Enzyme–inorganic nanoporous materials: Differential scanning calorimetric studies and protein stability

Enzyme–inorganic nanoporous materials: Differential scanning calorimetric studies and protein stability

Available online at www.sciencedirect.com Microporous and Mesoporous Materials 109 (2008) 223–232 www.elsevier.com/locate/micromeso Enzyme–inorganic...

491KB Sizes 0 Downloads 8 Views

Available online at www.sciencedirect.com

Microporous and Mesoporous Materials 109 (2008) 223–232 www.elsevier.com/locate/micromeso

Enzyme–inorganic nanoporous materials: Differential scanning calorimetric studies and protein stability Akhilesh Bhambhani, Challa V. Kumar

*

Department of Chemistry, U-60, University of Connecticut, 55 North Eagleville Road, Storrs, CT 06269-3060, USA Received 27 February 2007; received in revised form 20 April 2007; accepted 25 April 2007 Available online 6 May 2007

Abstract Direct assessment of the thermodynamic stabilities of enzymes bound to solids is essential to understand the factors that control bound enzyme stability. Here, the first reports of the thermodynamic stabilities of enzymes/proteins which are bound to an inorganic solid [a-Zr(HPO4)2 Æ H2O, abbreviated as a-ZrP] are described. The thermal denaturation of hen egg white lysozyme (Lys), met-myoglobin (Mb) and met-hemoglobin (Hb) bound to a-ZrP occurs over a wide range of temperatures (50–100 C). This is in contrast to the behavior of the free enzyme/proteins in the solution, which indicated sharp transitions at their respective denaturation temperatures. Denaturation of the bound protein depended on the scan rate and the denaturation process was kinetically controlled. At rapid scan rates (2 C/min), for example, the thermal profiles of the intercalated proteins became sharper while the free proteins indicated little or no changes. Careful analysis of the calorimetric data provided a clear distinction between the moving-boundary model and the uniform distribution model for protein binding. Calorimetric data also revealed that a distribution of thermodynamic states or kineticallyslow forming states are important in the denaturation. While the thermal denaturation of Mb bound to a-ZrP indicated a significant extent of reversibility, Lys and Hb did not. The solid stabilized a fraction of the intercalated protein, and efforts will be focused to maximize this portion. Improved thermal stabilities are important for biosensor or biocatalysis applications of enzyme–inorganic materials.  2007 Elsevier Inc. All rights reserved. Keywords: Lysozyme; Met-myoglobin; Met-hemoglobin; Zirconium (IV) phosphate; Differential scanning calorimetry; Protein–inorganic materials

1. Introduction Enzymes are excellent biocatalysts with high specificity, selectivity, and efficiency [1]. The exciting possibility of using enzymes in industry or laboratory is severely limited due to the high price and poor stability of most enzymes. Enzymes bound to solid supports can overcome some of these limitations, but binding to solids often distorts the enzyme structure and decreases catalytic efficiency [2,3]. Improved performance of immobilized enzymes is essential to achieve economical and widespread use of enzymes [4]. One approach to improve the performance of bound enzymes is to enhance their thermodynamic stability. Increased stability is important for a greater retention of *

Corresponding author. Tel.: +1 860 486 3213. E-mail address: [email protected] (C.V. Kumar).

1387-1811/$ - see front matter  2007 Elsevier Inc. All rights reserved. doi:10.1016/j.micromeso.2007.04.048

bound enzyme structure, longer shelf life and resistance to denaturation under the operational conditions [5,6]. Direct assessment of the thermodynamic stabilities of immobilized enzymes is sparsely reported in the literature [7], but these data are essential to design rational approaches to improve enzyme stability. Stabilizing the native state and destabilizing the denatured state of the immobilized enzymes, for example, can provide a general approach to improve the thermal stabilities of enzymes. The free energy gap (DGD ) between the native state (N) and the denatured state (D) controls enzyme stability. Increasing this gap will increase bound enzyme stability in an exponential manner. This is because the equilibrium constant for the denaturation process (Kd) is related to the denaturation free energy (DGD ) by the relation DGD ¼ RT lnðK d Þ. Therefore, even small increments in DGD will enhance the enzyme stability to a large extent.

224

A. Bhambhani, C.V. Kumar / Microporous and Mesoporous Materials 109 (2008) 223–232

However, detailed thermodynamic studies of enzymes bound to well-defined solid supports are needed to design strategies for improving bound enzyme stability. In this report, we describe our first attempts to characterize the thermodynamic denaturation profiles of a small set of representative proteins, which are intercalated in layered metal phosphate. The layered solid, a-Zr(HPO4)2 Æ nH2O (abbreviated as a-ZrP) was chosen for the current studies due to our continued interest in this material for biocatalytic applications [8]. a-ZrP and its analogs are lamellar materials where a layer of ZrIV ions is sandwiched between two adjacent layers of phosphate (a-ZrP) [9]. Each phosphate bridges three ZrIV ions, and each ZrIV is octahedrally coordinated to six oxygen atoms from the phosphate groups. The remaining OH groups, one from each phosphate, are oriented perpendicular to Zr(IV) plane, and this array of OH functions on each side of the platelets are expected to provide strong contributions to enzyme–solid interactions. These lamellar solids are topologically uniform, chemically inert (at pH 7 or below) and thermally stable. Previous work showed that a number of proteins and enzymes can be intercalated in the galleries of a-ZrP, under mild conditions (pH 7, 25 C, 0.5 h), by the exfoliation method (Chart 1) [10–18]. Intercalation was achieved by treating a-ZrP with tetrabutylammonium (TBA, pH 7) chloride, followed by exposing the exfoliated platelets to enzyme solutions. Surprisingly, the intercalated biocatalysts retained their structure and activities to a significant extent [10]. In two cases, the immobilized biocatalysts exhibited enzymatic activities well above room temperature (>85 C) [13], and these observations suggest that bound proteins are stabilized by the solid. However, there have been no direct measures of the thermodynamic stabilities of these systems. Here, we build on our earlier work by documenting the differential scanning calorimetric studies of a few proteins intercalated in the galleries of a-ZrP. Lysozyme (Lys), met-myoglobin (Mb), and met-hemoglobin (Hb) are chosen for the current studies, as representative examples. Lysozyme is a hydrolytic, antibacterial enzyme with potential to hydrolyze cells walls [19]. Even though Hb and Mb do not function as enzymes in biological systems, their ability to oxidize a number of biological substrates, in the presence of peroxide, is well known [20]. Current results show that a significant fraction of the bound protein is stabilized by a-ZrP and the extent of stabilization depended on the loading.

