Protein Expression and Purification

The hCA IX-c was expressed in recombinant BL21(DE3) competent cells as previously described.²³ Briefly, E. coli cells that contain the plasmid encoding hCA IX-c were grown in 2 L of Luria broth, supplemented with 100 μg/mL ampicillin, to an OD₆₀₀ between 0.6 and 1.0, at which point hCA IX-c expression was induced by the addition of isopropyl β-D-1-thiogalactopyrano-side for ~4 h at 37 °C in the presence of 1 μM zinc sulfate. Cells containing hCA IX-c were harvested, resuspended using a glass homogenizer, and enzymatically lysed overnight at 4 °C. Expressed hCA IX-c was purified by a simple two-step process that involves (1) affinity chromatography followed by (2) size-exclusion chromatography. Affinity separation of hCA IX-c followed the same protocol that has been utilized for hCA II.²³ A gravity-fed column containing agarose resin coupled to the inhibitor p-(aminomethyl)benzenesulfonamide [p-AMBS (Sigma)] was used for affinity purification followed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) of eluents to estimate hCA IX-c purity. SDS–PAGE from affinity chromatography displayed two distinct bands that migrated at ~100 and 33 kDa (Figure S2A). The higher-MW band was not identified; the band that migrated at ~33 kDa was determined to be the monomeric hCA IX-c through mass spectroscopic analysis (data not shown). Separation of the confirmed monomer of hCA IX-c was performed using size-exclusion chromatrography on an Äkta Pure 150 Fast Protein Liquid Chromatography (FPLC) purification system (GE healthcare), equipped with a prepacked Superdex 75 10/300 GL gel filtration column (exclusion range of 3–70 kDa; GE Healthcare Biosciences AB, Uppsala, Sweden). The column was equilibrated with 50 mM Tris-HCl (pH 7.8), 100 mM NaCl buffer. The protein was eluted at a flow rate of 0.5 mL/min, yielding 1 mL fractions. Protein fractions were detected by absorbance at 280 nm, and data acquisition and processing were performed using the FPLC UNICORN software. The final protein concentration was estimated to be 12 mg/mL, as determined by UV/vis spectroscopy at 280 nm using an extinction coefficient of 36565 M⁻¹ cm⁻¹ (calculated from the amino acid sequence using ExPASy software). The hCA II used for this study was expressed and purified as previously described by Mahon et al.²⁴ to a final concentration of 25 mg/mL.

Protonography

Protonography was used to screen for hCA IX-c activity. Protonography is a newly developed technique used for qualitatively screening CA activity via monitoring proton production in an SDS gel.²⁵ In brief, the SDS gel, or protonogram, is stained with bromothymol blue (a widely used pH indicator) following electrophoresis. The dye appears blue in its deprotonated form, while its color changes to yellow as it is protonated. This is directly proportional to the number of protons produced in a solution. Therefore, in the case of CA activity, the production of H⁺ following CO₂ hydration causes a drop in pH (in areas of the gel containing CA) until the color transition point of the dye is reached (pH 6.8), resulting in gel bands containing active CA to change to a yellow color.²⁶ Here, wells of a 12% SDS gel were loaded with purified hCA II (positive control), BSA (negative control), and the hCA IX-c, mixed with Laemmli loading buffer without 2-mercaptoethanol (BME) and without boiling. The gel was run at 90 V for 30 min and then continued at 150 V until completion was reached. Following electrophoresis, the gel was soaked in 2.5% Triton X-100 for 1 h on a shaker at ~70 rpm. The gel was then soaked in 100 mM Tris (pH 8.2) containing 10% isopropanol for 10 min followed by incubation in 0.04% bromothymol blue in 100 mM Tris (pH 8.2) for 45 min. The gel was then rinsed and stored in ddH₂O. To initiate CA-mediated CO₂ hydration, dry ice was added to ddH₂O surrounding the gel, and color changes of CA active bands were monitored. Bands pertaining to active hCA II and hCA IX-c were visible after 45 s. To test the reannealing potential of the hCA IX-c, we repeated these experiments using Laemmli buffer containing 5 mM BME and 8 M urea and heated the samples at 100 °C for 15 min prior to running SDS–PAGE.

