PC2 (polycystin-2) forms a Ca2+-permeable channel in the cell membrane and its function is regulated by cytosolic Ca2+ levels. Mutations in the C-terminal tail of human PC2 (HPC2 Cterm) lead to autosomal dominant polycystic kidney disease. The HPC2 Cterm protein contains a Ca2+-binding site responsible for channel gating and function. To provide the foundation for understanding how Ca2+ regulates the channel through the HPC2 Cterm, we characterized Ca2+ binding and its conformational and dynamic responses within the HPC2 Cterm. By examining hydrogen–deuterium (H–D) exchange profiles, we show that part of the coiled-coil domain in the HPC2 Cterm forms a stable helix bundle regardless of the presence of Ca2+. The HPC2 L1EF construct contains the Ca2+-binding EF-hand and the N-terminal linker 1 region without the downstream coiled coil. We show that the linker stabilizes the Ca2+-bound conformation of the EF-hand, thus enhancing its Ca2+-binding affinity to the same level as the HPC2 Cterm. In comparison, the coiled coil is not required for the high-affinity binding. By comparing the conformational dynamics of the HPC2 Cterm and HPC2 L1EF with saturating Ca2+, we show that the HPC2 Cterm and HPC2 L1EF share a similar increase in structural stability upon Ca2+ binding. Nevertheless, they have different profiles of H–D exchange under non-saturating Ca2+ conditions, implying their different conformational exchange between the Ca2+-bound and -unbound states. The present study, for the first time, provides a complete map of dynamic responses to Ca2+-binding within the full-length HPC2 Cterm. Our results suggest mechanisms for functional regulation of the PC2 channel and PC2’s roles in the pathophysiology of polycystic kidney disease.

INTRODUCTION

ADPKD (autosomal dominant polycystic kidney disease) is one of the most common genetic kidney disorders, characterized by the progressive development of renal cysts, ultimately resulting in kidney failure [1,2]. Mutations in two genes, PKD1 and PKD2, account for almost all ADPKD cases [3]. These two genes encode polycystin-1 and PC2 (polycystin-2) respectively. The molecular mechanism linking protein function to disease pathogenesis is not well understood, although intracellular Ca2+ dysregulation caused by PC2 malfunction can lead to cyst development [4]. Understanding the connection between PC2 function and Ca2+ is crucial to our understanding of the pathogenesis of ADPKD.

Full-length PC2 is an integral membrane protein with six transmembrane helices and two cytosolic termini (Figure 1A). As a member of the TRP (transient receptor potential) channel family, PC2 also contains a pore-forming loop between the fifth and sixth helices (Figure 1A, labelled ‘P’). PC2 can oligomerize to form a Ca2+-permeable channel, and its channel activation is also regulated by cytoplasmic Ca2+ levels. The probability of the open conformation of the PC2 channel is Ca2+-dependent and follows a bell-shaped curve relationship with respect to increasing Ca2+ concentration [5].

The domain topology of full-length PC2 and its C-terminal tail

Figure 1
The domain topology of full-length PC2 and its C-terminal tail

(A) Domain topology of the full-length human PC2. PC2 contains six transmembrane helices and two cytosolic termini. The HPC2 Cterm contains a Ca2+-binding EF-hand and coiled-coil domain. Helices are numbered, and N- and C-termini are labelled. P, pore-forming loop. (B) Different domains are included in the HPC2 Cterm (top) and HPC2 L1EF (bottom) constructs. The residue numbers mark the start and finish residues of each construct, in the context of full-length human PC2 proteins.

Figure 1
The domain topology of full-length PC2 and its C-terminal tail

(A) Domain topology of the full-length human PC2. PC2 contains six transmembrane helices and two cytosolic termini. The HPC2 Cterm contains a Ca2+-binding EF-hand and coiled-coil domain. Helices are numbered, and N- and C-termini are labelled. P, pore-forming loop. (B) Different domains are included in the HPC2 Cterm (top) and HPC2 L1EF (bottom) constructs. The residue numbers mark the start and finish residues of each construct, in the context of full-length human PC2 proteins.

The C-terminal cytoplasmic tail of human PC2 (HPC2 Cterm) is one of the important components for functional regulation in the PC2 channel [57]. Many mutations, affecting multiple regions in the HPC2 Cterm, were shown to lead to ADPKD (see the Autosomal Dominant Polycystic Kidney Disease Mutation Database at http://pkdb.mayo.edu), suggesting that the HPC2 Cterm is key in both channel regulation and disease pathogenesis. The HPC2 Cterm contains one Ca2+-binding site, with a dissociation constant (KD) of approximately 22 μM [9]. The identified Ca2+-binding site is located within the EF-hand domain of the HPC2 Cterm (Figure 1, green), and is crucial for functional regulation of PC2 channel [10]. However, the measured Ca2+-binding affinity of the isolated EF-hand domain was notably weaker than the full-length HPC2 Cterm [7,11,12], suggesting that the rest of the HPC2 Cterm possibly affects Ca2+ binding within the EF-hand. The HPC2 Cterm also contains a downstream coiled-coil domain (Figure 1, blue) that mediates oligomerization of the HPC2 Cterm [9,13], whereas the isolated EF-hand is monomeric in solution [11,14]. To understand how the downstream coiled-coil domain and the oligomerization of the HPC2 Cterm affect Ca2+-binding of each individual EF-hand in the HPC2 Cterm, we have designed a monomeric construct, HPC2 L1EF, that includes the EF-hand and also its N-terminal linker 1 region (Figure 1B). Characterizing Ca2+ binding and its subsequent structural and dynamic effects in both the HPC2 Cterm and HPC2 L1EF can help us to understand the molecular mechanism of how PC2 channels react to different cytosolic Ca2+ levels.

Although isolated domains of the HPC2 Cterm have been studied previously [1215], the full-length HPC2 Cterm has yet to be structurally characterized. This is largely due to its high conformational flexibility, contributing to susceptibility to aggregation and degradation [9,10], which limits the application of traditional structural approaches. In contrast, HDX-MS (hydrogen–deuterium exchange mass spectrometry) allows local conformational monitoring of protein dynamics under physiological conditions [1618]. In the HPC2 Cterm, the changes in local structural stability in response to Ca2+ can be reflected in the H–D (hydrogen–deuterium) exchange profiles of the backbone amide protons [1921]. MS can identify relevant peptide sequences and accurately measure the mass increase due to deuterium uptake, and hence provide a peptide-specific H–D exchange profile [22,23]. HDX-MS has become an emerging approach for conformational characterization when traditional approaches are limited [24]. Using HDX-MS, we can characterize the H–D exchange profiles of the HPC2 Cterm at different Ca2+ levels, linking dynamic responses to Ca2+ binding.

In the present study, we characterized the conformational and dynamic responses to Ca2+ binding within the PC2 C-terminal tail. Using a combination of biophysical approaches, we show that, although the full-length HPC2 Cterm and the isolated HPC2 L1EF have similar Ca2+-binding affinities, they display different Ca2+-dependent conformational dynamics profiles. These results bridge the structural information obtained previously from isolated domain fragments to the full-length HPC2 Cterm. Such knowledge can serve as the foundation to better understand the PC2 channel and its regulation by the HPC2 Cterm.

MATERIALS AND METHODS

Protein constructs and cloning

The HPC2 Cterm construct (Ile704–Val968) was amplified using PCR from human PC2 cDNA and cloned into pET-28(a+) (Novagen) with an N-terminal His tag. The HPC2 L1EF construct (Ile704–Pro797) was synthesized by Genscript with codon optimization according to its original amino acid sequence, and cloned into pET-28(a+) with an N-terminal His tag. The domain topology of the constructs is shown in Figure 1(B).

Recombinant protein purification

The HPC2 Cterm and HPC2 L1EF constructs were expressed and purified by nickel-affinity chromatography and gel-filtration chromatography (GE Healthcare) as described previously [9] with buffer containing 150 mM KCl, 25 mM Tris/HCl, 20 mM imidazole and 20 mM CaCl2, at a pH of 7.4. Protein purity in each sample was judged to be over 90% by SDS/PAGE and MS analysis.

Use of ITC to characterize Ca2+-binding properties of HPC2 L1EF and Cterm

The Ca2+-binding properties of HPC2 L1EF and the HPC2 Cterm were characterized by measuring the heats associated with Ca2+ binding, using a Nano ITC Low Volume instrument (TA Instruments). Purified protein samples were placed in the sample cells, and CaCl2 solution was titrated into the cell through the titration syringe. To prepare protein samples for ITC (isothermal titration calorimetry) experiments, the gel-filtration elution fractions from the purification process were exchanged into buffer containing 150 mM KCl, 25 mM HEPES and 20 mM imidazole, at a pH of 7.4. We detected micromolar levels of residual Ca2+ in the final ITC samples despite extensive buffer exchange. The protein concentrations of ITC experiments were determined by amino acid analysis.

