Expanding the Concepts in Protein Structure-Function Relationships and Enzyme Kinetics: Teaching using Morpheeins

Abstract

A morpheein is a homo-oligomeric protein that can exist as an ensemble of physiologically significant and functionally different alternate quaternary assemblies. Morpheeins exist in nature and utilize conformational equilibria between different tertiary structures to form distinct oligomers as a means of regulating their function. Notably, alternate morpheeins are not misfolded forms of a protein; they are differently assembled native states that contain alternate subunit conformations. Transitions between alternate morpheein assemblies involve oligomer dissociation, conformational change in the dissociated state, and reassembly to a different oligomer. These transitions occur in response to the protein's environment, e.g., effector molecules, and represent a new model of allosteric regulation. The unique features of morpheeins are being revealed through detailed characterization of the prototype enzyme, porphobilinogen synthase, which exists in a dynamic equilibrium of a high activity octamer, a low activity hexamer, and two dimer conformations. Morpheeins are likely far more common than previously appreciated. There are, however, both intellectual and experimental barriers to recognizing proteins as morpheeins. These barriers derive from the way we were taught and continue to teach about protein folding, protein purification, protein structure-function relationships, and enzyme kinetics. This article explores some of these limitations and encourages incorporation of morpheeins into both introductory and advanced biochemistry classes.

It is becoming increasingly apparent that protein dynamics are important to protein function at many structural levels and this paper addresses protein dynamics at the level of quaternary structure assembly. Decades ago it was proposed and widely accepted that one amino acid sequence produced one structure [1]. As the number of protein crystal structures proliferated, there arose a fairly rigid, crystalline view of protein structures. However, the work of many laboratories in the past two decades has revealed the importance of dynamics in protein structure-function relationships (e.g. [2-5]). For a given protein, dynamic behavior has been documented at the level of 2°, 3°, and 4° structure. For heteromeric assemblies of proteins, it is now common to think of dynamic ensembles that associate and dissociate, for instance in response to cellular signals. However, even in this dynamic context, a given protein is generally illustrated as having just one stable native structure. We believe that this view is unnecessarily limiting.

The goal of this article is to encourage teachers of protein structure-function relationships to introduce the concept that the quaternary structure of a homomeric assembly need not be invariable. Protein oligomers can come apart, the individual components can change conformation, and the protein can reassemble differently. The remarkable aspect of this phenomenon is that different assemblies can have different functions. We have introduced the term “morpheeins” to describe this dynamic assembly/disassembly behavior. Morpheeins are homo-oligomeric proteins whose function is controlled through a reversible transition between alternate non-additive quaternary structure assemblies [6]. A defining characteristic of morpheeins is that the interconversion between species involves homo-oligomers dissociating (disassembling) and there being a conformational change in the dissociated form that does not allow reassociation to the same oligomer. The alternate oligomers are not required to have different subunit stoichiometry. The required oligomer disassembly step is what differentiates the morpheein model for allosteric regulation from the classic Monod, Wyman, Changeux (MWC, or concerted) and Koshland, Nemethy, Filmer (KNF, or sequential) models [6-8]. In the case of the prototype morpheein, which is porphobilinogen synthase, the conformational change allows reassociation to an alternate oligomer of different stoichiometry. This article discusses a new idea in protein structure function relationships, specifically morpheeins, and suggests how morpheeins can be incorporated into graduate, undergraduate, and perhaps even high school education in biochemistry

What is a morpheein?