¹⁸O Mass Spectrometry

The kinetic rates of the hCA IX-c were obtained by measurement of the exhaustion of ¹⁸O from species of CO₂ at chemical equilibrium by way of membrane inlet mass spectrometry as described by Tu et al.^27,28 In brief, the reaction proceeds via the continuous measure of various isotopic species of CO₂ diffusing across a dissolved gas-permeable membrane. The membrane is submerged in the reaction solution and connected by glass tubing to a mass spectrometer (Extrel EXM-200).²⁸ The catalyzed exchange and uncatalyzed exchange of ¹⁸O between CO₂ and water at chemical equilibrium were measured in an unbuffered solution at a total substrate concentration of 25 mM bicarbonate. The reaction solution was maintained at 25 °C, and the ionic strength of the solution was normalized at 0.2 M by adding Na₂SO₄. The catalytic mechanism of isotope-labeled substrates is described as follows: (1) the dehydration of ¹⁸O-labeled HCO₃⁻ (eq 3) and (2) the protonation of the zinc-bound ¹⁸O-labeled hydroxide, forming H₂¹⁸O, which is then released into solution (eq 4).

The catalytic rate for the interconversion of CO₂ and HCO₃⁻ at chemical equilibrium is defined as R₁ (eq 5). The rate constant for the maximal interconversion of CO₂ to HCO₃⁻ is defined as k_cat^ex. K_eff^CO₂ is the effective binding constant for binding of either CO₂ or HCO₃⁻ to CA. Therefore, the ratio k_cat^ex/K_eff^CO₂ is equivalent to the catalytic efficiency (k_cat/K_M) of hydration under steady-state conditions.²⁸

The rate of proton transfer is obtained by measuring the diffusion of ¹⁸O-labeled water from the enzyme to the solvent (R_H2O). R_H2O is dependent on the donation of protons to the ¹⁸O-labeled zinc-bound hydroxide by the proton-shuttling residue (His64) as the second independent step of CA catalysis (eq 2) and reiterated in terms of labeled species in eq 4.²⁷ The rate constant for proton transfer to the zinc-bound hydroxide is defined in eq 6 by k_B, where (K_a)_His64 and (K_a)Zn_H2O are the ionization constants of the proton donor (represented by His64 from hCA II) and the zinc-bound water molecule, respectively.

Equation 5 and 6 were fitted to the data using a nonlinear least-squares methods in Enzfitter (Biosoft) and also in a log scale and fitted using nonlinear least-squares methods in GraphPad. Results are listed in Figure S3B and Table 1.

Differential Scanning Fluorimetery

Differential scanning fluorimetery (DSF) was used to assay hCA IX-c conformational stability in terms of melting temperature (T_M) against a pH range of 3.0–9.5. Samples of purified hCA II and hCA IX-c at a concentration of 0.25 mg/mL were buffer exchanged into a citrate-phosphate buffer adjusted to the desired pH. A citrate-phosphate buffer system was used for varying the pH while keeping the ionic strength constant.²⁹ The pH values selected for these experiments were 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, and 9.5. Prior to data collection, samples were incubated with 0.01% Sypro-Orange dye (no. S6651, Invitrogen Inc.) for ~30 min on ice. Melting curve assays were conducted in a quantitative PCR (qPCR) instrument (RG-3000, Corbett Research) with the temperature increasing from 30 to 99 °C, increasing at a rate of 0.1 °C/6 s. Solutions containing only buffer were also assayed in the same manner to use for background subtraction during data processing. The T_M was defined as the maximal value of the first derivative (dRFU/dT; change in fluorescence/change in temperature) of the signal that is produced in terms of relative fluorescent units (RFU). Each experiment was performed in triplicate. Results are summarized in Figure 1A and Table S2.

Open in a separate window

Figure 1

Stability of hCA IX-c and hCA II at a range of pH values. (A) T_M of hCA IX-c (red) and hCA II (blue) over a pH range from 3.0 to 11.0 as determined by DSF. (B) CD spectra of hCA IX-c and hCA II at pH 7.0 and 3.0.

