- Email: [email protected]
Sulfhydryl proteins and the engram
MedicolHypofheses(1990) 31,309-312 0 Longman Group UK Ltd 19%
Sulfhydryl
Proteins
and the Engram
G. W. RAFTER Department of Biochemistry, WV 26506, USA
West Virginia University
School of Medicine,
Morgantown,
Abstract - A scheme, which is selectional not instructional, is described for the molecular basis of the engram. It uses as reactants intermediates formed during the catabolism of neuronal membrane sulfhydryl proteins. It is proposed that these compounds undergoing intermolecular association generate a large number of diverse structures, some of which are selected for further complex formation. The basis of selection is the binding of a particular compound to membrane lipid, which enhances its reaction with another bound compound. Because the composition and configuration of membrane lipids are influenced by environmental factors, the scheme implies that incoming sensory signals are matched with endogenously generated chemical structures.
its biological context. More recently, the synapse as the physical basis of memory has held sway. In this hypothesis connections between synapses of cells are specified genetically, but the final wiring pattern depends on selective stabilization during neural activity (2). A form of stabilization, sometimes called synaptic plasticity, is long-term potentiation that is provoked by brief stimulation of certain presynaptic axons, and which leads to an increase in the effectiveness of synaptic transmission (3). In these systems the involvement of a small number of neurons is implied, which is in striking contrast to the large systems approach. According to one author (4) the boundary between the two approaches can be drawn at the level of about six neurons. Put another way ‘no conceivable neuron or set of neurons, no matter how diffuse it’s synaptic inputs, can evaluate the enormous amount of
Introduction
How the brain acquires and stores information is one of the most baffling and important problems in science. Because chemical structure and biological functions are intimately linked, it is axiomatic that chemical change must accompany the process. Thus far, the molecular nature of the engram or memory trace has eluded investigators. During the 1960’s much research time and money was spent to implicate protein and nucleic acid in the process, but the results were inconclusive. In a thoughtful retrospect of this period, Irwin (1) attributes some of these failures to the investigator’s isolation of the process from
Date received Date accepted
30 January 1989 27 February 1989
309
310
MEDICAL
neural activity shown to be involved in retrieval of even a simple form discrimination’ (5). This was concluded from the finding that between 5 to 100 X lo6 neurons were activated as detected by double-labeled metabolic maps during presentation of a visual cue to a split-brain cat. Likewise, the current mechanistic molecular approach to learning is not consonant with a system which is selective and degenerate and that accounts for the distributed nature of memory and its ability to recall events associatively in temporal sequence (6). In this paper a scheme utilizing catabolism of sulfhydryl proteins is proposed, which appears to be more in keeping with the biological aspects of higher brain function. Some considerations
in formulating
the proposal
Some of the ideas that influenced the selection of sulfhydryl protein metabolism as a molecular mechanism for learning or memory follow. Memory has been defined as a process that carries stored events in the brain forward in time (7). This means that generation of ‘memory molecules comprises aspects of both positive and negative entropy. Previously, processes characterized by nonlinear chemical kinetics and cross catalysis have been described in which a state of greater order (dissipative structures) can be generated from a state of lower order (8). The energetic rationale for this result is that entropy production of open systems can be divided into two parts: entropy flow due to exchanges with the environment and entropy production due to irreversible processes within the system. While the entropy production of the system is always zero or positive, the entropy flow can have.either sign. That neurons generate a diversity of chemical structures, some of which are selected for by incoming or reentrant signals, fits well with the molecular mechanism just described. Both Pringle (9) and Jerne (10) have previously discussed selective mechanisms for memory. The sulfhydryl proteins are uniquely suited for this role; the total number of isomeric structures formed from the oxidation and association of two light and two heavy immunoglobulin chains has been calculated to be 4.8 x 1016 (11). It is proposed that the participants in the learning process are not unique to it, but comprise membrane sulfhydryl-disulfide proteins with recognizable biological function, for example, cell surface receptors (12, 13) which because of