A morpheein is as a homo-oligomeric protein that can exist as ensemble of physiologically significant and functionally different alternate quaternary assemblies (Fig. 1). The various assemblies illustrated in Fig. 1 are called morpheein forms. The functional difference between/among morpheein forms allows morpheeins to serve as a structural basis for allosteric regulation of protein function [6]. An ensemble of morpheein forms can occur when a protein has more than one stable native conformation in a dissociated state, and these alternate stable native conformations can assemble into functionally distinct quaternary structures. As illustrated schematically in Figure 1, the different monomeric conformations exist in dynamic equilibria with each other, and conversion between different quaternary assemblies proceeds via dissociation to the fundamental units prior to structural rearrangement. The fundamental units can be single subunits but may be larger. In the exemplary case of porphobilinogen synthase, the fundamental unit is a homo-dimer, which can exist in two different conformations, and the alternative oligomers are a high activity octamer and a low activity hexamer.

Figure 1. A two-dimensional schematic of an equilibrium of morpheein forms.

Four possible conformations of the monomer are shown in cyan (A), green (B), lilac (C), and pink (D). A protein that functions as a morpheein need only adopt two of these monomer conformations. Following a “rule of engagement” where oligomerization must occur by an association of a dashed line with a thick line, the subunits can self-engage to monomers of A or oligomerize to dimers of B, trimers of C, or tetramers of D. Monomer A is comparable to what have previously been termed “auto-inhibited” conformations of some proteins that are active as dimers.

Morpheeins challenge a central dogma

A primary barrier to recognizing morpheeins from experimental results is the teaching that proteins fold into a fixed and unique quaternary structure assembly. Although protein flexibility and conformational change are common themes in protein function, in general we do not presume that homo-oligomers readily dissociate, undergo a conformational change, and then reassociate (reassemble) to a structurally and functionally distinct assembly. The general assumption that a protein has one correct stable assembly follows the “one sequence, one structure, one function” paradigm (Fig. 2). Admittedly, this paradigm is increasingly being challenged as we learn about the functional importance of intrinsic disorder [9] and conformational diversity [10]. Nevertheless, once a protein is established to have a particular quaternary assembly, such as a tetramer, the protein is generally presumed to exhibit all of its physiologically relevant functions as a tetramer. In other words, protein conformational changes are generally implied to occur within the context of a single stable quaternary assembly. One example of this is the standard illustration of the MWC [8] vs. KNF [7] models for allostery, both of which are routinely illustrated in terms of a fixed oligomer, and which will be discussed in more detail later.

Figure 2. Morpheeins challenge protein folding paradigm.

The similarities and differences of the classic protein folding paradigm (left), versus the morpheein model (right).

Morpheeins do not challenge the rules that govern protein folding

The folding of a polypeptide chain is driven by a combination of entropic and enthalpic forces that generally result in the burial of hydrophobic residues, the distribution of hydrophilic residues between the surface and the interior, and formation of inter- and intra-molecular forces such as a network of hydrogen bonds in well-defined secondary structure elements. A folded polypeptide is commonly thought of as a stable, low energy unit; the morpheein model does not contradict this. The conformational changes that occur during morpheein interconversions do not involve denaturation and de novo refolding of the polypeptide chain; rather, subtle rearrangements of secondary structure elements have a net impact on the tertiary structure of the monomer, resulting in assembly to a different oligomer. For example, in porphobilinogen synthase, the prototype morpheein described below, each subunit contains 330 amino acids and the bulk of the secondary and tertiary changes that occur in the hexamer to octamer transition are in the N-terminal 24 amino acids.