Differential Scanning Calorimetry

DSC experiments were performed to confirm the thermal stability and calculate the thermodynamic parameters of hCA IX-c and hCA II at various pH values. Experiments were performed using a MicroCal VP-DSC instrument (Malvern Instruments, Worcestershire, U.K.) with a cell volume of 250 μL for each of the sample and buffer cells. Samples of each enzyme (10–20 μM) were dialyzed against a citrate-phosphate buffer corresponding to pH values of 7.0, 4.0, 3.0, and 2.0 prior to data collection. The DSC thermograms were collected in a passive feedback mode from 25 to 90 °C with a 3 min prescan thermostat, a scan rate of 60 °C/h, and a 10 s filtering period between data points. The final dialysis buffer at each pH was added to both the sample and buffer cells for an initial run to be used for buffer subtraction during data processing. Analysis of thermographs was performed using the MicroCal VP-DSC Analysis software provided by Malvern Instruments. Curves were fitted using a non-two-state model to obtain calorimetric values with one transition point defined to obtain a single peak fit. T_M values for both enzymes were obtained from the midpoints of the DSC peaks, indicating a two-state transition. The calorimetric enthalpies of unfolding were calculated by integrating the area under the peaks in the thermograms after buffer subtraction.

The denaturation enthalpies (ΔH°_m) were calculated at a given temperature using the Kirchoff’s law equations (eq 7).³⁰

The van’t Hoff enthalpy (ΔH^vH) of unfolding was calculated assuming a two-state reversible model, such that the ideal narrow peak of unfolding would result in ΔH°_m ≈ ΔH^vH. Thermograms displaying buffer-subtracted data and a non-two-state model fit are given in Figure S3, and T_M and enthalpy values are summarized in Table 2. No baseline corrections were performed to preserve raw data quality in thermograms.

Circular Dichroism

Circular dichroism (CD) was utilized to identify any major shifts (unfolding) in secondary structure in hCA IX-c and hCA II as a function of pH. We utilized a citrate-phosphate buffer (similar to that used in the experiments described above) to generate the desired pH for CD experiments. Experiments were performed using an Aviv model 430 circular dichroism spectrometer using a cuvette with a cell path length of 0.1 cm and an incubation temperature of 25 °C. Optics were continuously purged with nitrogen gas during data collection such that oxygen concentrations were less than 7 ppm. Ten scans were performed for each hCA sample in the far-UV wavelength range of 260–190 nm at intervals of 1 nm. The resulting plots were averaged and smoothed, and the CD signal from buffer at each pH was subtracted prior to further processing and analysis. The mean residue molar ellipticity was calculated using a concentration of 0.5 mg/mL for all samples.

Crystallization, Data Collection, and Processing

Crystals of hCA IX-c were grown using the sitting-drop vapor diffusion method in a 96-well IntelliPlate (Hampton Research) as described by Diáz-Torres et al.³¹ Premade crystal screens were used to screen 96 different conditions (Crystal Screen HT, Hampton Research) following incubation at 17 °C. Several conditions produced crystals of hCA IX-c; however, the largest crystals (0.1 mm × 0.2 mm × 0.3 mm), which were used for data collection, formed in 0.1 M Tris-HCl (pH 8.5) and 8% (w/v) PEG 8000 (Figure S2B). Crystals were observed after 2 weeks. Prior to data collection, crystals were cryoprotected with 20% glycerol and stored in liquid nitrogen.

X-ray diffraction data were collected at the Cornell High Energy Synchrotron Source (CHESS) on beamline F1 using a wavelength of 0.9177 Å. The data sets were collected using an ADSC Quantum 270 CCD detector at a crystal-to-detector distance of 150 mm with a 0.5° oscillation angle and an exposure time of 1 s per image. A total of 360 images were collected. The data were indexed, integrated, and scaled using HKL2000.³² Data were scaled to the orthorhombic P2₁2₁2₁ space group (unit cell parameters a = 57.9 Å, b = 102.7 Å, c = 109.0 Å, and α = γ = β = 90°) and to a high resolution of 1.60 Å with a completeness of 100.0%, and an R_sym of 10.0%. A summary of other statistics is provided in Table 3.

