CHAPTERS

DIFFERENTIAL SCANNING

CALORIMETRY OF LF

Introduction

With the advent of recombinant DNA technology, it has become possible to

express and purify industrially and pharmaceutically important proteins from

heterologous systems in large quantities (Walsh and Headon, 1994) A large

number of recombinant proteins are in the process of clinical trials. The well

documented example of the use of recombinant proteins is that of insulin (Walsh

and Headon, 1994). However, degradation and instability of these proteins

presents unique difficulties in making them commercially successful. Thus there

is a need to address various aspects of protein instability. Stability of protein

activity in vitro reflects the molecular integrity of the constituent polypeptide

chain(s) and resistance to adverse influences such as high temperature and

denaturants. A fully functional protein can lose its biological activity by

unfolding of its tertiary structure to a disordered polypeptide as a result of which it

is no more functional. Such unfolding is called denaturation.

Instability of the proteins have been divided into two classes 1.e., chemical

instability and physical instability. Chemical instability involves modification of

the proteins yielding an entirely new chemical entity via different reactions such

as hydrolysis, deamidation, oxidation, racemization and disulfide linkages etc.

Deamidation of side chain amide linkages in Gln or Asp residues yields to form a

free carboxylic acid. A large number of proteins such as human growth hormone,

epidermal growth factor, prolactin (Graf et al., 1971) adrenocorticotrophic

hormone (Graf et al., 1971), trypsin, lysozyme (Ahem et al., 1985) and

ribonuclease (Zale et al., 1986) have been reported to undergo deamidation.

Deamidation plays a central role as a timer in protein turnover (Johnson et al.,

1989). Oxidation of side chains of His, Met, Cys, Tyr and Trp residues results in

loss of biological activity of many proteins such as gastrin (Morley et al., 1965),

calcitocin (Riniker et al., 1968), subtilisin, lysozyme, a- chymotrypsin (Ray et al.,

1962), trypsin (Holeysovsky et al., 1968), etc. Factors that influence the rate of

oxidation include temperature, pH, buffer medium used, and oxygen tension.

Sulfhydryl groups and disulfide bonds and their interrelationships are an important

factor affecting the properties of a majority of proteins. Interchange of disulfide

102

bonds can result in incorrect pairing, leading to an altered three dimensional

structure and hence loss of catalytic activity. Except Gly, all aminoacids are

chiral at carbon bearing the side chain. Racemization of aminoacids in proteins

can generate non metabolizable forms of aminoacids (D-enantiomers) or create

peptide bonds inaccessible to proteolytic enzymes. At high temperature ~

elimination from cysteine residues destroys sulfide bonds and inactivates the

protein. Residues such as Cys, Ser, Thr, Phe and Lys are capable ofundergoing ~

elimination resulting in the inactivation of proteins. ~ elimination is influenced by

pH, temperature and the presence of metal ions. Proteins such as phosvitin (Sen et

al., 1977), antifreeze glycoprotein (Lee et al., 1977) and lysozyme (Nashef et al.,

1977) undergo ~ elimination at high temperature, inactivating the protein.

Physical instability involves change in the structure of protein by processes such

as denaturation, adsorption to the surface and precipitation etc. Proteins, because

of their polymeric nature and their ability to adopt different structures can undergo

a variety of structural changes independent of chemical modification.

Denaturation refers to an alteration of the tertiary and often secondary structure of

proteins (Havel et al., 1986; Tombs, 1985). Denaturation is caused by a variety of

conditions such as increase or decrease in temperature, increase in pH, addition of

organic solvents or other denaturants. Denaturation can be reversible or

irreversible. Reversible denaturation is defined as unfolding caused by an

increase in temperature which can be reversed by subsequent lowering of

temperature. Irreversible denaturation is an unfolding process which does not

allow the native structure to be regained simply by lowering the temperature.

Partial unfolding intermediates whose solubility is less than the native state or ·the

unfolded state may lead to inactivation of the protein by aggregation.

Amphiphilic regions in the protein are the sites which associate readily and bring

about aggregation (Brems et al., 1986). CD studies of these aggregates have

shown that they retain a large amount of secondary structure while as the tertiary

structure is greatly disrupted. Interferon y below pH 4.5 and in the absence of

NaCl undergoes aggregation (Arakawa et al., 1985). Large aggregates of proteins

decrease the solubility of proteins which results in precipitation. Thus,

precipitation is the macroscopic equivalent of aggregation. Temperature, pH,

103

presence or absence of additives are the factors responsible for precipitation.

