- Email: [email protected]
Antibody structure
TIBS - August 1976
174 synaptic vesicles. A comparison with the isolated synaptic terminals (synaptosomes, Table I) shows, however, that all these components are far more concentrated (4-22 times) in the terminal region than in the perikaryon. These observations suggest that the function of the perikaryon is to synthesize all the components of the cholinergic system eventually needed at the presynaptic nerve terminal, including even acetylcholinesterase. Preliminary studies with ligatured [23,24] or excised lengths [25] of axons leave no doubt that acetylcholine [23], vesicle membrane protein [24] and choline acetyltransferase [25] are transported, at varying rates, and possibly by different mechanisms down the axon from cell body to terminal. Thus the axon besides conducting information from the cell body to the terminal by a propagated electrochemical disturbance, the action potential, also serves as a conveyor belt transporting components needed for transmitter synthesis and storage from their place of synthesis in the cell body to their site of action in the terminal.
I5 Dowdall, M. J., Boyne, A. F. and Whittaker, V. P.
References Otsuka, M., Konishi, S. and Takahashi, T. (1972) Proc. Jup. Acad. 48, 747-751
von Euler, U.S. and Gaddum, J. H. (1931) J. Physiol. 72, 7487
Cleugh, J., Gaddum, J. H., Mitchell, A. A., Smith, A. W. and Whittaker, V. P. (1964) J. Physiol. 170, 69-85
4 Duffy, M. J., Mulhall, D. and Powell, D. (1975) J. Neurochem.
25. 305-307
5 Hughes, J., Smith, T. W., Kosterlitz, H. W., Fothergill, L.A., Morgan, B.A. and Morris, H. R. C1975) Nature 258. 577-579 Dowdall, M. J. and Zimmermann, H. (1976) Exp. Brain Res. 24, 8-9
Dowdall, M. J. and Simon, E.J. (1973) J. Neurothem. 21,969-982 Yamamura, H.I. and Snyder, S.H. (1973) J. Neurochem.
10
11 12 13
21, 1355-1374
Dowdall, M.J., Wlchtler, K. and Henderson, F.M. (1975) Abstr. 5th Inr. Meet. Sot. Neurothem., Barcelona, p. 122 Whittaker, V.P. and Dowdall, M.J. (1975) in Cholinergic Mechanisms (Waser, P. G., ed.), pp. 2342, Raven Press, New York Whittaker, V.P., Essman, W.B. and Dowe, G.H.C. (1972) Biochem. J. 128, 833-846 Nagy, A., Baker, R. R., Morris, S. J. and Whittaker V. P. (1976) Bruin Res. (in the press) Whittaker, V. P., Dowdall, M. J., Dowe, G. H.C., Facino, R.M. and Scotto, J. (1974) Bruin Res.
(1974) Biochem. J. 140, 1-12 16 Schmidt, R. (1976) Diplomarbeit, Mainz 17 Schmidt, R., Zimmermann, H. and Whittaker, V. P. (1976) Exp. Bruin Res. 24, 19-20 18 Whittaker, V. P. (1973) in The Scientifi:c Basis of Medicine Annual Reviews (Gilliland, I. and Peden, M., eds), pp. 17-31, Athlone Press, London 19 Barker, L.A., Dowdall, M. J. and Whittaker, V.P. (1972) Biochem. J. 130, 1063-1080 20 Zimmermann, H. and Dowdall, M. J. (1976) Exp. Brain Res. 24,23
21 Dowdall, M.J., Fox, G., Wachtler, K., Whittaker, V.P. and Zimmermann, H. (1976) Cold Spring Harbor Symp. Quant. Biol. 40, 65-81 22 Whittaker, V. P., Fox, G. and Fiore, L. (1976) Exp. Brain Res. 24,22
23 Zimmermann, Abstr.
4th
Int.
H. and Whittaker, Meet.
Ini.
Sot.
V.P. (1973) Neurochem.,
Yokyo, p. 245 24 Ulmar, G. and Whittaker, V. P. (1974) Brain Res. 71,155-159 25 Davies, L. (1976) Unpublished observations 26 Heilbronn, E. (1975) in Cholinergic Mechanisms (Waser, P.G., ed.), pp. 343-364, Raven Press, New York 27 Raftery, M.A., Bode, J., Vandlen, R., Chao, Y., Deutsch, J., Duguid, J. R., Reed, K. and Moody, T. (1975) in Biochemistry of Sensory Functions (Jaenicke, L., ed.), pp.‘541-563, Springer Verlag, Berlin/Heidelberg/New York
75, 115-131
14 Stadtler, H. (1976) Unpublished observations
‘llre acetylcholine receptor Any survey of the electromotor system of the Torpedo would be incomplete without a mention of another application: the electric organ as a source material for the isolation of the acetylcholine receptor. A protein (AChR) with the properties expected of the recognition site of the acetylcholine receptor has been isolated by several groups from the cholinoceptive electrocyte cells of the electric tissue. (For further details the reader is referred to two recent reviews [26,27].) It is not quite clear yet whether this protein also accounts for the acetylcholine-stimulated ionophore properties seen in the physiological receptor. One of the most interesting developments in this work has been the discovery that the injection of puritied Torpedo AChR into rabbits induces a condition similar to that of the human disease myasthenia gravis characterized by muscle weakness. At first this was thought to be impaired neuromuscular transmission due to a simple masking of the rabbit muscle AChR by the rabbit’s own antibody to AChR, but further work has shown that the condition is more complex and more like the human disease in that the immune response of the lymphatic system is involved. Presumably a primary step in the aetiology of human myasthenia gravis is the release, from damaged postsynaptic membranes, of AChR, but what triggers this release, possibly a virus infection, is as yet obscure.
Antibody structure Robert Huber The crystal structures of a human IgG antibody molecule Kol and a human Fc fragment have been determined at 0.4 and 0.35 nm resolution, respectively, by isomorphous replacement. These molecular structures, in comparison to the known structures of antigen-binding fragments, bring us closer to a coherent view of antibody structure. From these structures it is hypothesized that antibody molecules are inherently flexible but become rigid upon interaction with antigen, which triggers the formation of all longitudinal contacts. Antibody molecules serve a dual function: to recognize foreign cells and macromolecules and to trigger the events leading to their elimination. Specific recognition requires surface structures complementary to the antigen and therefore needs a huge variety of antibody molecules. In contrast, the effector functions would most economically be performed if all antibody molecules were identical. An antibody molecule consists of two identical light- and two identical heavy chains [5,6]. Each of these polypeptide chains can be subdivided into a constant region (c) with an amino acid sequence identical to other chains of the same..class and a variable region whose sequence varies between antibodies specific for difR.C. is at the Max-Planck-Institut fur Biochemie, 8033 Marrinsried, und Physikalisch-Chemisches Institut der Technischen Universitiit Miinchen, G.F.R.
