Size exclusion chromatography (SEC) provides critical information about the purity and oligomeric state of protein therapeutic products. SEC analysis is important for process development, formulation optimization, release testing, stability, comparability, and forced degradation studies SEC coupled with light scattering (LS), ultra-violent (UV), and refractive index (RI) detection offers detailed molecular weight information about proteins, protein complexes, and glycosylated proteins (1–3). The current study uses a tetra detection system (laser light scattering, viscometer, RI, UV) coupled with software that integrates data from all detectors to analyze a model PEGylated protein. The detection system provides accurate molecular weight, intrinsic viscosity, and hydrodynamic volume with minimal input of known parameters. High molecular weight (HMW) impurities observed in different stress and process samples were characterized by this method. This study examines the major and minor species present in samples of a model PEGylated protein encountered during biopharmaceutical development. 

A model PEGylated protein was prepared for use in these studies. The protein of approximately 19 kDa was expressed in 

, purified, and then modified with methoxypolyethylene glycol aldehyde (mPEG) with a molecular weight of approximately 20 kDa. Samples representing process intermediates, PEGylation with different lots of mPEG, and force-degraded material were analyzed. 

The Viscotek TDAmax system (Malvern Instruments) was used for the SEC analysis of the protein conjugates. This tetra detection system (TDA305) includes photodiode array UV, differential RI, right angle light scattering (RALS), low angle static light scattering (LALS) and four-capillary differential viscometer detectors. The system also includes a GPCmax VE 2001 liquid chromatography autosampler with an isocratic pump and degasser equipped with a Superdex 200 column (10  × 300 mm, GE Healthcare). The detectors were in the following order: UV, LS, RI, and viscometer. OmniSEC 4.7 software (Malvern Instruments) was used for the acquisition and analysis of SEC data. 

Determination of the detector offsets, asymmetric band broadening correction, and detector constants was completed using bovine serum albumin (BSA) monomer peak in a mobile phase of PBS (150 mM phosphate and 100 mM NaCl) at 0.5 mL/min flow rate. 0.185 (1) and 0.66 (2) were used as (

  at 280 nm) of BSA, respectively. The same process was followed with the unmodified protein for comparative analysis. The 

 was set at 0.185 (experimentally confirmed) and the extinction coefficient (

  at 280 nm) was experimentally determined. An isocratic elution profile was used at a flow rate of 0.5 mL/min. Buffer solutions were vacuum filtered through a 0.2 µm nylon filter and degassed before use. Samples were run under the same chromatographic conditions. Injection volumes were adjusted to accommodate samples with lower protein concentrations (injection volume range = 20–100 µL).  

The light-scattering intensity of polymer molecules in solution is proportional to the molecular weight and concentration of the sample (4): 

where, Rθ is the excess intensity of light scattered at angle θ, 

 is the second virial coefficient and K is an optical constant, which is proportional to 

. Because the concentration of the sample solution in SEC is highly dilute, A

 is the particle scattering function, which represents the angular dependency of light-scattering intensity and is related to radius of gyration (Rg) of the polymer molecule. Depending on the size and dissymmetry of the polymer molecule, 

 can take on values that are equal to, or less than unity. From these equations, intensity of the static light-scattering signal (LS) can be simplified to:  

 is the instrument calibration constant and 

In tetra detection SEC, viscosity of the polymer solution is simultaneously measured by a four-capillary viscometer at every elution point of the chromatogram with the detector signal of (VIS):  

where, [η] is intrinsic viscosity and C is the concentration of solution. The viscometer detector combined with a concentration detector, such as UV or RI, provides local intrinsic viscosity. The molecular size of the polymer at every elution volume can be calculated using the intrinsic viscosity.  

The concentration detectors used in SEC tetra detection are UV and RI detectors. The UV absorbance detector works like a UV–Visible spectrophotometer. The response of the detector is as follows:  

  form of the extinction coefficient. The response of the RI detector may similarly be expressed as:  

To determine the molecular weight and the size of the target molecule, 

 values or concentrations of complexes must first be determined. 

 is assumed to be constant for native protein species. For heterogeneous protein conjugates, the 

 value will change as the chemical composition varies across the SEC separation. A true 

 value can only be obtained by direct determination of discrete 

 values at each chromatographic volume increment. The response of a RI detector is the sum of the responses of each species (protein + PEG) as shown (1):  

When the conjugate complex is detected by two different concentration detectors, UV and RI, then two equations are obtained with two unknowns and this then can lead us to the percentage of both protein and PEG at every elution point: 

 is calculated for each chromatographic volume increment with these equations and used to obtain a corrected concentration distribution. The corrected 

 for each chromatographic volume increment is then used to calculate the correct M

 and size. In our calculations, we used 0.185 mL/g and 0.113 mL/g for 

 values of protein and PEG were as decribed. 