Catalytic Activity of hCA IX-c

Kinetic parameters of hCA IX-c were assayed using ¹⁸O mass spectroscopic analysis. The activity of hCA IX-c was compared to previously published kinetic data for hCA II,¹⁵ the catalytic domain of wild-type hCA IX (hCA IX_W),⁴³ hCA IX containing the PG domain (hCA IX_PG),⁹ and hCA XII,⁴⁴ a CA that is also expressed in tumors. The maximal k_cat/K_M of CO₂ hydration for hCA IX-c was determined by measuring a pH profile, with data appearing as a bell-shaped curve fit to eq 5, and was determined to be 13 ± 1 μM⁻¹ s⁻¹ (Table 1). This is ~10-fold lower than those of both hCA II (120 μM⁻¹ s⁻¹) and hCA IX_PG (150 μM⁻¹ s⁻¹), ~5-fold lower than that of hCA IX_W (55 μM⁻¹ s⁻¹), and ~3-fold lower than that of hCA XII (34 μM⁻¹ s⁻¹) (Table 1). It is unclear why there is a difference in catalytic efficiency between hCA IX-c and hCA IX_W, although it is possible that the oligomeric state of hCA IX_W (which is dimeric, because of the presence of the intermolecular disulfide bridge at Cys174) may in part cause an increase in catalytic efficiency via a pseudocooperativity mechanism, a phenomenon observed in other types of multimeric enzymes.⁴⁵ This is further supported by measurements of the catalytic efficiency of the hCA IX catalytic domain expressed in Sf9 cells (hCA IX_SF9), which also exists as a dimer and displays a catalytic efficiency similar to that of hCA IX_W.⁹ These data rule out a contribution to the catalytic efficiency by the N-linked glycosylation at Asn346 (full-length hCA IX numbering) in hCA IX_SF9. Similar conclusions were reached by Li et al. when studying full-length hCA IX in MDA-MB-231 human breast cancer cells.⁴⁶

Interestingly, the presence of the PG domain seems to have little effect on the pK_a of the Zn-bound solvent when comparing hCA IX_PG to both hCA IX_W and hCA IX-c. This result contradicts previous data that suggest the PG domain acts as an “internal buffer” and affects the pH-dependent catalytic efficiency.⁷ It should be noted, however, that activity assays previously reported on hCA IX_PG utilized an esterase activity assay to determine catalytic parameters, which often shows differences in kinetic parameters compared to values determined via ¹⁸O mass spectrometry.⁴⁴ In the case of hCA XII, upon comparison of results from both types of activity assays, there was an increase of ~3-fold in values for k_cat/K_M determined by esterase activity assays over ¹⁸O mass spectrometry while measuring CO₂ hydration.⁴⁴ This suggests that the value previously reported for hCA IX_PG may be directly comparable to those for hCA IX_W if the experiments were to be repeated with ¹⁸O mass spectrometry.

The kinetic profile of hCA IX-c shows that enzymatic activity, in terms of CO₂ hydration, appears to be retained under low-pH conditions (as low as pH 5.5) (Figure S3B). This phenomenon is not observed in hCA II or hCA XII, each of which shows a rapid decrease in activity at pH <6.0.^44,47 At high pH, there is a large decrease in hCA IX-c activity (Figure S3B). The reason for this is unknown as there were no observable structural changes between hCA IX-c (crystals grown at pH 8.5 and presumably inactive) and the previously published 3IAI (crystals grown at pH 4.0). An overlay of active site residues of both hCA IX-c and 3IAI displays no major structural changes; however, there are noticeable yet subtle differences observed in Tyr7. It is unclear why this difference exists; however, it is possible this may be a result of the structure of hCA IX-c being determined at a higher pH, where it shows reduced activity, and thus a slight rearrangement of residues. This observation also differs from that seen in hCA II and XII where an increase in overall activity is observed as the pH is increased.^44,47 As previously mentioned, the residues that contribute to the active site architecture in hCA IX, relative to II and XII, are highly conserved with only a few variations occurring in residues located toward the opening of the catalytic site (Table S1). It is possible that these few variations in active site residues in hCA IX-c contribute to the difference in pH-dependent activity under more basic conditions, perhaps through altering the pK_a of the active site microenvironment to favor catalysis at a lower pH. However, more work will be required to provide definitive conclusions.

Previously published data suggest that the presence of the PG domain is necessary for the enzyme to retain its activity at low pH.⁹ This conclusion is supported by the large number of acidic residues contained in the PG domain that cause an apparent overall reduction in the theoretical pI of hCA IX from 5.4 (5.1 for the wild type) to 4.5 (Table 4). In addition, previous activity assays show that hCA IX_PG retains its catalytic activity at pH <6.0.⁷ However, the overall fold and amino acid composition of the catalytic domain alone appear to be sufficient for the enzyme to retain activity at acidic pH values. This further correlates with previous experiments that assayed full-length hCA IX activity in MDA-MB-231 cells that determined the catalytic efficiency of hCA IX was comparable to that of hCA IX_W.⁴⁶

