immunological and conformational characterization of a phosphorylated immunodominant epitope on the...
TRANSCRIPT
Vol. 187, No. 2, 1992
September 16, 1992
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Pages 783-790
IMMUNOLOGICAL AND CONFORMATIONAL CHARACTERIZATION OF A
PHOSPHORYLATED IMMUNODOMINANT EPITOPE ON THE PAIRED HELICAL
FILAMENTS FOUND IN ALZHEIMER’S DISEASE
Emma Lang*, Gyorgyi I. Szendrei*, Virginia M.-Y. Lee+ and
Laszlo Otvos, Jr.**
*The Wistar Institute of Anatomy and Biology, 3601 Spruce Street, Philadelphia, PA 19104
+Department of Pathology and Laboratory Medicine, University of Pennsylvania,
Philadelphia, PA 19104
Received July 27, 1992
The immunological recognition K
attern of one of the most commonly used monoclonal antibodies, PHF-1, whit detects the paired helical filaments of Alzheimer’s disease, exhibits a high degree of similarity with the recognition of a pol clonal antibody, anti-T3P, raised GALIVYKS(Phospho)PVVSGD,
a ainst a synthetic phosphopeptide, correspon ing to amino acids 389-402 of the f
microtubule-associated protein z. A panel of 16 synthetic non-phosphorylated and phosphorylated peptides, excised from different regions of z and thereof, were used to show that PHF-1 is indeed directed against t E
eptide analogs e T3 fragment.
Circular dichroism spectroscopy shows that the phos K limited propensity to form intramolecular B-pleated s
horylated peptide exhibits a eets, and alteration is found
in the reverse-turn structure that dominates the middle section of the molecule. The shift in the turn-forming amino acids may also allow a stacking procedure, may interfere with microtubule assembly, and, consequently, may be accountable for deposit formation. 0 1992 Academic Press. Inc.
The brains of AD patients are characterized by abundant fibrous lesions, i.e.
senile plaques, neurofibrillary tangles, and neuropil threads (l-3). Although not
restricted to AD, the burden of neurofibrillary tangles and senile plaques correlates
well with the severity of the dementia (4,5). Tangles represent dense accumulations
of ultrastructurally distinct PHFs (6,7). Recently it was shown that PHFs are
comprised of abnormally phosphorylated forms of 2, a group of low molecular
weight microtubule-associated proteins (8). Polyclonal antisera, raised against non-
phosphorylated (T3) and Ser 3g6 phosphorylated (T3P) forms of a synthetic
‘To whom correspondence should be addressed.
The abbreviations used are: AD, Alzheimer’s disease; PHFs, paired helical filaments; mAbs, monoclonal antibodies; CaMk, calmodulin-dependent kinase; CD, circular dichroism; TFE, trifluoroethanol.
0006-291 X/92 $4.00 Copyright 0 1992 by Academic Press, Inc.
783 All rights of I-eproducrion in ritzy form reserved.
Vol. 187, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
tetradekapeptide corresponding to amino acids 389-402 of z (H-Gly-Ala-Glu-Ile-Val-
Tyr-Lys-Ser*Pro-Val-Val-Ser-Gly-Asp-NHT) distinguished between z and PHF in
their cross-reaction pattern. The antiserum to T3P recognized PHF but not normal
human 2, and the antibody to T3 recognized z but not PHF. Two mAbs have been
developed so far that are specific for phosphorylated forms of X, found in PHF.
While the recognition site of mAb AT8 is identified between amino acids 195 and
209, featuring at least two abnormally phosphorylated serine residues (9), the epitope
for PHF-1, the most commonly used mAb (lo), remains an enigma. In this paper we
show that mAb PHF-1 is directed to a region almost identical to T3P by using a panel
of synthetic peptides and phosphopeptides that span different regions of the protein.
We have also shown that phosphorylation of 2-camk, a synthetic peptide
corresponding to amino acids 408-421 of 2 (LSNVSSTGSIDMVD), on one of its
serine residues (Ser4’6) that is located in the proximity of T3 and T3P, changed the
dominant B-turn structure to P-pleated sheets, the characteristic conformation of
PHFs (11). This serine in vivo is phosphorylated by the Ca2+ CaMk, and a similar
p-turn -+ B-pleated sheet-type secondary structural transition was found after
phosphorylation of the whole protein by CaMk (12). Since secondary structural
prediction (13) of peptide 389-402 resembles that of peptide 408-421, it is conceivable
that the two peptides undergo a similar conformational transition, leading to
structures that allow the formation of B-pleated sheets and irregular assembly of
PHF. We also report here this conformational transition which we traced by using
CD spectroscopy.