ferent antigens (see Fig. 1 for details of antibody structure and its fragments and list of abbreviations used in this article). Antibody function is located on different parts of this molecule with recognition lying in a fragment consisting of a light chain and half of the heavy chain - the Fab fragment. Effector function on the other hand resides in the C-terminal half of the heavy chain with the inter-heavy chain disulphide bond intact - the Fc part. There is, however, co-operativity, and effector functions are triggered by antigen binding. How this signal is transferred is open to debate and in this article I will discuss a hypothetical model based on our current knowledge of three-dimensional antibody structure. Several crystal structures of immunoglobulin fragments (Fab fragments and Bence-Jones proteins) have been analysed during the last few years [3,4,7-g], and pro-
TIBS - August 1976
175
vide a structural basis of antibody specilicity. These molecules are composed of globular. domains formed by the variable and constant segments. Within the domains the chain folds into two disulfidelinked pleated-sheet segments, forming a barrel and covering a hydrophobic interior. There are tertiary structural differences between V and C domains. V parts (VH and VL in Fab or VL and VL in Bence-
Jones proteins) and C parts (CH~ and CL in Fab or CL and CL in Bence-Jones proteins) aggregate in a characteristic fashion, using their V faces and C faces, respectively. The V dimers form a cavity lined with hypervariable segments into which haptens bind. From the recent crystalstructure analyses of a complete human IgG antibody Kol [lo] and a human Fc fragment [I 1,121 in comparison with the papain
HH
ckbohydrate
fltf coo-
antigen-binding fragments, a coherent view of antibody structure begins to emerge. The immunoglogulin Kol molecule
The C” carbon atom positions of the (Fab), part of the Kol molecule are shown in the stereogram, Fig. 2. The Fab arms subtend an angle of 125” with their tips about 14.6 nm apart. Apart from slight differences in domain tertiary structure, the VL-VH and CL-Cnl dimers appear to be indistinguishable between Kol and the mouse Fab fragment McPC603 [l]*. In particular, the geometry of aggregation within the V and C dimers is very similar. There is a difference, however, in the relative orientation of the V and C dimers which can be described by a rotation about an axis through the switch peptides that is, the segments connecting V and C parts at residues 118-l 19 (heavy chain) and 109110 (light chain). The McPC 603 Fab fragment [l] and the Kol Fab part as seen down the switch peptide rotation axis are compared in Fig. 3. VL and VH as well as CL and CH1 are related by local diads. They subtend an angle (Fab angle) of 120”
Fig. 1. Structure of antibody (IgG) molecule and enzyme cleavage products. VL-variable harf of light chain, CL-constant halfof light chain, Vn-variable part of heavy chain, Ctrl, CHZ, cH3 - the three constant homology regions of the heavy chain. Fab - antigen - binding fragment consisting of light chain and half of the heavy chain ( VL, Vn, CL, CJ) Fe - C - terminal half of the heavy chain with the inter-heavy chain disulftde bond intact. Hinge peptide: the segment connecting Cnl and Cn2 and containing the inter-heavy chain disulfde linkage; switch peptides: the segments connecting V and C parts comprising residues at 110 (light chain) and 119 (heavy chain).
Kol Fig. 2. Stereo-drawing of the Kol (Fab), part. C’ atom positions were derived by fitting the 1UcPC603~ V- and C-dimers to the Kol electron-density map. The coordinates of the C-terminal residues of heavy and light chain ana’the hinge peptide were obtained by interpretation of the 0.4nm Fourier map.
Fig. 3. Comparison of J4cPC603 Fob fragment and Kol Fab part. The Fab angle changes by 50°. There is a VI&HI contact in the fragment but not in the intact molecule. NH, CH, NL, CL: N-and C-terminus of heavy and light chain, respectively.
176
TIBS - August 1976
Tf‘heFc fragment structure
in the McPC 603 fragment and 170” in the Kol protein, demonstrating the major change in quarternary structure - ‘bending the elbow’. It is already evident from Fig. 3 that in the McPC603 molecule some residues of VH and CH1 come close to each other. These are the segments 8-10 in VH and 152 and 207-208 in CH~. Through the widening of the Fab angle from 120” to 170” in Kol there is no contact between VH and CH1. Also, VLand CL have no contact, nor do the two Fab arms. No interpretable electron-density is found in the Kol Fourier map which could be assigned to the Fc part [lo] leading to the conclusion that the Fc stem is disordered in the Kol crystals, C-terminal to the hinge peptide: There is a large cylindrial space in the crystal lattice around the three-fold screw axis, where the Fc parts must be located. Its volume is sufftcient to accommodate Fc, but its dimensions do not allow the Fc parts to obey the two-fold symmetry of the rest of the molecule. Fc is either statistically disordered or wagging thermally. A ‘minimum disorder’ Kol model has been constructed on the assumption that the deviation of Fc from the crystal symmetry as dictated by the crystal dimensions and the (Fab), packing should be minimal. If the disorder is due to temperature movement this model represents a snap-shot (Fig. 4). It is clear that there are no contacts between (Fab), and Fc in the minimum disorder model except the covalent linkage through the hinge peptide - the segment connecting Cnl and CH~ - which must have a rather extended conformation.
A stereo-plot of the C” carbon atom positions of the Fc fragment is shown in Fig. 5. The molecule looks like ‘Mickey Mouse’with the compact CH~ dimer forming the head and the CH~ domains the ears. The CH3 dimer pairing is closely similar to the CH~-Cr pairs found in Fab fragments. Both, Cn2 and CH~ show the common immunoglobulin fold. They are connected by a loosely-folded segment which is susceptible to proteolytic attack [13,14]. The longitudinal CHA-CHA contact is of the same size as the VH-CH~ contact in Fab fragments. It comprises the segments homologous to those mediating the VHCHI contact in Fab fragments described above. The carbohydrate moiety [15] covers a large part of the C-face of the CH2 domain, obviously shielding apolar amino acid residues. Fc was the first glycoprotein analysed and it was a surprise to find the carbohydrate moiety spatially fixed. The function of the carbohydrate in immunoglobulins is unclear but it might be involved in secretion, although recent experiments do not confirm this [16,17]. Its rigid conformation and close association with the protein points to a general stabilizing effect. Domaindomain interactions The lateral V-V (VWVL, VL-Vr) and CL-CL) interactions c-c (CL-W) observed in Fab fragments [8,9] and Bence-Jones proteins [7,4] have been described in great detail elsewhere and will not be considered here. The relative
“R
L
.
1
KOL FLEXIBLE
.
MODEL
Fig. 4. Stereo-drawings of the ‘minimum disorder’ model of the Kol molecule. This should be regarded as a snapshot of the flexible molecule. There are no contacts between (Fab), and Fe. The structure of the Fe part was derived from the Fc fragment, although we believe that Fc is flexible between Cn2 and Cn3 in the Kol molecule. The hinge peptide is drawn as observed in Kol (Fab), and the Fc fragment. In order to make the covalent linkage, the hinge, which is bent back in between the Cn2 domains, has to be pulled out. Ch. : carbohydrate.
