Ca²⁺ in the Cleft: Fast and Fluorescent

Calcium ion (Ca²⁺) is a universal second messenger that governs a vast array of biological phenomena, including muscle contraction, neuronal transmission, fertilization, aging, cell death and hormone secretion¹. Perturbations of intracellular Ca²⁺ signaling underlie a host of pathologic states, including the myocardial diseases of ischemia/reperfusion injury, arrhythmia, myocyte hypertrophy, and heart failure².

Despite the pivotal role of Ca²⁺ homeostasis in myocyte physiology and disease, large gaps exist in our understanding of its compartmentalization within subcellular microdomains and its trafficking among them. A vast literature has glossed over the finely tuned handling of intracellular Ca²⁺, its sites of storage, release, action, and reuptake, focusing instead on crude measures of bulk Ca²⁺ concentration. In so doing, the intricacies of spatiotemporal handling of Ca²⁺ – occurring at the scales of milliseconds and nanometers – are blurred, and our understanding is incomplete.

Appreciation of the pivotal role of Ca²⁺ in heart function dates to the late 19^th century, when Sidney Ringer discovered that this cation is absolutely required for cardiac mechanical function³. Cardiac contraction is triggered by influx of a small amount of Ca²⁺ through voltage-gated L-type Ca²⁺ channels (LTCC) embedded in the cell surface membrane. This Ca²⁺ influx, in turn, triggers release of much larger amounts of Ca²⁺ from sarcoplasmic reticulum (SR) stores through ryanodine receptors (RyRs), a process termed Ca²⁺-induced Ca²⁺ release (CICR). Thus, LTCC Ca²⁺ influx gates CICR; in the converse sense, Ca²⁺ released by CICR feeds back to control LTCC influx⁴. The end result is an elaborate, finely tuned, and self-regulated cascade of events that controls every heart beat.

This intricate dynamics of this process occurs within, and is facilitated by, the microarchitecture of the junctional cleft, the region within the myocyte separating the SR and intramyocyte invaginations of the cell membrane (T-tubules)⁵. This zone, also termed the dyad, represents a volume of ~10⁻³ fL in cardiac muscle⁶, with ~12 nm separating the LTCC within the T-tubule from the RyR where SR Ca²⁺ release takes place⁷.

Under steady-state conditions, Ca²⁺ entering through the LTCC is extruded back out of the cell by the Na⁺/Ca²⁺ exchanger⁸. SR Ca²⁺ stores are replenished by the SERCA pump. Ca²⁺ concentrations in cellular microdomains, and their time-dependent changes, are also governed by Ca²⁺ buffering proteins, including calsequestrin and S100A1, and Ca²⁺ storage and release in mitochondria.

As complex as this picture is already, evidence is emerging that dyads differ in their structure and molecular architecture⁹. In a variety of heart diseases, T-tubule and dyadic junction remodeling are seen, which very likely participates in disease pathogenesis and distorted Ca²⁺ release⁷. For all these reasons, it is of obvious importance to glean a more sophisticated understanding of Ca²⁺ handling within the junctional cleft, ideally with the spatial and temporal resolution required by the underlying biology.

Ca²⁺ sensors

Using Ca²⁺-sensitive bioluminescent probes and fluorescent indicators, whole cell Ca²⁺ concentration ([Ca²⁺]_i) during both systole and diastole have been studied extensively. The first experiments visualizing intracellular Ca²⁺ transients were reported in frog heart¹⁰ and canine Purkinje fibers¹¹ using the Ca²⁺-sensitive bioluminescent protein aequorin. As aequorin is membrane impermeant, these and subsequent experiments were performed by injecting protein purified from the jellyfish Aequorea directly into cells. The associated technical challenges, coupled with modest signal-to-noise ratio afforded by the sensor, limited the utility of this approach.

Next, development of small molecule chemical fluorescence indicators with high Ca²⁺ affinity and fast kinetics, such as Fura-2¹², as well as use of acetoxymethylated esters for non-invasive cell loading, were breakthroughs that expanded the feasibility of Ca²⁺ imaging in single cardiomyocytes¹³. Later, the advent of confocal microscopy and fast, high-contrast fluorescein- and rhodamine-based Ca²⁺ indicators, such as Fluo-3, led to characterization of Ca²⁺ sparks, the result of single-dyad Ca²⁺ events¹⁴.

Whereas development of these diffusible indicators greatly facilitated our understanding of cardiac EC coupling, they afforded neither the submicron spatial resolution nor the rapid on- and off-kinetics required for a complete understanding of this intricate Ca²⁺ biology. It is estimated that Ca²⁺ ions diffuse within myocytes at a rate of ~100 μm/s¹⁵ and typically traverse a distance of ~1.8 μm¹⁶. Thus, conventional fluorescent Ca²⁺ indicators substantially underestimate local peak Ca²⁺ levels and report a low-resolution profile of local Ca²⁺ gradients¹⁷.

The next leap in Ca²⁺ imaging technology was introduction of genetically encoded Ca²⁺ indicators (GECIs). GECIs offer advantages not achievable with conventional chemical Ca²⁺ probes, including precise targeting within cellular micro(nano)-compartments, prolonged imaging over days and months, and simultaneous multicolor Ca²⁺ imaging¹⁸. Among the most popular protein calcium sensors are the GCaMPs, which are variants of enhanced green fluorescent protein (cpEGFP) coupled to a C-terminal Ca²⁺ sensor (calmodulin) and a calmodulin-binding M13 peptide at the N terminus¹⁹. Upon Ca²⁺ binding, the calmodulin moiety interacts with the M13 peptide, triggeringstructural reorganization in proximity to the cpEGFP, deprotonation of the chromophore, and increased fluorescence²⁰. GCaMPs are rapidly emerging as valuable tools for in situ monitoring of Ca²⁺ activity in neurons²¹ and heart²².

