Designing Proteins to Combat Disease: Cardiac Troponin C as an Example

Abstract

Throughout history, muscle research has led to numerous scientific breakthroughs that have brought insight to a more general understanding of all biological processes. Potentially one of the most influential discoveries was the role of the second messenger calcium and its myriad of handling and sensing systems that mechanistically control muscle contraction. In this review we will briefly discuss the significance of calcium as a universal second messenger along with some of the most common calcium binding motifs in proteins, focusing on the EF-hand. We will also describe some of our approaches to rationally design calcium binding proteins to palliate, or potentially even cure cardiovascular disease. Considering not all failing hearts have the same etiology, genetic background and co-morbidities, personalized therapies will need to be developed. We predict designer proteins will open doors for unprecedented personalized, and potentially, even generalized medicines as gene therapy or protein delivery techniques come to fruition.

Calcium – The universal second messenger

Both prokaryotes and eukaryotes universally utilize the coordination of Ca²⁺ to modulate biological functions. Typically the “resting” free intracellular concentration of Ca²⁺ within a living cell is approximately four orders of magnitude lower than that of the surrounding aqueous or interstitial environments [1]. Cells have developed a plethora of ion pumps, exchangers, pores, buffers and organelles necessary to maintain a low free intracellular Ca²⁺ concentration, generally in the range of tens to hundreds of nM. The myriad of protein systems used to work against this large chemical gradient and maintain intracellular Ca²⁺ at very low levels is energetically costly [2]. Ultimately, changes in both the extracellular, and especially the intracellular Ca²⁺ concentrations, can have a profound impact on cellular behavior [3].

The first demonstration that Ca²⁺ is essential for a biological process was discovered by Dr. Sydney Ringer circa 1883 [4]. In a series of experiments (and insightful mishaps in 1882), Ringer and his colleagues discerned Ca²⁺ was a necessary component of physiological solutions required to sustain cardiac muscle contraction. Very soon afterwards, extracellular Ca²⁺ was demonstrated to strongly influence several cellular behaviors and functions [3]. However, it took almost a century (and much debate) before the work of many biochemists, cell biologists and physiologists studying muscle clearly showed intracellular Ca²⁺ was a bona fide biological signal [5]. Today, it is widely acknowledged that every physiological and cellular change in our bodies beginning from the initiation of life to death, is either directly controlled or at a minimum, influenced by changes in intracellular Ca²⁺ [3, 6].

There are a vast number of plasmalemmal and intracellular receptors that upon receiving a stimulus trigger a rise in intracellular Ca²⁺ [2]. Depending on the integration of all the stimuli that a cell is receiving, the intracellular Ca²⁺ level may: 1) remain constant; 2) adjust the “baseline”; 3) pulse with wave fronts that may propagate through the cell (spiral, chevron, etc. that can collide with interference patterns); and 4) globally rise in amplitude with a particular duration and frequency of occurrence. The shape, amplitude and duration of the intracellular Ca²⁺ signal depends on the stimulus, its signal strength, frequency, cell type and integration with other signaling processes. In addition to the mechanisms that increase intracellular Ca²⁺, the ability of the cell to extrude Ca²⁺ and/or refresh the intracellular Ca²⁺ stores also plays a central role in shaping the specific Ca²⁺ signature of a stimulus [7]. There are also receptors and signaling pathways that instead of raising intracellular Ca²⁺, work just the opposite, and lower available Ca²⁺ within the cell [8]. Additionally, there are numerous Ca²⁺ sensors and buffers within different cells that also play a pivotal role in defining the amplitude and duration of the Ca²⁺ signal. Due to the expansive combination of receptors and Ca²⁺ handling machinery, there is a vast and wide range of Ca²⁺ signals that can be generated by different cell types and even within a particular cell. The contractile cells of the heart (myocytes) are an excellent example of a cellular system that routinely modulates its intracellular Ca²⁺ to dictate functional outcomes [9].

