol) of aged aB-crystallin was placed in a glass tube and dried in vacuo. The peptide was dissolved in 50% aqueous pyridine (50/~l). Phenylisothiocyanate (2 /zl) was then added, nitrogen was introduced into the tubes, and the mixture was incubated for 20 rain at 55°C. When the coupling step was complete, excess reagent, solvents, and by-products were removed with n-butyl acetate and were dried completely in vacuo. Anhydrous trifluoroacetic acid (40/~l) was added to the residue, nitrogen was blown into the tube, and the mixture was incubated at 55°C for 5 min. The trifluoroacetic acid was removed, water (20 /zl) and n-butyl acetate (150 /xl) were added to the residue, and the contents were mixed thoroughly by vortexing. The two phases were separated by centrifugation, and the upper layer, which contained the anilinothiazolinone (ATZ) derivative, was extracted with n-butyl acetate and was then dried. The peptide-containing aqueous layer was dried and the procedure was repeated for four cycles. The ATZamino acids were hydrolyzed to free amino acids in 6 N HC1 at 150°C for 4 h, and were applied to the aminoacid analyzer. The remaining peptide was applied to the protein sequencer and an aliquot of the peptide was analyzed for the D / L ratio of Asp, as described below. 2.6. Determination of o / r ratio of amino acids Amino-acid contamination was controlled by baking all glassware at 500°C for 4 h. The peptide samples were lyophilized in tubes and were hydrolyzed with gas-phase 6 N HCI for 7 h at 108°C. After hydrolysis, the samples were dried again in vacuo prior to derivatization. The hydrolyzed samples were dissolved in 0.13 M borate buffer (pH 10.4) and were incubated briefly with o-phthalaldehyde (OPA) and n-tert-butyloxycarbonyl-L-cysteine (Boc-L-Cys) to form diastereoisomers [11]. The O / L ratio of the amino acids was determined using RP-HPLC (Shimadzu LC-9A) with a Nova-Pak ODS column (3.9 mm × 300 mm, Waters,
159
N. Fujii et al. / Biochimica et Biophysica A c t a 1204 (1994) 1 5 7 - 1 6 3
T2O
E ¢0 cO
0 e~
Vi ®
,5
/l
1"11
T8
~o
go
go
'(
(~
go go Retention Time (rain)
T14-16
?'o
go
g'G
Fig. 1. An elution profile of tryptic peptides of aged human aB-crystallin. Column, TOSOH ODS 80TM. Eluent: (A) 0.1% trifluoroacetic acid in water; (B) 0.1% trifluoroacetic acid in acetonitrile. Gradient: 0% B to 40% B in 90 min. Flow rate, 1 ml/min. Detection at 230 nm. Each peak was identified by amino-acid composition, sequence analysis, and FAB-MS. The tryptic peptides containing A s p / A s n residues are circled.
Tokyo, Japan), using fluorescence detection (344 nm of excitation wavelength and 433 nm of emission wavelength). Elution was carried out with a linear gradient of 7-47% acetonitrile, 3% tetrahydrofuran in 0.1 M acetate buffer (pH 6.0) in 120 min at a flow rate of 0.8 ml/min, at 30°C [12]. 2. 7. Fast atom bombardment mass spectrometry
Fast atom bombardment (FAB) mass spectra were recorded on a double-focussing mass spectrometer equipped with a cesium ion gun (JMS-HXllOHF, JEOL, Tokyo, Japan), as previously described [13]. 2.8. Nomenclature of peptides
The peptides are indicated by prefixes corresponding to the type of cleavage by which they are produced: T, tryptic digestion; Th, thermolytic digestion; V8, S. aureus V8 proteinase digestion; Ed, Edman degradation.