Samples prepared with different lots of mPEG were analyzed by the tetra detector system. The unmodified model protein was also tested for comparison. Figure 1 contains a representative chromatogram noting the signal from all four detectors. Table I lists parameters obtained for the main peak. Small shifts in the M

 can be detected between lots. As noted in the table, the relative difference in M

 for the mPEG accounts for the variability in the M

 of the PEGylated protein. While the mPEG is labeled as 20 kDa, this represents a polydisperse polymer with a broad population of mPEG molecules with different chain lengths. Orthogonal testing by mass spectrometry and nuclear magnetic resonance (NMR) indicates the M

 of the material used for PEGylation ranges from approximately 20100 Da to 21500 Da (7, 8, 9). The protein content percentage for each main peak is approximately 50% as would be expected for this protein modified primarily by a single molecule of 20 kDa mPEG. 

Figure 1: Tetra detector chromatogram of the PEGylated protein main peak. Detector signals shown include refractive index (red), UV at 280 nm (purple), viscometer (blue), and right angle static light scattering (green). All signals were normalized and the logMW data is shown in black to highlight the homogeneity of this native peak.  

The intrinsic viscosity is measured directly whereas the hydrodynamic size is calculated from Einstein's viscosity equation, which relates intrinsic viscosity and M

 to the hydrodynamic size of an equivalent sphere. The intrinsic viscosity for the model PEGylated protein (0.228–0.240 dL/g) is similar to free PEG (0.38 dL/g) and greater than protein alone (0.034 dL/g), as expected. The measured intrinsic viscosity for BSA (0.039 dL/g) is comparable to the published value (0.037 dL/g) (5). The hydrodynamic size of the PEGylated protein increases as the mPEG of larger size was used for modification. The hydrodynamic radius for the main peaks (5.17–5.35 nm) are similar to those measured (by SEC with calibrated PEG standards) for proteins with a similar M

 modified with a single 20 kDa mPEG (6). 

Table I: Comparison of model PEGylated proteins modified with different lots of mPEG1.  

Repeat injections differ by less than 1% for all parameters noted in Table I. These highly reproducible results were obtained with separate calibrations using BSA or the unmodified protein.  

Figure 2: Staggered overlay of the refractive index signal (a) and right angle static light scattering (RALS) signal (b) for the five stressed samples. It is clear that different applied stress conditions will impact both the concentration of oligomer and aggregate species, as well as their distribution. Note that the light scattering signal is a function of size, not necessarily abundance, especially for large aggregates.  

A process intermediate and several forced degradation samples were analyzed by the tetra detection system (see Figure 2). Generally, two primary peaks were detected, a soluble aggregate peak near the void of the column eluting near 8 mL and a soluble oligomeric peak eluting at 10 mL, near the main peak. For the oligomer and aggregate peaks M

 information was obtained from static light scattering and the copolymer method in the OmniSEC software. Analysis of the integrated peaks is shown in Table II, along with the repeatability and precision of the measurement for the oligomer peak. Data are tight across two days of testing for a peak at 0.34% (peak area by RI). Stress by agitation, heat or extended room temperature incubation produces samples with HMW impurities with different molecular weights. The agitated sample contains a large oligomeric peak whose molecular weight is indicative of a dimer of the PEGylated protein. In addition to molecular weight, the hydrodynamic radii can be determined by this analysis. 

Table II: Comparison of impurity peaks detected in model PEGylated protein samples.  

The oligomer and aggregate peaks were less homogeneous than the native conjugate. The heterogeneity is numerically depicted in the polydispersity value (M

). Visual examination of the light-scattering signal for the HMW impurities indicates differences between samples and heterogeneity in this peak. The aggregate peaks demonstrated greater variability in molecular weight and polydispersity. The intrinsic viscosity of the aggregate was unable to be reliably measured by this method. This is likely due to the small weight fraction present of these species and the inability to produce adequate viscosity data.  