Calcium Signaling and Cardiac Muscle Contraction

It was the intense and deep study of the three types of muscle (cardiac, skeletal and smooth) that began to shed light on the universal significance and control mechanisms of Ca²⁺ signaling. Whereas the basic mechanisms that control the myocyte’s intracellular Ca²⁺ transient are well understood and can be found in most biochemical and nearly all physiological textbooks, there are specific details and nuances that are still being discovered [10]. In general, intracellular free Ca²⁺ rises and then falls in a highly coordinated and timely manner across every cardiac myocyte to control the rhythmic contraction and relaxation cycle of the heart. For example, it is generally thought that during ventricular filling (relaxed heart), the free concentration of Ca²⁺ within the myocytes is lowered and then maintained at ~70 to 125 nM. This low concentration of Ca²⁺ is below the threshold required to initiate contraction. Upon an action potential, cytosolic Ca²⁺ rapidly increases, reaching free concentrations of 0.3 to 0.5 µM and as high as ~1.5 µM during maximal stimulation [9]. The amplitude of the Ca²⁺ signal within the myocyte under normal conditions is sufficient to initiate less than half the maximal contraction of the myocyte or muscle preparation [11]. Thus, there is a large contractile reserve that can be tapped into to increase cardiac contractility (or vice versa).

The myocyte can alter its contractility through several neural-hormonal influences that change the Ca²⁺ signal in both amplitude and duration [9]. The archetype of this phenomenon is the “fight or flight” response on the heart. Interestingly, as early as 1928 the effects of adrenaline on the heart was demonstrated to be dependent on extracellular Ca²⁺ and suggested to function via changes in intracellular Ca²⁺ [3]. Both the amplitude and duration of the Ca²⁺ signal can also be adjusted by the frequency of stimulation (i.e. the force-frequency response) [12]. Ultimately, the behavior of nearly all the electrical and Ca²⁺ handling proteins in the heart can be post-translationally modified to tune the Ca²⁺ signal [13]. Depending on the composition of the electrical and Ca²⁺ handling proteins (as occurs in different species), the duration of the Ca²⁺ transient can be drastically different (although changes in amplitude are less pronounced) [9]. Besides contraction, cytosolic and more spatially constrained Ca²⁺ signals (as occur in caveolae) also influence energetics, transcription, hypertrophy and many other cellular functions [14]. Ultimately, all of these Ca²⁺-dependent processes require a Ca²⁺ sensor (i.e. Ca²⁺ binding protein) to either translate or relay the Ca²⁺ signal into a cellular response. There is plenty of evidence that at least the contractile response of cardiac muscle can change without corresponding alterations in the Ca²⁺ signal [13]. Thus, the heart can also modulate contractility via altering the Ca²⁺ sensing machinery of the myocyte [15]. Through protein engineering of Ca²⁺ binding proteins, we are attempting to modulate the response of the heart to Ca²⁺ in order to palliate, and potentially cure, various cardiovascular diseases.

Classes of Ca²⁺ Binding Proteins

There is an ample and diverse array of Ca²⁺ binding proteins used by cells to decode the information carried by the Ca²⁺ signal. Depending on the location, concentration and type of Ca²⁺ sensors in the cell, different cellular outcomes can arise. Although there are other types, three major classes of Ca²⁺ binding motifs have evolved independently to sense Ca²⁺ signals: the annexin fold, the C2 domain, and the EF-hand.

The Annexin Fold

The annexin family of proteins in eukaryotes bind negatively-charged phospholipids in a reversible, Ca²⁺-dependent manner. The roles for annexins in the cell are varied, including Ca²⁺ homeostasis, endocytosis, exocytosis, cell migration, and cellular scaffolding [16]. Several of the annexins are expressed in cardiac myocytes, where they play a role in modulating the Ca²⁺ signal [17]. Annexins have a highly variable N-terminal region that regulates membrane association, but all family members share a conserved "core" Ca²⁺-binding, C-terminal domain consisting of either four or eight repeats that form α-helices and bind Ca²⁺ with moderate affinity. The variable N-terminal domain is used in a number of protein-protein interactions including dimerization, and can also be post-translationally modified by myristoylation or phosphorylation. The effects of such post-translational modifications vary widely between annexins, as do the free Ca²⁺ concentrations required for their activity, allowing each annexin isoform to be tuned to respond to specific Ca²⁺ signals. Much is to be learned on how precisely this is accomplished. There are at least 160 different proteins that contain the annexin fold [16].

