If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
* This work was supported, in whole or in part, by National Institutes of Health Grants NS-049134, DK46828, and GM28222. The on-line version of this article (available at http://www.jbc.org) contains supplemental “Experimental Procedures” and Fig. S1. 1 Recipient of a National Institutes of Health Ruth L. Kirschstein National Research Service Award.
The structure and unfolding of metal-free (apo) human wild-type SOD1 and three pathogenic variants of SOD1 (A4V, G93R, and H48Q) that cause familial amyotrophic lateral sclerosis have been studied with amide hydrogen/deuterium exchange and mass spectrometry. The results indicate that a significant proportion of each of these proteins exists in solution in a conformation in which some strands of the β-barrel (i.e. β2) are well protected from exchange at physiological temperature (37 °C), whereas other strands (i.e. β3 and β4) appear to be unprotected from hydrogen/deuterium exchange. Moreover, the thermal unfolding of these proteins does not result in the uniform incorporation of deuterium throughout the polypeptide but involves the local unfolding of different residues at different temperatures. Some regions of the proteins (i.e. the “Greek key” loop, residues 104–116) unfold at a significantly higher temperature than other regions (i.e. β3 and β4, residues 21–53). Together, these results show that human wild-type apo-SOD1 and variants have a partially unfolded β-barrel at physiological temperature and unfold non-cooperatively.
is the most common motor neuron disease in the United States, and the occurrence is predominantly sporadic and without a known cause. Approximately 10% of cases are familial, however, and 20% of these familial cases are caused by mutations in the gene that encodes the antioxidant enzyme copper-zinc superoxide dismutase (SOD1) (
). Over 90% of these mutations encode single amino acid substitutions but some encode insertions, deletions, and C-terminal truncations. A large amount of experimental evidence has shown that mutations of SOD1 induce ALS by imparting a toxic function to the protein that is hypothesized to be an increased propensity to misfold and self-assemble into oligomeric structures (
There now are >120 known ALS mutations of SOD1 that give rise to a similar clinical pathology. ALS mutations have diverse effects on the properties of the SOD1 polypeptide. Some decrease the thermal stability of folded SOD1, whereas others increase thermal stability (
). The destabilization of a protein's native fold or a reduction in cooperativity can render hydrophobic residues or hydrogen bond donors and acceptors more available for non-native intermolecular interactions that initiate and propagate aggregation (
In this work, we describe our investigation of the solution structure of three ALS variants of SOD1 (A4V, G93R, and H48Q) and human wild-type (hWT) SOD1 by measuring their rates of amide hydrogen/deuterium exchange (HDX) with mass spectrometry (
). We were able to document the exchange of hydrogen on the backbone amide and not the more rapidly exchanging functionalities such as alcohol or amine groups. Our goal was to test the hypothesis that ALS-linked mutations that are located in different regions of the protein can lead to a common region of structural perturbation. This similarity has been observed for various pathogenic variants of lysozyme that cause amyloidosis (
Recent work from our laboratory has shown that SOD1 can convert to amyloid fibrils under physiologically relevant conditions and that a small amount of disulfide-reduced hWT apo-SOD1 can initiate fibrillation in apo and partially metalated forms of SOD1 (
). We have also observed that more stable forms of SOD1 (i.e. disulfide-intact apo-SOD1) that do not initiate aggregation do, however, become incorporated into the propagating fibril after initiation.
We are therefore interested in identifying regions of disulfide-intact apo-SOD1 (WT and ALS mutants) that might facilitate the intermolecular interactions that occur during the propagation of fibrillation. We are also interested in studying the structural properties of SOD1 during thermal unfolding as opposed to chaotrope-induced unfolding. The high throughput capabilities of mass spectrometry enabled us to examine the structure and unfolding of all four proteins quickly using a standard protocol.