HYPOTHESES
their complex interaction during catabolism take on a new, unique biological role. The proposal
In presenting the proposal reference is made to insulin, which was used as a simple model protein to explore the hypothesis. Insulin upon reduction undergoes intermolecular sulfhydryl-disulfide exchange and ultimately forms insoluble complexes (14). The reaction proceeds with loss of free energy, because once the restrictive influence of the disulfide bonds are removed, the molecule seeks a more favorable minimum free energy state. When first synthesized and folded from a precursor molecule, it presumably was at a minimum free energy state, but this was lost in the case of insulin after removal of the connecting peptide by limited proteolysis. Similar reactions of multichain neuronal membrane proteins, as well as single chain proteins, which have undergone limited proteolysis, are suggested to form a large number of intermolecular complexes. The kinetically availability of the sulfhydryl and disulfide groups of the reactants would be expected to affect the products formed. Experimentally, this result is seen in the effect of urea on complex formation. Insulin incubated with reduced glutathione forms a high molecular weight soluble complex of about 70 000 while in the presence of urea the complex does not form (Fig. 1). Likewise, during the oxidation of the reduced B-chain of insulin a large series of higher molecular compounds are formed, while only one compound, the oxidized B-chain, is formed in the presence of urea (15). It appears that with of conformational constraints, inremoval tramolecular disulfide bond formation is favored over intermolecular disulfide bond formation. The kind and number of products formed will also depend, in part, on the kind and number of reactants. This, in turn, depends on factors that influence their formation such as availability of Ca++ to activate neutral proteases and the redoxpotential of the tissue. Thus the association of intermediate catabolic products of a small number of neuronal membrane proteins could conceivably generate a large number of discrete molecular species. And since their formation is constrained by the conformation of reactants, as well as by the metabolic state of the cell, the process is chaotic rather than random. But how is order (information) extracted from this chaotic process? It is proposed that some of the catabolic products are selected as reactants
SULFHYDRYL
PROTEINS
311
AND THE ENGRAM
0.3
0.3
0.2
0.2 0
is
% a
a? 0.1
0.1
4.0
9.6 ml
4.8
9.6 ml
Fig. 1 Gel filtration of insulin incubated with reduced glutathione in the presence of urea. Mixtures contained in 1.0 ml of 0.05 M Tris-HCI buffer, oH 7.7 2 me of bovine insulin (Sigma Chem Co., St. Louis, MO.)-and 6.5 mM glutathione. They were incubated 3 hrs at 30°C and then passed through a Sephadex G-75 column (0.9 x 15 cm) with 1% acetic acid and fractions collected. O--O 0.36 gms of urea added at end of incubation. o--o 0.36 gms of urea present during incubation.
Fig. 2 Gel filtration of insulin incubated with reduced glutathione inthe presence of lipopolysaccharide. Conditions as described in Fig. 1 except mixtures were incubated 2.5 hrs. l --0 no additions. o--o 100 Dg of lipopolysaccharide (E. coli, Difco Lab., Detroit, MI.).
for synthesis of new complexes, while others not selected are degraded to their constituent amino acids. It is proposed that the basis for the selection process is the binding of the compound to a membrane lipid and its subsequent reaction with another bound compound to form a new structure. For the compounds selected, because of their structure, the lipid serves as a concentrating media, which promotes the association reaction as it is very concentration dependent (16). In the insulin system this result is seen as enhancement of the formation of a high molecular weight complex by the addition of bacterial lipopolysaccharide to mixtures (Fig. 2). Because this effect of lipopolysaccharide is not seen in 1 hr reaction mixtures, it appears that some intermediate compound, not monomeric insulin, is the immediate precursor for the synthesis of the high molecular weight complex. It is the ‘selective value’ of each reactant with respect to its further anabolism that results in self organization in the proposed process (17). In turn, the process is influenced by environmental factors such as incoming or reentrant sensory signals which altered the composition and configuration of neuronal membrane lipids (18). In summary, the proposed scheme suggests that sensory signals are matched, using membrane lipid as an inter-
mediary, with endogenously structures.