The C₂ Domain

Another type of Ca²⁺-binding motif is the C₂ domain, an 8-stranded antiparallel β-sandwich connected by variable loops that are thought to confer functional specificity. Three loops located at the top of the domain bind two to three Ca²⁺ ions cooperatively but with low affinity [18]. As for annexins, many proteins containing the C₂ domain also bind the cell membrane and in most cases their action is Ca²⁺-dependent. The C₂ domain is generally found in proteins involved in signal transduction or membrane trafficking. Phospholipase C and protein kinase C isoforms both contain C2 domains and are involved in several signaling cascades in the heart [19]. Throughout eukaryotes, more than 200 proteins have been found to contain the C₂ domain [20].

The EF-Hand Motif

By far, the most common Ca²⁺-binding motif in eukaryotes is the EF-hand (Figure 1) [21]. The canonical EF-hand is formed by a helix-loop-helix motif in which one Ca²⁺ ion is chelated by side chain oxygen atoms in a flexible 12-residue loop, and the flanking α-helices are oriented perpendicular to one another [22]. Most EF-hands occur in pairs connected by a short β-sheet, which contributes to the high cooperativity of the adjacent Ca²⁺ binding sites [23]. However, a small number of proteins contain an odd number of EF-hand motifs typically used for dimerization rather than Ca²⁺ binding [24].

Figure 1. The EF-hand is the most common Ca²⁺ binding motif used to decode the Ca²⁺ signal.

Panel A depicts the crystal structure of cardiac TnC with its three functional EF-hands bound by Ca²⁺ (grey balls). Each EF-hand is a simple helix-loop-helix, where the loop chelates Ca²⁺ with pentagonal bipyrimidal geometry mimicking Ca²⁺’s hydration shell. Panel B shows the most common amino acids found within the 12 residue EF-hand loop [22]. Note, the first and twelfth amino acids (red) are essential for Ca²⁺ binding, whereas an Asp at position 5 (green) increased Mg²⁺ binding.

The Ca²⁺ binding loop of the classical EF-hand coordinates Ca²⁺ with seven oxygens in a pentagonal bipyrimidal arrangement, mimicking the hydration shell around the solvated ion [23]. Residues at positions 1, 3, 5, 7, 9, and 12 in the loop work together to non-covalently chelate the Ca²⁺ ion. In most EF-hands, aspartate occupies position 1, whereas a glutamate in position 12 provides bidentate oxygen ligands (Figure 1). In addition, a highly conserved glycine in position 6 allows the loop to bend around the bound Ca²⁺ ion. The identity of the intervening amino acids can be highly variable; proteins that are identical at positions 1, 3, 6 and 12 may still have vastly different affinities [23].

Although the EF-hand motif is generally considered to be highly selective for Ca²⁺, the loop is also capable of binding other divalent ions and heavy metals with low to extremely high affinity. Furthermore, environmental conditions such as pH, ionic strength and temperature may or may not alter cation binding [23, 25]. In the cell, the most prevalent competitor for EF-hand sites is Mg²⁺, which is present at concentrations of up to 2 mM [26]. In contrast to Ca²⁺, Mg²⁺ has a smaller ionic radius and has a strong preference for coordination by six ligands in an octahedral geometry. Many EF-hand loops are flexible enough to adjust to the smaller Mg²⁺ but bind with 10³–10⁴ lower affinity compared to Ca²⁺. The lower Mg²⁺ affinity is primarily caused by a slower rate of Mg²⁺ association to the EF-hand due to a substantially slower rate of Mg²⁺ desolvation compared to that of Ca²⁺ [3]. In order for the cations to bind a protein, the water shell around the ion must be entirely or partially replaced by the chelating residues of the protein. Mg²⁺ is chelated by the same ligands as Ca²⁺ at positions 1, 3, 5, and 7 in the loop, however the glutamate at position 12 only provides a single oxygen ligand or is displaced entirely by a water molecule [22, 27]. Some EF-hand proteins undergo a large conformational change when Ca²⁺ binds, but not upon Mg²⁺ binding, and it has been suggested that this lack of bidentate coordination from the glutamate at position 12 is the cause [27].