To account for the residual Ca2+ bound to the protein sample and in the buffer, the ITC experiments were conducted by pairing each Ca2+–protein titration experiment with another matching Ca2+ chelator–protein experiment. When protein samples were prepared, half of the sample was used for the regular Ca2+ titration experiment, and the Ca2+ chelator EDTA was added to the other half for subsequent ITC experiments. In addition, separate ITC experiments were performed to determine the thermodynamic binding parameters of chelator EDTA under our experimental conditions.

ITC baseline corrections were performed using the NanoAnalyze software (TA Instruments). Data were analysed in Mathematica (Wolfram Research) by scripts previously developed and validated in the laboratory [11]. Briefly, isotherms from the ‘protein-only’ and ‘protein+EDTA’ experiments were fitted simultaneously to the predefined binding model, the identical binding sites model, in order to derive the thermodynamics parameters of the binding interaction. The identical binding sites model describes the binding interaction with any number of identical and independent binding sites [11]. The best-fit values for the thermodynamic parameters of the binding model were obtained by a numerical minimization of the RSS (residual sum of squares) of the fitting results. The confidence intervals for each parameter were determined by plotting the RSS χ2 space around its best-fit value, and the intervals that define the 95% critical values were reported as the 95% confidence interval.

HDX-MS to characterize conformational dynamics in the HPC2 Cterm and HPC2 L1EF

The HDX-MS experiments were conducted using an HDX-enabled LC–MS/MS system (Waters Corporation), consisting of the HDX manager system, nanoAcquity UPLC and Synapt G2 Q-TOF Mass Spectrometry, specifically designed to measure protein backbone amide proton exchange rates, including software suites that automate the analysis of the complex HDX-MS data to extract peptide-specific exchange profiles. In the assessment of the H–D exchange profile of the HPC2 Cterm, the protein was labelled with 2H2O buffer containing 150 mM KCl, 10 mM imidazole and 5 mM Tris/HCl, at p2H of 7 (meter read=pH 6.6) under two different Ca2+ conditions (no added CaCl2 with 1 mM EDTA, and 10 mM CaCl2) (see Figure 3). The H–D exchange profiles were measured at exposure times of 15 s, 1 min, 5 min, 15 min, 1 h, 4 h and 20 h at 25°C (see Figure 3). In the comparison study of the HPC2 Cterm and HPC2 L1EF, the two proteins were labelled with 2H2O buffer containing four different levels of CaCl2 (0, 10 μM, 100 μM and 10 mM CaCl2) at multiple exposure times (15 s, 30 s, 5 min and 20 h) at room temperature (25°C) (see Figure 7). The protein concentrations during the deuterium labelling reactions were kept at approximately 2 μM for both the HPC2 Cterm and HPC2 L1EF. On the basis of the approximate Ca2+-binding affinities (KD ∼25 μM) for both the HPC2 Cterm and HPC2 L1EF, the Ca2+-bound proportions of the protein under 10 μM, 100 μM and 10 mM CaCl2 conditions were calculated to be 27%, 50% and 100% respectively. The labelling experiment for each time point was repeated in triplicate. Each repeat was then processed and analysed individually. Following variable exposure times to 2H2O labelling of the backbone amides, the exchange reaction was quenched by adding equal volume of pH 2.5 buffer at 0°C. The quenched mixture was then injected and passed through an online pepsin digestion column and the cleaved peptide mixture was separated by Acquity UPLC BEH C18 Column (Waters Corporation) by a linear gradient of water and acetonitrile mixture on the nanoAcquity UPLC at 0°C. The column elutes were then passed to the Synapt G2 Q-TOF instrument for mass analysis and sequence identification. Data were analysed with the Waters software packages PLGS and DynamX. For each experiment, the PLGS software identified and created a collection of peptides from the digestion reaction on the basis of their intact mass and MS/MS fragmentation pattern. Peptides with the highest score for certainty were selected for further quantification in DynamX software. In DynamX software, peptides present in no fewer than two replicates, shorter than 25 residues and with consistent signal-to-noise ratio were selected for deuterium-uptake analysis. The deuterium-uptake level was quantified by the increase in the centroid mass of each peptide, normalized by its theoretical maximum mass increase (number of exchangeable backbone amide protons in the peptide N−1). The variance of the HDX-MS analysis was assessed by the standard deviation (S.D.) of the deuterium-uptake level of multiple repeats for each time point. For each labelling time point, the labelling reaction, pepsin cleavage and the subsequent analysis were repeated three times. The differences among different Ca2+ states were also assessed by one-way ANOVA/Tukey test, based on the averaged deuterium-uptake level and the S.D.

The level of back-exchange for each peptide was assessed on the basis of the relationship between the maximum deuterium-uptake level measured (normally the plateau level at longer labelling time points) and the theoretical maximum level (number of residues −1) [25]. The level of back-exchange is affected by multiple factors, mainly the temperature (T), solvent pH and the duration between quenching and MS analysis [26]. Therefore the quenching experiment was carried out at 0°C and pH 2.5. The subsequent pepsin digestion reaction was also kept at pH 2.5, and the peptide separation and elution on the reverse-phase column was kept at 0°C to minimize the level of back-exchange after the quenching reaction.

The observed H–D exchange rate kobs is a combination of the H–D exchange rate of free and unprotected amide hydrogen kfree and the transition rates between the folded state and unfolded state (kunfold and kfold), on the basis of the following relation [27,28]:

Based on the mechanistic model, the observed H–D exchange rate kobs can be described as:

 
formula
1

Under the zero Ca2+ condition, kunfoldkfold, kfree and eqn (1) is simplified to:

 
formula
2

Therefore the measured exchange rate kobs is determined largely by kfree, the free amide proton exchange rate. Under saturating CaCl2 conditions, the majority of the protein is in the Ca2+-bound holo state, in which the kfoldkunfold and kfree, thus eqn (1) can be simplified to:

 
formula
3

Under sub-saturating Ca2+ concentrations, the conformational dynamics in converting between the unfolded and folded states (kunfold and kfold) were also altered due to the lowered Ca2+ concentrations. Under such circumstances, the observed H–D exchange rate kobs is a composite of all three rates kfree, kunfold and kfold.

NMR spectroscopy to characterize HPC2 L1EF

NMR experiments were performed on a Varian INOVA 600 MHz instrument as described previously [14]. The NMR experiments were performed with 500 μM 15N-labelled HPC2 L1EF in buffers containing different levels of CaCl2 (0, 100 μM, 1 mM and 10 mM), at pH 7.4, with 2 mM [2H11]Tris/HCl, 150 mM KCl, 10 mM imidazole, 1 mM TCEP [tris-(2-carboxyethyl)phosphine], a protease inhibitor cocktail (Roche) and 5% (v/v) 2H2O, with 5 mM sodium azide added as a preservative. Proton chemical shifts were referenced indirectly to 3-(trimethylsilyl)[2,2,3,3-2H4]propionic acid at 1H 0.00 p.p.m., with indirect dimensions referenced based on their relative gyromagnetic ratios.

On the basis of the approximate binding affinity of 25 μM for HPC2 L1EF, the Ca2+-bound proportions of the protein under 100 μM, 1 mM and 10 mM CaCl2 conditions were calculated to be approximately 19%, 95% and 100% respectively.

RESULTS

Combining three biophysical approaches, HDX-MS, ITC and protein NMR spectroscopy, we compared the full-length HPC2 Cterm and the isolated HPC2 L1EF (Figure 1B) under both Ca2+-saturating and -non-saturating conditions. Specifically, we monitored the conformational dynamics of the HPC2 Cterm and HPC2 L1EF at different Ca2+ levels to study the structural and dynamic effects of Ca2+ binding within the C-terminal tail.

Domain-dependent responses to different Ca2+ levels within the HPC2 Cterm

Using HDX-MS, we examined the responses to Ca2+ binding within the C-terminal tail of PC2, by quantifying the H–D exchange rates of the HPC2 Cterm and HPC2 L1EF under different Ca2+ concentrations. More than 95% of the protein sequence for both constructs was covered by peptides recovered from the pepsin digestion and each residue location can be represented by multiple overlapping peptides (Figure 2, designated by the blue and red bars under the protein sequences). For the majority of the examined peptides, the maximum deuterium incorporation levels are approximately 50% (Figure 3, all panels), indicating some level of back-exchange after the labelling experiment was quenched. Because the quenching and post-quenching analysis conditions were kept the same for the different Ca2+ states examined, the back-exchange for the same peptide at different Ca2+ levels can be normalized among the experiments. The study was based on comparing the H–D exchange profiles of each peptide among different Ca2+ states, therefore the back-exchange correction was not carried out for each individual peptide during the analysis.