arrangement of the two CH3 domains observed in the Fc fragment is closely similar to the C&HI aggregation found in McPC 603 [ 111.The CH2 domains have no contact with each other except the covalent linkage through the hinge. It is an intriguing question why CH~ forms neither the familiar C-C nor V-V contact. The tertiary structure of the CH~ domain does not resemble a C- but also lacks some features of a V-domain. This might prevent dimerization. In the presence of the spatially fixed carbohydrate moiety, formation of the C-C interface is impossible. Furthermore, some of the hydrophobic residues in CH~ which lie close to the local diad and are probably important for the contact, are replaced by polar groups in CH2. Formation of a V-V contact of CH2 is also prohibited as several of the contact residues [4], conserved in V segments, are exchanged for polar residues. In the Kol protein and the ‘minimum disorder’ model derived from it, no longitudinal contacts are observed between VH and CH~, VL and CL, CH~ and CH~. There are also no contacts between the Fab arms. In view of this lack of stabilization along the chain, it is difficult to assume that the Kol molecule is rigid in solution; rather, that flexibility exists between V and C dimers (‘elbow’ V-C-flexibility), between the Fab arms ((Fab),-flexibility) and between Fc and (Fab),(Fc-(Fab),-flexibility). As mentioned before, flexibility between Fc and (Fab), could explain the weak, uninterpretable electron-density for Fc in the Kol crystals. The relative positions of V- and Cdimers and Fab arms are frozen in the Kol crystal lattice dictated by crystal packing requirements. Even the CH2 segments might sit flexibly on the CH3 dimers (CH2CH3-flexibility) in the Fc part in the Kol protein in contrast to their fixed position in the fragment. The model that I would like to suggest for the Kol protein and for other unliganded (without antigen) IgG molecules is a highly flexible one. In obvious contrast the Fab fragments [8,9] and the closely similar Bence-Jones protein dimer [7] appear to be rigid molecules. In the three structures, V- and Cdimers subtend a similar Fab (‘elbow’) angle of about 120”. Crystal packing is different in the three crystal structures and would be unlikely to have the same effect on the Fab angle if it were variable. Indeed there exists a longitudinal contact between VHand Cul in these fragments as described before. In the Fc fragment CH2 and CH3 form a contact as described. The linking peptide is not in a completely extended conformation but is irregularly bent between residues Lys (338) and Gly (341)**. Loosening of the Cn2-CH3 con-
TIBS - August 1976
tact and a stretching out of the linking peptide appear possible. It is remarkable that homologous segments mediate the close V-C contacts in the Fab fragments and the close CH2-CH3 contact in the Fc fragment. Of further interest is the fact that the contact segments described for C”l and C”3 when projected into a V segment coincide with the first and third hypervariable segments forming the antigen-binding surface in V dimers. This suggests the presence of a general binding area in immunoglobulin domains. The residues forming in this area must of course be variable in V domains to meet the requirements of complementarity to various antigens, but conserved in Cdomains for inter-domain interaction. Signal transfer
In an allosteric, cooperative immunoglobulin model, the signal transfer through the domains from the general binding area (top contact) to the opposite end of the domains (bottom contact) would occur along homologous segments in all domains. I tried to construct a hypothetical liganded antibody molecule by combining the fragment molecules. It is well documented that Fab fragments show no gross structural change upon hapten binding [8,9] indicating that the fragment structure is characteristic for liganded molecules. Assuming that the Fc fragment structure is also characteristic for a liganded antibody molecule and starting with the Kol (Fab), I set the Fab angle to 120”, characteristic for Fab fragments, superimpose the hinge as seen in Kol and Fc and search for the allowed relative azimuthal angle. An important aspect of this model (Fig. 6) is that all longitudinal inter-domain contacts are closed, resulting in a rigid structure. (Fab), bending is prohibited by the C~l-c~2 interaction. It is clear that in view of the (Fab),-flexibility of the Kol molecule the choice of the relative arrangement of the Fab arms was arbitrary. By a small. (Fab), rearrangement the formation of a lateral contact of the Fab arms appears possible. Antigen-binding may cause a stiffening of the flexible antibody molecule by formation of the longitudinal inter-domain contacts described. In the usual terminology of allosteric the flexible cooperative interactions, model would represent the tense state (T) which changes to the relaxed state (R) upon antigen binding. Antigen-binding triggers the formation of the longitudinal interdomain contacts. The T state quarternary structure is maintained only in the intact molecule. Fragment structures adopt a conformation characteristic for
L
oc
?? c.h.
?
c
Fig. 5. Stereo-drawing of the Co: carbon positions and the centres of the carbohydrate hexose units of the Fc fragment. The carbohydrate structure was assumed to be as found in a myeloma protein j16J. In Fc the carbohydrate chain appears to be longer. ?? C.h. : approximate centres of carbohydrate hexose units.
the R-state of the intact molecule. Segmental flexibility between Fab and Fc parts of antibody molecules is experimentally established by spectroscopic and hydrodynamic observations [19,20]. Electron micrographs also show variable angles between the Fab arms of antibody molecules [ 18,211. Hydrogen-exchange experiments as well as experiments with limited proteolysis of IgM and their Fab fragments may be interpreted in terms of a less flexible fragment structure [22,23]. There are several observations of structural changes in antibody molecules upon antigen-binding. Circular polarization of fluorescence changes in complete antibodies upon antigen binding, but not in Fab fragments [24,25]. An antigen-induced volume contraction has been found by small angle X-ray scattering in complete antibodies, but not in Fab fragments [26,27]. Antibodies appear to lose flexibility upon antigen binding [28,29] and to gain conformational stability against denaturation [30]. Further-
more, there are observations on IgM molecules of differences in thermodynamic parameters for haptens and antigens, indicating a conformational change in Fc induced by antigen binding [3 11. Immunoglobulins have a number of effector functions of which the binding of complement has been studied on a molecular level. It is well documented that the binding site for Cl, the first component of the complement system is located on the CH~domains [32-351. CH~, Fc and IgG have intrinsic complement binding properties which are, however, greatly enhanced by specific antigen-binding or heat-aggregation [36,37]. It has been suggested that this might be due to the proximity of several complement-fixation sites and/or to structural changes [31,37] (for review see [38,39]). The rigid, liganded antibody molecule which I propose could enhance Cl binding for several reasons: (a) In the unliganded, flexible antibody molecule the C 1 binding site might be transiently occluded by the Fab arms and pos-
??
B
KOL
RIGID
.
MODEL
Fig. 6. Stereogram of a hypothetical liganded IgG molecule. The molecule is T-shaped with Fc forming stem. Fc and (Fab), are in close contact. The hinge peptide is folded back in between the Cn2 parts. C.h.: carbohydrate.
the
TIBS - August 1976
178 sibly also by CH~. The long extended hinge peptide in the unliganded Kol molecule allows the Fab arms far-reaching wagging motions to cover part of Cn2. (b) Protein-protein complex formation requires freezing out of rotational and translational degrees of freedom. A component with internal flexibility is therefore less favourable for complex formation than a rigid molecule. (c) The formation of the longitudinal contacts might induce a structural change in Cn2 thereby improving Cl binding. Little is known in molecular terms about other effector functions located on Cn3, as cytophilic activity [35], requiring interaction with Fc receptors on macrophages. A detailed comparison of the structures of Cn3 and CH~ might illuminate this problem.