Precipitation of insulin is well established. Over expression of recombinant

proteins often leads to aggregation into insoluble complexes termed inclusion

bodies.

There are several approaches to solve the problem of instability in proteins of

commercial relevance. Techniques like chemical modification of specific

aminoacid residues, protein engineering and the use of carefully designed

cosolvent additives have been employed to increase the thermostability of

proteins. Modification of certain reactive sites in proteins and peptides by

polyehtylene glycol and polyoxy ethylene may also increase the stability.

Conversion of Lys to homoarginine via guanidium has also been shown to

stabilize numerous proteins (Qaw et al., 1986; Cupo et al., 1980). Methylation of

basic aminoacid residues has been found to increase the thermostability of

proteins. Cross linking of certain proteins with bis imidates, bis succinimides or

bis maleimides has been reported to increase the thermostability of certain

proteins (Wong and Wong, 1992) such as HRPO with the bis - succinimides

suberic acid and ethylene glycol (Ryan, 1994) and beef heart creatine kinase with

dimethylsuberimidate (Sheehan, 1990).

Site directed mutagenesis refers to aminoacid substitutions at specific sites in a

protein. Using protein engineering approach the stability of many proteins such as

lysozyme (Matsumura et al., 1989), cytochrome c (Das et al., 1989), luciferase

(Kajiyama et al., 1993), and subtilisin (Bryan et al., 1992) has been enhanced.

The idea behind this approach is to improve interior interactions leading to

thermal stability, to introduce disulfide bridges to stabilize the native

conformation, to improve a helix stability, to design ion binding sites which will

increase thermal stability, and to replace sites for chemical degradation such as

deamidation (Manning et al., 1989). Protein stability may also be increased in the

presence of certain low molecular weight substances such as salts, polyols and

sugars (Fagain, 1995; Kaushik and Bhat, 1998; Kaushik and Bhat, 1999). In the

presence of low concentration of additives (salts, polyols and sugars) more water

molecules pack around the protein, in order to exclude more hydrophobic additive

which results in increased stability. While at higher concentrations of the

104

additive, the hydrophobic additive begins to denature the protein. Polyalcohol

material such as glycerol and sugars (sorbitol, xylitol, lactitol, trehalose and

DEAE- dextran) are known to stabilize proteins. Ionic compounds (salts) are also

known to stabilize proteins by decreasing reversible denaturation via the non

specific binding to the protein. Binding of ions have been reported to increase

thermal stability for subtilisin, ribonuclease, thermolysin and alkaline phosphatase

(Chlebowski et al., 1977; Dahlquist et al., 1976). Ion binding has also been

employed to control physical instability phenomena such as aggregation and

precipitation in the case of insulin (Palmieri et al., 1988).

Anthrax toxin proteins P A, LF and EF are highly thermolabile. The proteins have

to be kept frozen at -70°C and lose their activity within 1-2 weeks when kept at

4°C. PA and LF when stored at 37°C lose their activity in 48 hrs (Radha et al.,

1996; Smita, V. and Pankaj, unpublished results). Since PA is the major

component of vaccine against anthrax and LF is the major cytolytic factor of

anthrax toxin complex, it is of utmost importance to preserve the conformational

stability of these proteins. To enhance the thermostability of P A certain additives

such as MgS04, Sodium citrate, trehalose, xylitol and sorbitol were used in our

laboratory (Radha et al., 1996). It was found that in presence of 3M MgS04, 82

% of the activity of P A was retained after incubation at 3 7°C while in the presence

of sodium citrate (1 M), trehalose (1.5 M) 68 and 75% of the activity was retained

respectively (Radha et al., 1996). Similar experiments were also conducted to see

the effect of additives on the thermostability of LF. Glycine (1 M), sodium citrate

(1M), xylitol (1.5 M) and MgS04 (2 M) were found to be effective in the

thermostabilizing the lethal factor (Smita V., unpublished results).