We have chosen to study these three ALS variants based upon several criteria, including the location of each amino acid substitution in the three-dimensional structure of the folded protein, and based upon how the variants fit into our previous classification of ALS mutations into “metal-binding region” and “wild type-like” (
). The A4V substitution occurs at the dimer interface (Fig. 1), distal to the metal-binding region; this WT-like substitution does not diminish the metal-binding abilities of SOD1. There are many ALS substitutions like this, but we chose A4V because it is the most common substitution that is encountered in SOD1-linked ALS cases in North America. The G93R substitution is also a WT-like protein, but the substitution occurs distal to the dimer interface; the effects of substitutions at Gly93 are important for two more reasons. First, Gly93 is the most highly substituted residue in SOD1-linked ALS. Six different mutations result in the substitution of Gly93. Second, Gly93 is thought to function as a “β-barrel plug” that is critical to maintaining the native structure of folded SOD1. The H48Q substitution involves a histidine that normally plays a role in binding the copper ion found at the active site; this metal-binding region substitution greatly diminishes the ability of SOD1 to coordinate Cu2+ at this site and will likely prevent the protein from fully maturing into its holo form.
Because each of the three selected ALS variants is routinely isolated from yeast expression systems with different stoichiometric equivalents of Cu2+ and Zn2+ cofactors (
), it is necessary to demetalate each variant to study the apo forms of these proteins. Studying the metal-free protein is important because apo-SOD1 represents the least stable form of SOD1 and possibly the state that is most prone to aggregation in vivo; moreover, studying the proteins that are at the same state of metalation (e.g. metal-free) allows a valid comparison of how each amino acid substitution affects the structure of the protein separately from how each substitution might affect metal coordination.
We report that similar regions of the β-barrel are locally unfolded in both hWT and ALS variant SOD1 proteins. For example, each ALS variant protein, as well as the hWT protein, incorporates deuterium rapidly at residues 21–53 at physiological temperatures, whereas other regions of the β-barrel (i.e. residues 7–20) remain protected from deuteration even at elevated temperature. These locally unfolded β-strands could participate in intermolecular interactions that initiate or propagate the aggregation of SOD1.
Measuring the thermal unfolding of hWT SOD1 with global HDX and mass spectrometry yielded results that are quantitatively consistent with thermal unfolding experiments measured with DSC (
); an analysis of thermal unfolding of SOD1 with both methods revealed that the protein unfolds cooperatively near 52 °C. Interestingly, in global melt experiments, hWT SOD1 does not appreciably exhibit a fully exchanged form until the incubation temperature reaches 52 °C. In contrast, all three mutant proteins examined in this work indicate that a substantial population has undergone complete exchange at a temperature ∼6 °C below their respective melting temperatures; this complete exchange of hydrogen with deuterium at temperatures so far below the Tm suggests that the ALS mutant proteins unfold more non-cooperatively than the hWT protein.
Measuring the HDX of local regions of SOD1 with acid quenching and proteolysis showed, however, that the thermal unfolding of SOD1 is not entirely cooperative: residues 21–53 of all four proteins begin to unfold at a lower temperature than residues 104–116 or 7–20. In addition, other regions such as 117–144 are never “folded” (even at 10 °C), as demonstrated by the rapid deuteration seen in all three proteins.
Structural Rigidity in the Greek Key Loop (Residues 104–116) and β2-Strand (Residues 7–20) of hWT and ALS Variant SOD1
Despite the fact that the intrinsic rate of amide hydrogen exchange is predicted to change by a factor of 100 as temperature is increased from 10 to 50 °C (
), the rate of HDX of residues 7–20 does not change for hWT SOD1 across this temperature range. This result indicates that, throughout the temperature range, this region of the hWT apoprotein remains structured with its residues folded into a stable hydrogen-bonding arrangement. Residues 7–20 and 104–116 seem to unfold “last” and therefore may function as linchpins or anchor points for the folded SOD1 protein. This finding is consistent with NMR structural studies on the monomeric Q133M2SOD protein that have shown that this region is characterized by low root mean square deviation values for both protein backbone and residue heavy atoms (
In both G93R and H48Q SOD1, residues 7–20 display a rate of HDX that is similarly independent of temperature as in hWT SOD1, i.e. the degree of deuterium incorporation is constant from 10 °C up to 37 °C, where a gradual sigmoidal decrease in protection begins. In contrast, the A4V variant exhibits similar temperature independence below 30 °C, but above 30 °C, deuterium begins to be more rapidly incorporated into residues 7–20. The low thermostability of A4V and the proximal location of the substitution to residues 7–20 explain this temperature dependence; the A4V protein has the lowest Tm value of the four proteins studied in this work: hWT Tm = 52.5 °C, H48Q Tm = 47.35 °C, G93R Tm = 44.3 °C, and A4V Tm = 40.5 °C (
hWT and ALS Variant SOD1 Have a Partially Unfolded β-Barrel at 37 °C
At physiological temperature, it must be remembered that there is no subpopulation of hWT apo-SOD1 that is fully deuterated at residues 7–20 or 104–116 (e.g. locally unfolded at residues 7–20 or 21–53). The hWT SOD protein that is becoming fully deuterated at residues 21–53 is therefore unfolding at residues 21–53 at a rate sufficient for the incorporation of deuterons, whereas other residues such as 7–20 and 104–116 are not unfolding at rates that allow the incorporation of deuterons into the backbone. We therefore refer to hWT SOD1 as being “partially unfolded.” Because the region that is partially unfolded includes two β-strands (β3 and β4) and because other β-strands (β2) are not partially unfolded and protected from HDX, we describe the hWT SOD1 protein as having a partially unfolded β-barrel at physiological temperature.