3. Results
tryptic peptides of aB-crystallin shown in Fig. 2. The tryptic peptides containing A s p / A s n residues are indicated by circles. The peptides T1 and Tla were searched for by FAB-MS (Fig. 3), since they are known to be NHz-terminally blocked by an acetyl group [14]. The D / L ratio of the respective aspartyl residues in aged aB-crystallin indicated that each aspartyl and asparginyl residue was racemized to a different extent (Table 1). Two tryptic peptides, T3a(1) (D/L ratio of Asp: 0.46) and T4'(1) (D/L ratio of Asp: 0.57) were extremely racemized. All other aspartyl residues were not racemized to a great extent (Table 1). Peptide T4 was separated into 3 peaks, as shown in Fig. 1. FAB-MS revealed that the mass of peptides T4'(1) and T4'(2) were the same (M + H += 1512.7), and that they were 16 mass units heavier than T4
I0 20 30 ac-M-D-I-A-I-H-H-P-W-I-R-~R-P-F-F-P-F~H-S-P-S-R-~L--F-D--(~-F-F-G-E-H-L-L,,I 1"1a ~', ~ I ] ~3aEd-1-~.-+I3W+d-2-40 50 60 E-S-D-L-F-P-T-S-T-S-L-S-P-F~Y-L-R-p-p-s-F-L-R4A-P-S-W-F-D-T-G-L--S-E-M-
i
3.1. Analysis of aB-crystallin from the aged human lens
In order to determine the D / L ratio of the A s p / A s n residues in the otB-crystallin, the purified aB-crystallin was subjected to tryptic digestion and analyzed by RP-HPLC. Fig. 1 shows a typical separation of the tryptic peptides of aB-crystallin. To confirm their purity, each peak was again subjected to RP-HPLC, under different gradient conditions. The peaks were assigned by amino-acid composition, sequence analysis, and FAB mass spectrometry (MS). The symbols on the respective peaks in Fig. 1 correspond to the numbers of
"~',+-2
~[
I
70 80 90 100 R+L-E-K~-D-R~-F-S-V-N-L--D-V-K~H-F-S-P-E-E-L--K~V-K~V-L-G-D-V- I-E-V-H-G-
~
110
120
130
E-E-R.~Q-D-E-H-G-F-I-S-R.[E-F-H-R4K-y-R~I-p-A-D--V-D-P-L-T-I.:T-S-S-L_ 1"11-~.-~-------~2-----------~-~T13-~T14-15 ]~ L_ T 1 6 ~
S_SID40_G_V_L_T_V_N_G_P_R IIS_~v_s _G_p_E_RIT_1_1_64~_T_RIE_E_K_P_A_V_~_O_A_
Fig. 2. The primary structure of human aB-crystallin ([2],[14]). a B crystallin was digested with trypsin (T). The tryptic peptides, T3a and T16 were further digested with S. aureus V8 proteinase (V8) and thermolysin (Th), respectively. T3a peptide was also characterized by Edman degradation (Ed).
160
N. Fujii et al. / Biochimica et Biophysica Acta 1204 (1994) 157-163
Table 2 Characterization of the T4 peptide of aB-crystallin from the aged human lens
100.
1430.8
Ca)
80.
58
40
v
Tryptic aB
Sequences
D/L of Asp
Linkage
M+H+
T4'(1) T4'(2) T4
APSWFD62TGLSEM(O)R APSWFD62 TGLSEM(O)R APSWFD62 TGLSEMR
0.57 0.04 0.09
/3 a a
1512.7 1512.7 1496.7
28
The linkage a and /3 correspond to the a-aspartyl and /3-aspartyl residues, respectively. M(O); methionine sulfoxide.
0 ¢-
t~
"e0.
d
.Q
< 180
1
(b)
1161.6
88
60
|
tt
1G~0
1708
, 1608
, l~Hfl
I 20Bfl
M/Z
Fig. 3. Assignments of peptides T1 and Tla. (a) Partial FAB mass spectrum of peak eluting at 67.1 rain, showing an intense signal at m / z 1430.8, which corresponded to peptide T1 (theoretical value 1430.7). (b) Partial FAB mass spectrum of peak eluting at 68.4 rain, showing an intense signal at m / z 1161.6, which corresponded to peptide T la (theoretical value 1161.6).