TBA

Lys

Lys Lys

Chart 1. Exfoliation of a-ZrP stacks with tetrabutylammonium chloride (TBA), protein binding (Lys = lysozyme) and re-assembly.

2. Experimental 2.1. Protein sources Lysozyme (hen egg white), Mb (horse heart) and Hb (bovine) were from Sigma Chemical Co. (St. Louis, MO). The protein samples were >99% pure, as tested by SDS PAGE experiments. All protein solutions were prepared in potassium phosphate buffer (10 mM K2HPO4, pH adjusted to 7.2) unless stated otherwise. Synthesis of Zr(IV) phosphate was carried out by following a reported method [9]. The FTIR spectra and the powder XRD patterns of the ˚ for a-ZrP) matched with those sample (d-spacing of 7.6 A reported [21]. 2.2. Exfoliation of the layered solids and protein intercalation Exfoliated a-ZrP (2%) suspensions were prepared by mixing 0.1 g of a-ZrP in 5 ml of distilled water with stoichiometric amounts of tetrabutylammonium hydroxide (40% by weight, in water). The protein/a-ZrP samples were prepared by exposing these suspensions to protein solutions, at ambient conditions [9,18]. Any free protein left was removed by centrifugation followed by re-suspension in the phosphate buffer. The intercalated products were characterized by XRD, FTIR (Table 1) and activity studies, prior to the calorimetric scans. The solid was re-suspended in phosphate buffer (0.01% by weight of a-ZrP) for calorimetric studies, and the protein/a-ZrP suspensions scattered <5% of incident light above 400 nm. 2.3. XRD studies Protein/a-ZrP suspensions (50–100 ll) were spotted on glass slides and air-dried overnight. A Scintag Model 2000 diffractometer was used to record the XRD patterns of the protein/a-ZrP samples (scan rates of 2 C/min, nickel filtered Cu Ka radiation). The interlayer separations were measured from the 0 0 l reflections (l = 1, 2, etc. Table 1) and these values matched with those reported [8,9]. 2.4. Electron microscopy Scanning electron micrographs were imaged on a Leo/ Zeiss 982 digital field emission electron microscope interfaced with a Tracor Northern 30 mm2 beryllium window detector and a model 2000 spectral analysis system [16]. Aqueous suspensions of a-ZrP (6 mM), Lys/a-ZrP (29 lM Lys, 3 mM a-ZrP), Mb/a-ZrP (50 lM Mb, 10 mM a-ZrP) and Hb/a-ZrP (9 lM Hb, 6 mM a-ZrP) were dried in a Savant integrated speedvac system. These concentrations refer to samples from which the free protein has been removed. Dry powders were mounted on aluminum stubs with the aid of carbon coated double-sided tape. Samples were then coated with Au nanoparticles and SEM images captured in digital format.

A. Bhambhani, C.V. Kumar / Microporous and Mesoporous Materials 109 (2008) 223–232 Table 1 Observed d-spacings from the powder X-ray diffraction and the FTIR data of the protein/a-ZrP materials ˚ ˚ Protein Protein size /A d-spacing/A Amide I/Amide II band*cm

No protein Lysozyme Myoglobin Hemoglobin

– 32 · 32 · 55 (40) 30 · 40 · 40 (37) 53 · 54 · 65 (57)

7.6 47 54 63

225

Ref.

Free

Bound

– 1650/1520 1650/1530 1653/1520

– 1650/1520 1650/1530 1653/1520

[9] [8] [8] [8]

The d-spacings correlate well with the known diameters of the corresponding proteins. Protein size is estimated from the known crystal structures, and the average dimensions are given in parentheses.

2.5. Spectral measurements The absorption spectra were recorded on a Perkin– Elmer Lambda 3B spectrophotometer or HP 8453 diode array spectrometer in glass or quartz cuvettes (1 cm path length). a-ZrP suspensions of appropriate concentrations were used as reference samples to compensate for the light scatter by a-ZrP (<5% above 400 nm). Upon exfoliation with tetrabutyl ammonium hydroxide, a-ZrP suspensions became translucent and scattered less light. 2.6. CD studies A JASCO model 710 spectropolarimeter (Jasco Inc., Easton, MD) was used to record the CD spectra from 190 to 300 nm. All protein/a-ZrP suspensions were diluted as needed for the CD measurements. Losses in the CD signal, due to light scattering by a-ZrP, were corrected by placing an appropriate suspension of a-ZrP in front of the cuvette containing the free protein solutions. This way, the free protein spectrum has been modulated to an extent nearly similar to that of the bound protein. 2.7. Protein activity assay Protein-inorganic materials were further characterized in activity studies, prior to calorimetric studies. Lysozyme activity was monitored by a reported procedure, after minor modifications [22]. Lys (60 lM) was incubated with various concentrations of glycol chitin (0.05% to 0.30 wt%), at 40 C for 1.25 h. To 1.5 ml of the solution, 2 ml of 1.5 mM potassium ferricyanide (in 0.5 M sodium carbonate) was added, heated in boiling water for 30 min, cooled to room temperature and the product absorption at 420 nm recorded. Bound enzyme activity was determined in a similar way (7 lM Lys/1.5 mM a-ZrP). Peroxidase-like activities of heme proteins were followed by using H2O2 as the oxidant and guaiacol as the substrate [23]. The product absorbance at 470 nm was monitored as a function of time. Sample spectra were recorded at the end of the reaction and product formation confirmed. Protein/ a-ZrP samples contained 0.5–2 lM Mb (or Hb), 1 mM aZrP, 3 mM guaiacol, and the reactions initiated by adding H2O2 (0.5 mM, final concentration). Lineweaver–Burke plots were constructed using these data, and Km as well

as Vmax values have been calculated from these plots in the following equation [24]: 1=V ¼ K m =ðV max ½SÞ þ 1=V max