14, 5312-5315 26 Pilz, I., Krathy, O., Licht, A. and Sela, M. (1973) Biochemistry 12, 4998-5005
27 Pilz, I., Karthy, O., Licht, A. and Sela, M. (1975) Biochemistry 14, 13261333 28 Tumerman, L.A., Nezlin, R. S. and Zagyanski, Y.A. (1972) FEBS Lett. 19,29&292 29 Warner, C. and Shumaker, V. (1970) Biochemistry 9, 451-459 30 Cathou, R.E and Warner, T.C. (1970) Biochemistry 9, 3 149-3 155 31 Brown, J.C. and Koshland, M.E. (1975) Proc. Nat. Acad. Sci. U.S.A. 72, 5111-5115 32 Connell, G. E. and Porter, R. R. (1971) Biochem. J. 124, 53P 33 Colomb, M. and Porter, R. R. (1975) Biochem. J. 145, 177-183 34 Kehoe, J. M., Bourgois, A., Capra, J. D. and Fougereau, M. (1974) Biochemisfry 13, 24992504 35 Yasmeen, D., Ellerson, J. R., Dorrington, K.J. and Painter, R. H. (1976) J. Immunol. 116, 518-
526 36 Sledge, C. R. and Bing, D. H. (1973) J. Biol. Chem.
References 1 Davies, D.R., Padlan, E.A. and Segal, D.M. (1975) Annu. Rev. Biochem.44,639~667 2 Palm, W. and Hilschmann, N. (1975) HoppeSevler’ , s Z. Phvsiol. Chem. 356. 167-191 3 Epp, O., Colman, P.M., Fehlhammer, H., Bode, W., Schiffer, M., Huber, R. and Palm, W. (1974) Eur. J. Biochem. 45, 513-524
4 Epp, O., Lattman, E. E., Schiffer, M., Huber, R. and Palm, W. (1975) Biochemistry 14,49434952
5 Rutishauser, R., Cunningham, B.A., Bennet, C., Konigsberg, W. H. and Edelman, G. M. (1970) Biochemistry 9, 3 171-3 181 6 Fleischmann, J. B., Pain, R. H. and Porter, R. R. (1962) Arch. Biochem. Biophys. Suppl. 1, 174 7 Schiffer, M., Girling, R.L., Ely, K.R. and Edmundson, A. B. (1973) Biochemistry 12, 462(r 463 1 8 Amzel, L. M., Poljak, R. J., Saul, F., Varga, J. M. and Richards, F. F. (1974) Proc. Nut. Acad. Sci. U.S.A. 71, 1427-1430
9 Segal, D. M., Padlan, E. A., Cohen, G. H., Rudikoff, S., Potter, M. and Davies, D. R. (1974) Proc. Nat. Acad. Sci. U.S.A. 71,42984302
10 Colman, P.M., Deisenhofer, J., Huber, R. and Palm, W. (1976) J. Mol. Biol. 100,257-282 11 Deisenhofer, J., Colman, P.M., Huber, R., Haupt, H. andschwick, G. (1976) Hoppe-Seyler’s Z. Physiol. Chem. 357,435445
12 Huber, R., Deisenhofer, J., Colman, P.M., Matsushima, M. and Palm, W. (1976) Nature submitted 13 Connell, G. E. and Porter, R. R. (1971) Biochem. J. 124, 53P 14 Kehoe, J. M., Bourgois, A., Capra, J. D. and Fougereau, M. (1974) Biochemistry 13, 2499-2504 15 Kornfeld, R., Keller, J., Baendger, J. and Kornfeld, S. (1971) J. Biol. Chem. 246, 3259-3268 16 Melchers, F. (1973) Biochemisfry 12, 1471-1476 17 Weitzman, F. and Scharff, M.D. (1976) J. Mol. Biol. 102, 237-252
18 Valentine, R. C..and Green, N. M. (1965) J. Mol. Biol. 27,615617 19 Yguerabide, J., Epstein, H.F. and Stryer, (1970) J. Mol. Biol. 51, 573-590
,man, J. and Pecht, I. (1975) Proc. Nat. Acad. Sci. U.S.A. 72,2775-2779 25 Jaton, J.C., Huser, H., Braun, D.G., Givol, W., Pecht, I. and Schlessinger, J. (1975) Biochemisfry
L.
20 Noelken, M. E., Nelson, C. A., Buckley, C. E. and Tanford, C. (1965) J. Biol. Chem. 240, 218224 21 Feinstein, A. and Rowe, A. J. (1965) Nature 205, 147-149 22 Ashman, R. F., Kaplan, A.P. and Metzger, H. (1971) Immunochemistry 8,627~641 23 Ashman, R. F. and Metzger, H. (1971) Immunochemistry 8, 613-656 24 Schlessinger, J., Steinberg, I. Z., Givol, D., Hoch-
37 Goers, J. W., Shumaker, V.N., Glovsky, M.M., Rebek, J. and Miiller-Ehcrhard, H. J. (1975) J. Biol. Chem. 250,49 1884925 38 Cathou, R. E. and Dorrington, K. J. (1975) in Subunits in Biological Systems (Timasheff, S. N.
and Fasman, G. D., eds), pp. 91-224, Part C, Vol. 7, Marcel Dekker, New York/Base1 39 Nisonoff, A., Hoper, J. E. and Spring, S. B. (1975) in The Antibody Molecule (Dixon, Jr, F. J. and Kunkel, H. G., eds), Academic Press, New York/ San Francisco/London
Notes *Amino-acid-sequence numbers in the Kol Fab part refer to the McPC 603 molecule [ 1] and to the Rei protein VL dimer [2-4]. **In the Fc fragment the amino-acid-sequence numbers are based on the Eu amino-acid sequence [51.
248, 2818-2823
Catabolite inactivation in yeast Helmut Holzer Glucose and its catabolites not only cause repression of enzyme synthesis in some organisms but also inactivation of certain enzymes. This ‘catabolite inactivation’ in yeast is described and discussed.
A ‘glucose effect’, defined as the repression of synthesis of certain enzymes following the addition of glucose to cells growing in media with a carbon source metabolically not closely related to glucose, was first described in 1942 for bacteria by Epps and Gale [l] and Monod [2,3], and in 1947 for yeast by Spiegelman et al. [4,5]. Based on the finding that not glucose itself, but rather catabolites which are formed rapidly from glucose and accumulate in the cell, are responsible for repression of enzyme synthesis, Magasanik et. al. [6,7] subsequently termed the effect ‘catabolite repression’. The observation that all the enzymes repressed by glucose ‘are capable of converting -their substrates to metabolites which the cell can also obtain independently and more readily by the metabolism of glucose’ [6] provided an explanation for the usefulness of ‘catabolite repression’ in cell economy. Spiegelman and Reiner [5] found that addition of glucose to yeast cells grown on galactose not only repressed the synH.H. is Professor of Biochembtry at the University of Freiburg, Freiburg D-7800, G.F.R.
thesis of the galactose metabolizing system ‘galactozymase’ but also rapidly inactivated this system. In the following years glucose-induced inactivation of other yeast enzymes was discovered: a-glucoside permease [8,9], cytoplasmic malate dehydrogenase [10-l 21, fructose- 1,6-bisphosphatase [ 131, phosphoenolpyruvate carboxykinase [ 14,151 and isopropylmalate synthase [ 161.In this article the term ‘catabolite inactivation’ is proposed for the phenomenon of glucose-induced inactivation of enzymes*. This term is derived from ‘catabolite repression’ to show the similarities in both phenomena, but it must be kept in mind that very different mechanisms are being discussed - ‘catabolite inactivation’ involves inactivation of already existing enzymes; ‘catabolite repression’ refers to inhibition of synthesis of enzymes. ‘Galactozymase’
‘Galactozymase’ activity was defined by Spiegehnan and Reiner [5] as CO, evolution during anaerobic incubation of Saccharomyces with galactose. Galactozymase activity decreased by approximately
174 synaptic vesicles. A comparison with the isolated synaptic terminals (synaptosomes, Table I) shows, however, that all these components are far more concentrated (4-22 times) in the terminal region than in the perikaryon. These observations suggest that the function of the perikaryon is to synthesize all the components of the cholinergic system eventually needed at the presynaptic nerve terminal, including even acetylcholinesterase. Preliminary studies with ligatured [23,24] or excised lengths [25] of axons leave no doubt that acetylcholine [23], vesicle membrane protein [24] and choline acetyltransferase [25] are transported, at varying rates, and possibly by different mechanisms down the axon from cell body to terminal. Thus the axon besides conducting information from the cell body to the terminal by a propagated electrochemical disturbance, the action potential, also serves as a conveyor belt transporting components needed for transmitter synthesis and storage from their place of synthesis in the cell body to their site of action in the terminal.
