动态多模式CD光谱用于研究抗体**的组方筛选A new approach for optimising biotherapeutic development using Chirascan

A new approach for optimising biotherapeutic development usingChirascan Dynamic Multi-mode Spectroscopy (DMS)
Executive summary
A stable, correctly-folded protein is an absolute requirement for asuccessful biotherapeutic and the stability of a desiredconformation is determined by suitable formulation. Chirascandynamic multi-mode spectroscopy (DMS) is a new technology fromApplied Photophysics that uses multiple spectroscopic probes tomonitor changes in secondary structure as a function of temperatureand to determine the thermodynamics of unfolding. Chirascan DMSwill:
• Prove that a protein is initially correctly folded and hencebiologically active
• Distinguish between unfolding and aggregation steps, which is ofpotential relevance to future immunogenicity
• Identify the structural components that change and the order inwhich they denature, giving valuable information onconformation
• Determine quantitatively the Tm and enthalpy values as indicatorsof stability
In this study, DMS is applied to the denaturation of a monoclonalantibody under different pH conditions to show the potential of thetechnology in biotherapeutic development.
Introduction
The stability of a protein therapeutic is conventionally monitoredby calorimetric techniques, particularly differential scanningcalorimetry (DSC), yielding thermodynamic parameters such as theenthalpy and the mid-point of a conformational transition. Thethermodynamic parameters are indicative of the relative stabilitiesof proteins in different conditions but calorimetric techniquescannot tell you how the conformation changes, whether or not aconformational change leads to aggregation, or if the protein is inits desired conformation prior to heating.
Knowledge of secondary structure will confirm whether or not theprotein is correctly folded initially and observing the secondarystructure as a function of temperature will tell you how itsconformation changes on heating. However, the only common method todetermine protein secondary structure in solution is circulardichroism (CD) and monitoring CD as a function of temperature canbe a time-consuming business1.
Chirascan dynamic multi-mode spectroscopy combines the benefits ofspectroscopic and calorimetric measurements into a single, rapid,information-rich measurement that generates results from a completeexperiment in about an hour.
1Norma J. Greenfield, Nature Protocols, Vol.1 No.6, 2006, 2527.
- 1 -
Experimental
Chirascan DMS was used for all the experiments. Samples of amonoclonal antibody (mouse anti-human IgG-Fc sub-class IgG2a(Abazyme, MA, USA)) at concentrations of 0.25mg/ml and 0.5mg/mlwere buffered at pH7 and pH4.2 respectively. Protein/buffersolution in a 1mm cuvette was heated from 4oC to 76oC at a rate of1oC per minute; the sample temperature was recorded using anin-sample thermocouple. CD and absorbance spectra were measuredduring heating such that new scans commenced at intervals of 1oC.The duration of each of the two experiments was less than 75minutes.
Sample
mouse anti-human IgG
mouse anti-human IgG
Buffer
pH7 (45mM Na phosphate)
pH4.2 (45mM phosphate 25mM ascorbate)
Volume
300μlitres
300μlitres
Protein conc.
.25 mg/ml
.5mg/ml
Protein used
75μg
150μg
T range
4oC – 76oC
4oC – 76oC
T rate
1oC / minute
1oC / minute
λ range
260nm – 198nm
260nm – 205nm
Step-size
1nm
1nm
Duration
<75 minutes
<75 minutes
Cell pathlength
1mm
1mm
Optical probes
CD and absorbance
CD and absorbance
Table 1 Summary of experimental parameters
- 2 -
Results and discussion – pH7
Figure 1 CD spectra Figure 2 CD-temperature profiles
The CD data for the mAb at pH7 are displayed in figures 1 and 2 asa function of wavelength and temperature. There are 73 individualCD spectra in figure 1 with a new scan started every 1oC, and 64CD-temperature profiles, one for each integer wavelength from 197nmto 260nm. No significant change in the CD signature occurs belowabout 35oC. In the range 35oC<T<60oC a rapid changeaccompanied by loss of amplitude of the positive peaks at 202nm and233nm takes place (indicated by the red arrows in figure 1) with aconcurrent increase in the amplitude of the negative peak at 216nm.A second transition, indicated by the blue arrows, occurs above60oC, where the most significant change is the loss of amplitude ofthe peak at 216nm.
Figure 3 shows the absorbance as a function of temperature – notethat at approximately 60oC there is an abrupt change in theapparent absorbance of the sample even for those traces thatrepresent wavelengths at which there is no absorbing species(λ>250nm). In fact, such change in the absence of a chromophoreis due to light scattering caused by the onset of aggregation andtherefore the second transition seen in the CD data is accompaniedby aggregation.