ð1Þ

In Eq. (1), Km is the Michaelis constant, Vmax is the maximum reaction velocity, and [S] is the substrate concentration. 2.8. DSC measurements Thermal denaturation experiments were performed on a Calorimetry Sciences Corporation (CSC, Utah) 6100 Nano II differential scanning calorimeter (DSC). The calorimeter cell had a volume of 0.299 ml, and the system interfaced to a personal computer (IBM-compatible). In a series of DSC scans, both the cells were first loaded with buffer, equilibrated at 20 C for 10 min and scanned from 20 to 100 C at a scan rate of 2 C/min. The buffer versus buffer scan was repeated once, and upon cooling, the sample cell was emptied, rinsed and loaded with the protein solution (free or bound to a-ZrP). The sample and reference solutions were degassed for at least 5 min, prior to loading, and then the cells filled carefully while avoiding bubble formation. A constant pressure of 3 atm was maintained over the samples to prevent evaporation during the scan. A background scan, which was recorded by loading buffer in both cells, was subtracted from each sample scan. Each scan was also normalized with respect to protein concentration. a-ZrP itself did not produce any thermal transitions in 10–120 C. Reversibility of the transitions was examined by recording the DSC trace during the second heating cycle and all adsorbed samples indicated only irreversible transitions on these timescales. However, prolonged cooling of the samples, after thermal denaturation indicated significant recovery of the protein structure [11,14]. The peak transition temperature Tm, the full-width at half-maximum (FWHM), the temperature where the denaturation begins (Td), and the observed area under the transition (integral CpdT) are model independent. Since protein denaturation, after intercalation, is not reversible on the time scales of the DSC scans, the estimates of the area under the curve include all contributions from all thermally induced processes [25]. This many not be limited to the denaturation process and may include the formation of intermediate states.

226

A. Bhambhani, C.V. Kumar / Microporous and Mesoporous Materials 109 (2008) 223–232

The excess molar heat capacities of the samples were calculated using the mean molecular mass of 14.3 kDa for Lys, 16.7 kDa for Mb, and 64.5 kDa for Hb, respectively. The partial specific volume of the protein was calculated from the amino acid sequence as 0.73 ml/g, by following a method described in the literature [26]. Each value represents the average of at least three separate measurements, and the transition temperatures of the immobilized biocatalysts were reproducible within ±3 C. Since the pKa of the buffer is a function of temperature, this contribution to the thermodynamic parameters was also assessed. For example, the temperature coefficient of the pKa of phosphate buffer is 0.0028 per C, while that of citrate is near zero [27]. Using Mb as a model protein, therefore, we examined the denaturation in phosphate as well as citrate buffers (5 mM citrate, pH adjusted to 7.2) while maintaining the same ionic strength. The thermodynamic parameters (DH, DS and Tm) of free Mb, in 5 mM citrate buffer, matched with the corresponding parameters in phosphate buffer (@ scan rates of 1 or 2 C/min). Therefore, the temperature dependence of the phosphate buffer pKa did not influence the protein denaturation to a significant extent. 3. Results and discussion Differential scanning calorimetry provides the most direct experimental data to evaluate the energetics of protein denaturation [28], and it allows continuous measurement of partial heat capacity of a system (Cp) as a function of temperature [29]. Even a minor perturbation in protein structure, such as mutation of a single residue, can have a dramatic effect on protein stability [30]. DSC,

therefore, provides a sensitive and quantitative method to assess thermal stabilities of biomolecules. Quantitative descriptions of the thermal denaturation of the biomolecules can be constructed from these data. Thermal denaturation studies of a few representative biocatalysts, intercalated in the galleries of a-ZrP, are reported here for the first time. The samples were prepared and characterized as reported earlier [8], but a brief description follows. 3.1. Enzyme intercalation and SEM data The biocatalysts are intercalated by the exfoliation method (Chart 1) and intercalation was confirmed by powder XRD (Table 1). These powder patterns matched with those reported. After the intercalation, the samples retained their lamellar structure as evidenced from the scanning electron micrographs (Fig. 1a–d, size bar of 1 lm). 3.2. Activity and spectral studies Prior to calorimetric studies, the intercalated proteins were characterized by activity, FTIR and CD studies. The hydrolase activity of lysozyme and the peroxidase-like activity of the heme proteins indicated a significant retention of enzyme activities. The Km and Vmax values for Lys/a-ZrP (0.5 mM and 0.6 lM/s, respectively) and Hb/ a-ZrP (0.11 mM and 42 nM/s, respectively) matched with those reported earlier [8]. A significant retention of protein secondary structure, subsequent to intercalation, was confirmed by FTIR (Table 1) and circular dichroism (CD) spectra (Supplementary Material #1). These data matched with those reported ear-

Fig. 1. Scanning electron micrographs of (a) a-ZrP (30,000·), (b) Lys/a-ZrP (30,000·), (c) Mb/a-ZrP (30,000·) and (d) Hb/a-ZrP (20,000·), respectively. The size bar of 1 lm is shown as the thick line in the micrographs. The layered structure is clearly indicated in all the micrographs.