generated
chemical
Conclusions
That the synthesis of the postulated intermolecular complexes continues throughout the life of the organism, acquiring information as a consequence of the ‘valued information’ of some structures for further reaction, prompts several interesting questions. Since all biological molecules with the exception of DNA turnover in some finite time, then how is information retained in some cases for a lifetime? That the complexes become supramolecular aggregates and are sequestered in the lipid environment of a neuronal membrane would be expected to greatly reduce their susceptibility to proteolysis. Nevertheless, since generation of dissipative structures is a dynamic process, to retain this property some turnover would be expected. Then does the synthetic process possess aspects of selfreplication making the overall system redundant? A second question arises as a consequence of the commonality in the biochemical components among species of the proposed mechanism. Then what is the molecular basis of the differences in cognitive skills between species? Differences in
312 the variety and number of structures, from which selection is made for further anabolism, is suggested to be important to explain these In other words, the ‘valued differences. information’ of the reactants. Finally, because of the commonality of components in different species, some degree of cognitive skills is expetted in a variety of species.
References 1. Irwin L N. Fulfillment and frustration. The confessions of a behavioral biochemist. Perspect Biol Med 21: 476, 1978. J-P, Danchin A. Selective stabilization of 2. Changeux developing synapses as a mechanism for the specification of neuronal networks. Nature 264: 705, 1976. M B. Synatic memory molecules. Nature 3. Kennedy 335: 170, 1988. 4. Rose D. Is brain research dead? Trends Neurosciences 10: 196, 1987. 5. John E R, Tang Y, Brill A B. Young R. Ono K. Doublelabeled metabolic maps of memory. Science 233: 1167, 1986. V B. The Mindful Brain. The 6. Edelman GM, Mountcastle MIT Press, Cambridge, MA, 1978. imagery and memory, p. 1669 7. Holstead W C. Thinking, in Handbook of Physiology, Neurophysiology. Vol 111. Am Physiol Sot. Washington, DC, 1960.
MEDICAL HYPOTHESES Priaoaine I. Stengers I. Order out of Chaos. Bantam Book;. Toronto, i984. Prinele J W S. On the oarallel between learninr and 9. evol;tion. Behaviour 3: 174, 1951. lo, Jertfe N K. Antibodies and learning: Selection versus instruction. p. 200 in The Neurosciences, (G C Quarton, T Melnecheck. F 0 Schmitt, eds) Rockefeller U Press. New York.1967. 11. Wang Z-X. Ju M, Tsou C-L. Numbers of ways of joining S H groups to form multi-peptide chain proteins. J Theor Biol 124: 293, 1987. 12. Malbon C C, George S T, Moxham C P. Intramolecular disulfide bridges: avenues to receptor activation? Trends Biochem Sci 12: 172, 1987. K. Murase H, Murayama M, Matsuda M. 13. Kamikubo Miura K. Evidence for disulfide bonds in membranebound and solubilized opeoid receptors. J Neurochem 50: 503, 1988. F, Goldberger R F. Anfinsen C P. 14. Givol P, DeLorenzo Disulfide interchange and the three-dimensional structure of oroteins. Proc Natl Acad Sci 53: 676, 1965. R C, Carpenter F H. Oxidation of the 15. Paynovich sulfhydryl forms of insulin A-chain and B-chain. lnt J Peptide Protein Res 13: 113. 1979. protein modifications and 16. Light A. Protein solubility, orotein folding. Bio Techniaues 3: 298. 1985. of matter and the evolutien of 17. Eigen M. Selfirganization biological macromolecules. Naturwissenschaften 10: 465. 1971. M, Governa M, Gratton E, Fiorini R, 18. Valentino Bertoli E. membrane Curatola G, Increased heterogeneity in stimulated human granulocytes. FEBS Letters 234: 451, 1988.