In order to rationally design strategies to improve the protein stability it is

important to have a knowledge of the transition temperature (Tm) for protein

denaturation. Several spectroscopic techniques like UV spectroscopy, CD,

fluorescence etc. have been used to determine this parameter and calculate the free

energy· changes associated with the unfolding process. Differential scanning

calorimetry (DSC) is also a powerful and sensitive technique to characterize the

temperature induced conformational changes in proteins. DSC has been used very

successfully to evaluate transition temperature (Tm) of proteins and other

105

thermodynamic parameters to a very high accuracy. Protein stability curves can

be determined as a function of various parameters such as pH. In this study, the

differential scanning calorimetry of LF was carried out to determine the thermal

stability as a function of pH. The results in the correlation of thermal stability

with the % activity retained upon incubation of the protein at 37°C for several

hours as a function of pH have also been presented.

Experimental Methodology

Differential Scanning Calorimetry

DSC studies on the thermal denaturation of proteins have played a crucial role in

characterizing different proteins and revealing important informations on the

protein structure such as domain interactions, protein stability and protein folding.

A differential scanning calorimeter consists of two cells : the reference cell and

the sample cell. Reference is the buffer in which the protein is diluted and the

sample is the protein solution in the buffer. Prior to performing the experiment, a

baseline is plotted using buffer in both the cells. Both the cells are heated at a

constant scan rate and the output is the difference in the heat capacity between the

two cells. After the baseline is drawn, buffer from the sample cell is removed and

replaced by protein solution and the temperature scan started again. Heat capacity

of the reference (solvent) is always higher than that of sample protein solution due

to lower water content in the sample cell. When the protein solution in the sample

cell gets denatured at a particular temperature, the temperature of the cell falls due

to endothermic transition and the cell tries to absorb the heat from outside to

maintain the temperature. Output is plotted as the heat absorbed associated with

the denaturation of the protein. Partial heat capacity of the protein is the

difference in the heat capacities of the sample and the reference. The output of

DSC experiment is given as excess heat capacity ( CP ex) versus temperature where

cp ex is the partial heat capacity of the protein solution measured against the

reference. Integration of Cpex with respect to temperature gives the excess

enthalpy value.

106

where T 0 is the temperature at which the protein is in the native state and T is the

temperature at which the protein is denatured.

The area enclosed by the transition and the baseline gives the total enthalpy

change for the denaturation process and is referred to as calorimetric enthalpy.

DSC studies also gives an idea about the transition temperature (Tm) of proteins

and Gibb's free energy. By these parameters stability of proteins can be

determined. Thermodynamically, stability of the protein is the difference between

the Gibb's energy of its denatured state and that of its native state and protein

stability curves can be plotted. Scanning calorimetry also gives information on

domain structure and domain interactions in proteins (Privalov, 1979) which can

further reveal information on cooperative interaction linked partial unfolding

reaction of protein subunits. DSC experiments were carried out by using Model

MC2 differential scanning calorimeter from Microcal Inc., Massachusetts, USA.

To perform calorimetric experiments as a function of pH, recombinant LF

was purified as described in chapter 2. All DSC experiments were performed

within a few days of purifying the protein. The protein was concentrated to a

concentration of 2 mg I ml and dialyzed extensively against different buffers (10

mM Mes pH 5.5, 10 mM Mops pH 6.5, 10 mM Hepes pH 7.0 and 10 mM Tris pH

7.8). Any precipitate in the protein sample was removed by centrifugation at 4°C

for 30 minutes at 20, 000 x g. Samples were degassed for 30 minutes at ambient

temperature in a vaccum station connected to a pump and having a magnetic

stirring and then loaded with a gas tight Hamilton syringe in the cells. The cells

were pressurised wi~h 1.5 atm pressure from a N2 cylinder to avoid the formation

of bubbles at high temperatures Initially, the dialysate buffer was loaded in both

reference and sample cells and after a baseline scan had been conducted over

temperature range to be used for the protein, the contents of the sample cell were

gently removed and refilled with the dialyzed protein solution. Results were

107

obtained at a scan rate of 60° I hr. However at pH 5.5 and 7.0, 30° I hr. scan rate

was also used. The dependence of the molar heat capacity on temperature was

analyzed using the ORIGIN software (Microcal Inc., USA).

Activity assay of LF

Lethal factor was incubated at 37°C in buffers Mes (10 mM pH 5.5), Mops (10

mM pH 6.5), Hepes (10 mM pH 7.0) and Tris (10 mM pH 7.8). Samples were

removed after 6, 12, 24, 36, 48, 60 and 72 hrs. and activity of LF (1 Jlg I ml)

alongwith P A (1 Jlg I ml) was determined using macrophage lysis assay as

described earlier. After 3 hrs., viability was determined using the MTT dye and%

viability was plotted against time of incubation.