Figure 3 Absorbance-temperature profiles
- 3 -
Results and discussion – pH4.2
Figure 4 CD spectra Figure 5 CD-temperature profiles
The CD data for the mAb at pH4.2 are displayed in figures 4 and 5as a function of wavelength and temperature. There are 73individual CD spectra in figure 4 with a new scan started every1oC, and 56 CD-temperature profiles in figure 5, one for eachinteger wavelength from 205nm to 260nm inclusive. No significantchange in CD occurs below 50oC and the transition appears to bebiphasic, indicated by the red arrows.
Inspection of figure 6 shows no increase in apparent absorbance atthose wavelengths where there is no chromophore which suggests thatthere is no aggregation associated with the unfolding event, incontrast to the same sample at pH7.
Figure 6 Absorbance-temperature profiles
- 4 -
Results and discussion – analysis of the data
A data reduction technique known as singular-value decompositionwas used to identify the number of independent species representedin the CD data: it was three for both pH7 and pH4.2, although asnoted above, the CD-temperature profiles at pH4.2 do not obviouslyshow this. Triphasic denaturation models were therefore used infitting the data.
The data across all wavelengths were used in non-linearleast-squares refinement of the models. Constraints were applied toensure that the calculated values and profiles remain physicallymeaningful. The results with errors in brackets are given in thetable below.
mAb pH7
mAb pH4.2
Tm1
51.6(2) oC
56.2(1) oC
ΔH1
210(5 )kJ/mol
453(10) kJ/mol
Tm2
64.3(1) oC
62.4(1) oC
ΔH2
446(9) kJ/mol
477(4) kJ/mol
Table 2 Transition mid-points and enthalpies
If a higher Tm is taken as an indicator of better long-termstability of the protein, then Tm1 indicates that the protein atpH4.2 will be more stable. A larger enthalpy of transitionindicates a narrower transition and therefore a later onset of lossof conformation, which is a desirable characteristic; ΔH1 indicatesthat the protein at pH4.2 has the sharper transition.
The structural information inherent in the data permits furtherinvestigation of the behaviour of the protein. Reference to theresults of the refinements in figures 7 and 8 illustrates thefollowing points:
• The CD signature of the secondary structure is a very goodindicator of how the protein is folded and can be used to judge ifthe protein is correctly folded at the outset of the measurement.The familiar CD signature of a mAb is seen at the outset of bothexperiments, indicating that the protein is correctly folded atboth pH7 and pH4.2.
• The CD signature identifies the main secondary-structuralcomponents (α-helix, β-sheet etc.), which of them change understress, how they change and the order of their changing. Monoclonalantibodies are largely β-sheet proteins; the protein at pH7 andpH4.2 show a similar initial change to an intermediate but athigher temperatures, the behaviour is different. At pH7, furtherunfolding leads to a collagen-like CD signature and is accompaniedby aggregation (seen from the absorbance spectrum); at pH4.2,further unfolding leads to an extended conformation with noapparent associated aggregation.
• Most therapeutic antibodies will need to be active atphysiological pH and at body temperature and it is interesting tonote that at pH7 the monoclonal antibody is starting to unfold at37 degrees. This is best seen by reference to Species 1 (nativeconformation) in figure 8d below.
- 5 -
Figures 7a-7d mAb at pH4.2 – triphasic model – top: the model andresidual surfaces, bottom: the model species and their relativeconcentration profiles
Figures 8a-8d mAb at pH7 – triphasic model – top: the model andresidual surfaces, bottom: the model species and their relativeconcentration profiles
- 6 -
Conclusion
Dynamic multi-probe spectroscopy is a powerful addition to therange of biophysical tools available in biotherapeutic development.In addition to thermodynamic parameters which are indicative of therelative stability of proteins in different conditions, detailedinformation about the behaviour of the secondary structure andaggregation is readily available.
The technique is not restricted to far-UV secondary structureanalysis – it can be used equally well in the near-UV to study thechange in the chiral environment of the aromatic chromophores ofamino-acid residues, giving insights into the stability of theprotein’s tertiary structure.
If you are interested in knowing more about Chirascan dynamicmulti-mode spectroscopy, contact APL to register your interest.
Applied Photophysics Limited
203-205 Kingston Road, Leatherhead, KT22 7PB, United Kingdom.
Tel:+44 (0) 1372 386537 or USA 1-800 543 4130 ● Fax: +44 (0) 1372386477
E-mail: sales@photophysics.com ● Web:www.photophysics.com
Applied Photophysics was established in 1971 by the RoyalInstitution of Great Britain.
- 7 -