(M + H ÷= 1496.7), indicating that methionyl residues of T4'(1) and T4'(2) were oxidized to methionine sulfoxide. It was clear that the separation of T4 and T4' on the RP-HPLC was due to the oxidation of methionine. Sulfoxide is a chiral moiety with respect to the sulfur atom. Under this condition, however, the diastereoisomers, constituted from the S- and R-forms of Table 1 D/L ratio of aspartyl residues in tryptic peptides of ~B-crystallin from the aged human lens Peptide
Sequences
D//L
of
Asp T1 Tla T3a(1) T3a(2) T4 T4'(1) T4'(2) T6 T7 T10 T12 T16
acMD 2 IAIHHPWlR acMD2IAIHHPW LFD 25QFFGEHLLESD 36LFPTSTSLSPF LFD 25QFFGEHLLESD 36LFPTSTSLSPF APSWFD 62TGLSEMR APSWFD62TGLSEM(O)R APSWFD62 TGLSEM(O)R D7aR FSVNTSLDSOVK VLGD96VIEVHGK QDt°gEHGFISR IPAD 127VD 129PLTITSSLSSD 14°GVLTV NI46GPR
0.00 0.00 0.46 0.02 0.09 0.57 0.04 0.00 0.00 0.00 0.07 0.01
Methionyl residues of T4'(1) and T4'(2) were oxidized to methionine sulfoxide M(O). The detailed characterization of the T4, T4', T3a and T16 peptides are shown in Tables 2, 3 and 4, respectively.
L-methionine sulfoxide, could not be separated by RPHPLC. The ratio of T4' to T4 was estimated to be appprox. 1.0. The masses and amino-acid compositions of T4'(1) and T4'(2), were the same, but the D / L ratio of Asp was different. The D / L value (D/L ratio of Asp: 0.57) of the T4'(1) peptide, indicated that the separation of T4' into T4'(1) and T4'(2) was not caused by the diastereomeric effect. Had this not been the case, the L-Asp-containing peptide (L-Asp content: 100%) would need to be distinguished from the D-Asp-containing peptide (D-Asp content: 100%), chromatographically. Therefore, the linkage types of the Asp residues in T4'(1) and T4'(2), were analyzed by a protein sequencer. The analysis of T4'(2) detected phenylthiohydantoin(PTH)-Asp-62, while that of T4'(1) was stopped at Phe-61, which indicates a- and/3-type linkages, respectively. The separation of T4'(1) and T4'(2) on the RP-HPLC (Fig. 1) was almost certainly due to this difference. Table 2 summarizes the modifications, such as oxidation, racemization, and isomerization that were observed in the T4 peptide. Truscott et al. [15] have reported oxidation of methionine in the senile cataract human lens. The present study also demonstrated the oxidation of Met-68 to methionine sulfoxide in aB-crystallin from the aged human lens. T3a was also separated into two peaks (Fig. 1): highly racemized T3a(1) (D/L ratio of Asp: 0.46) and non-racemized T3a(2) ( D / L ratio of Asp: 0.02). Peptide T3a contained two Asp residues, at positions 25 and 36 (Fig. 2). In order to determine which Asp residue in T3a(1) was susceptible to racemization, and to clarify the linkage type of Asp in T3a(1) and T3a(2), both peptides were further digested with S. a u r e u s V8 proteinase, and separated into two peptides, T3aV8-1 (containing Asp-25) and T3aV8-2 (containing Asp-36). The non-racemized T3a(2) was completely digested with V8 proteinase and the T3a(2)V8-1 and T3a(2)V8-2 peptides were isolated by RP-HPLC (data not shown). These peptides were applied to the protein sequencer, and it was confirmed that the linkages of Asp-25 and Asp-36 were both a-type. After the hydrolysis of these peptides, no racemization was observed in either Asp residue (Table 3). The racemized T3a(1) was barely
N. Fujii et aL / Biochimica et Biophysica Acta 1204 (1994) 157-163
161
Table 3 Characterization of the T3a peptide of aB-crystallin from the aged human lens
Table 5 D//L ratio of aspartyl residues in the tryptic peptides of aB-crystallin from the young human lens
Tryptic aB
Sequences
D//L Linkage of ASp
Peptide
T3a(1)Ed-1 T3a(1)Ed-2 T3a(2)V8-1 T3a(2)V8-2
L/F/D25/Q F27FGEHLLESD36LFPTSTSLSPF LFD 25QFFGEHLLE SD36LFPTSTSLSPF
0.00 0.92 0.02 0.00
T1 T3a T4 T4' T6-7 T7 T10 T12 T16
a /3 a a
T3a(2) peptides were further digested with Staphylococcus aureus V8 proteinase, and separated into two peptides, T3a(2)V8-1 (containing D 25) and T3a(2)V8-2 (containing D36). T3a(1) was not digested with the enzyme. The separation of D 2s and D 36 of T3a(1) was performed using Edman degradation. The linkage a and /3 correspond to the a-aspartyl and /3-aspartyl residues, respectively.