Rates of Proton Transfer

The R_H2O (eq 6), corresponding to the rate of release of H₂¹⁸O from the active site, correlates directly to the rate of proton transfer. We determined this using a pH profile with data fit to eq 6 that gave a rate constant for intramolecular proton transfer (k_B) of 1.8 ± 0.1 μs⁻¹ for hCA IX-c (Table 1). This was only slightly higher compared to the k_B of hCA II and hCA IX_W, suggesting that hCA IX-c uses a similar proton transfer mechanism and shuttle at residue His64 (200). The proton transfer rates of hCA XII were also comparable but produced the lowest k_B value (0.4 μs⁻¹) (Table 1). As mentioned above, hCA IX_PG activity was determined using an esterase activity assay, so rates of proton transfer have currently not been reported.

Structural comparison of the positions of the ordered waters and surrounding residues between hCA IX-c and hCA II (PDB entry 3KS3) shows they are in very similar positions. It should be noted that the previous structure of hCA IX, 3IAI, was crystallized with AZM, and therefore, the ordered water network could not be compared. The stabilization of these waters occurs via interactions formed between Oγ and Nα of Thr199 (332) with the Zn–OH/H₂O and deep water (DW; important for the escape of HCO₃⁻ from the active site⁴⁷), respectively, Oγ of Thr200 (333) interacting with W1, Oη of Tyr7 (143) interacting with W3a, and Nδ2 of Asn62 (198) and Oε1 of Gln67 (203) interacting with W3b (Figure 3B). Small deviations (~0.2 Å) in bonding distance between W2 and W3b and between Zn–OH/H₂O and “deep-water” (DW) in hCA IX-c were observed compared to hCA II (Figure 3C). This is most likely a result of the substitution of N67Q (N203Q; hCA II to hCA IX) (Figure 3C). The replacement of an Asn with a Gln presents a small perturbation of Oε1 toward W3b, which may induce a slight shift in its position in hCA IX relative to hCA II and further influence the positioning of DW. This observation may explain the slight difference in the rate of proton transfer (Δk_B = 1.0 μs⁻¹) between hCA IX-c and hCA II, as both waters are predicted to influence the Zn–OH pK_a (Table 1).⁴⁷ However, both hCA IX-c and hCA II utilize the same proton shuttling residue in His64 (200), indicated by the similar Zn–H₂O pK_a (Table 1 and Figure 3C). Replacement of a Lys with a Glu at position 170 (from hCA II to hCA IX, respectively) provides another explanation for the small change in proton transfer rate. Previous studies have shown that residues at position 170 influence the pK_a of His64 by causing a significant reduction from 7.2 to 6.3.⁴⁸ The reduction in the His pK_a enhances its ability as a proton donor in the dehydration direction of CA-mediated catalysis and therefore influences rates of proton transfer (eq 2).⁴⁸ The pK_a of the Zn–H₂O during proton transfer in hCA XII is much higher than those of the other isoforms (pK_a = 8.0). This alone would suggest the use of an alternative proton shuttling residue other than His64 (200). However, previous experiments using site-directed mutagenesis confirmed that this is not the case.⁴⁴ Most likely, the higher Zn–H₂O pK_a is attributed to more basic residues in the active site (compared to hCA IX and II), causing a shift in pK_a.⁴⁴ Specifically, along with Lys170 and His64, hCA XII has a Lys at position 67 [Asn in hCA II and Gln in hCA IX (Table S2)]. The influence of these residues may cause an increase in the pK_a of His64 and, because of their proximity to the ordered water network, may also act as poor proton transfer residues; a similar mechanism is utilized by hCA III, whereby position 64 is occupied by a Lys.⁴⁹ This also explains the increase in the net positive charge of the hCA XII active site (theoretical pI of 8.6) compared to those of the other hCAs (Table 4).