Interpretation of the results depends on many variables. These measurements used the same 

 inputs for protein and polymer. The assumption is that the 

 values of the HMW impurities are the same as that of the monomeric PEGylated protein. Confirming the 

 values would require purifying individual HMW impurities. This would be challenging, considering the transient nature of certain HMW impurities described elsewhere for a protein of similar M

 (10, 11). We noticed a qualitative loss of aggregate when testing several stressed samples. In addition, confirming the homogeneity of the peaks themselves would require thorough orthogonal testing. The chromatography could be further optimized using an alternate column or mobile phase to improve separation between different HMW components (12).  

A tetra detection system was used to analyze a model PEGylated protein and its associated HMW impurities. To our knowledge, this study is the first comparison of impurity peaks in a PEGylated protein by this methodology. The sensitivity of this system is high and permits detection of HMW impurities as low as 0.34%. The outputs of this type of analyses can be used for comparison of batches during process development, stressed samples and assessing mechanisms of aggregation. 

The authors would like to thank Dr. Alok Sharma and Dr. Stan Bastiras for helpful comments on the manuscript and Dr. Thomas Vanden Boom and Dr. Dave Schwinke for suggestions and support of this work. 

 is senior group leader at Hospira, Inc., Global Biologics R&D, 275 North Field Drive, Lake Forest, IL 60045, tel. 224.212.6363, fax 224.212.7905, 

 is an associate research ccientist at Hospira, Inc. 

 is a product manager at Malvern Instruments Inc. 

 5.  Cantor and Schimmel, "Other Hydrodynamic Techniques," in 

Biophysical Chemistry, Part II: Techniques for the study of biological structure and function

 (W.H. Freeman and Company, NY, 1st ed., 1980), p. 650-653. 

Articles

Peer-Reviewed Research

Reliably detecting low amounts of high molecular weight impurities during process development and characterization of biopharmaceutical products.

Analysis of PEGylated Protein by Tetra Detection Size Exclusion Chromatography

Five of the top seven and six of the top twenty biological products in 2010 (in terms of revenue) were monoclonal antibodies (mAbs) (1). Several new mAb products and mAb biosimilars are in development, and all require extensive characterization to obtain the necessary approvals for clinical trials and to eventually be released onto the market. 

 The most recent regulatory document covering characterization of mAbs was published by the European Medicines Agency (EMA) in July 2009 (2). This EMA guideline entitled 

Development, Production, Characterization and Specifications for Monoclonal Antibodies and Related Products

 states that the mA should be characterized thoroughly (2). This characterization should include the determination of physicochemical and structural properties, purity, impurities and quantity of the mAb, in line with International Conference on Harmoniaation (ICH) guideline Q6B (3). 

ICH Q6B provides a uniform set of internationally accepted principles for characterization of biotechnological products to support new marketing applications. The document suggests that analyses be performed to provide the following information for biological or biopharmaceutical products: 

Sulfhydryl group(s) and disulfide bridges

Extinction coefficient (or molar absorptivity)

The techniques used to provide the analytical data required by these guidelines are discussed below. 

AMINO ACID COMPOSITION AND            DETERMINATION OF THE EXTINCTION COEFFICIENT 

ICH Q6B states that "The overall amino acid composition is determined using various hydrolytic and analytical procedures, and compared with the amino acid composition deduced from the gene sequence for the desired product, or the natural counterpart, if considered necessary. In many cases amino acid composition analysis provides some structural information for peptides and small proteins, but such data are generally less definitive for large proteins. Quantitative amino acid analysis data can also be used to determine protein content in many cases." 

Amino acid composition analysis using reverse-phase high-performance liquid chromatography (RP–HPLC) with pre-column derivatization or ion exchange chromatography with post-column derivatization are now routinely used to determine the amino acid composition of biopharmaceutical products. Provided that the molar ratios of amino acids per mole of protein and the overall molecular weight of the protein are known, the protein content can be calculated from the amount of each robust amino acid detected in a hydrolysate of the product. Also, if the optical density (at 280 nm) for the solution from which the aliquot(s) for amino acid analysis were taken is known, the Beer–Lambert law (absorbance [OD] = ε.c.l; where ε is the extinction coefficient, c is the concentration in M/L, and l is the path length in cm) allows calculation of the extinction coefficient for the product, meaning that spectrophotometry at 280 nm can be used to quantify the product in solution. 

For a mAb, the EMA guideline and ICH Q6B request a deduction of the amino acid sequence from the DNA sequence and confirmation of the DNA-derived sequence experimentally by appropriate methods (e.g., peptide mapping, amino-acid sequencing, mass spectrometry analysis). In addition, the variability of the 

terminal amino acid sequences (whether a free amino acid or a pyroglutamic acid is present) and the 

-terminal amino-acid sequences (e.g. the presence or absence of 

-terminal lysine(s) on the heavy chain) should be analyzed. 