A. Bhambhani, C.V. Kumar / Microporous and Mesoporous Materials 109 (2008) 223–232

lier [8]. Lysozyme indicated a significant distortion of its structure while Mb and Hb show essentially no major changes. These differences between the bound lysozyme and heme proteins are also noticeable in the thermal denaturation studies presented below. 3.3. Differential scanning calorimetry The DSC profiles of free proteins are recorded first (Fig. 2) and these indicated sharp transitions at the corresponding, reported, denaturation temperatures [31–33]. After subtracting the baseline, the experimental thermograms were integrated to estimate the enthalpy changes associated with the denaturation transition. The areas under the curve (DHdenaturaiton) are model independent parameters and the observed values are collected in Table 2. Note that in the case of free lysozyme, the denaturation was reported at pH 7.0 (Hepes buffer) and our values at pH 7.2 (phosphate buffer) are comparable, with only minor differences. The denaturation curves of Mb did not indicate any signs of exothermic transitions due to protein aggregation, under our experimental conditions. While the estimation of DH from the DSC data is model independent, integrated area under the transition, calculation of the entropy change (DSdenaturation) requires the assumption that the denaturation is reversible. This assumption is valid for the free Lys and Mb, under our

a

1ºC/ min

6

b C p (kcal/ K.mol)

Cp (kcal/ K.mol)

pH conditions, but the denaturation of Hb is only partially reversible [29]. Accordingly, the entropy changes have been estimated for the free proteins with this assumption, and these data matched well with reported values (Table 2). Next, the DSC thermograms of intercalated proteins were examined to assess their thermal stabilities. Thermodynamic data on proteins which are immobilized on solid surfaces are very limited[7] and current studies are an attempt to narrow this gap. The protein/a-ZrP biocatalysts were suspended in phosphate buffer, subsequent to removal of the free protein by centrifugation, and bound protein denaturation has been monitored by DSC. The thermograms, thus obtained, are overlaid over those of the corresponding free proteins (Fig. 3). Several differences between the thermograms of the free and bound protein are noteworthy. Intercalated biocatalysts underwent thermal denaturation over a wider temperature range than the native protein, denaturation of the intercalated proteins commenced at temperatures much lower than the corresponding free proteins, and fraction of the bound protein denatured well above the denaturation temperature of the native protein, all the way up to 100 C. Therefore, a fraction of the bound protein is more stable than the free protein, which is consistent with our earlier studies [13]. The broad transitions, observed here, may imply the presence of a heterogeneous population of the bound protein states. This is surprising for two reasons.

1ºC/ min 15

Lys

5 4 3 Lys/ZrP 2

227

Mb

10

5

Mb/ZrP

1 0 20

40

60

80

100

Temperature / o C

0 20

40

60

80

100

Temperature / o C

Fig. 2. The molar heat capacity profiles of (a) Lys, 30 lM (dashed line) and Lys, 54 lM/a-ZrP, 3 mM (thick line) and (b) Mb, 50 lM (dashed line) and Mb, 50 lM/a-ZrP, 10 mM (thick line) in 10 mM K2HPO4 pH 7.2, at a scan rate of 1 C/min.

Table 2 Thermodynamic parameters and the full-width at half-maximum (FWHM) of free and intercalated proteins (scan rate of 1 C/min) Protein

Buffer

pH

DH (kJ/mol)

Tm (K)

DS (kJ/K mol)

FWHM (K)

Lys Lys* [31] Lys/a-ZrP Mb Mb* [33] Mb/a-ZrP Hb Hb* [34] Hb/a-ZrP

Phosphate Hepes Phosphate Phosphate Phosphate Phosphate Phosphate 0.9% Saline Phosphate

7.2 7.0 7.2 7.2 7.0 7.2

427 ± 8 439 ± 8 700 ± 140 360 ± 33 298 600 ± 130 636 ± 8 514, 368, 191(±4) 2134 ± 251

347.8 ± 0.1 346.7 ± 0.1 326 ± 2 356 ± 1 342.7 355.1 ± 1.5 340 ± 1 340.5, 335.8, 328.8 328.1

1.2 ± 0.04 1.3 – 1.0 ± 0.08 1.0 – 1.8 ± 0.2 – –

9 ± 1.8 8.0 50 ± 6 4.6 ± 0.6 10 53.0 ± 5 10 – 50 ± 5

The literature values are indicated by

7.0 7.2 *

and in some cases, the values were extracted from the published thermograms.

228

A. Bhambhani, C.V. Kumar / Microporous and Mesoporous Materials 109 (2008) 223–232

a

15

b

Lys, 30 µM

3

Lys 54.4 µM/α -ZrP, 3 mM

0

Cp (Kcal/ Kmol) (Mb)

C p (kcal/K.mol)

C p (kcal/K.mol)

5

o

2

1.0 C/min

o

1.5 C/min 1

2.0 oC/min

0 40

60

80

Temperature / o C

100

10

40

60

80 o

Temperature / C

100

15

10

Mb/ 2oC/min Mb/α -ZrP (1oC/min) Mb/ α -ZrP (2 oC/min)

5

0 20

5

Cp (Kcal/ Kmol) (Mb/ZrP)

10

20 Mb/ 1oC/min

2.0oC/min 1.5oC/min 1.0oC/min

c

0 40

60

80

100

o

Temperature / C

Fig. 3. Thermograms of (a) free lysozyme at scan rates of 1 C/min, 1.5 C/min and 2 C/min. Only minor changes are evident. (b) The thermograms of Lys/a-ZrP (54 lM Lys, 3 mM a-ZrP, 10 mM K2HPO4, pH 7.2) at scan rates of 1, 1.5 and 2 C/min. (c) Thermograms of Mb/a-ZrP (52 lM Mb, 10 mM aZrP, 10 mM K2HPO4, pH 7.2) at scan rates of 1 and 2 C/min, as marked. The left Y-scale in panel 2 C is for free Mb while the right is for bound Mb.