EF-hand proteins play three major roles in the cell: structural integrity, Ca²⁺ buffering and/or Ca²⁺-dependent regulation (Figure 2). The well-known EF-hand containing proteins calmodulin (CaM) and troponin C (TnC) can play all three of these roles in muscle. Certain EF-hand proteins have Ca²⁺ and/or Mg²⁺ sensitivities below resting ion levels and thus are nearly always occupied by Ca²⁺ or Mg²⁺ [23]. In some of these cases, through cation dependent protein-protein interactions, the EF-hand protein remains bound and an integral, structural member of a larger protein system, as exemplified by the C-terminal domain of TnC [28]. Ca²⁺ buffering proteins help to control the free Ca²⁺ in a cell by their relatively slow Ca²⁺ exchange, while the regulatory proteins relay or transduce a Ca²⁺ signal into action. The archetype Ca²⁺ buffer, alpha parvalbumin, also has a relatively high affinity for both Ca²⁺ and Mg²⁺, but its structure does not appreciably change between the apo and Ca²⁺-bound states (Figure 2) [29, 30]. Thus, alpha parvalbumin is generally thought to only buffer Ca²⁺. This lack of protein-protein interaction may not be the case for the close relative of alpha parvalbumin, oncomodulin (a.k.a. beta parvalbumin) [31]. In contrast, the regulatory proteins, such as the N-terminal domain of TnC and both domains of CaM, generally undergo a large conformational change upon Ca²⁺ binding, which allows these proteins to interact with additional target proteins acting as “Ca²⁺-dependent switches”. Whereas CaM has hundreds of potential targets in a cell, TnC only binds a single target, troponin I (TnI) [32, 33]. Depending on the concentration of these Ca²⁺-dependent switches, they can also become substantial “sinks” or buffers themselves for rises in cellular Ca²⁺. For instance, the concentration of CaM in the myocyte is in the 10 µM range (4 Ca²⁺ binding sites), whereas TnC is in the 70 µM range (3 Ca²⁺ binding sites in cardiac isoforms and 4 in the skeletal isoforms). These non-negligible concentrations of switches in the heart have a substantial influence on the shape of the free Ca²⁺ transient [34]. Thus, the various cation binding properties of EF-hands allow proteins to perform structural, buffer and switch-like functions.

Figure 2. Ca²⁺-dependent switches and buffers in muscle.

Unlike parvalbumin, TnC and CaM expose a hydrophobic “pocket” upon Ca²⁺ binding that is then utilized to bind target proteins, TnI in the case of TnC and as an example, myosin light chain kinase (MLCK) for CaM. Only the regulatroy, N-terminal domain is depicted for TnC. Yellow amino acids are shown in yellow, whereas all other amino acids are shown in red with the target in blue.

Ca²⁺ Binding Properties of EF Hand Proteins

Of all the Ca²⁺ binding motifs, the EF-hand has been the most extensively studied. Utilizing this simple motif, different EF-hand proteins bind Ca²⁺ with affinities that can differ by over a million fold (nM to mM) [23]. Within the Ca²⁺ binding loop alone, there are hundreds of unique EF-hand sequences that can be found in nature (Figure 1B) [35]. Adding the variability of the helices to the mix, the number of unique EF-hand sequences grows into the thousands. To date, there are no algorithms that can predict the Ca²⁺ binding properties of EF-hand proteins, making it challenging to re-engineer or modulate function in a specific manner.

Due to the high variability within the loops, it was initially thought that the amino acid composition of the loop alone dictated the Ca²⁺ affinity. Although the loop sequence had a strong impact on the Ca²⁺ binding properties of isolated EF-hand peptides, one cannot simply exchange the Ca²⁺ binding loops between EF-hand containing proteins and expect the chimeric protein to adopt the Ca²⁺ binding properties of the parent protein [36–38]. However, this is not to say that the loop has no impact on the overall Ca²⁺ binding properties of the proteins. For instance, moderate to high affinity Ca²⁺ binding requires an Asp residue at the 1^st and a Glu residue at the 12^th positions of the loop (Figure 1) [39]. We and others have demonstrated that having an Asp residue at the 5^th position increases Mg²⁺ affinity for that EF-hand, decreasing overall Ca²⁺ sensitivity via increased Mg²⁺ competition (Figure 1) [36, 40, 41]. Depending on the protein, altering the other chelating residues can also have a subtle to moderate effect on the Ca²⁺ affinity [42, 43]. In the end, residues throughout the entire protein (yes, even outside of the EF-hand) can have a profound influence on Ca²⁺ binding [44].