Peptide mapping (i.e., analysis of specific protease digests of the mAb followed by analysis of the products using on-line RP-HPLC with ultraviolet [UV] and electrospray mass spectrometric detection [LC/ES–MS]) provides molecular weight information for the peptides released from the mAb by the protease of choice. The data obtained are able to provide mapping confirmation (or otherwise) of the DNA-derived sequence but does not provide confirmation of the sequence of the light and heavy chains of the mAb. To provide confirmation at the primary amino acid sequence level, a number of protease digests combined with on-line RP–HPLC with tandem MS/MS (LC/ES–MS/MS) analysis are required. ICH Q6B does not actually request sequencing at the primary amino acid level. There is a move, however, initiated by some of the regulatory bodies, to provide confirmation of the primary protein sequence for new mAb products and also provide data showing comparability of the the primary amino acid sequence of mAb biosimilar products to a reference product. 

As mass spectrometric-based sequencing is unable (in most cases) to differentiate between isoleucine and leucine (these amino acids have the same molecular weight), automated 

-terminal sequencing of purified peptides is required for unambiguous assignment of these two amino acids, particularly within the variable regions of the light and heavy chains. 

Terminal amino acid analysis is performed to identify the amino- and carboxy-terminal amino acid sequence(s) of the light and heavy chains of a mAb. If the product exhibits more than one terminal amino acid sequence, the relative amounts of the termini should be determined. 

-terminal sequencing following separation of the light and heavy chains using sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and blotting onto polyvinylidene fluoride (PVDF) membrane is used routinely for analysis of the 

-termini of mAb light and heavy chains. Automated 

-terminal sequencing uses Edman chemistry which requires a free amino functionality at the 

-terminus of a protein for labeling prior to cleavage of the 

-terminal and subsequent amino acids. This means that for a number of mAbs (which have a pyroglutamic acid at the 

-terminus and therefore no free amino group), 

-terminal sequencing using this method will not provide sequence information. In this case pyroglutaminase can be used to enzymatically remove the 

-terminal pyroglutamic acid residue before 

There is no fully reliable method analogous to Edman chemistry for determining the 

-terminal amino acid sequence for a biological product. Information relating to the 

-terminal sequence of a peptide or protein can be obtained using carboxypeptidase digestion and/or mass-spectrometric mapping strategies. In the latter approach, intactness of a protein 

-terminus or the presence of ragged ends can be assessed using data obtained from a peptide map and intact molecular weight analysis of the product. 

SULFHYDRYL GROUP(S) AND DISULFIDE BRIDGES

During a peptide mapping/sequencing analysis, free sulfhydryl groups and disulfide bridges should also be considered. Peptide mapping post-digestion with a specific protease, followed by analysis using on-line LC/ES–MS or LC/ES–MS/MS prior to/and following reduction, provides the data necessary for a full assessment of disulfide bridges and free thiols within the mAb. Care must be taken when digesting proteins containing free thiols at basic pH as the free thiol can initiate scrambling within the digest, and the result obtained may therefore not be consistent with the true disulfide bridge/free thiol pattern within the product. 

As mAbs are glycoproteins, it is also necessary to characterize the glycan portion of each product. Typically, mAbs have one 

-linked glycan consensus sequence (asparagine–Xxx–serine or threonine, where Xxx can be any amino acid except proline) within each heavy chain located in the constant fragment (Fc) region. The light chain component of a mAb is not normally glycosylated. Since additional 

-linked glycan consensus sequences may be present within the heavy chains, the presence or absence of glycosylation should be confirmed. ICH Q6B requests that "For glycoproteins, the carbohydrate content (neutral sugars, amino sugars and sialic acids) is determined. In addition, the structure of the carbohydrate chains, the oligosaccharide pattern (antennary profile) and the glycosylation site(s) of the polypeptide chain is analyzed, to the extent possible." 

As shown in Figure 1, monosaccharide composition analysis is usually carried out to determine the carbohydrate content of a mAb. Liquid chromatography and gas chromatrography (often with mass spectrometric detection) are commonly used to define the carbohydrate content of a mAb. The method(s) used should allow identification and quantitation of the levels of the neutral sugars (fucose, mannose and galactose), amino sugars (

-glycolylneuraminic acid) within the product. 