The prototype morpheein: porphobilinogen synthase

The best-characterized morpheein is the enzyme porphobilinogen synthase (PBGS), which catalyzes the condensation of two molecules of δ-aminolevulinic acid to form porphobilinogen (the first common step in tetrapyrrole biosynthesis, e.g. heme, chlorophyll, vitamin B₁₂). PBGS exists as an equilibrium between a high activity octamer, a low activity hexamer, and two conformations of a dimer (Fig. 3 A); the interconversion between the higher order assemblies takes place via a conformational change at the dimer level. X-ray crystal structures have been solved for the hexameric (PDB: 1PV8) and octameric (PDB: 1E51) assemblies of human PBGS [11]. Each PBGS monomer is composed of an αβ-barrel with an extended N-terminal arm, and the conformation of this arm dictates higher-order oligomerization. The crystal structures show that the hexamer is comprised of three “detached dimers” in which the N-terminal arms extend outwards, away from the core, while the octamer is comprised of four “hugging dimers” in which the N-terminal arm of each monomer hugs the barrel of the adjacent monomer (Fig. 3 B). When the structure of a detached dimer is compared to that of a hugging dimer, the αβ-barrels superimpose almost perfectly (Fig. 3 C); differences in a few phi/psi angles of the backbone result in the different N-terminal arm positions that give rise to the alternate oligomers. In the octameric assembly, each subunit has three different surface-to-surface interactions with adjacent subunits in the oligomer. In the hexameric assembly, each subunit has two different surface-to-surface interactions with adjacent subunits in the oligomer. The surfaces that interact with each other exclusively in the octamer are solvent exposed in the hexamer. Not surprisingly, these surfaces are hydrophilic in character and their subunit-to-subunit interface is characterized by salt bridges and ordered water molecules. It is possible that hydrophilic interfaces in homo-oligomeric assemblies may portend the possible existence of morpheein forms.

Figure 3. The prototype morpheein, porphobilinogen synthase.

(A) The main quaternary assemblies of human porphobilinogen synthase are a low-activity hexamer (light and dark blue) and a high-activity octamer (light and dark pink). Each assembly is shown as viewed from the top with the subunits shown as spheres. (B) The morpheein equilibrium of human PBGS involves the hexamer, the octamer and two structurally distinct dimer conformations. The dissociated hexamer is shown as a detached dimer, which is in equilibrium with a hugging dimer, which associates to the octamer. The dimers are shown as cartoons. The detached and hugging dimers are shown in the context of the hexamer and octamer, respectively, with the remaining subunits shown as spheres. (C) The detached dimer (light and dark blue) superimposes almost perfectly with the hugging dimer (shades of pink), with the exception of the N-terminal arm (24 of a total 330 amino acids); the orientation of this arm guides higher order oligomerization. Note that the illustrated dimer conformations correspond to the asymmetric units of crystal structures PDB codes 1PV8 and 1E51 (hexamer and octamer respectively); we have presented alternative dimeric assemblies as the solution structures of the dimer configurations [12].

The alternate oligomeric assemblies of human PBGS have been observed in solution, and the position of the oligomeric equilibrium can be altered by manipulating factors such as pH, enzyme concentration, active site ligand concentration, or point mutations [6, 11-14]. When first revealed, the remarkable four-part human PBGS morpheein equilibrium was unprecedented and actually quite surprising since these authors were taught to think of quaternary structure assembly as both stable and additive. But, it is now established that the alternate morpheein forms of PBGS interconvert through a process that involves oligomeric disassembly, conformational change, and reassembly to an alternate oligomer. One of the remarkable aspects of this discovery is that the Jaffe laboratory had been working with the PBGS family of enzymes for nearly two decades prior to recognition of its morpheein forms. We knew this protein to be a highly thermostable homo-octamer (e.g. retaining 90 % activity after five days at 37 °C [15]), though both the mammalian and plant proteins had proven resistant to a high quality crystal structure determination. The solution of a quality crystal structure is dependent upon achieving a well-packed crystal that diffracts to a high resolution; this requires a homogeneous protein sample. The interchanging morpheein forms of PBGS are predicted to impair crystallization of a single morpheein form.

Morpheeins may be overlooked during traditional and tagged protein purification strategies

Prior to the introduction of covalent appendages to facilitate protein purification (purification tags), the standard approach for purification of an enzyme from a cellular lysate typically involved a multi-stage fractionation scheme that eliminated other proteins from the sample while retaining the protein of interest [16]. This approach is still in use and a reasonable hypothetical purification scheme is as follows: after disruption of the source cells/tissues, a centrifugation step is used to eliminate insoluble components, a salt cut is used to fractionate proteins with different solubility properties, and a series of column chromatography steps are used to separate the proteins into discrete fractions based on their surface charge, hydrophobicity, or size. Our favorite scheme uses an ammonium sulfate fractionation, a hydrophobic column, an anion exchange column, and a final size exclusion column into a buffer where the protein is known to be stable (e.g. containing required metal ions, reducing agents).