Many EF-hand proteins alter their structure upon Ca²⁺ binding, which alters their interaction with one or more target proteins (Figure 2). In general, mutations that hinder the ability of the protein to change its structure upon Ca²⁺ binding can drastically impair Ca²⁺ affinity [45]. Protein-protein interactions add another complication to predicting the Ca²⁺ binding properties of an EF-hand protein. Typically the binding of an EF-hand protein to its target protein enhances the overall Ca²⁺ sensitivity of the system, often by orders of magnitude [15]. However, there are several examples where target interactions actually decrease the Ca²⁺ sensitivity of the EF-hand protein, as can occur with CaM binding to IQ motif containing peptides [46]. Unlike traditional CaM binding motifs that require Ca²⁺-bound CaM to initiate an interaction, IQ motifs typically bind Ca²⁺-free CaM [47]. We are also learning that target protein interactions away from the direct EF-hand interaction can also have a large impact on how the EF-hand protein senses and releases Ca²⁺ [28]. Cardiac TnC is an excellent example that clearly demonstrates Ca²⁺ binding to a single EF-hand is a function of the state of the system rather than solely determined by the composition of the EF-hand protein alone. That is, the Ca²⁺ binding properties of TnC can change over an order of magnitude going from the isolated protein, to the Tn complex, to the thin filament, to myofibrils, to muscle and in most cases depending if there is a strongly bound myosin attached to actin (Figure 3) [15]. Understanding how the system behaves has helped us to engineer TnCs that when combined with gene therapy approaches may one day be used to treat various cardiovascular diseases [48, 49].

Figure 3. Comparison of Ca²⁺ dissociation rates from cardiac TnC in systems of increasing complexity with that of muscle relaxation.

The rate of Ca²⁺ dissociaiton from the single regulatory EF-hand of cardiac TnC is extremely state dependent [15]. This figure highlights the fact that the loop sequence alone is not the dominating factor that determines the Ca2+ binding properties of an EF-hand or its system.

Reformulating Cardiac Troponin C

Cardiovascular disease is the leading cause of death throughout the world. This seems preposterous considering we know so much about how the heart and vascular systems function in both health and disease, from in vivo performance to the minute workings of the isolated proteins. Considering the importance of the Ca²⁺ signal to cardiac function and performance, a great deal of effort has been put into altering the Ca²⁺ signal to halt or reverse disease [50, 51]. However, since the Ca²⁺ signal controls so many processes in the myocyte (contraction only being one), long term alteration of the Ca²⁺ signal leads to detrimental consequences. Much of this results from altering the Ca²⁺ signal via modulating specific steps in the beta adrenergic pathway. Although enhancing the beta adrenergic response can acutely improve the quality of life, overstimulation and long term use leads to excessive remodeling, hypertrophy, increased risk of arrhythmias and shortening of life span [52]. Interestingly, in the process of developing pharmaceuticals to modulate the L-type Ca²⁺ channel, it was discovered that some of these compounds sensitized the contractile apparatus to Ca²⁺ through increasing the Ca²⁺ sensitivity of TnC [53, 54]. Since then, several compounds have been discovered that enhance contraction of the heart via TnC. Unfortunately, none of these compounds have high affinity or specificity for cardiac TnC, can prolong relaxation and begin to bind and alter the function of other proteins such as CaM and phosphodiesterases, limiting their efficacy and making them potentially dangerous [55, 56]. However, a more targeted pharmaceutical designed against fast skeletal TnC shows promise as a therapeutic for a plethora of skeletal muscle weakness [57]. Alternatively, we have taken a direct route to enhance Ca²⁺ binding of TnC through protein engineering [25, 48, 49].