I5 Dowdall, M. J., Boyne, A. F. and Whittaker, V. P.
References Otsuka, M., Konishi, S. and Takahashi, T. (1972) Proc. Jup. Acad. 48, 747-751
von Euler, U.S. and Gaddum, J. H. (1931) J. Physiol. 72, 7487
Cleugh, J., Gaddum, J. H., Mitchell, A. A., Smith, A. W. and Whittaker, V. P. (1964) J. Physiol. 170, 69-85
4 Duffy, M. J., Mulhall, D. and Powell, D. (1975) J. Neurochem.
25. 305-307
5 Hughes, J., Smith, T. W., Kosterlitz, H. W., Fothergill, L.A., Morgan, B.A. and Morris, H. R. C1975) Nature 258. 577-579 Dowdall, M. J. and Zimmermann, H. (1976) Exp. Brain Res. 24, 8-9
Dowdall, M. J. and Simon, E.J. (1973) J. Neurothem. 21,969-982 Yamamura, H.I. and Snyder, S.H. (1973) J. Neurochem.
10
11 12 13
21, 1355-1374
Dowdall, M.J., Wlchtler, K. and Henderson, F.M. (1975) Abstr. 5th Inr. Meet. Sot. Neurothem., Barcelona, p. 122 Whittaker, V.P. and Dowdall, M.J. (1975) in Cholinergic Mechanisms (Waser, P. G., ed.), pp. 2342, Raven Press, New York Whittaker, V.P., Essman, W.B. and Dowe, G.H.C. (1972) Biochem. J. 128, 833-846 Nagy, A., Baker, R. R., Morris, S. J. and Whittaker V. P. (1976) Bruin Res. (in the press) Whittaker, V. P., Dowdall, M. J., Dowe, G. H.C., Facino, R.M. and Scotto, J. (1974) Bruin Res.
(1974) Biochem. J. 140, 1-12 16 Schmidt, R. (1976) Diplomarbeit, Mainz 17 Schmidt, R., Zimmermann, H. and Whittaker, V. P. (1976) Exp. Bruin Res. 24, 19-20 18 Whittaker, V. P. (1973) in The Scientifi:c Basis of Medicine Annual Reviews (Gilliland, I. and Peden, M., eds), pp. 17-31, Athlone Press, London 19 Barker, L.A., Dowdall, M. J. and Whittaker, V.P. (1972) Biochem. J. 130, 1063-1080 20 Zimmermann, H. and Dowdall, M. J. (1976) Exp. Brain Res. 24,23
21 Dowdall, M.J., Fox, G., Wachtler, K., Whittaker, V.P. and Zimmermann, H. (1976) Cold Spring Harbor Symp. Quant. Biol. 40, 65-81 22 Whittaker, V. P., Fox, G. and Fiore, L. (1976) Exp. Brain Res. 24,22
23 Zimmermann, Abstr.
4th
Int.
H. and Whittaker, Meet.
Ini.
Sot.
V.P. (1973) Neurochem.,
Yokyo, p. 245 24 Ulmar, G. and Whittaker, V. P. (1974) Brain Res. 71,155-159 25 Davies, L. (1976) Unpublished observations 26 Heilbronn, E. (1975) in Cholinergic Mechanisms (Waser, P.G., ed.), pp. 343-364, Raven Press, New York 27 Raftery, M.A., Bode, J., Vandlen, R., Chao, Y., Deutsch, J., Duguid, J. R., Reed, K. and Moody, T. (1975) in Biochemistry of Sensory Functions (Jaenicke, L., ed.), pp.‘541-563, Springer Verlag, Berlin/Heidelberg/New York
75, 115-131
14 Stadtler, H. (1976) Unpublished observations
‘llre acetylcholine receptor Any survey of the electromotor system of the Torpedo would be incomplete without a mention of another application: the electric organ as a source material for the isolation of the acetylcholine receptor. A protein (AChR) with the properties expected of the recognition site of the acetylcholine receptor has been isolated by several groups from the cholinoceptive electrocyte cells of the electric tissue. (For further details the reader is referred to two recent reviews [26,27].) It is not quite clear yet whether this protein also accounts for the acetylcholine-stimulated ionophore properties seen in the physiological receptor. One of the most interesting developments in this work has been the discovery that the injection of puritied Torpedo AChR into rabbits induces a condition similar to that of the human disease myasthenia gravis characterized by muscle weakness. At first this was thought to be impaired neuromuscular transmission due to a simple masking of the rabbit muscle AChR by the rabbit’s own antibody to AChR, but further work has shown that the condition is more complex and more like the human disease in that the immune response of the lymphatic system is involved. Presumably a primary step in the aetiology of human myasthenia gravis is the release, from damaged postsynaptic membranes, of AChR, but what triggers this release, possibly a virus infection, is as yet obscure.
Antibody structure Robert Huber The crystal structures of a human IgG antibody molecule Kol and a human Fc fragment have been determined at 0.4 and 0.35 nm resolution, respectively, by isomorphous replacement. These molecular structures, in comparison to the known structures of antigen-binding fragments, bring us closer to a coherent view of antibody structure. From these structures it is hypothesized that antibody molecules are inherently flexible but become rigid upon interaction with antigen, which triggers the formation of all longitudinal contacts. Antibody molecules serve a dual function: to recognize foreign cells and macromolecules and to trigger the events leading to their elimination. Specific recognition requires surface structures complementary to the antigen and therefore needs a huge variety of antibody molecules. In contrast, the effector functions would most economically be performed if all antibody molecules were identical. An antibody molecule consists of two identical light- and two identical heavy chains [5,6]. Each of these polypeptide chains can be subdivided into a constant region (c) with an amino acid sequence identical to other chains of the same..class and a variable region whose sequence varies between antibodies specific for difR.C. is at the Max-Planck-Institut fur Biochemie, 8033 Marrinsried, und Physikalisch-Chemisches Institut der Technischen Universitiit Miinchen, G.F.R.