Students are typically taught that after each separation step, the fraction(s) containing the protein of interest are identified by a PAGE procedure (SDS, native, or blot) or an activity screen, and the identified protein peak is pooled for the next purification step while the remainder of the sample is discarded. While this approach is effective, and has certainly withstood the test of time, it involves the assumptions that the surface charge, hydrophobicity, and/or size of a given enzyme are constants such that all of the physiologically relevant forms of the protein-of-interest will chromatograph identically. In the case of an ensemble of morpheein forms in equilibrium with each other, all of these assumptions may be false.

Unfortunately, this standard approach to protein purification disfavors the discovery of morpheeins. For instance, if the protein-of-interest is discovered in two chromatographic peaks by PAGE analysis, but the associated activity is only identified in one peak, the “other peak” is usually discarded as a misfolded form or a non-physiological aggregate. In the case of PBGS, the hexamer and octamer differ sufficiently in surface charge to be separated by ion exchange chromatography, and the size difference allows at least partial separation by gel filtration chromatography [14]. Prior to characterization of the inactive PBGS hexamer, this form was discarded rather than identified as a physiologically relevant form of the enzyme that can be activated under appropriate conditions. With any novel purification scheme, it is prudent to examine every peak carefully before discarding what at first may appear to be a misfolded form or aggregate of the protein of interest.

Among the wonders of molecular biology was the dramatic shift from traditional purification methods described above to the use of purification tags, such as the now ubiquitous GST-tag and His-tag. In this approach, the DNA sequence coding for the protein of interest is modified so that the resulting protein is produced with a “tag” of amino acids attached at one terminus. Depending on the system used, the tag can be a small number of residues, or larger than the protein of interest itself. These tagged proteins are then purified using a resin that specifically binds the tag, which can be either cleaved after purification (in the case of large tags) or left in place with the assumption that a few extra residues at the N or C terminus will not impact the function of the protein of interest. In fact, entire industries have been built on the development of technologies to introduce tags to proteins, to use these tags for protein purification, and also (sometimes) to remove tags post-purification. However, in the case of an equilibrium of morpheein forms, the use of a tag for purification will mask the potentially valuable chromatographic separation of morpheein forms. Furthermore, the use of a tag might also shift the morpheein equilibrium toward or away from one of the alternate oligomeric forms, potentially locking the protein in an inactive form which might be taken as “misfolded”.

Anomalous kinetic behavior can arise from the behaviors of an ensemble of morpheein forms

Non-Michaelis Menten kinetics

That an enzyme is a morpheein can explain deviations from Michaelis-Menten kinetics, kinetic hysteresis, and a protein concentration dependence to an enzyme's specific activity. In fact, observation of such anomalous kinetic behavior can be a clue that an enzyme may function as a morpheein. In the standard approach for determining kinetic constants such as K_m and V_max, we generally instruct that to get good kinetic values, one determines the rate while varying substrate concentration within the range of the K_m. But what if the protein exists as an equilibrium of morpheein forms where one assembly has a K_m value of 20 μM and the other assembly has a K_m of 5 mM, and both forms are present under assay conditions? Determination of both K_m values requires varying substrate concentration over at least three orders of magnitude, which is not currently a standard procedure. Furthermore, the results will not be described by standard Michaelis-Menten kinetics for a single enzyme species (Fig. 4 A). At low substrate concentration the activity results predominantly from the low K_m morpheein form; but as substrate concentration approaches the K_m of the higher K_m morpheein form, it becomes a significant contributor to the overall rate of product production. On first glance, this situation could be mistaken as indicative of negative cooperativity; Segel's classic text on enzyme kinetics describes just such a situation for a mixture of isozymes with different kinetic constants [17]. In either case, the appropriate equation for fitting the kinetic data is the sum of two hyperbolic equations. However, the apparent V_max obtained from such a fit for each morpheein form is actually the product of the mole fraction of that morpheein form and the true V_max for that morpheein form. To obtain the latter value, one requires an independent assessment of the mole fraction of each morpheein form [12, 18].