8.
Sulfhydryl
Proteins
and the Engram
G. W. RAFTER Department of Biochemistry, WV 26506, USA
West Virginia University
School of Medicine,
Morgantown,
Abstract - A scheme, which is selectional not instructional, is described for the molecular basis of the engram. It uses as reactants intermediates formed during the catabolism of neuronal membrane sulfhydryl proteins. It is proposed that these compounds undergoing intermolecular association generate a large number of diverse structures, some of which are selected for further complex formation. The basis of selection is the binding of a particular compound to membrane lipid, which enhances its reaction with another bound compound. Because the composition and configuration of membrane lipids are influenced by environmental factors, the scheme implies that incoming sensory signals are matched with endogenously generated chemical structures.
its biological context. More recently, the synapse as the physical basis of memory has held sway. In this hypothesis connections between synapses of cells are specified genetically, but the final wiring pattern depends on selective stabilization during neural activity (2). A form of stabilization, sometimes called synaptic plasticity, is long-term potentiation that is provoked by brief stimulation of certain presynaptic axons, and which leads to an increase in the effectiveness of synaptic transmission (3). In these systems the involvement of a small number of neurons is implied, which is in striking contrast to the large systems approach. According to one author (4) the boundary between the two approaches can be drawn at the level of about six neurons. Put another way ‘no conceivable neuron or set of neurons, no matter how diffuse it’s synaptic inputs, can evaluate the enormous amount of
Introduction
How the brain acquires and stores information is one of the most baffling and important problems in science. Because chemical structure and biological functions are intimately linked, it is axiomatic that chemical change must accompany the process. Thus far, the molecular nature of the engram or memory trace has eluded investigators. During the 1960’s much research time and money was spent to implicate protein and nucleic acid in the process, but the results were inconclusive. In a thoughtful retrospect of this period, Irwin (1) attributes some of these failures to the investigator’s isolation of the process from
Date received Date accepted
30 January 1989 27 February 1989
309
310
MEDICAL
neural activity shown to be involved in retrieval of even a simple form discrimination’ (5). This was concluded from the finding that between 5 to 100 X lo6 neurons were activated as detected by double-labeled metabolic maps during presentation of a visual cue to a split-brain cat. Likewise, the current mechanistic molecular approach to learning is not consonant with a system which is selective and degenerate and that accounts for the distributed nature of memory and its ability to recall events associatively in temporal sequence (6). In this paper a scheme utilizing catabolism of sulfhydryl proteins is proposed, which appears to be more in keeping with the biological aspects of higher brain function. Some considerations
in formulating
the proposal
Some of the ideas that influenced the selection of sulfhydryl protein metabolism as a molecular mechanism for learning or memory follow. Memory has been defined as a process that carries stored events in the brain forward in time (7). This means that generation of ‘memory molecules comprises aspects of both positive and negative entropy. Previously, processes characterized by nonlinear chemical kinetics and cross catalysis have been described in which a state of greater order (dissipative structures) can be generated from a state of lower order (8). The energetic rationale for this result is that entropy production of open systems can be divided into two parts: entropy flow due to exchanges with the environment and entropy production due to irreversible processes within the system. While the entropy production of the system is always zero or positive, the entropy flow can have.either sign. That neurons generate a diversity of chemical structures, some of which are selected for by incoming or reentrant signals, fits well with the molecular mechanism just described. Both Pringle (9) and Jerne (10) have previously discussed selective mechanisms for memory. The sulfhydryl proteins are uniquely suited for this role; the total number of isomeric structures formed from the oxidation and association of two light and two heavy immunoglobulin chains has been calculated to be 4.8 x 1016 (11). It is proposed that the participants in the learning process are not unique to it, but comprise membrane sulfhydryl-disulfide proteins with recognizable biological function, for example, cell surface receptors (12, 13) which because of