The sequence coverage of peptides from pepsin cleavage for the HPC2 Cterm and HPC2 L1EF

Figure 2
The sequence coverage of peptides from pepsin cleavage for the HPC2 Cterm and HPC2 L1EF

The highly reproducible peptides appearing in all samples tested were selected to quantify the level of H–D exchange at each labelling time point. (A) The peptides (blue bars of various lengths spanning the construct sequence) from a pepsin digestion reaction cover 99.2% of the HPC2 Cterm sequence and provide an average 5.41-fold redundancy. The redundancy was used to describe the averaged repeat numbers for each residue location from the overlapping peptides. (B) The peptides (red bars of various lengths spanning the construct sequence) from pepsin digestion reaction cover 99.1% of the HPC2 L1EF sequence and provide an average 10.36-fold redundancy for each residue location.

Figure 2
The sequence coverage of peptides from pepsin cleavage for the HPC2 Cterm and HPC2 L1EF

The highly reproducible peptides appearing in all samples tested were selected to quantify the level of H–D exchange at each labelling time point. (A) The peptides (blue bars of various lengths spanning the construct sequence) from a pepsin digestion reaction cover 99.2% of the HPC2 Cterm sequence and provide an average 5.41-fold redundancy. The redundancy was used to describe the averaged repeat numbers for each residue location from the overlapping peptides. (B) The peptides (red bars of various lengths spanning the construct sequence) from pepsin digestion reaction cover 99.1% of the HPC2 L1EF sequence and provide an average 10.36-fold redundancy for each residue location.

Following an increasing duration of exposure to 2H2O, peptides representing different domains of the HPC2 Cterm protein displayed different deuterium-uptake profiles, indicating a diversity of conformational stability. In Figure 3, we selected and presented the deuterium-uptake profiles of several peptides covering different regions of the HPC2 Cterm in the apo (1 mM EDTA; Figure 3, red curves) and near-saturating holo (10 mM CaCl2, ∼100% occupancy; Figure 3, blue curves) states. Analysing their different uptake patterns in the two states, we identified the differential effects of Ca2+ binding throughout the HPC2 Cterm protein.

The Ca2+-dependent responses of the HPC2 Cterm characterized by HDX-MS

Figure 3
The Ca2+-dependent responses of the HPC2 Cterm characterized by HDX-MS

Deuterium-uptake plots of different peptides covering different domains of the HPC2 Cterm under both saturated holo (blue) and Ca2+-free apo (red) conditions. In each peptide panel, the deuterium-uptake level was plotted along increasing 2H2O labelling time on a semi-logarithmic scale. The H–D exchange level was quantified by tracking the increase in the centroid mass for the isotope pattern, normalized by its theoretical maximum H–D exchange level. Results are means±S.D. for triplicates at each time point. The corresponding secondary structure and domain of the peptide represented was identified, with the start and end position of each peptide labelled by the residue number in the context of full-length human PC2. The 3D structure of the HPC2 Cterm was plotted on the basis of two separate structures (PDB codes 2Y4Q and 3HRN).

Figure 3
The Ca2+-dependent responses of the HPC2 Cterm characterized by HDX-MS

Deuterium-uptake plots of different peptides covering different domains of the HPC2 Cterm under both saturated holo (blue) and Ca2+-free apo (red) conditions. In each peptide panel, the deuterium-uptake level was plotted along increasing 2H2O labelling time on a semi-logarithmic scale. The H–D exchange level was quantified by tracking the increase in the centroid mass for the isotope pattern, normalized by its theoretical maximum H–D exchange level. Results are means±S.D. for triplicates at each time point. The corresponding secondary structure and domain of the peptide represented was identified, with the start and end position of each peptide labelled by the residue number in the context of full-length human PC2. The 3D structure of the HPC2 Cterm was plotted on the basis of two separate structures (PDB codes 2Y4Q and 3HRN).

Ca2+ binding causes a significant change in the H–D exchange patterns within the EF-hand. Under apo conditions, the EF-hand region is highly flexible, with the deuterium uptake of each of the peptide fragments reaching the maximum uptake level (50%) in under 1 min (Figure 3, red curves). In contrast, the EF-hand helices α2 and α3 (Figure 3, lower left corner) and loop 2 (Figure 3, lower right corner) exhibit notably lower uptake rates in the presence of saturating Ca2+ compared with their apo state, indicating a stable conformation upon Ca2+ binding. A similar increase in the stabilizing effect was also observed in EF-hand helix α4 (Table 1 and Figure 4B, regions of bright red). In comparison, the EF-hand helix α1 (Figure 3, upper left corner) and loop 1 (Figure 3, upper right corner) within the EF-hand domain only showed a moderate increase in local stability upon Ca2+ binding (Table 1 and Figure 4B, regions of light pink), suggesting a less stable local conformation. Previous structural characterization of the isolated EF-hand using NMR have identified one Ca2+-binding site within the EF-hand domain [12,14], which was hypothesized to be located within the loop 2 region. The H–D exchange profiles are consistent with this hypothesis. The increases in structural stability seen within loop 2, and helices α2, α3 and α4 (Table 1 and Figure 4, regions of bright red) suggest that Ca2+ binding is localized to this region. In contrast, we speculate that the smaller increase observed in the first helix α1 and the truncated loop 1 upon Ca2+ binding (Table 1 and Figure 4, regions of light pink) is mostly likely to be due to the longer-range tertiary interactions with loop 2 and the other helices in the EF-hand domain.

Deuterium-uptake patterns of the HPC2 Cterm in apo and holo states

Figure 4
Deuterium-uptake patterns of the HPC2 Cterm in apo and holo states

(A) The H–D exchange levels of the HPC2 Cterm after 1 min of 2H2O (‘D2O’) labelling in both holo and apo states are mapped on to the HPC2 Cterm structure. One unit of the trimeric HPC2 Cterm is shown for comparison. The colour gradient indicates the degree of deuterium exchange after 1 min of labelling time, with blue representing the lowest exchange level and red representing the highest exchange level. The residues without H–D exchange information are grey. The 3D structure of the HPC2 Cterm was plotted on the basis of two separate structures (PDB codes 2Y4Q and 3HRN). (B) The differences of the uptake levels between the apo and holo states were mapped on the HPC2 Cterm structure. The colour gradient indicates the uptake level differences between two states after 1 min of labelling time, with white representing no change and red representing the 30% increase in uptake levels.

Figure 4
Deuterium-uptake patterns of the HPC2 Cterm in apo and holo states

(A) The H–D exchange levels of the HPC2 Cterm after 1 min of 2H2O (‘D2O’) labelling in both holo and apo states are mapped on to the HPC2 Cterm structure. One unit of the trimeric HPC2 Cterm is shown for comparison. The colour gradient indicates the degree of deuterium exchange after 1 min of labelling time, with blue representing the lowest exchange level and red representing the highest exchange level. The residues without H–D exchange information are grey. The 3D structure of the HPC2 Cterm was plotted on the basis of two separate structures (PDB codes 2Y4Q and 3HRN). (B) The differences of the uptake levels between the apo and holo states were mapped on the HPC2 Cterm structure. The colour gradient indicates the uptake level differences between two states after 1 min of labelling time, with white representing no change and red representing the 30% increase in uptake levels.

Downstream of the EF-hand, the region between Val843 and Ala862 (Figure 3, lower middle panel) in the coiled-coil domain displayed the lowest deuterium uptake rate of the entire protein sequence (Figure 4A, regions of deep blue). This region of the coiled-coil domain is distinct from the EF-hand and exchanges slowly in both apo and holo states (Figure 4), with the apo state being slightly more flexible than the holo state (Figure 3, lower middle panel), suggesting that the local changes in the EF-hand could be relayed to the distal coiled-coil region. The rest of the identified coiled-coil domain (Ala873–Gly895), appeared very flexible, with rapid H–D exchange rates in both apo and holo states (Figure 4A). The crystal structure of the isolated coiled-coil domain (Val833–Gly895) forms a coiled-coil helix trimer bundle [13]. Our H–D exchange characterization revealed that, although the Val843–Ala862 region of the coiled-coil trimer was engaged in a stable interaction and protected the region from deuterium labelling (Figure 4A), the rest of the trimer complex showed a high level of flexibility and solvent accessibility. We speculate that such flexibility could lead to possible chemical exchange of the trimeric helix bundle and interaction with other proteins.

Different from the structured EF-hand and coiled-coil region, the rest of the HPC2 Cterm exhibited rapid H–D exchange profiles in both states, with their deuterium incorporation reaching a maximum level within 15 s of incubation. The rapid exchange patterns are consistent with the sequence-based prediction that the connecting linker regions were equally flexible under apo and holo conditions.