Thermostability of hCA IX-c

DSF, DSC, and CD were used to assay the stability of hCA IX-c at a range of pH values, and the results were compared to the stability of hCA II and previously published results for hCA XII. Stability in the case of DSF and DSC was determined by the two-state transition that can be correlated to the melting temperature (T_M) of the enzyme. DSF experiments were performed to assay a wide range of pH values from 3 to 11 (Figure 1A and Table S2). DSF allows for screening of a wide range of conditions while using minimal protein. The pH stability profiles for hCA IX-c indicate that the enzyme maintains a high level of stability throughout the pH values tested with a significant decrease in stability at pH <5.0. Despite this decrease in stability, hCA IX-c still shows some stability near physiological temperatures (~37 °C) even at a pH of 3.0 (Figure 2A and Table S2). In contrast, hCA II was shown to be stable only at pH ≥5.0 (Figure 1A). Values lower than pH 5.0 resulted in complete denaturation of hCA II, as shown by a loss of signal in the DSF experiments.

To confirm the DSF results, DSC was performed on selected pH values as it provides a higher level of accuracy in determining T_M values for enzymes. In addition, DSC was used to calculate thermodynamic parameters to assess the nature of unfolding of both hCA IX-c and hCA II. The major transition peak for each enzyme displayed a single unfolding dynamic (Figure S5). The transitions were calculated to be endothermic from eq 3 and were centered at the melting temperature (T_M), with a maximal heat capacity (C_p) occurring at the midpoint of the peak. A summary of the calculated T_M and enthalpies values is presented in Table 2. The van’t Hoff enthalpy (ΔH^vH) of unfolding was calculated assuming a two-state reversible model (based on previous data from the observation of unfolding transitions of hCA II),³⁰ such that an ideal narrow peak of unfolding would result in ΔH°_m ≈ ΔH^vH and larger transition peaks in lower ΔH^vH values (Table 2). For both hCA II and hCA IX-c at pH 7.0, the ΔH^vH is within the same order of magnitude as ΔH°_m. This relative equivalence confirms that the unfolding process is most likely a reversible two-state transition for both hCA IX-c and hCA II at neutral pH. These data correlate with protonography experiments that suggest that hCA IX-c (and hCA II) can refold once under stable conditions and can regain activity. At pH 4.0 and 3.0, hCA II did not exhibit unfolding, indicating that the protein did not undergo a transition at these pH values and, most likely, the enzyme was denatured prior to the experiment. However, hCA IX-c displayed a T_M transition at 51.9 °C for pH 4.0, a T_M of 38.4 °C for pH 3.0, and a T_M of 30.1 °C for pH 2.0. The thermogram at pH 3.0 had an unusual spike in signal prior to the unfolding peak, which likely indicates that a population of protein molecules were undergoing an unfolding transition at or below the starting recorded temperature of 28 °C (because of the 3 °C thermostat from 25 to 28 °C), and this initial transition could not be fitted using the buffer subtraction. Similarly, the peak obtained for hCA IX-c at pH 2.0 indicated that while 50% of the enzyme population was unfolded by 30.1 °C, unfolding began at or before data collection at 28 °C (Figure S5). Interestingly, the enthalpy values for hCA IX-c at pH 4.0 were equivalent (Table 2), indicating that the enzyme was able to refold at this pH. This is not mirrored in hCA II, which appeared to be unfolded prior to DSC measurements at pH ≤4.0. In addition, on the basis of previously published data, it is assumed that the same trend would be true for hCA XII, which shows denaturation at pH ~4.5 (Table 2).⁵⁰ Unlike the enthalpy values obtained for hCA IX-c at pH 7.0 and pH 4.0, the ΔH°_m and ΔH^vH values were not equivalent at pH 3.0 and 2.0 (Table 2). This indicates that in these lower-pH ranges, while the enzyme is able demonstrate some stability and can unfold at temperatures above ambient, this process is likely irreversible, and reannealing may not occur.

The T_M values determined from DSC show that hCA IX-c is the most stable isoform with a T_M of 59.3 °C at pH 7.0, compared to values of 56.3 and 53.0 °C for hCA II and XII, respectively (Table 2). As predicted, hCA II is no longer stable at pH <5.0, resulting in no determinable T_M for low pH values. Interestingly, hCA XII, which also exists in acidic tumor microenvironments, had reduced acid stability compared to that of hCA IX-c via a T_M of 45 °C at pH ~4.5 and was completely denatured at pH 3.0.⁵⁰ The presence of the PG domain does not seem to influence the stability of hCA IX. This is displayed by a T_M comparable to that of hCA IX-c at neutral pH (Table 2), implying that the presence of the PG domain in hCA IX does not contribute significantly to the overall stability of the enzyme (Table 2). However, it should be noted that hCA IX_PG has not been assayed over a broad pH range. It is currently unknown how this domain might influence the stability of the enzyme in an acid environment, although it is predicted that the majority of the acid stability of hCA IX, based on our current data, is a direct result of the conformation and chemical structure of its catalytic domain.