One is that the solid support is chemically uniform and topologically homogeneous at the molecular level [9], and secondly the CD spectra, especially those of Mb/a-ZrP and Hb/a-ZrP, are essentially superimposable with those of the corresponding free proteins (Supplementary Materials #1). Therefore, a heterogeneous population of the bound protein was not anticipated. Since most spectroscopic methods provide properties averaged over the ensemble, spectral methods may not resolve different thermodynamic states. The observed broadening of the DSC profiles could be due to a distribution of protein orientations, a varying number of protein neighbors for each bound protein molecule, and binding to defect sites or a heterogeneous lattice. Note that earlier studies indicated that the predominant orientation of Hb is with its long axis oriented perpendicular to the galleries [17]. Even though this later result is consistent with the powder XRD data, it does not rule out the existence of other orientations. A distinct possibility is that broad thermal transitions could be due to a kinetic control of the denaturation process where one or more intermediate states are formed during the thermal scan (discussed below). Analysis of the DSC data of the intercalated proteins indicated significant increases in the area under the curves (Table 2, apparent DH) when compared to those of the corresponding free proteins. Since the calorimeter measures the net heat absorbed or released, the greater heat change observed with the bound protein can be attributed, among other process, the formation of the intermediates states via kinetically controlled denaturation [35–37]. This later hypothesis was tested by examining the effect of scan rate on the thermograms. 3.4. Scan rate dependence The population of slow-forming intermediate states can be suppressed by higher scan rates, and this could result in considerable sharpening of the thermal transitions or decrease the net enthalpy change accompanying the dena-

turation. The thermal profiles of free and bound proteins, recorded at various scan rates, are presented in Fig. 3. The scan rate did not influence the thermal profiles of free lysozyme or free Mb to a significant extent (Fig. 3, some data not shown). The data confirm that the instrument does not have any significant lag in the temperature at these scan rates. The results are consistent with a single native state and the validity of the two-state model for the thermal denaturation of the free proteins. Even at higher scan rates there is no significant changes in the Tm values. These data are consistent with the absence of any slowforming intermediates in the denaturation. On the other hand, the thermograms of Lys/a-ZrP depended strongly on the scan rate, and the thermal profile was clearly resolved into two transitions at a scan rate of 2 C/min (Fig. 3b). This dependence on scan rate is remarkable. Consistent with the above hypothesis, the overall area under the scan decreased substantially at higher scan rates, and this could be due to a decrease in the population of the kinetically slow-forming states. The scan rate dependence of the DSC profiles clearly shows that the denaturation process involves kinetically slow steps, and faster scan rates resolved the two distinct transitions. Interestingly, increased rate decreased the fraction of the protein which denatured at higher temperatures and we conclude that the higher temperature transition arises from species populated by the kinetically slow steps. Note that the transition at the higher Tm (90 C) is substantially greater than that of the free protein (75 C). As in the case of lysozyme, Mb/a-ZrP samples also indicated scan rate dependence on their thermograms (Fig. 3c). The second, higher temperature transition appears to resolve to some extent at faster scan rates (@ 2 C/min), but the overall area under the transition is decreased substantially. These data clearly support the above hypothesis that bound proteins are undergoing kinetically slow conversion to intermediate states and faster scan rates may suppress formation of these states and decrease the overall enthalpy change recorded by the calorimeter. Please note that some of the broadening of the thermograms could also

A. Bhambhani, C.V. Kumar / Microporous and Mesoporous Materials 109 (2008) 223–232

result from a heterogeneity in the bound protein population, discussed above. Next, we examined the effect of loading on the thermal denaturation of the bound proteins to test if the broadening depended on the loading. 3.5. Effect of loading One important assumption made in the above interpretations is that protein–protein interactions are not significant and that denaturation involves individual protein molecules isolated from each other. Validity of this assumption was tested by recording the thermograms as a function of loading. Two extreme binding models can be distinguished from these studies. One of them is the moving boundary model [38] which proposes that proteins are bound at the solid surface in clusters, and the clusters grow as the loading is increased. Protein–protein interactions play an important role in controlling protein behavior within these clusters and protein denaturation will be independent of loading. The uniform distribution model, an alternative description, allows for maximum separation between adjacent protein molecules and protein–protein interactions become important only when binding begins to saturate. According to the later model, protein behavior may change as a function of loading, particularly when the inter-protein interactions become important at high loadings. The effect of loading may test the applicability of these two models to the current systems. The denaturation profiles strongly depended on lysozyme loading, and the data have been normalized to the amount of protein present in the sample (Fig. 4). At very low loadings (phosphate to protein ratio of 692:1, thin line, scan rate of 1 C/min), the lower temperature transition became much more prominent, and the overall area under the transition diminished substantially. As the loading increased, the contributions of the higher temperature transition became more prominent (Fig. 4, thick line, phosphate to protein ratio of 110:1, scan rate of 1 C/min) and the area under the curve increased. PO 3H2: Lys 444

C p (kcal/K.mol)

4

222 110 2

692 0 20

40

60

80

100

Temperature / 0 C

Fig. 4. Thermograms of Lys/a-ZrP as a function of the ratio of metal phosphate to Lysozyme (692:1–110:1 as phosphates to protein) at a scan rate of 1 C/min. Lys/a-ZrP (9 mM a-ZrP, 26 lM Lys, thin line, 692:1), Lys/a-ZrP (6 mM a-ZrP, 26 lM Lys, thin dotted line, 444:1), Lys/a-ZrP (3 mM a-ZrP, 27 lM Lys, thick dashed line, 222:1) and Lys/a-ZrP (3 mM a-ZrP, 54.4 lM Lys, thick line, 110:1).