Figure 4. Examples of atypical kinetics that can be associated with morpheeins.

(A) In a classic enzyme-catalyzed reaction, the catalytic rate increases as a hyperbolic function of substrate concentration (grey squares and dashed line) system. In the event that two kinetically distinct forms of an enzyme coexist, the data (black circles) cannot be fit to a single hyperbola (dotted black line). Instead, the data fit to a double hyperbola (solid line), where the sum of two Michaelis-Menten equations describes the contributions from both species. (B). In a classic enzyme-catalyzed reaction (dashed grey line) the product in creases linearly with time as long as substrate is saturating. For an enzyme that exists as an equilibrium of morpheein forms (solid black line), a lag time can reflect an inactive form dissociating, equilibrating, and re-associating into a more active form in the presence of substrate. Alternatively, in the presence of an inhibitor, a rapid initial rate can slowly approach a final steady state rate as the ensemble of morpheein forms reequilibrates to the less active assembly (not shown). (C) The activity per enzyme unit (specific activity) is a fixed property that does not vary with enzyme concentration for standard enzymes (dashed grey line). In a morpheein system in which a higher protein concentration favors a more active assembly, the specific activity increases as a function of enzyme concentration (solid black line). Alternatively, if the lower stoichiometry assembly is more active, one might see an inverse relationship between enzyme concentration and specific activity (not shown).

Kinetic hysteresis

Kinetic hysteresis is another potential indicator of an equilibrium of morpheein forms (Fig. 4 B). Kinetic hysteresis is a slow change in the rate of an enzyme catalyzed reaction in response to a rapid change in reaction conditions (e.g. addition of ligand, change in pH) [19]. This phenomenon has typically been attributed to slow conformational changes as well as experimental factors such as viscosity or mixing effects. We propose that another viable explanation for a slow approach to steady state may be that an enzyme (as it is added to the reaction mix) is predominantly in one morpheein form (e.g. low activity) and that turnover in the new environment triggers dissociation, conformational change, and reassembly to an altered morpheein form (e.g. high activity). Following a hysteretic approach to a new equilibrium of morpheein forms, the reaction proceeds at the expected linear rate until substrate becomes depleted. The slow hysteretic activation of one human PBGS construct has been demonstrated to arise from a hexamer to octamer transition, the time scale of which is about 0.017 per minute (half-life ∼ 1 h), depending on conditions (e.g. variant, pH, etc.) [14].

Enzyme-concentration dependent specific activity

Yet another characteristic that can be diagnostic of a morpheein is a specific activity that varies as a function of enzyme concentration (Fig. 4 C). The specific activity (typically expressed in μmoles product formed/unit time/mg enzyme) is generally taught to be a fixed property of an enzyme and is often a criterion applied to pooling column fractions during a purification (see above). In the case of a morpheein system such as PBGS where the higher-order oligomer is the more active form, the specific activity increases as a function of enzyme concentration (as in Fig. 4 C) since the octamer is favored over the hexamer under these conditions. This phenomenon has been documented for PBGS from plant and bacterial species [20].