To confirm that hCA IX-c maintains its conformation at low pH, we utilized CD to assay hCA IX-c and hCA II at neutral and acidic pH values of 7.0 and 3.0, respectively (Figure 1B). As shown in Figure 1B, the ellipticity trends for the hCA IX-c samples at pH 7.0 and 3.0 are fairly similar to both possessing minima at 208 nm, indicating that the structural dynamics and folding of hCA IX-c remain largely unchanged at low pH. Conversely, the spectra for hCA II show a large shift in ellipticity from pH 7.0 to 3.0, especially around 200 nm where the hCA II signal becomes negative with low pH. As a reference standard, we note that the spectra observed for the hCA II pH 7.0 sample are consistent with previously collected CD data for the enzyme.⁵¹ The difference in the spectra for hCA II at pH 3.0 indicates possible changes in β-sheet content after incubation in an acidic environment as the negative peak changes from 210 nm at pH 7.0 to 202 nm at pH 3.0. This further correlates to DSF and DSC data that suggest a complete denaturation of hCA II at pH <5.0.

It has been predicted that several structural elements contribute to the pH stable conformation of hCA IX-c. Similar to other hCAs, hCA IX-c contains an intramolecular disulfide bond (Cys156–Cys336; full-length hCA IX numbering). This intramolecular disulfide bond is also present in hCA XII, which may explain its higher stability at pH <5.0 compared to that of hCA II (Table 2). This is further supported by data from Boone et al., in which the same disulfide bond was engineered into hCA II and increased its stability under acidic conditions.³⁰ The more compact structure of hCA IX-c relative to that of hCA II may also contribute to stability. In that regard, Diaz-Torres et al. have proposed that CAs from extremophilic organisms have more compact structures that may underlie their enhanced stability.³¹ However, the most striking or unusual characteristic of hCA IX-c is that it contains a surface far more net negatively charged than that of hCA II or XII (Table 4). This is revealed by observing the predicted theoretical pI of hCA IX-c [pI = 5.4 (Table 4)]. The electrostatic potential of the hCA IX-c surface supports this by showing large regions of negatively charged residues compared to hCA II and XII (Figure S4). In addition, the ratio of negatively charged residues (Asp and Glu) to those most likely to have a positive charge at low pH (Arg, Lys, Asn, Gln, and His) is increased in hCA IX-c (0.73) relative to that in hCA II and XII where the ratio is reduced (0.53 and 0.44, respectively) (Table 4). The increase in net negatively charged residues on the surface of hCA IX-c, upon protonation as pH levels of the solution drop, delays the point at which excess positive charge induces a destabilizing effect.^52–54

Alternatively, the presence of net positive charges (mostly from Arg and Lys) on protein surfaces has been correlated with protein instability, because, as the pH becomes more acidic, these residues will become protonated and create charge–charge repulsion forces that will act to destabilize the protein.⁵⁵ Interestingly, hCA IX-c has two added positively charged residues (Lys210 and Lys258) compared to the wild type. This may imply that hCA IX-c may actually be less acid stable than wild-type hCA IX, further highlighting the acid resistant nature of the enzyme’s overall fold. The same is seen in hCA IX_PG because of the large number of Glu residues in the PG domain, which increases the ratio of net negatively charged residues from 0.73 to 1.12 (Table 4). However, the limited effect of the PG domain on hCA IX stability raises questions about this domain’s contribution to the conformational rigidity of the enzyme at low pH. It should be noted that the presence of counterions in a buffered solution could possibly contribute to the observed thermostability of hCA IX-c as seen in experiments by Goto et al.;⁵⁶ however, this has not been tested in the study presented here. Nonetheless, it is apparent that hCA IX-c exhibits an acid stability higher than those of hCA II and XII.

Funding

This research was supported in part by National Institutes of Health Grants GM 25154 and CA16584.

We acknowledge the staff at the Cornell High Energy Synchrotron Source (CHESS) for assisting with data collection and the University of Florida Interdisciplinary Center for Biotechnology Research (ICBR) for assisting with sequencing experiments. R.M. also thanks Drs. David Silverman and Claudiu Supuran for their valuable input and support over the years.