229

The changes in the DSC profiles could result from protein–protein interactions, change in protein orientation as a function of loading, number of protein neighbors, or binding to a heterogeneous support. Previous studies showed that the dominant orientation of lysozyme is with its Arg5–Arg14 helix axis perpendicular to the metal phosphate surface, but did not rule out contributions of other orientations or change in orientation as a function of loading [16]. Since lysozyme is known to form dimers at high concentrations, we expect the protein–protein interactions to be dominant at high loadings. It is true that surface heterogeneity, although expected to be minor in our case, can also result in such changes. If the binding lattice is homogeneous, the data are consistent with the uniform distribution model where the broadening of the transition is primarily due to protein– protein interactions which become important at high loadings. The data are not consistent with the moving-boundary model where the protein–protein interactions will be independent of loading and the thermograms are not expected to change as a function of loading (at constant scan rate). Thus, it appears that the individual lysozyme molecules are well separated at low loadings, and the dominant low-temperature transition is assigned to this species. At high loadings, the higher temperature transition became prominent and the fact that this species begins to dominate at slow scan rates implies that lysozyme begins to aggregate in the galleries, as the heating continues. The aggregate formation in the galleries is expected to be slow due to the slow diffusion of the protein on the solid surface. Therefore, slow scan rates and higher loadings increase the contributions of the high-temperature fraction. It will be interesting to test these conclusions by following the d-spacings of these materials by XRD, as a function of loading, and these will be the focus of future studies. 3.6. Reversibility Another important issue regarding protein denaturation is reversibility. The values of DG and DS for the denaturation can be obtained from the thermal data if the transition is reversible. Reversibility of the thermal transition, under the calorimetric conditions, was tested by subjecting the samples to the second heating cycle, and each cycle consisted of heating to 100 C followed by cooling to room temperature. Thermograms recorded during the second heating cycle indicated that the thermal transitions of Lys (bound) and Hb (free and bound) were irreversible (data not shown). Previous reports from this laboratory demonstrated that Hb bound to a-ZrP re-folds after thermal denaturation but the re-folding occurs over a longer period of time (>24 h) [11,14]. This time scale is much longer than the cooling time in the calorimeter (<2 h), and hence, these proteins do not refold to a significant extent under these conditions. This conclusion of irreversible denaturation was also confirmed from the CD spectra which were recorded after the first heating cycle, at room temperature

230

A. Bhambhani, C.V. Kumar / Microporous and Mesoporous Materials 109 (2008) 223–232

C p (Kcal/ Kmol)

20

1 oC/min Mb, 1 st scan

15

10 Mb/ α -ZrP, 1 st scan 5

0 20

Mb/ α -ZrP, 2 nd scan

Mb, 2 nd scan 40

60

80

100

Temperature / 0 C

Fig. 5. Thermograms for the 1st heating cycle (thin line) and 2nd heating cycle (dotted line) of Mb, 52 lM indicate that the thermal denaturation on these time scales in mostly irreversible. On the other hand, thermograms of Mb/a-ZrP (50 lM Mb, 10 mM a-ZrP) for the 1st heating cycle (thick line) and 2nd heating cycle (thick dashed line) indicate significant improvement in the reversibility. All samples were scanned at 1 C/min.

(Supplementary #2). The CD spectra showed only minor recovery of protein secondary structure. In contrast to the behavior of Lys and Hb bound to the solid, thermograms of bound Mb indicated a significant extent of reversibility (Fig. 5) on the calorimetric timescales. This was indeed surprising. The endothermic transition during the second heating cycle of free Mb is substantially less than during the first cycle (<26 ± 4% of the original area). Re-folding of free Mb is not quite efficient under these conditions, but that of the bound Mb has been significantly higher (Fig. 5). Analysis of these data indicated that recovery for bound Mb is up to 50% of the original area, and this is another novel observation. Recovery of the native structure at a solid surface, on short time scales (2 h), is a very interesting observation. In summary, the thermal denaturation of bound proteins indicated kinetically-slow forming intermediates, depended on the loading, and in the case of Mb/a-ZrP, the denaturation is reversible to a large extent. 4. Conclusions Even though the binding of proteins to solids is known for decades [39], the general guidelines for the binding of proteins with retention of structure, activity and improved stability are still not clear. Our understanding of protein– solid interactions is rudimentary. Stabilization of the native state, and destabilization of the denatured state is expected to increase the thermodynamic stability of the bound protein. Thermal denaturation of Lys, Mb and Hb bound to silica were previously reported [7], and the denaturation was not reversible. Current data show that a small fraction of the bound protein is stabilized by a-ZrP, and intercalated Mb indicated a significant extent of reversibility. The broad DSC profiles and the thermal behavior noted here is in sharp contrast to the FTIR (Table 1) and CD studies (Supplementary Materials #1), which clearly indicated significant structure retention for the intercalated

proteins. Particular attention may be paid to intercalated Mb and Hb whose CD spectra nearly matched with those of the free proteins. Even in these cases, the DSC studies show considerable heterogeneity of the bound protein population and/or kinetically resolvable intermediate states. At the least, the bound protein underwent a slow transformation to partially unfolded states, and this behavior is distinct from that of the free protein. Therefore, one important conclusion to be drawn is that the spectral characterization of solid-bound proteins can be deceptive unless they are also complemented by calorimetric studies. Since the native states of most proteins are stabilized by less than 84 kJ/mol relative to their denatured states, a relatively small number of interactions of moderate strength can distort the protein structure and decrease or increase its stability [40]. The charge, size, amino acid composition, and specially the structural stability (hard vs. soft) of the proteins are some of the various factors which influence the protein– solid interactions [41]. The binding of lysozyme, a ‘‘hard protein’’, to the hydrophilic a-ZrP is governed significantly by electrostatic interactions. The large pI value of Lys (11) indicates a significant net positive charge on the protein at pH 7.2, and this charge contributes to a significant interaction with the oppositely charged metal phosphate lattice. Therefore, even a hard protein such as lysozyme is distorted to a significant extent (Supplementary Material #1). Earlier studies showed that increased ionic strength can weaken these interactions and improve the bound protein structure [18]. In contrast to lysozyme, the so-called ‘‘soft’’ proteins such as Hb have low internal stabilities and are generally known to adsorb on all surfaces irrespective of electrostatic interactions, owing to a gain in conformational entropy on adsorption [42,43]. Because of the lower pI values of Hb and Mb (<7), the electrostatic interactions are not expected to be as strong as in the case of lysozyme. This weaker interaction is probably responsible for the better structure retention of these heme proteins (Supplementary Materials #1). In contrast to the spectral data, which provide the ensemble averages, thermal denaturation studies provided information regarding the heterogeneity of the samples (Figs. 2–5). There are, at the least, two distinct thermal transitions and the relative populations of these depended on the loading as well as the scan rate. The extensive broadening of the peaks is likely due to heterogeneity in conformation, orientation, surface sites, and slow population of intermediate states which denatured over a wide range of temperatures [44]. At faster scan rates, the slow forming states are not fully attained, while at increasing loadings the protein–protein interactions begin to influence the denaturation dynamics (Tables 2 and 3). One of the reasons for the broad thermal transitions is the kinetically controlled denaturation process. The decrease in the area under the thermal transition with increased scan rate can be readily explained by these