The results from the HDX analysis of the HPC2 Cterm region are largely consistent with those previous studies on the isolated fragments [10,12,29,30]. A large portion of the protein in solution exhibited a high level of flexibility and solvent accessibility even in the Ca2+-bound holo state (Figure 4A). Specifically, one region within the coiled-coil domain (Ala873–Gly895) displayed rapid H–D exchange profiles in both apo and holo states (Figure 4A). A previous study using molecular dynamics simulation has also suggested that this region of the coiled-coil trimer is highly dynamic and can accommodate interactions with an additional α-helix [31]. The flexible nature of the protein is also consistent with the difficulties encountered in the previous efforts to study the protein using crystallography [10].

The HPC2 Cterm and HPC2 L1EF share similar Ca2+-binding affinities

To understand how the coiled-coil domain affects Ca2+-binding in the EF-hand of the HPC2 Cterm, we designed the HPC2 L1EF construct to include the N-terminal linker 1 region and the EF-hand domain (Figure 1B). A comparison of the Ca2+-binding affinity between the HPC2 Cterm and HPC2 L1EF will help us to determine the importance of the coiled-coil domain in the protein stability and dynamics. On the basis of MS, chromatography and NMR analysis (Supplementary Figure S1), we verified that the HPC2 L1EF construct lacking the downstream coiled-coil domain is still monomeric in solution and stable for up to 48 h at room temperature. ITC experiments were first performed to characterize the Ca2+-binding thermodynamics of the HPC2 L1EF protein. In addition to the Ca2+-protein titration experiment (Figure 5A), a second titration experiment with both the Ca2+ chelator EDTA and HPC2 L1EF protein (Figure 5B) was set up, to more definitively determine the binding stoichiometry, ‘n’. The thermodynamic parameters of binding interaction were derived by simultaneous fitting of the two binding isotherms to the predefined binding model (Figure 5C).

ITC measurement of Ca2+-binding interactions in the HPC2 Cterm and HPC2 L1EF

Figure 5
ITC measurement of Ca2+-binding interactions in the HPC2 Cterm and HPC2 L1EF

(A) Raw heat measurement of 5 mM CaCl2 titrated into 200 μM HPC2 L1EF protein. (B) Raw heat measurement of 5 mM CaCl2 titrated into 197 μM HPC2 L1EF protein and 379 μM EDTA. (C) Simultaneously fitted isotherms of CaCl2 into HPC2 L1EF only (blue trace) and protein with EDTA (red trace) ITC experiments. (D) Raw heat measurement of 5 mM CaCl2 titrated into 550 μM HPC2 Cterm protein. (E) Raw heat measurement of 5 mM CaCl2 titrated into 539 μM HPC2 Cterm protein and 378 μM EDTA. (F) Simultaneously fitted isotherms of CaCl2 into the HPC2 Cterm only (blue trace) and protein with EDTA (red trace) ITC experiments.

Figure 5
ITC measurement of Ca2+-binding interactions in the HPC2 Cterm and HPC2 L1EF

(A) Raw heat measurement of 5 mM CaCl2 titrated into 200 μM HPC2 L1EF protein. (B) Raw heat measurement of 5 mM CaCl2 titrated into 197 μM HPC2 L1EF protein and 379 μM EDTA. (C) Simultaneously fitted isotherms of CaCl2 into HPC2 L1EF only (blue trace) and protein with EDTA (red trace) ITC experiments. (D) Raw heat measurement of 5 mM CaCl2 titrated into 550 μM HPC2 Cterm protein. (E) Raw heat measurement of 5 mM CaCl2 titrated into 539 μM HPC2 Cterm protein and 378 μM EDTA. (F) Simultaneously fitted isotherms of CaCl2 into the HPC2 Cterm only (blue trace) and protein with EDTA (red trace) ITC experiments.

Analysis of the ITC showed that the best-fit value for parameter ‘n’, the binding stoichiometry, was very close to 1 (Table 2), indicating that there is only one Ca2+-binding site, presumably the same Ca2+-binding site identified previously within the EF-hand domain [9]. Furthermore, the best-fit value for the binding affinity KD is 25.8 μM for HPC2 L1EF (Table 2). We applied the same strategies to define the Ca2+–HPC2 Cterm binding interaction (Figures 5D–5F). Interestingly, we found that the Ca2+-binding affinity of the monomeric HPC2 L1EF is similar in magnitude to the binding affinity of Ca2+ exhibited by the full-length HPC2 Cterm construct (KD ∼24.8 μM) under identical conditions. The similar binding affinities indicate that the coiled-coil domain and the oligomerization state contribute little to Ca2+ binding. For the HPC2 Cterm, the molar enthalpy change (ΔH) is found to be approximately −13 kcal/mol (1 kcal=4.184 kJ), slightly higher than the enthalpy change of HPC2 L1EF (ΔH of approximately −11 kcal/mol). The difference in molar enthalpy suggests that the HPC2 Cterm undergoes a slightly higher range of structural rearrangement upon Ca2+ binding. Because the HPC2 Cterm also includes the downstream coiled-coil domain, it is likely that the coiled-coil region also experiences a minor change in local structural stability, as we have observed in the HDX-MS characterization of the HPC2 Cterm (Figure 3, lower middle panel). Taken together, although the HPC2 Cterm contains the downstream coiled-coil domain, which mediates its oligomerization in solution, the downstream regions in the Cterm are not contributing to stabilizing the Ca2+-bound conformation of the EF-hand.

NMR studies of HPC2 L1EF

To explore the structural effects of Ca2+ binding to the HPC2 L1EF construct, NMR spectroscopy was used to characterize the HPC2 L1EF protein under different Ca2+ levels. In the spectra of the apo HPC2 L1EF protein (Figure 6, upper left panel), the majority of the resonances are clustered together, suggesting that the protein is in a largely unstructured state. In addition, most of the peaks in the 1H-15N HSQC spectra were broad and of relatively low intensity, suggestive of extensive conformational exchange line-broadening. In the presence of 100 μM CaCl2 (Figure 6, upper right panel), approximately 19% of the total HPC2 L1EF with bound Ca2+, the resonances in the HSQC spectrum exhibited an increase in dispersion. Specifically, three threonine residues (Thr751, Thr760 and Thr771), located in helices α3 and α4 of the EF-hand, are resolved and visible in the spectra (Figure 6, upper right panel). With 1 mM or 10 mM CaCl2 present, approximately 95% or 100% of the HPC2 L1EF Ca2+-binding sites are occupied. Under these Ca2+-saturating conditions, the HSQC spectra are characteristic of well-folded globular proteins (Figure 6, two lower panels). Furthermore, Gln768 and Leu770 located within the Ca2+-binding loop exhibited the characteristically downfield 1H chemical shift (Figure 6, lower panels) that is seen in other Ca2+-binding EF-hand motifs [21,3234]. The majority of the peaks are well resolved, with narrower line-widths (Figure 6, lower panels), than the zero Ca2+ and 100 μM CaCl2 conditions, indicative of a decrease in exchange broadening conformational motions.

NMR spectra of HPC2 L1EF under different Ca2+ levels

Figure 6
NMR spectra of HPC2 L1EF under different Ca2+ levels

Comparison of 1H-15N HSQC NMR spectra of 15N HPC2 L1EF under zero CaCl2 (green contours), 100 μM CaCl2 (blue contours), 1 mM CaCl2 (red contours), 10 mM CaCl2 (orange contours) conditions.

Figure 6
NMR spectra of HPC2 L1EF under different Ca2+ levels

Comparison of 1H-15N HSQC NMR spectra of 15N HPC2 L1EF under zero CaCl2 (green contours), 100 μM CaCl2 (blue contours), 1 mM CaCl2 (red contours), 10 mM CaCl2 (orange contours) conditions.

Differences in conformational dynamics for HPC2 L1EF and Cterm uncovered using sub-saturating Ca2+ levels

To detect possible domain–domain interactions and characterize the Ca2+-dependent dynamic responses, we examined the H–D exchange profiles of the monomeric HPC2 L1EF and the trimeric Cterm under both saturating (10 mM) and non-saturating (0, 10 and 100 μM) Ca2+ conditions by HDX-MS (Figure 7A). Although the HPC2 Cterm and HPC2 L1EF exhibit similar conformational dynamics in the EF-hand under zero CaCl2 and 10 mM CaCl2 conditions, the two constructs have different dynamic profiles (Figure 7B) under two non-saturating Ca2+ conditions (10 and 100 μM CaCl2).