HYPOTHESES
their complex interaction during catabolism take on a new, unique biological role. The proposal
In presenting the proposal reference is made to insulin, which was used as a simple model protein to explore the hypothesis. Insulin upon reduction undergoes intermolecular sulfhydryl-disulfide exchange and ultimately forms insoluble complexes (14). The reaction proceeds with loss of free energy, because once the restrictive influence of the disulfide bonds are removed, the molecule seeks a more favorable minimum free energy state. When first synthesized and folded from a precursor molecule, it presumably was at a minimum free energy state, but this was lost in the case of insulin after removal of the connecting peptide by limited proteolysis. Similar reactions of multichain neuronal membrane proteins, as well as single chain proteins, which have undergone limited proteolysis, are suggested to form a large number of intermolecular complexes. The kinetically availability of the sulfhydryl and disulfide groups of the reactants would be expected to affect the products formed. Experimentally, this result is seen in the effect of urea on complex formation. Insulin incubated with reduced glutathione forms a high molecular weight soluble complex of about 70 000 while in the presence of urea the complex does not form (Fig. 1). Likewise, during the oxidation of the reduced B-chain of insulin a large series of higher molecular compounds are formed, while only one compound, the oxidized B-chain, is formed in the presence of urea (15). It appears that with of conformational constraints, inremoval tramolecular disulfide bond formation is favored over intermolecular disulfide bond formation. The kind and number of products formed will also depend, in part, on the kind and number of reactants. This, in turn, depends on factors that influence their formation such as availability of Ca++ to activate neutral proteases and the redoxpotential of the tissue. Thus the association of intermediate catabolic products of a small number of neuronal membrane proteins could conceivably generate a large number of discrete molecular species. And since their formation is constrained by the conformation of reactants, as well as by the metabolic state of the cell, the process is chaotic rather than random. But how is order (information) extracted from this chaotic process? It is proposed that some of the catabolic products are selected as reactants
SULFHYDRYL
PROTEINS
311
AND THE ENGRAM
0.3
0.3
0.2
0.2 0
is
% a
a? 0.1
0.1
4.0
9.6 ml
4.8
9.6 ml
Fig. 1 Gel filtration of insulin incubated with reduced glutathione in the presence of urea. Mixtures contained in 1.0 ml of 0.05 M Tris-HCI buffer, oH 7.7 2 me of bovine insulin (Sigma Chem Co., St. Louis, MO.)-and 6.5 mM glutathione. They were incubated 3 hrs at 30°C and then passed through a Sephadex G-75 column (0.9 x 15 cm) with 1% acetic acid and fractions collected. O--O 0.36 gms of urea added at end of incubation. o--o 0.36 gms of urea present during incubation.
Fig. 2 Gel filtration of insulin incubated with reduced glutathione inthe presence of lipopolysaccharide. Conditions as described in Fig. 1 except mixtures were incubated 2.5 hrs. l --0 no additions. o--o 100 Dg of lipopolysaccharide (E. coli, Difco Lab., Detroit, MI.).
for synthesis of new complexes, while others not selected are degraded to their constituent amino acids. It is proposed that the basis for the selection process is the binding of the compound to a membrane lipid and its subsequent reaction with another bound compound to form a new structure. For the compounds selected, because of their structure, the lipid serves as a concentrating media, which promotes the association reaction as it is very concentration dependent (16). In the insulin system this result is seen as enhancement of the formation of a high molecular weight complex by the addition of bacterial lipopolysaccharide to mixtures (Fig. 2). Because this effect of lipopolysaccharide is not seen in 1 hr reaction mixtures, it appears that some intermediate compound, not monomeric insulin, is the immediate precursor for the synthesis of the high molecular weight complex. It is the ‘selective value’ of each reactant with respect to its further anabolism that results in self organization in the proposed process (17). In turn, the process is influenced by environmental factors such as incoming or reentrant sensory signals which altered the composition and configuration of neuronal membrane lipids (18). In summary, the proposed scheme suggests that sensory signals are matched, using membrane lipid as an inter-
mediary, with endogenously structures.