Unfortunately, there is no off-the-shelf “cook-book” that one can peruse to quickly whip up an EF-hand recipe that will bind Ca²⁺ with specific kinetics and thus altered affinities. One of the only reproducible and highly effective ways to alter Ca²⁺ binding to classical EF-hands is to replace the first or last chelating residues in the loop with another amino acid (typically Ala) [39]. Although this is an excellent method to “destroy and inactivate” an EF-hand, it is not a satisfactory approach to retain the switch-like properties or fine tune EF-hand Ca²⁺ binding. There have been several hypotheses put forth to explain the diversity of EF-hand Ca²⁺ sensitivities and kinetics. One of the earliest was the acid-pair hypothesis that suggested the number and specific geometry of negatively charged residues within the chelating positions of the loop controlled Ca²⁺ sensitivity in a predictable manner [36]. Subsequently, the “gateway” hypothesis, suggested that the size and chemical properties of the ninth chelating residue in the loop controlled the kinetics of cation exchange [58]. Utilizing CaM as a model system, we directly tested these hypotheses and disappointingly demonstrated these particular rules were insufficient to predict mutational outcomes in Ca²⁺ sensitivity, selectivity or kinetics of the N-terminal domain of CaM [41, 42]. Interestingly, these mutations tended to alter the rate of Ca²⁺ association more so than the rate of Ca²⁺ dissociation, similar to mutations that increased the negative surface charge of another EF-hand protein, calbindin [59]. The short coming of these theories lie in not considering the diversity and complexity of how the other parts of the EF-hand protein (and binding partners) influence the loop residues and thus Ca²⁺ chelation. It will be extremely challenging, if not impossible, to establish a set of simple rules that all EF-hand proteins follow to control their Ca²⁺ sensitivities and kinetics. However, within a protein family such as CaM and TnC, there may be some general principles that can be discerned and manipulated [6, 25, 40–42, 44, 48, 49, 60–64].

As mentioned above, Ca²⁺-dependent switches typically display a large structural change upon Ca²⁺ binding (Figure 2). In general, the apo (and often Mg²⁺ bound) structure of the protein is compact, burying several hydrophobic amino acids within its core. Occluding hydrophobic residues from the aqueous environment is thought to be energetically favorable considering “oil and water do not like to mix”. However, upon Ca²⁺ binding, a large hydrophobic surface area becomes exposed to the cytosol, which is thought to be energetically costly to Ca²⁺ binding [65]. Thus, minimizing this oily patch was predicted to enhance Ca²⁺ binding, potentially keeping the “pocket” open for longer periods of time – precisely what we wanted to do in order to test whether the kinetics of Ca²⁺ dissociation from TnC played a role in modulating the rate of muscle relaxation (yes it can) [66]. We individually mutated every hydrophobic residue throughout the N-terminal, regulatory domain of skeletal TnC and found that although the Ca²⁺ sensitivity could be altered over three orders of magnitude, surprisingly, there was no correlation between the Ca²⁺ affinity and the solvent accessibility of the mutated residues in either the absence or presence of Ca²⁺ (or the difference between the two states) [44]. Thus, the change in solvent exposure of a hydrophobic residue is not the sole determinant in how a residue affects Ca²⁺ binding. Instead, altering the ability of the protein to achieve its “open” state upon Ca²⁺ binding, as well as stabilizing the beta sheet, specific subdomains within the protein and the hinges for structural movement play a more central role in setting the Ca²⁺ sensitivity of TnC (Figure 4). Recent, molecular dynamics simulations have confirmed some of our original hypotheses [67].

Figure 4. Designing Ca²⁺ sensitizing mutations into the regulatory domain of cardiac TnC.

In the apo state, there is a cluster of hydrophobic residues in the NAD and BC subunits of the regulatory domain of cardiac TnC that interact with one another to keep the hydrophobic pocket closed. Upon Ca²⁺ binding the residues in the two domains begin to come apart and are far removed from one another upon TnI binding. Individually exchanging these hydrophobic residues with polar Gln help to increase pocket opening upon Ca²⁺ binding, in some cases (L48Q) substantially increasing overall Ca²⁺ sensitivity that carries through upon TnI binding [64].