Morpheeins provide a new structural basis for allostery

From a historical perspective, fixed oligomeric stoichiometry was a deliberate and simplifying assumption of the development of both the Monod, Wyman, Changeux (MWC, or concerted) and Koshland, Nemethy, Filmer (KNF, or sequential) models for allosteric regulation [7, 8]. With few exceptions this assumption has become a part of the fabric of how we think and teach about protein structure. The basic premises for each of these models are illustrated in terms of the relevant ensemble of structures in Figure 5 A and B; allosteric regulation in a morpheein system is introduced in Figure 5 C. Similarities in all three models are that the enzyme subunits can exist in both a less active and a more active conformation, and that a regulatory molecule influences which conformation the subunits assume. The MWC model requires that all subunits of one oligomer are in the same conformation, while the KNF model allows for an oligomer composed of mixed subunits; both models assume that the conformational change occurs within the context of the fixed oligomeric stoichiometry. The required oligomer disassembly step differentiates the morpheein model for allosteric regulation from the classic MWC and Koshland models [6-8]. In the morpheein model, dissociation of the subunits is required for the conformational change to occur. The new subunit conformation then drives assembly to a different oligomeric state. Like the MWC model, the morpheein model requires all of the subunits of one oligomer to exist in the same conformation. Most importantly, the morpheein model explains why all of the subunits of one oligomer must be in the same conformation. In the morpheein model for allosteric regulation, it is not possible to have an asymmetric assembly where one subunit is different from all the other subunits. They just don't fit together!

Figure 5. Morpheeins introduce a third model for allostery.

(A) In the Monod, Wyman, Changeux (MWC, or concerted) model of allostery, each oligomer exists in the inactive (composed of round subunits) or active (composed of square subunits) states. The oligomers are in equilibrium with each other, and binding of an allosteric activator (shown as jagged black squares), draws the equilibrium towards the active state. (B) In the Koshland, Nemethy, Filmer (KNF, or sequential) model of allostery, the individual subunits of the oligomers exist in the inactive (round) or active (square) state. Binding of an allosteric activator to a subunit within an oligomer induces a conformational change of an adjacent subunit to the active state. (C) In the example of the morpheein model, allosteric control involves dissociation of the oligomers. The inactive form is a trimer of pie-wedge subunits, the active form is a tetramer of square subunits; the trimers, pie-wedges, squares, and tetramers all coexist in equilibrium. The allosteric activator binds to the square subunits and draws the equilibrium towards the active tetramer. For example, in the case of plant PBGS, the allosteric activator is a magnesium ion [11].

We have introduced the interconversion of morpheein forms as a third general mechanism for allosteric regulation of protein function [6]. Furthermore, we have proposed that the morpheein mechanism for allosteric regulation may be relatively common, though under-recognized because oligomer dissociation runs counter to the traditional view of stable quaternary structure assemblies [6]. As before, we invite readers to review data on proteins they may have purified and characterized with an eye toward clues that these proteins may be regulated through the allosteric control of an ensemble of morpheein forms.

The existence of morpheeins expands the concept of auto-inhibition beyond the monomer – dimer situation

Figure 1 shows a two dimensional schematic describing the general phenomenon of morpheein equilibria. As the fundamental unit of the PBGS morpheein equilibrium is a dimer, this system is analogous to the equilibrium between tetramers and trimers. However, the general schematic also includes the previously established phenomenon wherein a protein can exist in a three-part equilibrium among an autoinhibited monomer that cannot dimerize, a conformation of the subunit that can dimerize, and the dimer. A prime example is the epidermal growth factor (EGF) receptor [21]. In that case the monomer has two possible structures; one is a closed form that cannot dimerize and the other is an open form whose most stable assembly is dimeric. The EGF binding site is present in the open form of the monomer, but not the closed form of the monomer. EGF binding stabilizes the open form of the receptor, thereby promoting dimer formation. Many cell surface receptors may function this way; this is a distinctly different mechanism than the one wherein the dimeric form of the receptor is tethered together by the effector molecule (Fig. 6). One can consider the autoinhibited monomer as the “self-associated” morpheein form, and can draw a correlation between the dimer and various domain swapped dimeric structures for which there are now many examples including ribonucleaseA and catalase [22].