Results and Discussion

Figure 5.1 - 5.5 present the differential scanning calorimetric data for the thermal

denaturation of the lethal factor as a function of pH ranging from pH 7.8 to 5.5.

At pH 7, wherein the toxicity assay for LF is carried out, the transition

temperature is 42°C and the onset of denaturation starts at ~30°C. This supports

the activity data wherein it has been observed that LF loses its efficacy when

incubated at 37°C for 48 hrs. (Smita, V., unpublished results). The noise observed

in the post transition zone at pH 7.0 indicates aggregate formation. The enthalpy

of transition (~Heal) has been found to be 127 kcal mol-1 at pH 7.0 and the ratio

of calorimetric and van't Hoff enthalpy nearly equal to unity, ie., ~Heal I~ Hv.H.

::::: 1, indicative of the presence of one structural domain in the protein molecule

unlike the protective antigen (P A), another component of anthrax lethal toxin,

which has been structurally observed to consist of 4 folding domains (Petosa et

al., 1997). Since the three dimensional structure of LF is not known as yet, the

calorimetric data obtained can throw some light on the domain organisation in the

LF molecule. No significant temperature scan rate dependence has been observed

for LF denaturation as evident by similar Tm and ~Heal values observed at the

scan rate of 60°/hr. and 300/hr.

108

Thermal Denaturation _Qrofil~_<?f~F _at J?li7_.8 __ _ -------------------------- -.-

-..­' 0 E

pH 7.8

•

t 20 kcal mol_,

0 10 "'~0 30 40 50 60 70 80 90 100

Temperature (°C)

A ----------

pH 7.8

15

--.. ....... ' 10 0 E

.......

~ ro ~ 5 -~c. u

0

10 20 30 40 50 60

Temperature (°C)

B ---------------------

Fig. 5.1 Thermal denaturation profile of LF at pH 7.8, Scan rate 60° C I hr. (A)

raw data after buffer - buffer base line subtraction and concentration

normalization. (B) analyzed curve after fitting the data.

Thermal Denaturation profile of LF at pH 7.0 scan rate 60°C I hr

-..­' 0 E

~a. u

-..--0 E

..-~ ro u ~ -X 4)a. u

20

15

10

5

0

-5

pH 7

Scan Rate 60°C/hr

5 kcal mol" 1

0 10 20 30 40 50 60 70 80 90 1 00

Temperature (°C) --- -

A

pH 7

Scan Rate 60°C/hr

0 1 0 20 30 40 . 50 60 70 80 90 100

Temperature (°C)

B

Fig. 5.2 Thermal denaturation profile ofLF at pH 7, Scan rate 60° C I hr. (A) raw

data after buffer - buffer base line subtraction and concentration nomtalization.

(B) analyzed curve after fitting the data.

-.­I

0 E

..-~ ro 0

.::t:. ->< Q)a.

(.)

pH 7

t 5 kcal mol ·1

0 10 20 30 40 so. 60 70 80 90 100

Temperature (°C)

A

20~------------------------------~--------~

15

-..-15 10 E

~c. (.)

0

10

pH7

20 30 40 50 60 70 Temperature (°C)

B

Fig. 5.3 Them1al denaturation profile ofLF at pH 7, Scan rate 30° C I hr. (A) raw

data after buffer - buffer base line subtraction and concentration normalization.

(B) analyzed curve after fitting the data.

Thermal Denaturation nrofile q_f_1.f atpl:J 5 .~ ·------·. -- -- -- ---------- ------<>-------- ··--------------- -------- . -- .. -

20~----------------------------------------~

15

..... I

0 10 E

0

pH 5.5

Scan Rate 60°C hr

-5~----r---~----.-----~--~----~----~--~

15

- 10 ..... I

0 E

..... ~

~ ~ -X Q)Q.

5

(.) 0·

20 30 40 50

Temperature (°C)

A

pH 5.5

Scan Rate 30°C/hr

·5•cr. ------~----~------~----~------~----~ 20 30 40 50

Temperature (°C)

B

Fig. 5.5 Thennal denaturation profile of LF at pH 5.5, (A) analyzed curve after

fitting the data at scan rate 60° C I hr. (B) analyzed curve after fitting the data ~t

scan rate 30° C I hr.