A. Bhambhani, C.V. Kumar / Microporous and Mesoporous Materials 109 (2008) 223–232 Table 3 Thermodynamic parameters obtained at a scan rate of 2/min from the DSC thermograms (DH, Tm, DS, FWHM) of free and bound proteins Protein

DH (kJ/mol)

Tm (C)

DS (kJ/K mol)

FWHM (C)

Lys-free Lys/a-ZrP Mb-free Mb/a-ZrP

460 ± 25 673 ± 200 460 ± 67 611 ± 250

348 ± 0.2 324 ± 2.3 355 ± 0.3 360 ± 1.0

1.3 ± 0.2 – 1.3 ± 0.2 –

7.5 ± 1.2 22.0 ± 5.7 6.4 ± 1.1 49.2 ± 6.3

slow-forming intermediate states. Similarly, the decrease in area at low loadings can be accounted for, where the protein–protein interactions promote the formation of these intermediates. In most cases, a significant fraction of the bound protein indicated a state which is more stable than the free protein. Calorimetric studies complement spectral studies, for a more complete description of proteins bound to solids. These are the very first studies of characterizing the denaturation thermodynamics of proteins (Lys, Hb and Mb) bound to a-ZrP. Since, a small fraction of the bound protein is stabilized, the free energy gap between the native and the denatured states is increased for this fraction. Efforts are underway to maximize the contributions of this fraction. We also conclude that calorimetric studies can provide data for the construction of energy level diagrams for the denaturation of bound proteins. These may form a strong basis for engineering more effective synthetic materials to maximize bound enzyme function. Acknowledgment We thank the National Science Foundation (DMR0300631) for generous financial support of this work. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.micromeso. 2007.04.048. References [1] L. Stryer, Biochemistry, W.H. Freeman, New York, 1995. [2] (a) L. Caoi, Carrier Bound Immobilized Enzymes: Principles, Application, and Design, Wiley-VCH, Weinhheim, Germany, 2005; (b) B. Kasemo, Biological surface science, Surf. Sci. 500 (2002) 656– 677; (c) G.S. Chaga, J. Biochem. Biophys. Method 49 (2001) 313–334; (d) K. Nakanishi, T. Sakiyama, K. Imamura, Jpn. J. Biosci. Bioeng. 91 (2001) 233–244; (e) L. Gorton, G. Marko-Varga, E. Dominguez, J. Emneus, in: S. Lam, G. Malikin (Eds.), Analytical Application of Immobilized Enzyme Reactors, Blackie Academic & Professional, New York, 1994, p. 1. [3] V.P. Torchilin, Immobilized Enzymes in Medicine, vol. 11, SpringerVerlag, Berlin, Fed. Rep. Germany, 1991. [4] (a) J. Fang, C.M. Knobler, Langmuir 12 (1996) 1368; (b) J.F. Kennedy, C.A. White, in: A. Wiseman (Ed.), Handbook of Enzyme Biotechnology, Ellis Horwood, Chichester, 1985, p. 147.

231

[5] R. Freitag (Ed.), Biosensors in Analytical Biotechnology, Academic Press, San Diego, 1996; A. Rosevear, J.F. Kennedy, J.M.S. Cabral, Immobilized Enzymes and Cells, Adam Hilger, Philadelphia, 1987. [6] (a) V.V. Mozhaev, M.V. Sergeeva, A.B. Belova, Y.L. Khmelnitsky, Biotechnol. Bioeng. 35 (1990) 653–659; (b) K. Mosbach, Immobilized enzymes, Methods Enzymol. 44 (1976) 397; (c) V. Hlady, J. Buijs, Curr. Opin. Biotechnol. 7 (1996) 72–77. [7] R.K. Sandwick, K.J. Schray, J. Colloid Interf. Sci. 115 (1987) 130; B.L. Steadman, K.C. Thompson, R.C. Midduagh, K. Matsuno, S. Vrona, E.Q. Lawson, R.V. Lewis, Biotechnol. Bioeng. 40 (1992) 8–15; C.E. Giacomelli, W. Norde, J. Colloid Interf. Sci. 233 (2001) 234. [8] (a) C.V. Kumar, A. Chaudhari, J. Am. Chem. Soc. 122 (2000) 830– 837; (b) C.V. Kumar, A. Bhambhani, N. Hnatiuk, in: K. Carrado, P. Dutta, S.M. Auerbach (Eds.), Handbook of Layered Materials, vol. II, Marcel Dekker, New York, 2004. [9] (a) G. Alberti, in: P.A. Williams, M.J. Hudson (Eds.), Recent Developments in Ion Exchange, Elsevier, New York, 1987, p. 233; (b) G. Alberti, U. Costantino, S. Allulli, N. Tomassini, J. Inorg. Nucl. Chem. 40 (1978) 1113–1117. [10] C.V. Kumar, G.L. McLendon, Chem. Mater. 9 (1997) 863. [11] (a) C.V. Kumar, A. Chaudhari, Chem. Mater. 13 (2001) 238–240; (b) C.V. Kumar, A. Chaudhari, Micropor. Mesopor. Mater. 47 (2001) 407–410. [12] (a) C.V. Kumar, B.B. Raju, Advances in Molecular and Supramolecular Photochemistry Series, vol. 8, Marcel Dekker, New York, 2001; (b) C.V. Kumar, A. Chaudhari, J. Am. Chem. Soc. 116 (1994) 403– 404; (c) C.V. Kumar, A. Chaudhari, Micropor. Mesopor. Mater. 41 (2000) 307–318. [13] C.V. Kumar, A. Chaudhari, Chem. Commun. (2002) 2382. [14] C.V. Kumar, A. Chaudhari, Micropor. Mesopor. Mater. 57 (2003) 181. [15] (a) V. Panchagnula, C.V. Kumar, J.F. Rusling, J. Am. Chem. Soc. 124 (2002) 12515; (b) C.V. Kumar, Z.J. Williams, J. Phys. Chem. 99 (1995) 17632– 17639; (c) C.V. Kumar, E. Asuncion, G.L. Rosenthal, J. Am. Chem. Soc. (1994) 116, 403; (d) C.V. Kumar, E.H. Asuncion, G.L. Rosenthal, Micropor. Mater. 1 (1993) 299–308. [16] A. Chaudhari, J. Thota, C.V. Kumar, Micropor. Mesopor. Mater. 75 (2004) 281. [17] A. Chaudhari, C.V. Kumar, Micropor. Mesopor. Mater. 77 (2005) 175. [18] A. Bhambhani, C.V. Kumar, Chem. Mater. 18 (3) (2006) 740. [19] B. Masschalck, C.W. Michiels, Crit. Rev. Microbiol. 29 (3) (2003) 191–214. [20] T. Stachyra, D. Guillochon, S. Pulvin, D. Thomas, Appl. Biochem. Biotechnol. 59 (1996) 231. [21] (a) G. Alberti, U. Constantino, in: M.S. Whittingham, A.J. Jacobson (Eds.), Intercalation Chemistry, Academic Press, New York, 1982; (b) G. Alberti, U. Constantino, S. Allulli, N.J. Tomassini, Inorg. Chem. 29 (1990) 2112; (c) M.B. Dines, P.M. DiGiacomo, Inorg. Chem. 20 (1981) 92. [22] T. Imoto, K. Yagishita, Agric. Biol. Chem. 35 (1971) 1154. [23] I. Hsiao, H. Uedaira, Agric. Biol. Chem. 53 (1989) 833. [24] A.G. Marshall, in: Biophysical Chemistry, Principles, Techniques, and Applications, Wiley & Sons, New York, 1978, p. 70. [25] M.L. Galisteo, P.L. Mateo, J.M. Sanchez-Ruiz, Biochemistry 30 (1991) 2061. [26] S.J. Perkins, Eur. J. Biochem. 157 (1986) 169. [27] S.B. Peterson, P. Fojan, E.I. Peterson, M.T.N. Peterson, J. Biomed. Biotechnol. 1 (2001) 62–69;