HDX-MS reveals different stabilizing effects of Ca2+ in the HPC2 Cterm and HPC2 L1EF

Figure 7
HDX-MS reveals different stabilizing effects of Ca2+ in the HPC2 Cterm and HPC2 L1EF

(A) The H–D exchange patterns were collected from the HPC2 Cterm and HPC2 L1EF samples under zero (red), 10 μM (green), 100 μM (pink) and 10 mM (blue) CaCl2 conditions. We selected a series of peptides that were reproducible and consistent across all the Ca2+ levels tested for comparative analysis. For each peptide, the deuterium-uptake level was plotted along increasing 2H2O labelling time on a semi-logarithmic scale. The H–D exchange level was quantified by tracking the increase in the centroid mass for the isotope pattern, normalized by its theoretical maximum deuterium exchange level. Results are means±S.D. for triplicates at each time point. (B) The stabilizing effects of two Ca2+ concentrations (10 μM and 10 mM) in the EF-hand domain of the HPC2 Cterm and HPC2 L1EF were mapped on to the structure of the EF-hand (PDB code 2Y4Q). On the basis of the deuterium-uptake curves, the stabilizing effects of Ca2+ binding were calculated by subtracting the apo state uptake from the Ca2+-present state(s). The differences are absolute values averaged across different time points and normalized by the maximum uptake level reached in the apo state. The colour gradient indicates the degree of differences between two states after 1 min of labelling time, with green representing no stabilizing effects and blue representing the 50% increase in stabilizing effects.

Figure 7
HDX-MS reveals different stabilizing effects of Ca2+ in the HPC2 Cterm and HPC2 L1EF

(A) The H–D exchange patterns were collected from the HPC2 Cterm and HPC2 L1EF samples under zero (red), 10 μM (green), 100 μM (pink) and 10 mM (blue) CaCl2 conditions. We selected a series of peptides that were reproducible and consistent across all the Ca2+ levels tested for comparative analysis. For each peptide, the deuterium-uptake level was plotted along increasing 2H2O labelling time on a semi-logarithmic scale. The H–D exchange level was quantified by tracking the increase in the centroid mass for the isotope pattern, normalized by its theoretical maximum deuterium exchange level. Results are means±S.D. for triplicates at each time point. (B) The stabilizing effects of two Ca2+ concentrations (10 μM and 10 mM) in the EF-hand domain of the HPC2 Cterm and HPC2 L1EF were mapped on to the structure of the EF-hand (PDB code 2Y4Q). On the basis of the deuterium-uptake curves, the stabilizing effects of Ca2+ binding were calculated by subtracting the apo state uptake from the Ca2+-present state(s). The differences are absolute values averaged across different time points and normalized by the maximum uptake level reached in the apo state. The colour gradient indicates the degree of differences between two states after 1 min of labelling time, with green representing no stabilizing effects and blue representing the 50% increase in stabilizing effects.

In the absence of Ca2+ (Figure 7A, red curves), the EF-hand and linker 1 are very flexible in both the HPC2 Cterm and HPC2 L1EF, with all of the representative peptides reaching their maximum H–D exchange levels in under 1 min (Figure 7A, red curves in all panels). At saturating CaCl2 level (10 mM), the two constructs both showed decreased rates in H–D exchange within the EF-hand (Figure 7A, blue curves in all panels), specifically in peptides covering EF-hand helices α2 and α3 (Figure 7A, two bottom panels). Different from the EF-hand, the linker 1 region remained flexible in both the HPC2 Cterm and HPC2 L1EF, with 10 mM CaCl2 (Figure 7A, top two panels, and Table 1). Furthermore, the EF-hand domain experienced similar levels of increased stability with 10 mM CaCl2 present, when the HPC2 Cterm and HPC2 L1EF constructs were compared (Figure 7B, bottom two panels, and Table 1). These observations are consistent with the similar Ca2+-binding affinities observed for both the HPC2 Cterm and HPC2 L1EF using ITC, indicating similar local stabilizing effects due to Ca2+ binding. On the basis of the lack of differences in the uptake patterns between HPC2 L1EF and HPC2 Cterm under Ca2+-saturating conditions (Figure 7A, blue curves), these results further indicate that there are no direct structural stabilizing effects or interactions imposed on the EF-hand from the coiled-coil region. Therefore we conclude that the coiled-coil domain, although necessary to form the trimeric HPC2 Cterm structure, is not directly engaged in interactions with the EF-hand.

Table 1
Different stabilizing effects in the HPC2 Cterm and HPC2 L1EF induced by Ca2+ binding

The differences in deuterium uptake between different states are quantified by subtracting the apo state uptake from the Ca2+-present state(s). The differences are averaged across different time points and normalized by the maximum uptake level reached in the apo state. S.D. are given within parentheses for the triplicate experiments, normalized to the maximum uptake level reached in the apo state.

 ΔΔ2H2O% (apo, 10 μM CaCl2ΔΔ2H2O% (apo, 100 μM CaCl2ΔΔ2H2O% (apo, 10 mM CaCl2
 HPC2 Cterm HPC2 L1EF HPC2 Cterm HPC2 L1EF HPC2 Cterm HPC2 L1EF 
Ile704–Leu715 linker 1 −1.0 (1.4) −1.5 (1.7) −5.7 (1.6) −2.7 (0.8) −6.4 (2.1)  0.0 (1.1) 
Lys716–Glu727 helix α1 −11.1 (1.0) −2.0 (2.1) −16.2 (2.3) −4.2 (0.7) −18.3 (1.5) −12.6 (0.9) 
Arg730–Phe738 loop 1 −8.9 (1.0) −2.6 (1.9) −15.0 1.7 −3.5 (0.5) −19.7 (1.1) −16.4 (0.9) 
Asp739–Glu756 helix α2 and α3 −27.2 (1.3) −8.4 (1.9) −30.1 (2.1) −19.0 (1.1) −37.4 (2.1) −35.3 (0.8) 
Phe759–Leu770 loop 2 −22.4 (1.5) −2.2 (1.3) −27.6 (1.8) −6.7 (0.8) −39.6 (1.5) −33.7 (0.7) 
Leu770–Leu782 helix α4 −15.3 (1.6) −3.8 (4.5) −22.9 (5.2) −7.8 (1.2) −36.3 (2.1) −31.3 (1.3) 
 ΔΔ2H2O% (apo, 10 μM CaCl2ΔΔ2H2O% (apo, 100 μM CaCl2ΔΔ2H2O% (apo, 10 mM CaCl2
 HPC2 Cterm HPC2 L1EF HPC2 Cterm HPC2 L1EF HPC2 Cterm HPC2 L1EF 
Ile704–Leu715 linker 1 −1.0 (1.4) −1.5 (1.7) −5.7 (1.6) −2.7 (0.8) −6.4 (2.1)  0.0 (1.1) 
Lys716–Glu727 helix α1 −11.1 (1.0) −2.0 (2.1) −16.2 (2.3) −4.2 (0.7) −18.3 (1.5) −12.6 (0.9) 
Arg730–Phe738 loop 1 −8.9 (1.0) −2.6 (1.9) −15.0 1.7 −3.5 (0.5) −19.7 (1.1) −16.4 (0.9) 
Asp739–Glu756 helix α2 and α3 −27.2 (1.3) −8.4 (1.9) −30.1 (2.1) −19.0 (1.1) −37.4 (2.1) −35.3 (0.8) 
Phe759–Leu770 loop 2 −22.4 (1.5) −2.2 (1.3) −27.6 (1.8) −6.7 (0.8) −39.6 (1.5) −33.7 (0.7) 
Leu770–Leu782 helix α4 −15.3 (1.6) −3.8 (4.5) −22.9 (5.2) −7.8 (1.2) −36.3 (2.1) −31.3 (1.3) 

At the other sub-saturating conditions (10 and 100 μM CaCl2), we observed the differential deuterium uptake patterns in the EF-hand between the HPC2 Cterm and HPC2 L1EF (Figure 7A). The EF-hand, in the context of HPC2 L1EF and partially bound to Ca2+, experienced a slight decrease in deuterium-uptake rates at 10 μM (green curves) and 100 μM (pink curves) CaCl2 levels (Figure 7A and Table 1) in most regions of the EF-hand. We only observed proportionally lower uptake rates in the regions directly involved in Ca2+ binding (Figure 7A, bottom right panel, and Table 1). In contrast, the EF-hand region in the trimeric HPC2 Cterm construct displayed a much slower deuterium-uptake pattern under non-saturating Ca2+ conditions (Figure 7B, left-hand panels). The stabilizing effects are proportional to the increased Ca2+ concentrations (Figure 7B, left-hand panels, and Table 1). Taken together, under non-saturating Ca2+-conditions, we observed higher levels of local stability within the EF-hand, residing in the trimeric HPC2 Cterm construct, compared with the EF-hand in the monomeric HPC2 L1EF construct.