Figure 6. Dimeric cell surface receptors.

(A) The insulin receptor is composed of monomers in its inactive state. The binding site for insulin exists at the dimerization interface, and the binding of a single insulin molecule tethers the monomers into a dimer, turning on the signal. The insulin receptor does not function as a morpheein. (B) The EGF receptor also exists as monomers in the inactive state, but the monomers equilibrate between a closed (auto-inhibited) structural form that cannot oligomerize, and an open structural form that can. Dimerization is mediated by insertion of the dimerization arm (shown in red) from one monomer into a pocket on the other monomer. This pocket only exists in the open structural form; in the auto-inhibited form, the dimerization arm self-associates with its own monomer. The receptor binding site is distant from the oligomeric interface. Binding of EGF stabilizes the open form and draws the oligomeric equilibrium to the dimer, turning on the signal. The EGF receptor functions as a morpheein.

Morpheeins are not misfolded states; rather, they are akin to functionally distinct chemically modified forms of a protein, but without the chemical modification

The idea that alternate non-additive quaternary assemblies are “misfolded” is particularly, if misleadingly, “obvious” if one form is inactive. However, the inactive state is a well known and physiologically relevant form of many proteins. A more common way to think about the transition between a more or less active form of a protein is to couple the structural transition to a chemical modification event, such as phosphorylation or adenylylation. In those cases, the chemical modification stabilizes one quaternary structure assembly and draws the ensemble of structures toward that stabilized form. In the case of morpheeins, no such chemical modification occurs. Rather, the alternate assemblies are chemically identical and the equilibrium between forms is shifted by environmental factors such as pH or non-covalent ligand binding. In the case of morpheeins, alternate assemblies may be considered metastable structures, similar to the metastable or multistable protein structures (including phosphofructokinase and β-galactosidase) suggested nearly 40 years ago [23, 24], which have been largely overlooked. A more recent example of metastable protein forms of the multi-drug resistance protein MDR-1 has recently been suggested to arise from variations in the kinetics of protein expression [25].

Morpheeins should not be confused with isozymes

Different morpheein forms of the same protein share some commonalities with isozymes. For instance, different morpheein forms catalyze the same chemical reaction but may be characterized by different K_m and V_max values. Consequently it is important to differentiate morpheeins from isozymes. A morpheein is a protein that can exist in functionally and structurally distinct morpheein forms. Isozymes are two different proteins (different primary sequence) that can catalyze the same reaction (usually with some functional distinctions).

Morpheeins are distinct from other examples of alternate quaternary structures such as the amyloid forming proteins and quasiequivalent virus coat proteins

The assembly of alternate oligomers plays a role in the pathologic process of amyloid plaque formation in Alzheimer's disease and prion infections like scrapie and Creutzfeld-Jacob disease. Like morpheeins, these examples also involve a single protein that can take on different conformations leading to alternate oligomeric states. The critical difference between morpheeins and amyloid-forming proteins is that the oligomeric interconversions of morpheeins are non-pathologic, reversible, and limited to oligomers of finite stoichiometry (Fig. 1). The proteins involved in amyloid plaque formation, however, assemble irreversibly into insoluble polymers of indefinite size.

The assembly of one protein into alternate quaternary assemblies has also been demonstrated for some icosahedral virus coat proteins where the structural variation has been called quasiequivalence [26, 27]. However, these quasiequivalent forms differ from morpheeins in that the coexistence of the different forms in a single assembly is necessary for function, the ratio of the alternate oligomers must be fixed to build the higher-order geometric viral coats, and there is no evidence for a dynamic interchange of subunits between alternate oligomeric forms. Furthermore, the assembly of the oligomers of the viral coat proteins does not occur spontaneously.