ferent antigens (see Fig. 1 for details of antibody structure and its fragments and list of abbreviations used in this article). Antibody function is located on different parts of this molecule with recognition lying in a fragment consisting of a light chain and half of the heavy chain - the Fab fragment. Effector function on the other hand resides in the C-terminal half of the heavy chain with the inter-heavy chain disulphide bond intact - the Fc part. There is, however, co-operativity, and effector functions are triggered by antigen binding. How this signal is transferred is open to debate and in this article I will discuss a hypothetical model based on our current knowledge of three-dimensional antibody structure. Several crystal structures of immunoglobulin fragments (Fab fragments and Bence-Jones proteins) have been analysed during the last few years [3,4,7-g], and pro-
TIBS - August 1976
175
vide a structural basis of antibody specilicity. These molecules are composed of globular. domains formed by the variable and constant segments. Within the domains the chain folds into two disulfidelinked pleated-sheet segments, forming a barrel and covering a hydrophobic interior. There are tertiary structural differences between V and C domains. V parts (VH and VL in Fab or VL and VL in Bence-
Jones proteins) and C parts (CH~ and CL in Fab or CL and CL in Bence-Jones proteins) aggregate in a characteristic fashion, using their V faces and C faces, respectively. The V dimers form a cavity lined with hypervariable segments into which haptens bind. From the recent crystalstructure analyses of a complete human IgG antibody Kol [lo] and a human Fc fragment [I 1,121 in comparison with the papain
HH
ckbohydrate
fltf coo-
antigen-binding fragments, a coherent view of antibody structure begins to emerge. The immunoglogulin Kol molecule
The C” carbon atom positions of the (Fab), part of the Kol molecule are shown in the stereogram, Fig. 2. The Fab arms subtend an angle of 125” with their tips about 14.6 nm apart. Apart from slight differences in domain tertiary structure, the VL-VH and CL-Cnl dimers appear to be indistinguishable between Kol and the mouse Fab fragment McPC603 [l]*. In particular, the geometry of aggregation within the V and C dimers is very similar. There is a difference, however, in the relative orientation of the V and C dimers which can be described by a rotation about an axis through the switch peptides that is, the segments connecting V and C parts at residues 118-l 19 (heavy chain) and 109110 (light chain). The McPC 603 Fab fragment [l] and the Kol Fab part as seen down the switch peptide rotation axis are compared in Fig. 3. VL and VH as well as CL and CH1 are related by local diads. They subtend an angle (Fab angle) of 120”
Fig. 1. Structure of antibody (IgG) molecule and enzyme cleavage products. VL-variable harf of light chain, CL-constant halfof light chain, Vn-variable part of heavy chain, Ctrl, CHZ, cH3 - the three constant homology regions of the heavy chain. Fab - antigen - binding fragment consisting of light chain and half of the heavy chain ( VL, Vn, CL, CJ) Fe - C - terminal half of the heavy chain with the inter-heavy chain disulftde bond intact. Hinge peptide: the segment connecting Cnl and Cn2 and containing the inter-heavy chain disulfde linkage; switch peptides: the segments connecting V and C parts comprising residues at 110 (light chain) and 119 (heavy chain).
Kol Fig. 2. Stereo-drawing of the Kol (Fab), part. C’ atom positions were derived by fitting the 1UcPC603~ V- and C-dimers to the Kol electron-density map. The coordinates of the C-terminal residues of heavy and light chain ana’the hinge peptide were obtained by interpretation of the 0.4nm Fourier map.
Fig. 3. Comparison of J4cPC603 Fob fragment and Kol Fab part. The Fab angle changes by 50°. There is a VI&HI contact in the fragment but not in the intact molecule. NH, CH, NL, CL: N-and C-terminus of heavy and light chain, respectively.
176
TIBS - August 1976
Tf‘heFc fragment structure
in the McPC 603 fragment and 170” in the Kol protein, demonstrating the major change in quarternary structure - ‘bending the elbow’. It is already evident from Fig. 3 that in the McPC603 molecule some residues of VH and CH1 come close to each other. These are the segments 8-10 in VH and 152 and 207-208 in CH~. Through the widening of the Fab angle from 120” to 170” in Kol there is no contact between VH and CH1. Also, VLand CL have no contact, nor do the two Fab arms. No interpretable electron-density is found in the Kol Fourier map which could be assigned to the Fc part [lo] leading to the conclusion that the Fc stem is disordered in the Kol crystals, C-terminal to the hinge peptide: There is a large cylindrial space in the crystal lattice around the three-fold screw axis, where the Fc parts must be located. Its volume is sufftcient to accommodate Fc, but its dimensions do not allow the Fc parts to obey the two-fold symmetry of the rest of the molecule. Fc is either statistically disordered or wagging thermally. A ‘minimum disorder’ Kol model has been constructed on the assumption that the deviation of Fc from the crystal symmetry as dictated by the crystal dimensions and the (Fab), packing should be minimal. If the disorder is due to temperature movement this model represents a snap-shot (Fig. 4). It is clear that there are no contacts between (Fab), and Fc in the minimum disorder model except the covalent linkage through the hinge peptide - the segment connecting Cnl and CH~ - which must have a rather extended conformation.
A stereo-plot of the C” carbon atom positions of the Fc fragment is shown in Fig. 5. The molecule looks like ‘Mickey Mouse’with the compact CH~ dimer forming the head and the CH~ domains the ears. The CH3 dimer pairing is closely similar to the CH~-Cr pairs found in Fab fragments. Both, Cn2 and CH~ show the common immunoglobulin fold. They are connected by a loosely-folded segment which is susceptible to proteolytic attack [13,14]. The longitudinal CHA-CHA contact is of the same size as the VH-CH~ contact in Fab fragments. It comprises the segments homologous to those mediating the VHCHI contact in Fab fragments described above. The carbohydrate moiety [15] covers a large part of the C-face of the CH2 domain, obviously shielding apolar amino acid residues. Fc was the first glycoprotein analysed and it was a surprise to find the carbohydrate moiety spatially fixed. The function of the carbohydrate in immunoglobulins is unclear but it might be involved in secretion, although recent experiments do not confirm this [16,17]. Its rigid conformation and close association with the protein points to a general stabilizing effect. Domaindomain interactions The lateral V-V (VWVL, VL-Vr) and CL-CL) interactions c-c (CL-W) observed in Fab fragments [8,9] and Bence-Jones proteins [7,4] have been described in great detail elsewhere and will not be considered here. The relative
“R
L
.
1
KOL FLEXIBLE
.
MODEL
Fig. 4. Stereo-drawings of the ‘minimum disorder’ model of the Kol molecule. This should be regarded as a snapshot of the flexible molecule. There are no contacts between (Fab), and Fe. The structure of the Fe part was derived from the Fc fragment, although we believe that Fc is flexible between Cn2 and Cn3 in the Kol molecule. The hinge peptide is drawn as observed in Kol (Fab), and the Fc fragment. In order to make the covalent linkage, the hinge, which is bent back in between the Cn2 domains, has to be pulled out. Ch. : carbohydrate.