Sequences
D//L of Asp
acMD 2IAIHHPWIR LFD 25QFFGEHLLESD 36LFPTSTSLSPF APSWFD62TGLSEMR APSWFD62TGLSEM(O)R D73RFSVN78LD8°VK FSVN7SLDS°VK VLGDq6VIEVHGK QDI°gEHGFISR IPAD127VD 129PLTITSSLSSD 14°GVLTV N146GPR
0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
The methionyl residue of T4' was oxidized to methionine sulfoxide
cleaved by the V8 proteinase under any of the reaction conditions. The first four amino-acid residues (Leu 23Phe24-Asp25-Gln26)from the N-terminus of the T3a(1) peptide could be analyzed by a protein sequencer and showed an a-type linkage of Asp-25. Therefore, the PTH-Asp-25 obtained from protein sequencer, was collected and hydrolyzed with 6 N HCI for 17 h at 108°C to free Asp-25 (this condition prevents racemization). The Asp-25 thus obtained was converted to BOC-LC y s / O P A derivative and RP-HPLC analysis confirmed that the Asp-25 was not racemized (Table 3). We then attempted to determine the D / L ratio of Asp-36 in the remaining peptide obtained by releasing the first four residues in T3a(1) using manual Edman degradation. When the degradation of T3a(1) had proceeded to the first four residues, the resulting ATZ-amino acids were hydrolyzed and the progress of the reaction was confirmed using the amino-acid analyzer. The sequence analysis of the remaining peptide proceeded from Phe27 to Ser-35, with a repetitive yield of 92%, and little PTH-Asp-36 was detected (below 10% of the theoretical value). The results indicated that the configuration of Asp-36 was /3-type. An aliquot from the remaining peptide, was hydrolyzed and the D / L value of Asp-36 was estimated to be 0.92 (Table 3). The racemization of Asp-36 was accompanied by isomerization, as in the case of Asp-62. Non-digestion with V8 proteinase may be caused by the presence of D-Asp or/3-linkage aspartic acid residue (/3-Asp) in the T3a(1) peptide. Since T16 contained three Asp residues and one Asn residue, the peptide was digested with thermolysin Table 4 D//L ratios and linkage types of the ASp/ASn residues in thermolytic peptides of the T16 peptide of aB-crystallin from aged human lens Peptide
Sequences
D//L of Asp
Linkage
T16Thl T16Th4 T16Th5
IPADI27VD129p LSSD14°GVLT VN146GPR
0.01 0.01 0.02
a a ot
T16 was further digested with thermolysin (Th). The linkage a and /3 correspond to the ct-aspartyl and /3-aspartyl residues, respectively.
M(O). in order to determine whether site-specific racemization and isomerization of A s p / A s n had occurred. The resulting peptides were isolated by RP-HPLC and applied to the protein sequencer to assign the peptide and to determine the linkage type of Asp/Asn. The D / L value of A s p / A s n was determined following hydrolysis of the peptides and no racemization was found (Table 4). The results are consistent with the ones recently reported [16].
3.2. Analysis of aB-crystallin from the young human lens We also analyzed the tryptic peptides of aB-crystallin obtained from the young human lens. None of the Asp residues were racemized (Table 5). Only a small amount (7%) of the T4' peptide (methionine sulfoxide-containing peptide) was detected. In contrast to the peptides of the aged lens, the T4' and T3a peptides of young aB-crystallin were all single peptides. We also confirmed that the linkages of Asp of the T4, T4' and T3a peptides of young aB-crystallin were all atype.