232

[28] [29] [30] [31]

[32] [33] [34] [35] [36] [37] [38]

A. Bhambhani, C.V. Kumar / Microporous and Mesoporous Materials 109 (2008) 223–232 T. Huang, J. Pawliszyn, Electrophoresis 23 (2002) 3504–3510; D. Frechet, R. Ehrlich, P. Remy, Nucl. Acid Res. 7 (1979) 1965–1980. J.M. Sturtevant, Ann. Phys. Chem. 38 (1987) 46. I. Jelesarov, H.R. Bosshard, J. Mol. Recognit. 12 (1999) 3. J.F. Culajay, S.I. Blaber, B.M. Khurana, Biochemistry 39 (2000) 7153. W. Pfeil, P.L. Privalov, Biophys. Chem. 4 (1976) 41; F. Delben, V. Crescenzi, Biochim. Biophys. Acta 194 (1969) 615; A. Michnik, Z. Drzazga, A. Kluczewska, K. Michalik, Biophys. Chem. 118 (2005) 93. L. Kelly, L.A. Holladay, Biochemistry 29 (21) (1990) 5062–5069. K.C. Cho, C.L. Choy, Biochim. Biophys. Acta 622 (1980) 320. Z. Drzazga, A. Michnik, M. Bartoszek, E. Beck, J. Therm. Anal. Calorim. 65 (2001) 575. S. Krok-kedracka, Z. Wasyleski, Eur. J. Biochem. 270 (2003) 4564– 4573. G. Zoldak, A. Zubrik, A. Musatov, M. Stupak, E. Sedlak, J. Biol. Chem. 279 (2004) 47601–47609. I.V. Sochava, T.V. Belopolskaya, O.I. Smirnova, Biophys. Chem. 22 (1985) 323. M. Casciola, U. Costantino, L. Di Corce, F. Marmottini, J. Inclusion Phenom. 6 (1988) 291.

[39] (a) B. Kasemo, Biol. Surf. Sci. 500 (2002) 656; (b) Y. Lvov, H. Mohwald, in: Protein Architecture; Interfacing Molecular Assemblies and Immobilization Biotechnology, Marcel Dekker, New York, 2000; (c) G.S. Chaga, J. Biochem. Biophys. Method 49 (2001) 313; (d) K. Nakanishi, T. Sakiyama, K. Imamura, J. Biosci. Bioeng. 91 (2001) 233. [40] (a) P.L. Privalov, S.J. Gill, Adv. Prot. Chem. 39 (1988) 191; (b) A. Kondo, T. Urabe, J. Colloid Interf. Sci. 174 (1995) 191. [41] K. Nakanishi, T. Sakiyama, K. Imamura, J. Biosci. Bioeng. 91 (2001) 233–244. [42] M.A. Bos, Z. Shervani, A.C.I. Anusiem, M. Giesbers, W. Norde, J.M. Kleijn, Colloid Surf. B 3 (1994) 91. [43] T. Sakiyama, K. Tanino, M. Urakawa, K. Imamura, T. Takahashi, T. Nagai, K. Nakanishi, J. Biosci. Bioeng. 88 (1999) 536. [44] (a) T.A. Horbett, J.L. Brash, Proteins at interfaces: current issues and future prospects, in: T.A. Horbett, J.L. Brash (Eds.), Physicochemical and Biochemical Studies, American Chemical Society, Washington, DC, 1987, pp. 1–33; (b) T.J. Lenk, T.A. Horbett, B.D. Ratner, K.K. Chittur, Langmuir 7 (1991) 1755.

Fear the Walking Dead 8 | Audi 80 2,6 l V6, Typ B4 sehr guter Zustand + wenig Kilometer | Hicham Smati 2019 Mp3