Figure 1: Schematic of the normal approaches employed to analyze the glycosylation of a monoclonal antibody (mAb). LC is liquid chromatography, MS is mass spectroscopy.  

Following monosaccharide composition analysis, the next stage of analysis moves into characterization of the oligosaccharides normally present on each heavy chain. An example of the methodology used for releasing, purifying and analyzing the 

-linked glycans from a mAb is shown in Figure 1.  

Reduction/alkylation and the specific protease digestion are intended to denature the mAb so that the relatively large enzyme PNGase F can do an efficient job at removing the 

-glycans are released, a simple reverse phase cartridge can be used to separate the hydrophilic glycans from the more hydrophobic peptides. The 

-linked glycan fraction can then be analyzed using mass spectrometry and/or liquid chromatography-based techniques. 

The matrix-assisted laser desorption ionisation time-of-flight (MALDI–TOF) mass spectrum obtained from analysis of the 

-glycans released from a mAb (using the methodology shown in Figure 1) is shown in Figure 2. The major signals observed are consistent with the commonly encountered mAb glycans G0F, G1F and G2F (see Figure 2). 

Figure 2: Raw data obtained from matrix-assisted laser desorption ionisation time-of-flight (Maldi-TOF) mass spectrometric analysis of the N-glycans released from a mAb. The N-glycans were permethylated prior to analysis.  

The high-pH anion-exchange chromatography with pulsed amperometric detection (HPAEC–PAD) trace obtained from analysis of the 

-glycans released from the same mAb product is shown in Figure 3. 

Figure 3: Raw data obtained from high-pH anion-exchange chromatography with pulsed amperometric detection analysis (HPAECâPAD) of the N-glycans released from a mAb.  

Both sets of data can be used to provide a relative quantitation of the 

-linked oligosaccharides observed and suggest the structures of the 

-linked glycans present. Glycosylation is the most common post-translational modification encountered by the regulatory authorities. The regulators are aware that glycosylation can be significantly affected by even simple changes in the manufacturing processes, such as a pH change or a change of growth media. The monosaccharide composition analysis data (in terms of levels of mannosylation, galactosylation and sialic acid) and 

-linked oligosaccharide profiling data obtained from the analysis of various batches of a mAb product are therefore used by the regulator as part of the assessment of whether a manufacturer has control of the manufacturing process of a mAb. Comparability of 

-glycan profiles from batch to batch suggests that the manufacturing process is under control. 

The EMA and ICH guidelines also recommend that the higher-order structure of the mAb should be characterized by appropriate physicochemical methodologies. ICH Q6B recommends the use of "circular dichroism, nuclear magnetic resonance (NMR), or other suitable techniques, as appropriate." Today, circular dichroism (CD) analysis is commonly used to assess the secondary structure of mAbs. The data obtained from CD analysis of a mAb is shown in Figure 4. Assessment of the raw data using CDsstr software allows the relative percentages of the various secondary structures to be estimated as shown in Table I. 

Figure 4: Raw data obtained from near ultraviolet (UV) analysis of a mAb.  

In addition to the structural characterization requirements outlined above, a number of additional physicochemical properties need to be assessed. 

Table I: The result of CDSSTR algorithm assessment of the near UV CD (circular dichroism) data shown in Figure 4. 

The author describes techniques that can be used to provide the analytical data required by ICH Q6B for characterization of monoclonal antibodies.

Structural Characterization of Monoclonal Antibodies

The characterization of particulates in biologics is a relatively new concern that presents unique challenges, compared with the characterization of particulates in classical (i.e., nonbiologic) injectable drug formulations. Dynamic imaging particle analysis represents an exciting new and increasingly popular method for the characterization of particulates in biologics. However, as with any other type of instrumentation, it is important to understand how dynamic-imaging instruments measure things and what factors affect the measurements to properly interpret the results. This article will discuss three of the primary factors to understand when using dynamic imaging particle analysis as a particle-characterization technique for biologics: resolution, thresholding, and image quality. 

ALL FIGURES ARE COURTESY OF THE AUTHOR 			

Characterization of subvisible particulates in parenterals has been a concern since 1936 and was formally addressed by 

 <788> in 1975 (1). At the time of its implementation, <788> was primarily concerned with foreign matter, such as rubber stopper pieces. The concern was largely mechanical in nature, because these were hard particles that might not be distributed through the blood system easily (2). Although biologics raise the same concerns, they also are subject to protein aggregation, whereby small particles combine to create larger ones. Aggregated proteins are soft particles, so they may be able to pass through restrictions that block hard particles. Aggregated proteins may be more difficult to detect than opaque particles because they are transparent. Although light-obscuration devices specified by <788> detect hard, opaque particles well, they do not always detect or properly characterize soft, transparent particles. This deficiency is well documented (2). 