We propose that the differences in the local stability between the HPC2 Cterm and HPC2 L1EF result from their different conformational dynamics of the folded state and the unfolded state at different Ca2+ levels. In the HDX experiment, we can measure the overall H–D exchange rate (kobs) directly. In the absence of Ca2+, the EF-hand regions in both the HPC2 Cterm and HPC2 L1EF are in the unfolded state, in which the observed exchange rate (kobs) is very similar to the H–D exchange rate for a freely exchangeable backbone amide proton (kfree). Therefore we observed the rapid uptake patterns in both constructs under the zero Ca2+ condition (Figure 7A, red curves in all panels). In contrast, when placed under the 10 mM CaCl2 condition, the majority of the EF-hand is in the Ca2+-bound state. In this case, the observed exchange rate kobs is determined by the local stability of the EF-hand when Ca2+-bound. Because the HPC2 Cterm and HPC2 L1EF share similar Ca2+-binding affinities, the two also obtain a similar level of structural stabilizing effects within the EF-hand upon Ca2+ binding. We therefore observed similar H–D exchange rates between the HPC2 Cterm and HPC2 L1EF, when they were under saturating Ca2+ conditions (Figure 7A, blue curves in all panels). However, under sub-saturating Ca2+ concentrations, the protein population is in a mixture of unbound and bound states, and the protein dynamics in converting between the unfolded and folded states (kunfold and kfold) are also altered, due to the lower Ca2+ concentration. Under such circumstances, the observed H–D exchange rate kobs is dependent on all three rates kfree, kunfold and kfold. Therefore the different dynamics in converting between the unfolded and folded state, kunfold and kfold, will contribute to the different observed H–D exchange profiles in the EF-hand between the HPC2 Cterm and HPC2 L1EF (Figure 7A, pink and green curves in all panels).

The different H–D exchange patterns indicate that the EF-hand in the trimeric HPC2 Cterm have different conformational dynamics of the folded state and the unfolded state compared with its monomeric form in HPC2 L1EF. Without evidence showing a direct interaction between the EF-hand and coiled-coil domain in the HPC2 Cterm, we deduce that such differences in molecular motions between the HPC2 Cterm and HPC2 L1EF are mostly likely to result from the restraints on the movements of the EF-hand when it is linked to the downstream large coiled-coil trimer complex. The differences in molecular motions under sub-saturating Ca2+ conditions are of particularly interest in understanding how the C-terminal tail of PC2 would respond to the changes in Ca2+ levels under physiological conditions, which are well below saturating levels [5,10].

DISCUSSION

Analysis of the H–D exchange profiles of the HPC2 Cterm in both apo and holo states has identified Ca2+-dependent structural responses in the HPC2 Cterm, which was structurally characterized as a whole for the first time. To understand further how the downstream coiled-coil domain affects the Ca2+ binding of the EF-hand, we compared the Ca2+ binding of both the trimeric HPC2 Cterm and the monomeric HPC2 L1EF proteins. The HPC2 L1EF protein shares similar Ca2+-binding affinity with HPC2 Cterm, indicating that the downstream coiled-coil domain is not required to stabilize the Ca2+-bound conformation of the EF-hand. However, the two constructs showed different profiles of H–D exchange under non-saturating Ca2+ conditions, in which the conformation exchange dynamics of the Ca2+-bound and Ca2+-unbound states would affect the channel function of the full-length PC2 protein.

The linker 1 region stabilizes the EF-hand domain and enhances its Ca2+-binding affinity

Previous studies have shown that the isolated and monomeric EF-hand binds to Ca2+ much more weakly than the full-length HPC2 Cterm [7,9,11]. Nevertheless, our ITC results indicate that HPC2 L1EF shared similar Ca2+-binding affinity with the full-length HPC2 Cterm (Table 2). A comparison of the ITC results of HPC2 L1EF with other previously characterized EF-hand constructs that do not include the N-terminal linker 1, has identified that the additional N-terminal linker 1 region largely increases the Ca2+-binding affinity and structural stability of the EF-hand [7,11,12]. A comparison of the previously published NMR results [12,14] has shown that additional residues on the N-terminus of the EF-hand would contribute to a longer helix 1 of the EF-hand, and the formation of a short antiparallel β-strand between the two loops. Compared with the previously characterized PC2 EF-hand constructs, the inclusion of the linker 1 region in HPC2 L1EF greatly enhances the Ca2+-binding affinity of the EF-hand. However, the linker itself does not contain a well-defined structural motif. In addition to the NMR results, the H–D exchange profiles of HPC2 L1EF also revealed that Ile704–Leu715 is a flexible region under both apo and holo conditions (Figure 7, top right panel). On the basis of the combined results from ITC, HDX-MS and NMR, we propose that the additional N-terminal L1 region, as a flexible linker region, contributes to a higher Ca2+-binding affinity by providing stabilizing capping effects for the helix 1 region in the EF-hand region, although the linker itself is not involved in direct structural interactions.

Table 2
Ca2+-binding thermodynamics of HPC2 L1EF

Fitting results for thermodynamic parameters associated with HPC2 L1EF and Ca2+-binding interactions based on identical binding sites model. 95% confidence intervals are given within parentheses.

Molar enthalpy (ΔH, kcal/mol) Dissociation constant (KD, μM) Binding stoichiometry (nResidual Ca2+ (Ro, μM) 
−10.9 (−11.5, −10.3) 25.8 (23.3, 31.6) 1.12 (1.05, 1.20) 94.0 (80.1, 107) 
Molar enthalpy (ΔH, kcal/mol) Dissociation constant (KD, μM) Binding stoichiometry (nResidual Ca2+ (Ro, μM) 
−10.9 (−11.5, −10.3) 25.8 (23.3, 31.6) 1.12 (1.05, 1.20) 94.0 (80.1, 107) 

The dynamics of conformational transitions provides a molecular mechanism to understand channel gating and regulation

Although the HPC2 Cterm and HPC2 L1EF have similar Ca2+-binding affinities, the HDX-MS study reveals that the EF-hand within the HPC2 Cterm has a decreased level of conformational dynamics between the folded state and the unfolded state compared with its monomeric form in HPC2 L1EF at non-saturating Ca2+ levels. Without evidence indicating any direct thermodynamic linkage between the EF-hand and the downstream coiled-coil domain, we propose that the slower exchange patterns of the EF-hand in the HPC2 Cterm could result from the restraints on the movements of each EF-hand when it is linked to the downstream coiled-coil trimer. Furthermore, we speculate that the molecular motions of the EF-hand, residing in the full-length PC2, would experience additional restraints from its N-terminus, while linking to the sixth transmembrane helix via linker 1. Hence we expect that the dynamics between the Ca2+-unbound and -bound states of the EF-hand would be decreased further in the full membrane-bound form under similar low levels of Ca2+.

As a Ca2+-regulated Ca2+-permeable channel, the open probability of PC2 is dependent on the cytoplasmic Ca2+ concentration [5,10,35]. Although the conformational changes due to Ca2+ binding within the EF-hand provide the structural basis to understand channel functional regulation by Ca2+, the dynamics of the conformational transitions of the C-terminal domain provide additional mechanistic information for understanding channel gating. In the present study, the measured Ca2+-binding affinities of both the HPC2 Cterm and HPC2 L1EF are lower than the optimal Ca2+ concentrations required to activate the channel. Nevertheless, as we have proposed, the EF-hand in the full-length PC2 channel may experience a lower level of conformational dynamics at sub-saturating Ca2+ levels due to the movement restraints, compared with its isolated and monomeric form. Such slow dynamics may provide the timeframe necessary to relay the conformational changes from the C-terminal tail to the pore-forming loop, leading to the opening of the PC2 channel, even under non-saturating Ca2+ conditions. The importance of protein motions and dynamics have long been implicated in understanding the biomolecular functions of proteins [36,37]. Recent studies of several membrane proteins have indeed revealed that protein dynamics are associated with different activation states and regulatory mechanisms for both ligand-gated ion channels and G-protein-coupled receptors [3840]. Therefore the dynamic characterization of the PC2 cytoplasmic C-terminal tail can provide the molecular foundation to understand further the regulation mechanism of the PC2 channel. In addition to the restraints posed by the transmembrane helix, structural interactions with other domains of the full-length PC2 channel or post-translational modifications could modify the dynamic profiles of the PC2 C-terminal tail allosterically. It is important to note that several post-translational sites, located in the linker region between the EF-hand and the coiled-coil domains, are known to change the Ca2+-dependent response of the PC2 channel [5,14,41,42]. In addition, the PC2 N-terminal tail and other intracellular loops are also involved in channel regulation [43,44].