MATERIALS AND METHODS
Peptide Synthesis - Pe amino-acid Pfp esters as coup ing reagents 14). Perphosphorylation of peptides was P
tides were s Y
nthesized on solid-phase using Fmoc-
carried out with polyphosphoric acid made in situ from P,O, and H3P04 (15). Serine residues to be selectively phosphorylated were incorporated into the peptide with their side chain h phosphorylated Y
droxyl grou a ter the pe ti i
s unprotected. Parts of the peptide-resins were e
earlier (16). Peptides (and chain assemblies were completed as described
cp R
with trifluoroacetic acid an osphopeptides) were detached from the solid support
they were purified by reversed-phase high performance liquid chromatography using a 0.1% aqueous trifluoroacetic acid-acetonitrile gradient system.
Enzyme-Linked Immunoadsorbent Assay - Binding of 0.04-5.00 ug amounts of the synthetic dilutions of P Ip
eptides (and phosphopeptides) (Table 1) was tested with 1:lOO F-l mAb, anti-T3
antibod 4;
on Linbro plates (Flow P olyclonal antibody, and anti-T3P polyclonal aboratories).
mouse RI’ conju A 1:lOOO dilution of goat anti-
ate was used as a secondary antibody. Color development was made with 0- pheny -diamine P and was measured at 450 nm.
Circular Dichroism Measurements - CD spectra were taken on a Jasco J-720 instrument at room temperature in a 0.2-mm pathlength cell. Double-distilled water and spectroscopy grade TFE were used as solvents. The was 0.24-0.32 mg/ml, determined each time by quantitative hi K
eptide concentration
f performance li
chromatography. Curve smoothing is accomplished by the a gorithm provide if uid by
Jasco. Mean residue ellipcity ([OIMR) is expressed in degrees/dmole by using a mean residue weight of 110.
784
Vol. 187, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Secondary Structural Prediction and Estimation - Secondary structural
E rediction was made using a modified Chou-Fasman algorithm (13). The INCOMB algorithm was used for secondary structure estimation. This program
determines the conformational wei hts of different secondary structures from a CD s ectrum B
(17). 7 Based on resu ts obtained from spectroscopical structure eterminations other than CD, we modified the basic curves for the clean secondary
structures as follows: a-helix: myoglobin in TFE (18); J3- loid peptide in 2% B-octyl-glucosidez; type I (III) B-turn: c
leated sheet: B-am
Ava] in TFE (19); t 23
e II B-turn: Boc-Pro-D-Ser-NH z clo Gly-Pro-Ser(O’Bu)- P cy ly-6-
polylysine at pH = . in water (21). H3 in TFE (20); and random:
RESULTS
The PHF-1 mAb recognized only the phosphorylated forms of T3 peptide
from the tested r fragments (Fig. 1). This recognition was highly specific, since
strongly related sequences as T3, T3Ala, or even T3PAla peptides were not
recognized, and suggests that the epitope comprises the phosphorylated KSPVVS
sequence (Table 1). The recognition pattern of PHF-1 is very similar to that of
polyclonal antibody anti-T3P (8), but not quite the same. Anti-T3P detects not only
peptide T3P, but also weakly T3PAla, while anti-T3 recognizes only T3, not T3Ala.
This indicates that the recognition site of anti-T3 is shifted slightly toward the C-
terminus of the peptide compared to that of anti-T3P (in addition to the difference
observed based on the state of phosphorylation). A similar shift in the binding of a
synthetic peptide (corresponding to the phosphoprotein of rabies virus) and its
monophosphorylated analog to the major histocompatibility complex was found
(22). It appears that PHF-1 and anti-T3 bind to the identical primary amino acid
sequence and are complementary, with PHF-1 recognizing the phosphorylated, and
anti-T3 recognizing the non-phosphorylated form of the peptide.
The CD spectra of T3 and T3P peptides were measured in mixtures of water
and TFE. (Peptides phosphorylated on both or on only the second serines showed
similar CD spectra to T3P in all conditions examined.) In water both peptides
showed a negative ellipticity band between 195 and 200 nm, with a negative
shoulder around 210 nm that signaled a mostly unordered conformer population.