arrangement of the two CH3 domains observed in the Fc fragment is closely similar to the C&HI aggregation found in McPC 603 [ 111.The CH2 domains have no contact with each other except the covalent linkage through the hinge. It is an intriguing question why CH~ forms neither the familiar C-C nor V-V contact. The tertiary structure of the CH~ domain does not resemble a C- but also lacks some features of a V-domain. This might prevent dimerization. In the presence of the spatially fixed carbohydrate moiety, formation of the C-C interface is impossible. Furthermore, some of the hydrophobic residues in CH~ which lie close to the local diad and are probably important for the contact, are replaced by polar groups in CH2. Formation of a V-V contact of CH2 is also prohibited as several of the contact residues [4], conserved in V segments, are exchanged for polar residues. In the Kol protein and the ‘minimum disorder’ model derived from it, no longitudinal contacts are observed between VH and CH~, VL and CL, CH~ and CH~. There are also no contacts between the Fab arms. In view of this lack of stabilization along the chain, it is difficult to assume that the Kol molecule is rigid in solution; rather, that flexibility exists between V and C dimers (‘elbow’ V-C-flexibility), between the Fab arms ((Fab),-flexibility) and between Fc and (Fab),(Fc-(Fab),-flexibility). As mentioned before, flexibility between Fc and (Fab), could explain the weak, uninterpretable electron-density for Fc in the Kol crystals. The relative positions of V- and Cdimers and Fab arms are frozen in the Kol crystal lattice dictated by crystal packing requirements. Even the CH2 segments might sit flexibly on the CH3 dimers (CH2CH3-flexibility) in the Fc part in the Kol protein in contrast to their fixed position in the fragment. The model that I would like to suggest for the Kol protein and for other unliganded (without antigen) IgG molecules is a highly flexible one. In obvious contrast the Fab fragments [8,9] and the closely similar Bence-Jones protein dimer [7] appear to be rigid molecules. In the three structures, V- and Cdimers subtend a similar Fab (‘elbow’) angle of about 120”. Crystal packing is different in the three crystal structures and would be unlikely to have the same effect on the Fab angle if it were variable. Indeed there exists a longitudinal contact between VHand Cul in these fragments as described before. In the Fc fragment CH2 and CH3 form a contact as described. The linking peptide is not in a completely extended conformation but is irregularly bent between residues Lys (338) and Gly (341)**. Loosening of the Cn2-CH3 con-
TIBS - August 1976
tact and a stretching out of the linking peptide appear possible. It is remarkable that homologous segments mediate the close V-C contacts in the Fab fragments and the close CH2-CH3 contact in the Fc fragment. Of further interest is the fact that the contact segments described for C”l and C”3 when projected into a V segment coincide with the first and third hypervariable segments forming the antigen-binding surface in V dimers. This suggests the presence of a general binding area in immunoglobulin domains. The residues forming in this area must of course be variable in V domains to meet the requirements of complementarity to various antigens, but conserved in Cdomains for inter-domain interaction. Signal transfer
In an allosteric, cooperative immunoglobulin model, the signal transfer through the domains from the general binding area (top contact) to the opposite end of the domains (bottom contact) would occur along homologous segments in all domains. I tried to construct a hypothetical liganded antibody molecule by combining the fragment molecules. It is well documented that Fab fragments show no gross structural change upon hapten binding [8,9] indicating that the fragment structure is characteristic for liganded molecules. Assuming that the Fc fragment structure is also characteristic for a liganded antibody molecule and starting with the Kol (Fab), I set the Fab angle to 120”, characteristic for Fab fragments, superimpose the hinge as seen in Kol and Fc and search for the allowed relative azimuthal angle. An important aspect of this model (Fig. 6) is that all longitudinal inter-domain contacts are closed, resulting in a rigid structure. (Fab), bending is prohibited by the C~l-c~2 interaction. It is clear that in view of the (Fab),-flexibility of the Kol molecule the choice of the relative arrangement of the Fab arms was arbitrary. By a small. (Fab), rearrangement the formation of a lateral contact of the Fab arms appears possible. Antigen-binding may cause a stiffening of the flexible antibody molecule by formation of the longitudinal inter-domain contacts described. In the usual terminology of allosteric the flexible cooperative interactions, model would represent the tense state (T) which changes to the relaxed state (R) upon antigen binding. Antigen-binding triggers the formation of the longitudinal interdomain contacts. The T state quarternary structure is maintained only in the intact molecule. Fragment structures adopt a conformation characteristic for
L
oc
?? c.h.
?
c
Fig. 5. Stereo-drawing of the Co: carbon positions and the centres of the carbohydrate hexose units of the Fc fragment. The carbohydrate structure was assumed to be as found in a myeloma protein j16J. In Fc the carbohydrate chain appears to be longer. ?? C.h. : approximate centres of carbohydrate hexose units.
the R-state of the intact molecule. Segmental flexibility between Fab and Fc parts of antibody molecules is experimentally established by spectroscopic and hydrodynamic observations [19,20]. Electron micrographs also show variable angles between the Fab arms of antibody molecules [ 18,211. Hydrogen-exchange experiments as well as experiments with limited proteolysis of IgM and their Fab fragments may be interpreted in terms of a less flexible fragment structure [22,23]. There are several observations of structural changes in antibody molecules upon antigen-binding. Circular polarization of fluorescence changes in complete antibodies upon antigen binding, but not in Fab fragments [24,25]. An antigen-induced volume contraction has been found by small angle X-ray scattering in complete antibodies, but not in Fab fragments [26,27]. Antibodies appear to lose flexibility upon antigen binding [28,29] and to gain conformational stability against denaturation [30]. Further-
more, there are observations on IgM molecules of differences in thermodynamic parameters for haptens and antigens, indicating a conformational change in Fc induced by antigen binding [3 11. Immunoglobulins have a number of effector functions of which the binding of complement has been studied on a molecular level. It is well documented that the binding site for Cl, the first component of the complement system is located on the CH~domains [32-351. CH~, Fc and IgG have intrinsic complement binding properties which are, however, greatly enhanced by specific antigen-binding or heat-aggregation [36,37]. It has been suggested that this might be due to the proximity of several complement-fixation sites and/or to structural changes [31,37] (for review see [38,39]). The rigid, liganded antibody molecule which I propose could enhance Cl binding for several reasons: (a) In the unliganded, flexible antibody molecule the C 1 binding site might be transiently occluded by the Fab arms and pos-
??
B
KOL
RIGID
.
MODEL
Fig. 6. Stereogram of a hypothetical liganded IgG molecule. The molecule is T-shaped with Fc forming stem. Fc and (Fab), are in close contact. The hinge peptide is folded back in between the Cn2 parts. C.h.: carbohydrate.
the
TIBS - August 1976
178 sibly also by CH~. The long extended hinge peptide in the unliganded Kol molecule allows the Fab arms far-reaching wagging motions to cover part of Cn2. (b) Protein-protein complex formation requires freezing out of rotational and translational degrees of freedom. A component with internal flexibility is therefore less favourable for complex formation than a rigid molecule. (c) The formation of the longitudinal contacts might induce a structural change in Cn2 thereby improving Cl binding. Little is known in molecular terms about other effector functions located on Cn3, as cytophilic activity [35], requiring interaction with Fc receptors on macrophages. A detailed comparison of the structures of Cn3 and CH~ might illuminate this problem.
14, 5312-5315 26 Pilz, I., Krathy, O., Licht, A. and Sela, M. (1973) Biochemistry 12, 4998-5005
27 Pilz, I., Karthy, O., Licht, A. and Sela, M. (1975) Biochemistry 14, 13261333 28 Tumerman, L.A., Nezlin, R. S. and Zagyanski, Y.A. (1972) FEBS Lett. 19,29&292 29 Warner, C. and Shumaker, V. (1970) Biochemistry 9, 451-459 30 Cathou, R.E and Warner, T.C. (1970) Biochemistry 9, 3 149-3 155 31 Brown, J.C. and Koshland, M.E. (1975) Proc. Nat. Acad. Sci. U.S.A. 72, 5111-5115 32 Connell, G. E. and Porter, R. R. (1971) Biochem. J. 124, 53P 33 Colomb, M. and Porter, R. R. (1975) Biochem. J. 145, 177-183 34 Kehoe, J. M., Bourgois, A., Capra, J. D. and Fougereau, M. (1974) Biochemisfry 13, 24992504 35 Yasmeen, D., Ellerson, J. R., Dorrington, K.J. and Painter, R. H. (1976) J. Immunol. 116, 518-
526 36 Sledge, C. R. and Bing, D. H. (1973) J. Biol. Chem.