Morpheeins as potential drug targets

The emergence of multi-drug resistant organisms drives the continued need to discover or develop antibiotics that function via novel mechanisms. The current golden age of structural genomics has yielded thousands of detailed X-ray crystal structures of proteins. Initially, there were expectations that this wealth of structural information would lead directly to drugs custom-designed to block the active sites of, and thereby inhibit, medically relevant enzymes. With a few exceptions, this approach has not been fruitful.

A critical disadvantage to active-site targeted inhibitors is the high level of conservation of enzyme active sites across species. Most of the classical antibiotics function by targeting enzymes that are specific to prokaryotes because targeting a shared enzyme would likely harm the human host. The discovery of morpheeins has revealed a previously unforeseen mechanism to target universally essential enzymes for species-specific drug design and discovery. An inhibitor directed to trap a component of a morpheein equilibrium is not targeted to the enzyme active site, but rather to a surface site that is unique to one oligomeric assembly; these surface sites are much more phylogenetically diverse than active sites. A morpheein-based inhibitor would function by binding to and stabilizing the inactive morpheein form of the enzyme, thereby shifting the equilibrium to favor that form (Fig. 7). One possible example is a recently discovered inhibitor that stabilizes a dimeric conformation of tumor necrosis factor α [28], for which the well characterized quaternary structure assembly is a trimer [29]. A second example is a newly described class of inhibitors for the human immunodeficiency virus integrase enzyme, which exists in an equilibrium of dimers, tetramers and higher-order oligomers. A recently discovered class of small peptide inhibitors termed “shiftides” functions by shifting the oligomeric equilibrium towards the inactive form [30-32]. It is not established that either tumor necrosis factor α or HIV integrase fulfill the criteria we have set forth for morpheeins, though the data are consistent with this model. We have recently described small molecule stabilization of the hexameric PBGS assembly as a new strategy for development of antibiotics or herbicides [33]. We propose that morpheeins provide an exciting new avenue for drug discovery and suggest that the concept of targeting morpheeins be introduced in drug discovery courses.

Figure 7. Targeting morpheeins for drug design/ discovery.

For a morpheein that exists as an inactive trimer formed of pie-wedge subunits and an active tetramer formed of square subunits, an inhibitor (yellow wedge) would function by binding to the pie-wedges or trimers, and drawing the equilibrium towards the inactive form.

What proteins besides PBGS might function as morpheeins?

It is unlikely that the morpheein paradigm for control of enzyme function (allosteric regulation) is limited to the prototype, PBGS. Our growing understanding of the dynamic quaternary structure equilibrium observed for PBGS serves as a basis for targeted searches of the literature to look for enzymes that display some of the kinetic hallmarks of an enzyme that can exist in multiple quaternary forms. One can also look for reports of an enzyme existing in more than one quaternary assembly (either in solution, or in crystal structures). Approximately two dozen such proteins have been identified from literature searches thus far (E.K. Jaffe, unpublished data), and ongoing structural and kinetic analyses of several of these are focused towards establishing them as morpheeins.

Key Concepts

• Morpheeins are homo-oligomeric proteins whose function is controlled through reversible transitions between alternate non-additive quaternary structure assemblies.

• Morpheeins can deviate from classical Michaelis-Menten kinetics.

• Morpheeins provide a structural basis for allostery. The required oligomer disassembly process differentiates the morpheein model of allostery from the classic MWC and Koshland models. The morpheein model also accounts for why all of the subunits of a given oligomer must be in the same conformation.

• Hydrophilic subunit interfaces may indicate that an oligomeric protein can readily dissociate and possibly adopt alternate morpheein forms.

• Morpheeins are distinct from isozymes because isozymes have different primary structures. Morpheeins are distinct from amyloid-plaque forming proteins because morpheein forms are of finite stoichiometry. Morpheeins are distinct from quasiequivalent virus coat proteins because alternate forms interconvert spontaneously in the absence of a scaffold or chaperone.

• Trapping of alternate morpheein oligomers provides a novel mechanism of drug action.

• The number and variety of proteins that function as morpheeins are unknown.