The decreased amplitude of the negative band between 195 and 200 nm in the CD
spectrum of T3P correlated well with preliminary nuclear magnetic resonance
studies, indicating an increased and altered turn structure. The secondary structure
estimations from the CD spectra of the non-phosphorylated and the phosphorylated
peptides in water also showed an increased turn propensity. Thirty-seven and forty-
seven percent of reverse-turn conformational weights were found for the non-
phosphorylated and the phosphorylated peptide, respectively.
20tvos, L., Jr., Szendrei, G.I., Lee, V.M.-Y., and Mantsch, H.H. (1992) Eur. J. Biochem., submitted.
785
Vol. 187, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Amount of peptide (pg)
-+----- E:. Fi ure 1. Enz me-linked immunoadsorbant assay of pe tide T3 (curve a) and
peptide 3P (curve ) usmg PHF-1 as a primary antibody. ;P 3 binds to the Linbro plate, since the same amounts were recognized by anti-T3.
The CD spectra of both peptides measured in different TFE-water mixtures
showed a continuous and similar spectral transition. Although the secondary
structural prediction of T3 suggested a high propensity of P-pleated sheet formation
at the N-terminal and the middle section, the peptide exhibits a CD spectrum in TFE
which can be described most accurately as a mixture of type C and type D spectra (23),
according to the classification of Woody (24) (Fig. 2). Types C and D spectra are
Table 1. Recognition of Synthetic Peptides and Phosphopeptides by Monoclonal Antibody PHF-1, and Polyclonal Antibodies Anti-T3 and Anti-T3P
Recognition of antibodies Peptide Sequence Amino Acids PHF-1” Anti-T3 Anti-T3P Human TDG16 TDG16P Tl T7-13 T7-13P T4
T3 389-402 T3P 389-402 T3PP 389-402
T3 0’) 389-402
T3Ala 389-402 T3AlaP 389-402
TPRHcamk 405-421 TPRHcamkP 405-421
40-55
4-55
189-207 192-204 192-204 291-322
DAGLKESPLQTPTEDG DAGLKES (Ph)’ PLQTPTEDG I’KSGDRSGYSSPGSPGTPG
GDRSGYSSPGSPG GDRSGYSS (Ph) PGSPG
CGSKDNIKHVPGGGSV- QIVYKPVDLSKVTSKC
GAEIVYKSPVVSGD GAEIVYKS (Ph) PVVSGD
GAEIVYKS (Ph) PVVS (Ph) GD GAEIVYKSPVVS (Ph) GD
GAEIVYKSPVVAGD GAEIVYKS (Ph) PVVAGD PRHLSNVSSTGSIDMVD
PRHLSNVSSTGS (Ph) IDMVD
N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A N/A
+ + + N/A + N/A
+ N/A N/A
N/A N;A N/A N/A
Bovine T2Ser 95-108 AGIGDTSNLEDQAA - N/A N/A T2Pro 95-108 AGIGDTPNLEDQAA - N/A N/A
IPh, Phospho-.
786
Vol. 187, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
is TSP.+ ” Fi ure 2. CD spectra of T3 peptides in TFE. Curve a (-1 is T3; curve b (-----1
bile addition of a IO-molar excess of Ca(ClO& to T3 (curve c, - - - - -) just slightly affected the secondar structure, the conformation of T3P became unordered after addition of Caz+ (curve d: -.-.-.-.-I.
characteristic of the type I (III) B-turns and distorted turn structures, respectively (25).
Based on this, T3 probably exists as a series of different B-turns, and this correlated
well with the secondary structure estimation from its CD spectrum in pure TFE.
Addition of the conformational weights of type I (III) and type II B-turns (based on
curve estimations calculated from 4 and 3 pure secondary structural components)
showed that the total B-turn content was 84 and 98%, respectively.
The CD spectrum of T3P in TFE exhibits bands blue-shifted and with
increased intensity compared to those of the non-phosphorylated analog. The most
remarkable new band in the CD spectrum of T3P was the negative shoulder at 220
nm, which may indicate the appearance of a conformer featuring B-pleated sheet
structure (23). Recently we found a B-turn + B-sheet transition upon
phosphorylation of another z fragment, peptide r-camk (11). z-camk was also
predicted to assume an extended structure, but this conformation became apparent
only after phosphorylation.