References 1 Davies, D.R., Padlan, E.A. and Segal, D.M. (1975) Annu. Rev. Biochem.44,639~667 2 Palm, W. and Hilschmann, N. (1975) HoppeSevler’ , s Z. Phvsiol. Chem. 356. 167-191 3 Epp, O., Colman, P.M., Fehlhammer, H., Bode, W., Schiffer, M., Huber, R. and Palm, W. (1974) Eur. J. Biochem. 45, 513-524
4 Epp, O., Lattman, E. E., Schiffer, M., Huber, R. and Palm, W. (1975) Biochemistry 14,49434952
5 Rutishauser, R., Cunningham, B.A., Bennet, C., Konigsberg, W. H. and Edelman, G. M. (1970) Biochemistry 9, 3 171-3 181 6 Fleischmann, J. B., Pain, R. H. and Porter, R. R. (1962) Arch. Biochem. Biophys. Suppl. 1, 174 7 Schiffer, M., Girling, R.L., Ely, K.R. and Edmundson, A. B. (1973) Biochemistry 12, 462(r 463 1 8 Amzel, L. M., Poljak, R. J., Saul, F., Varga, J. M. and Richards, F. F. (1974) Proc. Nut. Acad. Sci. U.S.A. 71, 1427-1430
9 Segal, D. M., Padlan, E. A., Cohen, G. H., Rudikoff, S., Potter, M. and Davies, D. R. (1974) Proc. Nat. Acad. Sci. U.S.A. 71,42984302
10 Colman, P.M., Deisenhofer, J., Huber, R. and Palm, W. (1976) J. Mol. Biol. 100,257-282 11 Deisenhofer, J., Colman, P.M., Huber, R., Haupt, H. andschwick, G. (1976) Hoppe-Seyler’s Z. Physiol. Chem. 357,435445
12 Huber, R., Deisenhofer, J., Colman, P.M., Matsushima, M. and Palm, W. (1976) Nature submitted 13 Connell, G. E. and Porter, R. R. (1971) Biochem. J. 124, 53P 14 Kehoe, J. M., Bourgois, A., Capra, J. D. and Fougereau, M. (1974) Biochemistry 13, 2499-2504 15 Kornfeld, R., Keller, J., Baendger, J. and Kornfeld, S. (1971) J. Biol. Chem. 246, 3259-3268 16 Melchers, F. (1973) Biochemisfry 12, 1471-1476 17 Weitzman, F. and Scharff, M.D. (1976) J. Mol. Biol. 102, 237-252
18 Valentine, R. C..and Green, N. M. (1965) J. Mol. Biol. 27,615617 19 Yguerabide, J., Epstein, H.F. and Stryer, (1970) J. Mol. Biol. 51, 573-590
,man, J. and Pecht, I. (1975) Proc. Nat. Acad. Sci. U.S.A. 72,2775-2779 25 Jaton, J.C., Huser, H., Braun, D.G., Givol, W., Pecht, I. and Schlessinger, J. (1975) Biochemisfry
L.
20 Noelken, M. E., Nelson, C. A., Buckley, C. E. and Tanford, C. (1965) J. Biol. Chem. 240, 218224 21 Feinstein, A. and Rowe, A. J. (1965) Nature 205, 147-149 22 Ashman, R. F., Kaplan, A.P. and Metzger, H. (1971) Immunochemistry 8,627~641 23 Ashman, R. F. and Metzger, H. (1971) Immunochemistry 8, 613-656 24 Schlessinger, J., Steinberg, I. Z., Givol, D., Hoch-
37 Goers, J. W., Shumaker, V.N., Glovsky, M.M., Rebek, J. and Miiller-Ehcrhard, H. J. (1975) J. Biol. Chem. 250,49 1884925 38 Cathou, R. E. and Dorrington, K. J. (1975) in Subunits in Biological Systems (Timasheff, S. N.
and Fasman, G. D., eds), pp. 91-224, Part C, Vol. 7, Marcel Dekker, New York/Base1 39 Nisonoff, A., Hoper, J. E. and Spring, S. B. (1975) in The Antibody Molecule (Dixon, Jr, F. J. and Kunkel, H. G., eds), Academic Press, New York/ San Francisco/London
Notes *Amino-acid-sequence numbers in the Kol Fab part refer to the McPC 603 molecule [ 1] and to the Rei protein VL dimer [2-4]. **In the Fc fragment the amino-acid-sequence numbers are based on the Eu amino-acid sequence [51.
248, 2818-2823
Catabolite inactivation in yeast Helmut Holzer Glucose and its catabolites not only cause repression of enzyme synthesis in some organisms but also inactivation of certain enzymes. This ‘catabolite inactivation’ in yeast is described and discussed.
A ‘glucose effect’, defined as the repression of synthesis of certain enzymes following the addition of glucose to cells growing in media with a carbon source metabolically not closely related to glucose, was first described in 1942 for bacteria by Epps and Gale [l] and Monod [2,3], and in 1947 for yeast by Spiegelman et al. [4,5]. Based on the finding that not glucose itself, but rather catabolites which are formed rapidly from glucose and accumulate in the cell, are responsible for repression of enzyme synthesis, Magasanik et. al. [6,7] subsequently termed the effect ‘catabolite repression’. The observation that all the enzymes repressed by glucose ‘are capable of converting -their substrates to metabolites which the cell can also obtain independently and more readily by the metabolism of glucose’ [6] provided an explanation for the usefulness of ‘catabolite repression’ in cell economy. Spiegelman and Reiner [5] found that addition of glucose to yeast cells grown on galactose not only repressed the synH.H. is Professor of Biochembtry at the University of Freiburg, Freiburg D-7800, G.F.R.
thesis of the galactose metabolizing system ‘galactozymase’ but also rapidly inactivated this system. In the following years glucose-induced inactivation of other yeast enzymes was discovered: a-glucoside permease [8,9], cytoplasmic malate dehydrogenase [10-l 21, fructose- 1,6-bisphosphatase [ 131, phosphoenolpyruvate carboxykinase [ 14,151 and isopropylmalate synthase [ 161.In this article the term ‘catabolite inactivation’ is proposed for the phenomenon of glucose-induced inactivation of enzymes*. This term is derived from ‘catabolite repression’ to show the similarities in both phenomena, but it must be kept in mind that very different mechanisms are being discussed - ‘catabolite inactivation’ involves inactivation of already existing enzymes; ‘catabolite repression’ refers to inhibition of synthesis of enzymes. ‘Galactozymase’
‘Galactozymase’ activity was defined by Spiegehnan and Reiner [5] as CO, evolution during anaerobic incubation of Saccharomyces with galactose. Galactozymase activity decreased by approximately