As the pH is changed from 7.0 to 7.8, there does not seem to be any appreciable

change in the thermodynamic parameters observed. This indicates that there is no

or little change in the protein conformation between the two pH values. It has

been further observed that there is sudden aggregation of the protein at pH 7.8

unlike at pH 7.0. The transition observed at all the pH values were found to be

irreversible. When pH is lowered to 6.5 (Fig. 5.4), there is a decrease in the Tm

value to 39.7°C and the 11 Heal to 75 kcal moi-l. The onset temperature of

denaturation has been observed to be marginally higher than at pH 7.0. The

aggregation however, occurs at much lower temperature of 55°C compared to at

pH 7.0. When the pH is futher lowered to 5.5, the Tm decreases to 39.3°C and

38.2°C at 60°/hr. and 30°/hr. scan rates respectively (Table 5.1 ). The 11 Heal

values decreased to 82 and 62 kcal mol-l respectively. The scan rate dependence

is due to the irreversible nature of the thermal transition. From the DSC transition

it is obvious that the protein is partially denatured at this pH especially at 30°C !hr.

scan rate.

Fig. 5.6 shows a plot ofTm ofLF versus pH of the buffer. It can be seen that the

protein is more stable at pH 7 and 7.8 compared to at pH 6.5 and 5.5. There is a

shift in stability as we move from pH 6.5 to 7.0. To establish any possible

correlation of the thermodynamic data as a function of pH by DSC with the

activity data on LF at the same pH values, LF was incubated at pH 7.8, 7.0, .6.5

and 5.5 for varied time periods upto 72 hrs. at 37°C and the activity monitored by

the usual J774A.l cytotoxicity assay as described earlier. Fig 5.7 presents the

results of the activity studies. It can be seen that there is a steep decrease in the

LF. cytotoxicity upto 36 hrs. of incubation followed by a gradual decrease

reaching~ 90 % viability (10 % activity) at 72 hrs. at pH 7.0. There is marginal

difference between the data of pH 7.0 and 7.8 just as observed in the DSC data.

However at pH 6.5 LF loses its activity rather slowly, for example at 48 hrs. 55 %

of the activity (cell lysis) is retained, compared to only about 20% at pH 7.0. But

at 72 hrs. the activity at all the pH's studied appears to be same, i.e., protein

become inactive and is unable to lyse the cells.

Interestingly, at pH 5.5 no cell lysis is seen 1.e., % cell viability is ~95 %

indicative of complete loss in biological function . This result correlates very well

109

Table 5.1

Thermodynamic parameters for the denaturation of LF

pH Tm 8. Heal 8. HvH (oC) (kcal mol-l) (kcal moi-l)

7.8 41.8 106 97

7.0 42.1 127 97

7.0(SR 30) 41.5 105 104

6.5 39.7 75 135

5.5 39.3 82 140

5.5(SR 30) 38.2 62 153

'SR' is scan rate and is 60°C /hr. unless specified in the table above . The error in Tm measurements is± 0.2oc and that in 8.H value is± 5 %.

Tm of LF as a function of pH

5 5.5 6 6.5 7 7.5 8

pH

Fig. 5.6 Transition temperature (Tm) of LF at different pH at scan rate 60°C/ hr.

0 ·--·-~ ·-> '2f(

Biological activity of LF

100

80

60

-+-pH 5.5 40 -pH6.5

-6--pH 7.0

-*-pH 7.8 20

0 12 24 36 48 60 72

Incubation Time (hrs.)

Fig. 5.7 Biological activity ofLF (after incubation in buffers of different pH) on J774A.l cells alongwith PA (l).!g I ml).

The values are means of experiments done in triplicate

with the DSC results showing considerably diminished !1 Heal values and a small

transition. The activity data thus support the DSC data obtained. A decrease in

Tm and !1 Heal values is likely to occur as a result of a decrease in the electrostatic

interactions in the protein as the pH is lowered. pi of LF is 6.01 . It has been

observed that proteins are maximally stable near their pi (Stigter and Dill, 1990).

Any change in the pH away from their pi leads to accumulation of net +ve or -ve

charge which can influence the balance of electrostatic interactions and lead to

lowering in the free energy of stabilization followed by partial or complete

denaturation. Our results are in accordance with this hypothesis. Similar results

have also been obtained with several proteins studied in the literature as a function

of pH (Privalov, 1979). Further studies on the conformational changes monitored

at these pH values by far and near UV Circular dichorism spectroscopy would

throw light on the differences in the secondary and tertiary structures as a function

of pH and its relationship with the biological activity of LF.

110