For the most part, the majority of foundational work that one would utilize to rationally design cardiac TnC has been collected with isolated proteins, or with proteins that negligibly change structure. TnC does not function in isolation, but is an integral part of the Tn complex anchored tightly to the thin filament. Thus, a functional cardiac TnC must be able to stably form the Tn complex, anchor to the thin filament and via changes in Ca²⁺ concentration, reversibly bind the C-terminal domain of TnI (which ultimately controls the position of tropomyosin and the strong binding of myosin to actin). The behavior of the reconstituted thin filament is one of the best and relatively simple biochemical systems where all these processes can be systematically manipulated. Furthermore, the Ca²⁺-dependent behavior of the thin filament is an excellent predictor of how a particular perturbation of the thin filament will alter mechanical force [64]. Using this system, we and others have demonstrated that post-translational modification of several of the thin filament proteins (even at several different locations within a particular protein) can either increase or decrease the Ca²⁺ sensitivity and kinetics of the thin filament and force development/relaxation in a like manner [61, 68–71]. Interestingly, many disease associated mutations aberrantly alter the Ca²⁺ binding properties of the thin filament [48, 62]. Thus, designer TnCs should change the force response of cardiac muscle without altering the normal regulation of the thin filament (primarily the beta adrenergic response, i.e. Ser 23/24 TnI phosphorylation).

Working with both CaM and skeletal TnC, we were able to build a substantial database of point mutations that altered the N-terminal Ca²⁺ binding properties of these proteins. Considering cardiac TnC is very similar in structure and amino acid sequence to CaM and skeletal TnC, we reasoned that analogous mutations in these protein systems would behave similarly. Excitingly, and with high fidelity, we were able to generate a series of cardiac TnC mutants with a broad range of Ca²⁺ sensitivities and kinetics [25, 48]. Through combining mutations, we were successful at designing TnCs that could be finely tuned with precise Ca²⁺ affinities that could be used to counteract the aberrant Ca²⁺ sensitivities brought on by disease associated mutations in the other troponin subunits (TnI and TnT). Interestingly, not all of the mutations that affected the isolated protein carried this property through to the reconstituted thin filament, highlighting the importance of studying systems with physiological significance that contain additional proteins, which too can influence Ca²⁺ binding [64].

Excitingly, we and others have recently demonstrated that a Ca²⁺ sensitizing mutation in cardiac TnC (L48Q) enhances the contraction of healthy and diseased cardiac myocytes, and more significantly, improves in vivo contraction therapeutically after a myocardial infarction [49, 72]. This can be achieved in mice without the detrimental consequences observed with other positive inotropes such as adversely affecting cardiac morphology, electrical activity, beta-adrenergic response, relaxation, diastolic function, intracellular Ca²⁺ or survival [49]. Studies are currently underway to test this gene therapy approach in a larger myocardial infarction model in sheep. Besides enhancing contraction, there are several cardiac diseases where decreasing cardiac contractility may be warranted, such as occurs in hypertrophic and restrictive cardiomyopathies. We have specifically engineered Ca²⁺ desensitized TnCs for these diseases as well [48]. With the expanding impact of viral gene delivery and gene editing (CRISPR) technologies as viable gene therapies, and the promise of alternative protein delivery systems, protein engineering is at the precipice of substantially impacting cardiovascular medicine.

Conclusion

This review highlights the universal nature of Ca²⁺ signaling and the importance of Ca²⁺ binding proteins to sense these signals and translate or relay cellular and physiological outcomes. Muscle research has been at the forefront of pushing these ideas to the broader biological fields. In an attempt to understand the governing rules that dictate Ca²⁺ binding to EF-hand proteins, we have designed several cardiac TnC constructs that have been demonstrated to have a therapeutic potential. By no means has our work been limited to TnC, as we have also designed several parvalbumins and CaMs that also have potential to help the failing heart and vasculature [73–75]. Considering the importance of Ca²⁺ sensing to all of life’s processes, it is not farfetched to begin to design Ca²⁺ binding proteins to physiologically and therapeutically augment any organ system.