The increased B-pleated sheet propensity was not reflected, however, in the
secondary structure estimation from the CD spectrum of T3P in TFE, which rather
indicated an alteration in the turn system. A type II + type I (III) B-turn transition
with altogether increased turn contribution was found for peptide T3P compared to
T3, based on the CD curve analysis in different TFE-water mixtures.
Vol. 187, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
The major difference between the conformation of the phosphorylated and
the non-phosphorylated peptides was indicated by the CD curves after addition of a
10 molar excess of Ca2+ ions (Fig. 2). While a characteristic loosened type I p-turn
conformation was found in peptide T3 (the contribution of type II turns decreased),
the entire p-turn system was destroyed in peptide T3P, resulting in the formation of
almost completely unordered structure. These findings are also supported by
nuclear magnetic resonance, when intramolecular salt bridge interaction was
observed. This clearly shows that the phosphate group plays an integral role in the
definition of the conformation, and that more than just the reverse-turn system is
affected.
DISCUSSION
Using T3, a synthetic peptide corresponding to amino acids 389-402 of the z
protein, we showed that the mAb PHF-1 binds to the same abnormal
phosphorylation site at Ser 396 that has been previously identified using antisera
raised to the phosphorylated and non-phosphorylated forms of this synthetic
peptide. Recently, we demonstrated that a consequence of abnormal
phosphorylation at Ser396 is an alteration in its electrophoretic mobility and a
reduction in its binding capacity to microtubules3. Other evidence from our
laboratory also suggests that the specific region of r containing phosphorylated Ser396
resists proteolysis when soluble z derived from PHFs is injected into rat braind.
These data suggests that phosphorylation at Ser 3% dramatically changes the local
conformation of the protein leading to alteration in its biochemical properties.
Indeed, the change in the conformation of T3 by phosphorylation may lead to the
direct formation of P-pleated sheets, resulting in the inability of z to bind to
microtubules and promoting the self aggregation of phosphorylated forms of z into
PHFs. Peptides T3 and the r-camk are present in all z isoforms and are located near
the microtubule binding repeats (residues 244 to 368) (26,27). The P-pleated sheet-
forming tendency, however, of the peptides in a phosphorylated form is different.
The phosphopeptide around Ser 416 forms intermolecular P-pleated sheets, as is
demonstrated by a dilution experiment, when at a very low concentration the
extended structure disappears (11). In contrast, the shoulder in the CD spectrum of
T3P (which we recognize as a low frequency contributor to P-pleated sheets) still
exists at a peptide concentration of 0.1 mg/ml. Addition of Ca2+ ions destroyed the
extended structure of both synthetic phosphopeptides, and led to the restoration of
SBramblett, G.T., Goedert, M., Jakes, R., Merrick, S.E., Trojanowski, J.Q., and Lee, V.M.-Y. (1992) Science, submitted.
‘%hin, R.-W., Bramblett, G.T., Lee, V.M.-Y., and Trojanowski, J.Q. (1992) Science, submitted.
788
Vol. 187, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
B-turns for the peptide corresponding to the CaMk site (ll), and to an unordered
conformation for T3P. In contrast, addition of Ca *+ did not considerably affect the
reverse-turn system of non-phosphorylated peptide T3.
Both PHF-specific mAbs are directed against regions of the r protein featuring
more than single abnormal phosphorylation sites. Both Ser199 and Ser*O* were
found to be abnormally phosphorylated in the protein fragment recognized by mAb
AT8 (9). Most recently, Ser404 as a new PHF-specific phosphorylation site was
reported (28) only 8 amino acids downstream of the KSPV site and 12 amino acids
upstream of the CaMk site. Concomitant abnormal phosphorylation of 3 serine
residues between amino acids 396 and 416 conceivably alters the conformation and
function of the region drastically. Further investigations are in progress to undo the
formation of B-pleated sheets in this area by using phosphorylated small peptides
(11) or other ionophoric reagents. Understanding this process fully may lead to the
development of measures to prevent or remove the formation of PHFs.
ACKNOWLEDGMENTS
The authors wish to thank Drs. Hildegund C.J. Ertl and John Q. Trojanowski, and Mrs. Shirley Peterson (editor) for their critical reading of the manuscri t. This work was supported by National Institutes of Health Grants AGO9215 and A e 10670.
1.
2.
3.
4. 5. 6. 7. 8.