generated
chemical
Conclusions
That the synthesis of the postulated intermolecular complexes continues throughout the life of the organism, acquiring information as a consequence of the ‘valued information’ of some structures for further reaction, prompts several interesting questions. Since all biological molecules with the exception of DNA turnover in some finite time, then how is information retained in some cases for a lifetime? That the complexes become supramolecular aggregates and are sequestered in the lipid environment of a neuronal membrane would be expected to greatly reduce their susceptibility to proteolysis. Nevertheless, since generation of dissipative structures is a dynamic process, to retain this property some turnover would be expected. Then does the synthetic process possess aspects of selfreplication making the overall system redundant? A second question arises as a consequence of the commonality in the biochemical components among species of the proposed mechanism. Then what is the molecular basis of the differences in cognitive skills between species? Differences in
312 the variety and number of structures, from which selection is made for further anabolism, is suggested to be important to explain these In other words, the ‘valued differences. information’ of the reactants. Finally, because of the commonality of components in different species, some degree of cognitive skills is expetted in a variety of species.
References 1. Irwin L N. Fulfillment and frustration. The confessions of a behavioral biochemist. Perspect Biol Med 21: 476, 1978. J-P, Danchin A. Selective stabilization of 2. Changeux developing synapses as a mechanism for the specification of neuronal networks. Nature 264: 705, 1976. M B. Synatic memory molecules. Nature 3. Kennedy 335: 170, 1988. 4. Rose D. Is brain research dead? Trends Neurosciences 10: 196, 1987. 5. John E R, Tang Y, Brill A B. Young R. Ono K. Doublelabeled metabolic maps of memory. Science 233: 1167, 1986. V B. The Mindful Brain. The 6. Edelman GM, Mountcastle MIT Press, Cambridge, MA, 1978. imagery and memory, p. 1669 7. Holstead W C. Thinking, in Handbook of Physiology, Neurophysiology. Vol 111. Am Physiol Sot. Washington, DC, 1960.
MEDICAL HYPOTHESES Priaoaine I. Stengers I. Order out of Chaos. Bantam Book;. Toronto, i984. Prinele J W S. On the oarallel between learninr and 9. evol;tion. Behaviour 3: 174, 1951. lo, Jertfe N K. Antibodies and learning: Selection versus instruction. p. 200 in The Neurosciences, (G C Quarton, T Melnecheck. F 0 Schmitt, eds) Rockefeller U Press. New York.1967. 11. Wang Z-X. Ju M, Tsou C-L. Numbers of ways of joining S H groups to form multi-peptide chain proteins. J Theor Biol 124: 293, 1987. 12. Malbon C C, George S T, Moxham C P. Intramolecular disulfide bridges: avenues to receptor activation? Trends Biochem Sci 12: 172, 1987. K. Murase H, Murayama M, Matsuda M. 13. Kamikubo Miura K. Evidence for disulfide bonds in membranebound and solubilized opeoid receptors. J Neurochem 50: 503, 1988. F, Goldberger R F. Anfinsen C P. 14. Givol P, DeLorenzo Disulfide interchange and the three-dimensional structure of oroteins. Proc Natl Acad Sci 53: 676, 1965. R C, Carpenter F H. Oxidation of the 15. Paynovich sulfhydryl forms of insulin A-chain and B-chain. lnt J Peptide Protein Res 13: 113. 1979. protein modifications and 16. Light A. Protein solubility, orotein folding. Bio Techniaues 3: 298. 1985. of matter and the evolutien of 17. Eigen M. Selfirganization biological macromolecules. Naturwissenschaften 10: 465. 1971. M, Governa M, Gratton E, Fiorini R, 18. Valentino Bertoli E. membrane Curatola G, Increased heterogeneity in stimulated human granulocytes. FEBS Letters 234: 451, 1988.
8.