4. Discussion
We found that the racemization of Asp residues occurs in both aA-and aB-crystallin from the aged human lens, and we found site-specific racemization of Asp residues in aA-crystallin [3]. Site-specific racemization was also found in the aB-crystallin of the aged human lens. Two specific positions at Asp-36 (D/L ratio of Asp: 0.92) and Asp-62 (D/L ratio of Asp: 0.57) were extremely racemized and the racemization of Asp at both sites was accompanied by isomerization (Tables 2 and 3). This is the first study to demonstrate the formation of fl-Asp and D-Asp residues by spontaneous isomerization and racemization of a-aspartyl residues in aB-crystallin during the aging.
162
N. Fujii et al. / Biochimica et Biophysica Acta 1204 (1994) 157-163
RP-HPLC clearly separated the peptides containing /3- Asp and a-Asp, and the peptide containing /3-Asp was eluted faster than that containing a-Asp in both the T4 and T3a peptides (Fig. 1). The tendency is consistent with the separation of synthetic peptides containing a-Asp and /3-Asp residues [17]. On the other hand, the resolution of diastereomeric peptides (O- and L-Asp-containing-peptides, and the S- and Rforms of L-methionine sulfoxide-containing-peptides) could not be achieved under these conditions. The racemization of the Asp residue occurs to a greater extent in/3-Asp than in a-Asp in both T3a and T4 (Tables 2 and 3). It has been known that/3-Asp and D-Asp residues are derived from L-Asn/Asp residues under physiological conditions through the formation of a succinimide intermediate [17-21] in small peptides. The intermediate may be hydrolyzed at either side of its two carbonyl groups, yielding both /3- and a-Asp residues in small peptides. The succinimide can racemize far more rapidly than open chain structures. It is well known that o-succinimide is prefer opened to /3-D-Asp, tO a-D-Asp [19]. The rate of succinimide ring formation was found to be influenced by the neighboring amino-acid residue. The cyclization reaction to form succinimide takes place most easily in the Asn-Gly sequence in small peptides because there is less steric hindrance of the Gly residue [19-21]. In the case of aB-crystaUin, however, the isomerized and racemized Asp residues originated from L-Asp adjoined to bulky Leu and Thr, respectively, which may cause difficulty in forming a succinimide intermediate. Yet, despite the fact aB-crystallin contains both Asp140-Gly and Asn146-Gly sequences, none of the Asp and Asn residues in the T16 peptide were racemized or isomerized (Table 4). Additionally, it is known that /3-Asp residues are derived from the L-Asp residue in glucagon (Asp-Tyr) [22] and calmodulin (Asp-Gin, Asp-Thr) [23]. These observations indicate that the succinimide formation in the protein must depend on the higher order structure rather than on the primary structure. The molar ratio of /3-Asp to a-Asp residues in synthetic peptides has been reported to be 3 to 1 [19]. However, this was not the case for either T3a or T4, which may indicate that, under physiological conditions, the racemization and isomerization reactions do not reach equilibrium among the/3-Asp, a-Asp, and succinimidyl forms in the aged lens. We predicted the secondary structure of aB-crystallin using Robson algorithms (data not shown). The result suggests that Asp-36 and Asp-62 may exist in a /3-turn structure and that the following sequences are in a random coil structure. These sites may be located in the flexible surface region of the aB-crystallin protein. In particular, the fact that methionine was oxidized in the T4 peptide indicates that the surrounding region of Asp-62 may be exposed outside of the pro-
tein. Shapira et al. recently suggested that the racemized aspartic-acid residue of myelin basic protein in brain may exist in a/3-turn region [24]. The racemization and isomerization of Asp residues in aB-crystallin is likely to take place through a succinimide intermediate, which produces the D-Asp and /3-Asp residues observed in this study. Regardless of the mechanism of racemization and isomerization, this posttranslational modification of the amino-acid residues of aB-crystallin is likely to occur spontaneously in the aging process. This is likely because similar alternations were not observed in young aB-crystallin and such modifications would be expected to change properties such as solubility and conformation of the protein. Since aggregated, insoluble a-crystallin protein is reported to be more racemized than soluble a-crystallin [25], it is possible that the insolubilization of crystallins is closely related to the configuration of a-crystallin in the aging lens. If acrystallin have a protective role for some lens proteins from aggregation in transparent eye lens [9], the disruption of their structure by the modification of aBcrystaUin itself, such as the racemization and isomerization of Asp residues, and the oxidation of Met residue, could have wider effects on lens supramolecular organization.