9.
10.
11.
12.
13. 14.
REFERENCES
Henderson, V.W., and Finch, C.E. (1989) J. Neurosurg. 70,335-353. Trojanowski, J.Q., Schmidt, M.L., Otvos, L., Jr., Gur, R.C., Gur, R.E., Hurtig, H., and Lee, V.M.-Y. (1989) Ann. Med. 21,113-116. Braak, H., Braak, E., Grundke-Iqbal, I., and Iqbal, K. (1986) Neurosci. Lett. 65, 351-355. Blessed, G., Tomlinson, B.E., and Roth, M. (1968) Br. J. Psychiatr. 114, 797-811. Wilcock, G.K., and Esiri, M.M. (1982) J. Neurol. Sci. 56, 343-356. Kidd, M. (1963) Nature 197,192-193. Terry, R.D. (1963) J. Neuropathol. Exp. Neurol. 22, 629-642. Lee, V.M.-Y., Balin, B.J., Otvos, L., Jr., and Trojanowski, J.Q. (1991) Science 251, 675-678. Biernat, J., Mandelkow, E.-M., Schroter, C., Lichtenberg-Kraag, B., Steiner, B., Berling, B., Meyer, H., Mercken, M., Vandermeeren, A., Goedert, M., and Mandelkow, E. (1992) EMBO J. 11,1593-1597. Greenberg, S.G., Davies, I’., Schien, J.D., and Binder, L.I. (1992) J. Biol. Chem. 267,564-569. Lang, E., Szendrei, G.I., Elekes, I., Lee, V.M.-Y., and Otvos, L., Jr., (1992) Biochem. Biophys. Res. Commun. 182, 63-69. Steiner, B., Mandelkow, E.-M., Biernat, J., Gustke, N., Meyer, H.E., Schmidt, B., Mieskes, G., Soling, H.D., Drechsel, D., Kirschner, M.W., Goedert, M., and Mandelkow, E. (1990) EMBO J. 9,3539-3544. Argos, I’., Hanei, M., and Garavito, R.M. (1978) FEBS Lett. 93, 19-24. Atherton, E., and Sheppard, R.C. (1985) J. Chem. Sot. Chem. Commun. 165- 166.
789
Vol. 187, No. 2, 1992 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
15.
16.
17.
18. 19.
20. 21. 22. 23.
24.
25.
26. 27. 28.
Otvos, L., Jr., Tangoren, LA., Wroblewski, K., Hollosi, M., and Lee, V.M.-Y. (1990) J. Chromatogr. 512,265-272. Otvos, L., Jr., Elekes, I., and Lee, V.M.-Y. (1989) Int. J. Pept. Protein Res. 34, 129- 133. Perczel, A., Tusnady, G., Hollosi, M., and Fasman, G.D. (1991) Protein Eng. 4, 669-679. Jackson, M., and Mantsch, H., (1991) Biochim. Biophys. Acta 1078,291-235. Hollosi, M., Kover, K.E., Holly, S., Radics, L., and Fasman, G.D., Biopolymers (1987) 26, 1555-1572. Hollosi, M., Perczel, A., and Fasman, G.D., Biopolymers (1990) 29, 1549-1564. Greenfield, N., and Fasman, G.D. (1969) Biochemistry 8,4108-4116. Larson, J.K., Otvos, L., Jr., and Ertl, H.C.J. (1992) J. Virol. 66,3996-4002. Lee, V.M.-Y., and Otvos, L., Jr. (1992) In Peptides: Chemistry and Biology (Proceedings of the Twelfth American Peptide Symposium) (J.A. Smith, and J.E. Rivier, Eds.), pp. 109-112, ESCOM, Leiden. Woody, R.W. (1974) In Peptides, Polypeptides and Proteins (Blout, E.R., Bovey, F.A., Lotan, N., and Goodman, M., Eds.), pp. 338-360, Wiley, New York. Gierasch, L.M., Deber, C.M., Madison, V., Niu, C.-H., and Blout, E.R. (1981) Biochemistry 20, 4730-4738. Goedert, M., Spillantini, M.G., and Potier, M.C. (1989) EMBO J. 8,393-399. Cleveland, D.W. (1990) Cell 60, 701-702. Kanemaru, K., Takio, K., Miura, R., Titani, K., and Ihara, Y. (1992). J. Neurochem. 58, 1667-1675.
790