In this issue of the journal, Despa et al²³ describe two novel Ca²⁺ sensors specifically targeted to the junctional cleft. One harbors GCaMP2.2, a T203V mutation of GCaMP2 with intensified brightness and dynamic range²¹. The second, GCaMP2.2_Low, is a variant with reduced Ca²⁺ affinity, harboring a D133E mutation. Each construct was targeted to the dyadic cleft by means of covalent coupling to the N-terminus of FKBP12.6, a protein which binds with high affinity and selectivity to RyRs. Both fusion proteins retained high affinity for RyRs (K_d ~ 15-45 nmol/L), albeit somewhat lower than native FKBP12.6 (1 nmol/L)²⁴. A striated pattern of fluorescence, competed away by FKBP12.6, was observed in cultured adult rat ventricular cardiomyocytes, suggesting that both FKBP12.6-tagged sensors target selectively to the junctional cleft.

What have these new probes taught us?

GCaMP2.2_Low manifests baseline fluorescence and a dynamic range similar to GCaMP2.2, but its affinity for Ca²⁺ is 10-fold lower (K_d 5 μmol/L vs. 450 nmol/L). These investigators capitalized on these differences to quantify the high [Ca²⁺] expected in the junctional cleft. Prior to this work, estimates of Ca²⁺ transient amplitudes near Ca²⁺ release sites ranged between 10 μM⁸ to 100 μM²⁵, and peak amplitude was estimated to occur 5 10 ms after depolarization, with recovery occurring within 15-50 ms^{8, 26}. In contrast, the global Ca²⁺ transient within the cytoplasm is many-fold lower, reaching a maximum of ~1 μM within ~70-100 ms⁸. Thus, GCaMP2.2_Low, with its low Ca²⁺ affinity, could provide reliable measures of Ca²⁺ transients within the junctional cleft, whereas GCaMP2.2, with its high Ca²⁺ affinity, is optimal for measuring diastolic [Ca²⁺]_I.

With these newly engineered probes in hand, these investigators first calibrated the targeted and untargeted sensors in the myocyte environment, which reduced affinities by ~2-fold. However, FKBP12.6-targeting did not alter the Ca²⁺ affinity of either. Next, they set out to measure Ca²⁺ dynamics, reporting that [Ca²⁺]_Cleft reaches higher levels (1.3 μM) with faster kinetics than global [Ca²⁺]_Bulk (0.5 μM at peak). However, estimates of peak [Ca²⁺]_Cleft reported with this sensor are much smaller than previous estimates, and the time to [Ca²⁺]_Cleft peak (46 ms) is also much slower (5 10 ms). GCaMPs manifest significantly slower response kinetics (typically τ_on 20 ms-1.4 s and τ_off 0.4-5 s) compared with synthetic BAPTA-based dyes (e.g. Oregon Green 488 BAPTA 5N, τ_on <1 ms)²⁷. As acknowledged by the authors, the very rapid rises and falls in [Ca²⁺]_Cleft almost certainly exceed the kinetics of the GCaMP sensor. In this respect, it is highly likely that the low Ca²⁺ affinity, cleft-targeted GCaMP2.2_Low significantly under-estimated peak Ca²⁺ transients.

That being said, these novel sensors are ideally suited to measuring diastolic [Ca²⁺]_Cleft. As Ca²⁺ leak from the SR has been implicated in a variety of disorders, including heart failure²⁸, such measures are of great interest. To address this, the authors compared the signal emitted by GCaMP2.2 ([Ca²⁺]_Bulk) with that emitted by GCaMP2.2-FKBP12.6 ([Ca²⁺]_Cleft), noting that [Ca²⁺]_Cleft was nearly 2-fold higher than [Ca²⁺]_Bulk (194 μM versus 100 μM). Next, by selectively interrupting Ca²⁺ flux through LTCC versus RyRs, the authors went on to provide evidence that this resting diastolic gradient of Ca²⁺ stems from RyR leak, rather than sarcolemmalCa²⁺ flux.

The future is fast and bright

Whereas these ingenious, cleft-targeted probes have provided novel insights already, challenges remain, especially given their slow kinetics. Recently, a new generation of fast GCaMPs was introduced²⁷. One of them, GCaMP6f, is reported to have fast kinetics (the fastest among all currently known GCaMPs), high affinity, and robust dynamic range; it was successfully implemented to measure single neuron synapse events²⁹. Shang et al³⁰ fused GCaMP6f with junctin or triadin and reported high resolution images of junctional Ca²⁺ at individual dyads, so-called “Ca²⁺ nanosparks”. Nevertheless, GCaMP6f is still not fast enough to accurately track the extremely rapidly rising and fast-decaying Ca²⁺ gradient in the junctional cleft³¹.

Whereas GCaMP2.2_Low – FKBP12.6 may not provide accurate measures of junctional cleft Ca²⁺ transients during excitation, GCaMP2.2 – FKBP12.6 is an ideal sensor to measure [Ca²⁺]_Cleft during diastole. Because heart failure and cardiac arrhythmias are often associated with dyssynchronous Ca²⁺ leak in diastole, this probe is likely to prove useful in dissecting underlying mechanisms. Also, high diastolic [Ca²⁺]_Cleft may play an important role beyond governing E-C coupling by regulating local signaling cascades (e.g. calcineurin or CaMKII)³². As a result, these clever new tools are likely to be informative in multiple domains of cardiovascular biology.