The MS and proteomics-based approaches in HDX-MS enable the dynamic characterization of small amounts of protein samples (picomole to nanomole range) in a mixture of conformational states. Therefore the HDX-MS method will be readily applied to full-length membrane proteins reconstituted in liposome or detergent [45,46]. Thus these strategies will eventually be expanded to characterize the full-length PC2 channel or other TRP channel family members for further understanding of conformational dynamics in different channel states and after pathological mutation manipulations.

In summary, the addition of HDX-MS to ITC and NMR has enabled us to identify dynamic properties of the HPC2 Cterm and HPC2 L1EF associated with conformational changes at different Ca2+ levels. Although the HPC2 Cterm and HPC2 L1EF apparently share similar Ca2+-binding profiles, they display different Ca2+-dependent profiles in conformational dynamics that were revealed by H–D exchange patterns. In the trimeric full-length HPC2 Cterm protein, the Ca2+-binding EF-hand exhibited higher levels of local stability and slower H–D exchange profiles compared with the monomeric isolated HPC2 L1EF under non-saturating Ca2+-conditions. The dynamic differences between the HPC2 Cterm and HPC2 L1EF under non-saturating Ca2+ conditions is important in predicting how the PC2 channel functions at physiological Ca2+-levels. The present study provides the first dynamic characterization of the full-length PC2 C-terminal tail, which revealed its Ca2+-dependent and highly flexible conformation. The present study not only expands our understanding of how PC2 might be regulated by cellular Ca2+ level changes via its C-terminal tail, but also shows the importance of dynamic properties in elucidating the link between structure and function of PC2 protein.

AUTHOR CONTRIBUTION

Yifei Yang, Barbara Ehrlich, Elias Lolis and Michael Hodsdon designed the experiments. Yifei Yang conducted the experiments. Yifei Yang analysed the data and prepared the figures. Yifei Yang, Barbara Ehrlich, Elias Lolis and Michael Hodsdon wrote the paper.

Dr J. Patrick Loria is thanked for advice and guidance on the ITC and NMR studies and for a critical reading of the paper before submission. Dr James Howe and Dr Enrique M. De La Cruz are thanked for a critical reading of the paper before submission.

FUNDING

This work was funded by the National Institutes of Health [grant numbers R01 DK087844 (to B.E.E. and M.E.H.) and R21RR032351 (to M.E.H.)]. Y.Y was supported by a predoctoral fellowship from the Chinese Scholarship Council (CSC)–Yale World Scholars programme.