References [1] De Jong, W.W., Terwindt, E.C. and Bloemendal, H. (1975) FEBS Lett. 58, 310-313. [2] Dubin, R.A., Ally, A.h., Chung, S. and Piatigorsky J. (1990) Genomics, 7, 594-601. [3] Fujii, N., Muraoka, S., Satoh, K., Hori, H. and Harada, K. (1991) Biomed. Res. 12, 315-321. [4] Groenen, P.J.T.A., Van den IJssel, P.R.L.A., Voorter, C.E.M., Bloemendal, H. and De Jong, W.W. (1990) FEBS Lett. 269, 109-112, [5] Bhat, S.P. and Nagineni, C.N. (1989) Biochem. Biophys. Res. Commun. 158, 319-325. [6] Dubin, R.A., Wawrousek, E.F. and Piatigorsky, J. (1989) Mol. Cell. Biol. 9, 1083-1091. [7] Kato, K., Shinohara, H., Kurobe, N., tnaguma, Y., Shimizu, K. and Oshima, K. (1991) Biochim. Biophys. Acta. 1074, 201-208. [8] Klemenz, R., Frohli, E., Steiger, R.H., Schafer, R. and Aoyama, A. (1991) Proc. Natl. Acad. Sci. USA 88, 3652-3656. [9] Horwitz, J. (1992) Proc. Natl. Acad. Sci. USA 89, 10449-10453. [10] Allen, G. (1981) in Sequencing of Proteins and Peptides (Work T.S. and Burdon R.H., eds.), pp. 174-221, North-Holland, Amsterdam. [11] Bruckner, H. and Wittner, R. (1989) J. Chromatogr. 476, 73-82. [12] Hashimoto, A., Nishikawa, T., Oka, T., Takahashi, K. and Hayashi, T. (1992) J. Chromatogr. 582, 41-48. [13] Ishibashi, Y., Kikuchi, T., Wakimasu, M., Mizuta, E. and Fujino, M. (1991) Biol. Mass. Spectrom. 20, 703-708. [14] Kramps, J.A., De Man, B.M. and De Jong, W.W. (1977) FEBS Lett. 74, 82-84. [15] Truscott, R.J.W. and Augusteyn, R.C. (1977) Biochim. Biophys. Acta. 492, 43-52.
N. Fujii et al. / Biochimica et Biophysica Acta 1204 (1994) 157-163
[16] Groenen, P.J.T.A., van Dongen, M.J.P., Voorter, C.E.M., Bloemendal, H. and De Jong, W.W. (1993) FEBS Lett. 322, 69-72. [17] Murray, E.D., Jr. and Clarke, S. (1984) J. Biol. Chem. 259, 10722-10732. [18] McFadden, P.N. and Clarke, S. (1982) Proc. Natl. Acad. Sci. USA 79, 2460-2464. [19] Geiger, T. and Clarke, S. (1987) J. Biol. Chem. 262, 785-794. [20] Stephanson, R.C. and Clarke, S. (1989) J. Biol. Chem. 264, 6164-6170.
163
[21] Cross, R.T. and Schirch, V. (1991) J. Biol. Chem. 266, 2254922556. [22] Ota, I.M., Ding, L. and Clarke, S. (1987) J. Biol. Chem. 262, 8522-8531. [23] Ota, I.M. and Clarke, S. (1989) J. Biol. Chem. 264, 54-60. [24] Shapira, R., Wilkinson, K.D. and Shapira, G. (1988) J. Neurochem. 50, 649-654. [25] Muraoka, S., Fujii, N., Ueda, Y., Mitsui, Y., Satoh, K. and Harada, K. (1991) Biomed. Res. 12, 61-64.