The abundance of hWT apo-SOD1 that has a partially unfolded β-barrel remains constant between 30 and 45 °C (e.g. the ratio of signal intensity of lighter and heavier peaks in Fig. 4 is a constant 3:1). The static nature of this bimodal mass distribution suggests that the local unfolding of the β-barrel at residues 21–53 is slower than the intrinsic rate of amide HDX for an unstructured polypeptide (∼103 min−1) at physiological temperature.
Similar to hWT SOD1, all three mutant SOD1 proteins examined contain a subpopulation that undergoes partial unfolding of the β-barrel so as to allow for complete isotopic exchange of residues 21–53. Nonetheless, the dynamics of this region in all three mutants is clearly distinguishable from that of the hWT protein. For example, residues 21–53 of G93R exchange similarly to hWT SOD1 below 37 °C, but the intensity for fully exchanged peptide dramatically increases in G93R above 37 °C such that the locally unfolded G93R protein is the predominant species at 40 °C. The temperature at which this type of transition occurs for each protein correlates with, but does not equal, the melting transition temperature for each protein that was measured previously with DSC (
). However, at this temperature, the unfolding process is sufficiently slow that the completely deuterated form never exceeds the protected form after 300 min in D2O. At higher temperatures, the localized unfolding of residues 21–53 becomes more pronounced in A4V SOD1. At 37 °C, for example, ∼50% of the A4V proteins are completely deuterated; at 40 °C, this fraction increases to ∼90%.
The metal-binding region mutant H48Q exhibits behavior in this region that is similar to that of G93R, A4V, and hWT SOD1. Approximately ∼15% of H48Q is fully deuterated at residues 21–53 at 37 °C; between 40 and 45 °C, a majority of H48Q SOD1 proteins are completely deuterated at residues 21–53.
We point out that a recent investigation by Agar et al. (
) into the structure of hWT and ALS mutant SOD1 proteins (including A4V) failed to detect the type of local unfolding that we detected herein with hWT, G93R, and H48Q apo-SOD1 and that we detected in our previous study of A4V apo-SOD1 (
). As already mentioned, this discrepancy could be due to the fact that the study by Agar et al. involved metalated A4V (1.62 eq of zinc and 0.22 eq of copper) and that the local unfolding that we observed in the β-barrel (the bimodal mass distribution that we observed for residues 21–53) might occur only in metal-free A4V SOD1. In addition to being metalated, the A4V protein was also studied at 4 °C in the study by Agar et al. Local unfolding of apo-A4V SOD1 has been detected at 4 °C (
); however, this temperature might be too low to observe local unfolding in the metalated form of A4V, which is expected to be more structurally rigid and thermostable than A4V apo-SOD1. Unfortunately, because we do not have access to the original mass spectra of the peptides from any protein, we cannot independently analyze the symmetry of the molecular ions of the peptides from the study by Agar et al.
The results presented here provide clues as to the regions of SOD1 that are sampling unfolded states at physiological temperature and that are potentially able to interact with other SOD1 subunits and/or be restructured during fibrillation. Specifically, subpopulations of hWT and mutant SOD1 show complete exchange in segment 21–53 at 37 °C, suggesting that the conformational flexibility in this region could allow, perhaps be necessary, for the ability of soluble SOD1 to self-assemble into oligomeric structures such as amyloid protofibrils. Preventing the local unfolding of the β-barrel of hWT or ALS variant SOD1 at residues 21–53 could be an important aim in the design of pharmacological agents that prevent the oligomerization and fibrillation of SOD1 in vivo (