Abbreviations

     
  • ADPKD

    autosomal dominant polycystic kidney disease

  •  
  • H–D

    hydrogen–deuterium

  •  
  • HDX-MS

    hydrogen–deuterium exchange mass spectrometry

  •  
  • HPC2 Cterm

    C-terminal cytoplasmic tail of human PC2

  •  
  • HPC2 L1EF

    N-terminal linker 1–EF-hand region of human PC2

  •  
  • ITC

    isothermal titration calorimetry

  •  
  • PC2

    polycystin-2

  •  
  • RSS

    residual sum of squares

  •  
  • TRP

    transient receptor potential

References

References
1
Gabow
P.A.
Autosomal dominant polycystic kidney disease
N. Engl. J. Med.
1993
, vol. 
329
 (pg. 
332
-
342
)
[PubMed]
2
Ong
A.C.
Harris
P.C.
A polycystin-centric view of cyst formation and disease: the polycystins revisited
Kidney Int.
2015
, vol. 
88
 (pg. 
699
-
710
)
[PubMed]
3
Cornec-Le Gall
E.
Audrezet
M.P.
Meur
Y.L.
Chen
J.M.
Ferec
C.
Genetics and pathogenesis of autosomal dominant polycystic kidney disease: 20 years on
Hum. Mutat.
2014
, vol. 
35
 (pg. 
1393
-
1406
)
[PubMed]
4
Chebib
F.T.
Sussman
C.R.
Wang
X.
Harris
P.C.
Torres
V.E.
Vasopressin and disruption of calcium signalling in polycystic kidney disease
Nat. Rev. Nephrol.
2015
, vol. 
11
 (pg. 
451
-
464
)
[PubMed]
5
Cai
Y.
Anyatonwu
G.
Okuhara
D.
Lee
K.B.
Yu
Z.
Onoe
T.
Mei
C.L.
Qian
Q.
Geng
L.
Wiztgall
R.
, et al. 
Calcium dependence of polycystin-2 channel activity is modulated by phosphorylation at Ser812
J. Biol. Chem.
2004
, vol. 
279
 (pg. 
19987
-
19995
)
[PubMed]
6
Koulen
P.
Cai
Y.
Geng
L.
Maeda
Y.
Nishimura
S.
Witzgall
R.
Ehrlich
B.E.
Somlo
S.
Polycystin-2 is an intracellular calcium release channel
Nat. Cell Biol.
2002
, vol. 
4
 (pg. 
191
-
197
)
[PubMed]
7
Celic
A.
Petri
E.T.
Demeler
B.
Ehrlich
B.E.
Boggon
T.J.
Domain mapping of the polycystin-2 C-terminal tail using de novo molecular modeling and biophysical analysis
J. Biol. Chem.
2008
, vol. 
283
 (pg. 
28305
-
28312
)
[PubMed]
8
Reference deleted
9
Yang
Y.
Keeler
C.
Kuo
I.Y.
Lolis
E.J.
Ehrlich
B.E.
Hodsdon
M.E.
Oligomerization of the polycystin-2 C-terminal tail and effects on its Ca2+-binding properties
J. Biol. Chem.
2015
, vol. 
290
 (pg. 
10544
-
10554
)
[PubMed]
10
Celic
A.S.
Petri
E.T.
Benbow
J.
Hodsdon
M.E.
Ehrlich
B.E.
Boggon
T.J.
Calcium-induced conformational changes in C-terminal tail of polycystin-2 are necessary for channel gating
J. Biol. Chem.
2012
, vol. 
287
 (pg. 
17232
-
17240
)
[PubMed]
11
Keeler
C.
Poon
G.
Kuo
I.Y.
Ehrlich
B.E.
Hodsdon
M.E.
An explicit formulation approach for the analysis of calcium binding to EF-hand proteins using isothermal titration calorimetry
Biophys. J.
2013
, vol. 
105
 (pg. 
2843
-
2853
)
[PubMed]
12
Allen
M.D.
Qamar
S.
Vadivelu
M.K.
Sandford
R.N.
Bycroft
M.
A high-resolution structure of the EF-hand domain of human polycystin-2
Protein Sci.
2014
, vol. 
23
 (pg. 
1301
-
1308
)
[PubMed]
13
Yu
Y.
Ulbrich
M.H.
Li
M.H.
Buraei
Z.
Chen
X.Z.
Ong
A.C.
Tong
L.
Isacoff
E.Y.
Yang
J.
Structural and molecular basis of the assembly of the TRPP2/PKD1 complex
Proc. Natl. Acad. Sci. U.S.A.
2009
, vol. 
106
 (pg. 
11558
-
11563
)
[PubMed]
14
Petri
E.T.
Celic
A.
Kennedy
S.D.
Ehrlich
B.E.
Boggon
T.J.
Hodsdon
M.E.
Structure of the EF-hand domain of polycystin-2 suggests a mechanism for Ca2+-dependent regulation of polycystin-2 channel activity
Proc. Natl. Acad. Sci. U.S.A.
2010
, vol. 
107
 (pg. 
9176
-
9181
)
[PubMed]
15
Schumann
F.
Hoffmeister
H.
Bader
R.
Schmidt
M.
Witzgall
R.
Kalbitzer
H.R.
Ca2+-dependent conformational changes in a C-terminal cytosolic domain of polycystin-2
J. Biol. Chem.
2009
, vol. 
284
 (pg. 
24372
-
24383
)
[PubMed]
16
Resetca
D.
Wilson
D.J.
Characterizing rapid, activity-linked conformational transitions in proteins via sub-second hydrogen deuterium exchange mass spectrometry
FEBS J.
2013
, vol. 
280
 (pg. 
5616
-
5625
)
[PubMed]
17
Li
J.
Rodnin
M.V.
Ladokhin
A.S.
Gross
M.L.
Hydrogen–deuterium exchange and mass spectrometry reveal the pH-dependent conformational changes of diphtheria toxin T domain
Biochemistry
2014
, vol. 
53
 (pg. 
6849
-
6856
)
[PubMed]
18
Zhang
H.
Cui
W.
Gross
M.L.
Mass spectrometry for the biophysical characterization of therapeutic monoclonal antibodies
FEBS Lett.
2014
, vol. 
588
 (pg. 
308
-
317
)
[PubMed]
19
Kawasaki
H.
Kretsinger
R.H.
Analysis of the movements of helices in EF hands
Proteins
2012
, vol. 
80
 (pg. 
2592
-
2600
)
[PubMed]
20
Gifford
J.L.
Walsh
M.P.
Vogel
H.J.
Structures and metal-ion-binding properties of the Ca2+-binding helix–loop–helix EF-hand motifs
Biochem. J.
2007
, vol. 
405
 (pg. 
199
-
221
)
[PubMed]
21
Grabarek
Z.
Structural basis for diversity of the EF-hand calcium-binding proteins
J Mol Biol.
2006
, vol. 
359
 (pg. 
509
-
525
)
[PubMed]
22
Englander
J.J.
Del Mar
C.
Li
W.
Englander
S.W.
Kim
J.S.
Stranz
D.D.
Hamuro
Y.
Woods
V.L.
Jr
Protein structure change studied by hydrogen–deuterium exchange, functional labeling, and mass spectrometry
Proc. Natl. Acad. Sci. U.S.A.
2003
, vol. 
100
 (pg. 
7057
-
7062
)
[PubMed]
23
Hu
W.
Walters
B.T.
Kan
Z.Y.
Mayne
L.
Rosen
L.E.
Marqusee
S.
Englander
S.W.
Stepwise protein folding at near amino acid resolution by hydrogen exchange and mass spectrometry
Proc. Natl. Acad. Sci. U.S.A.
2013
, vol. 
110
 (pg. 
7684
-
7689
)
[PubMed]
24
Rajabi
K.
Ashcroft
A.E.
Radford
S.E.
Mass spectrometric methods to analyze the structural organization of macromolecular complexes
Methods
2015
, vol. 
89
 (pg. 
13
-
21
)
[PubMed]
25
Wales
T.E.
Eggertson
M.J.
Engen
J.R.
Considerations in the analysis of hydrogen exchange mass spectrometry data
Methods Mol. Biol.
2013
, vol. 
1007
 (pg. 
263
-
288
)
[PubMed]
26
Iacob
R.E.
Engen
J.R.
Hydrogen exchange mass spectrometry: are we out of the quicksand?
J. Am. Soc. Mass Spectrom.
2012
, vol. 
23
 (pg. 
1003
-
1010
)
[PubMed]
27
Krishna
M.M.
Hoang
L.
Lin
Y.
Englander
S.W.
Hydrogen exchange methods to study protein folding
Methods
2004
, vol. 
34
 (pg. 
51
-
64
)
[PubMed]
28
Skinner
J.J.
Lim
W.K.
Bedard
S.
Black
B.E.
Englander
S.W.
Protein hydrogen exchange: testing current models
Protein Sci.
2012
, vol. 
21
 (pg. 
987
-
995
)
[PubMed]
29
Yu
Y.
Ulbrich
M.H.
Li
M.-H.
Buraei
Z.
Chen
X.-Z.
Ong
A.C.M.
Tong
L.
Isacoff
E.Y.
Yang
J.
Structural and molecular basis of the assembly of the TRPP2/PKD1 complex
Proc. Natl. Acad. Sci. U.S.A.
2009
, vol. 
106
 (pg. 
11558
-
11563
)
[PubMed]
30
Petri
E.T.
Celic
A.
Kennedy
S.D.
Ehrlich
B.E.
Boggon
T.J.
Hodsdon
M.E.
Structure of the EF-hand domain of polycystin-2 suggests a mechanism for Ca2+-dependent regulation of polycystin-2 channel activity
Proc. Natl. Acad. Sci. U.S.A.
2010
, vol. 
107
 (pg. 
9176
-
9181
)
[PubMed]
31
Zhu
J.
Yu
Y.
Ulbrich
M.H.
Li
M.H.
Isacoff
E.Y.
Honig
B.
Yang
J.
Structural model of the TRPP2/PKD1 C-terminal coiled-coil complex produced by a combined computational and experimental approach
Proc. Natl. Acad. Sci. U.S.A.
2011
, vol. 
108
 (pg. 
10133
-
10138
)
[PubMed]
32
Biekofsky
R.R.
Martin
S.R.
Browne
J.P.
Bayley
P.M.
Feeney
J.
Ca2+ coordination to backbone carbonyl oxygen atoms in calmodulin and other EF-hand proteins: 15N chemical shifts as probes for monitoring individual-site Ca2+ coordination
Biochemistry
1998
, vol. 
37
 (pg. 
7617
-
7629
)
[PubMed]
33
Capozzi
F.
Casadei
F.
Luchinat
C.
EF-hand protein dynamics and evolution of calcium signal transduction: an NMR view
J. Biol. Inorg. Chem.
2006
, vol. 
11
 (pg. 
949
-
962
)
[PubMed]
34
Chazin
W.J.
Relating form and function of EF-hand calcium binding proteins
Acc. Chem. Res.
2011
, vol. 
44
 (pg. 
171
-
179
)
[PubMed]
35
Kuo
I.Y.
Keeler
C.
Corbin
R.
Celic
A.
Petri
E.T.
Hodsdon
M.E.
Ehrlich
B.E.
The number and location of EF hand motifs dictates the calcium dependence of polycystin-2 function
FASEB J.
2014
, vol. 
28
 (pg. 
2332
-
2346
)
[PubMed]
36
Baldwin
A.J.
Kay
L.E.
NMR spectroscopy brings invisible protein states into focus
Nat. Chem. Biol.
2009
, vol. 
5
 (pg. 
808
-
814
)
[PubMed]
37
McHaourab, Hassane
S.
Steed
P.R.
Kazmier
K.
Toward the fourth dimension of membrane protein structure: insight into dynamics from spin-labeling EPR spectroscopy
Structure
2011
, vol. 
19
 (pg. 
1549
-
1561
)
[PubMed]
38
Durr
K.L.
Chen
L.
Stein
R.A.
De Zorzi
R.
Folea
I.M.
Walz
T.
McHaourab
H.S.
Gouaux
E.
Structure and dynamics of AMPA receptor GluA2 in resting, pre-open, and desensitized states
Cell
2014
, vol. 
158
 (pg. 
778
-
792
)
[PubMed]
39
Manglik
A.
Kim
T.H.
Masureel
M.
Altenbach
C.
Yang
Z.
Hilger
D.
Lerch
M.T.
Kobilka
T.S.
Thian
F.S.
Hubbell
W.L.
, et al. 
Structural insights into the dynamic process of β2-adrenergic receptor signaling
Cell
2015
, vol. 
161
 (pg. 
1101
-
1111
)
[PubMed]
40
Vafabakhsh
R.
Levitz
J.
Isacoff
E.Y.
Conformational dynamics of a class C G-protein-coupled receptor
Nature
2015
, vol. 
524
 (pg. 
497
-
501
)
[PubMed]
41
Streets
A.J.
Wessely
O.
Peters
D.J.
Ong
A.C.
Hyperphosphorylation of polycystin-2 at a critical residue in disease reveals an essential role for polycystin-1-regulated dephosphorylation
Hum. Mol. Genet.
2013
, vol. 
22
 (pg. 
1924
-
1939
)
[PubMed]
42
Cantero Mdel
R.
Velazquez
I.F.
Streets
A.J.
Ong
A.C.
Cantiello
H.F.
The cAMP signaling pathway and direct protein kinase A phosphorylation regulate polycystin-2 (TRPP2) channel function
J. Biol. Chem.
2015
, vol. 
290
 (pg. 
23888
-
23896
)
[PubMed]
43
Feng
S.
Okenka
G.M.
Bai
C.X.
Streets
A.J.
Newby
L.J.
DeChant
B.T.
Tsiokas
L.
Obara
T.
Ong
A.C.
Identification and functional characterization of an N-terminal oligomerization domain for polycystin-2
J. Biol. Chem.
2008
, vol. 
283
 (pg. 
28471
-
28479
)
[PubMed]
44
Feng
S.
Rodat-Despoix
L.
Delmas
P.
Ong
A.C.
A single amino acid residue constitutes the third dimerization domain essential for the assembly and function of the tetrameric polycystin-2 (TRPP2) channel
J. Biol. Chem.
2011
, vol. 
286
 (pg. 
18994
-
19000
)
[PubMed]
45
Shukla
A.K.
Westfield
G.H.
Xiao
K.
Reis
R.I.
Huang
L.Y.
Tripathi-Shukla
P.
Qian
J.
Li
S.
Blanc
A.
Oleskie
A.N.
, et al. 
Visualization of arrestin recruitment by a G-protein-coupled receptor
Nature
2014
, vol. 
512
 (pg. 
218
-
222
)
[PubMed]
46
Rey
M.
Forest
E.
Pelosi
L.
Exploring the conformational dynamics of the bovine ADP/ATP carrier in mitochondria
Biochemistry
2012
, vol. 
51
 (pg. 
9727
-
9735
)
[PubMed]