Further complicating the characterization of biologics is the fact that the aggregates can be amorphous and range from strandlike to circular shapes. Because light-obscuration devices calculate size based on an assumption of spherical particles, the size measurements can be highly inaccurate. Finally, because these biologics frequently will be delivered through prefilled syringes, the presence of silicone droplets also can cause inflated particle counts and an overall mischaracterization of the biologic. 

Recognizing the above cited limitations of light-obscuration techniques, researchers have begun to look for alternative methods for characterizing subvisible particulates in biologics. One area that has shown great promise is in dynamic imaging particle analysis (2). Dynamic imaging particle analysis systems capture digital microscope images of particles in the biologic as they pass through a flow cell. Each particle captured can be measured using standard image-analysis algorithms. Unlike light-obscuration systems, imaging systems can make various measurements, both morphological and spectral, on each particle, even when the particle is transparent. With all these different measurements available for each particle, one can achieve a detailed description that includes particle shape. This description enables automatic differentiation between particle types, such as protein aggregates and silicone droplets. A block diagram of a typical system is shown in Figure 1. 

 Figure 1: Block diagram of FlowCAM (Fluid Imaging Technologies) dynamic imaging particle-analysis system. LED is light-emitting diode.			

To demonstrate how dynamic imaging particle analysis can characterize subvisible particulates in biologics, the author analyzed a sample protein-based therapeutic using the FlowCAM (Fluid Imaging Technologies) particle imaging system. Figure 2 shows the result of the analysis, including summary statistics and representative particle images. This particular sample contained a large amount of silicone droplets. To enable the system to differentiate between various types of particles automatically, one can build a digital filter to distinguish the silicone droplets from other particulates. The filter is created by selecting a few examples of silicone droplets and instructing the software to find similar particles. The system uses a sophisticated method known as statistical pattern recognition to classify every particle automatically as being one of the class (i.e., silicone droplets) or not (3). 

Figure 2: Results of a FlowCAM run with protein-based biologic sample, including (left) graphs and summary statistics and (right) representative particle images including, protein aggregates and silicone droplets.			

In this example, out of the 2844 original particles in the run, 1788 were identified as silicone droplets. Because 

 <788> addresses particles larger than 10 µm and larger than 25 µm, the particles were sorted into two categories (see Table I). One can see a dramatic difference between the particle counts for the two categories once the silicone droplets have been removed. If the silicone droplets are considered to be a harmless byproduct of the delivery method, then excluding them from the analysis will yield different results than would be found using light obscuration according to <788>. 

Table I: Number of particles found in each size category before and after silicone droplets were removed.			

The first factor to consider when using dynamic imaging particle analysis is resolution. Unlike the human eye, which sees the world as continuous, digital images are sampled versions of the real world. This means that any given image is divided into discrete picture elements or pixels that form the whole. Thus, the imaging sensor has a discrete number of samples of the object being captured. The easiest way to visualize this is to project the sensor geometry onto the object being captured, as shown in Figure 3. In this diagram, the overall system magnification is 200%, so that the object's image is two times larger than the actual object. A more meaningful way to express this is in terms of a calibration factor equal to the size of one pixel of the sensor projected onto the object. In this case the calibration factor would be 2.5 µm/pixel. Therefore, one pixel of the image covers 2.5 µm in length or width. Since a pixel is the smallest unit of a digital image, this means that, in theory, the system can measure down to 2.5 µm. 

Figure 3: Conceptual diagram of object and image space in an optical system.			

However, the system cannot really measure to that small a level because of sampling. The Nyquist–Shannon sampling theorem states that in the conversion of a continuous signal to a discrete signal, one can only resolve the original signal by sampling at two times the rate of the original sample. To use the previous example, one can only resolve a 2.5 µm signal by sampling at a minimum rate of 1.25 µm. Using the same example, with a 2.5-µm minimum pixel size, the smallest object the system can possibly resolve is 5 µm in size. In terms of particle analysis, the term "resolve" would apply only to counting a particle, and not to any higher-level description of the particle, such as shape. Indeed, the higher the order of the measurement desired, the more resolution is needed (3). 

As a result of diffraction limitations inherent to microscopy-based systems and sampling theory as discussed above, dynamic particle imaging systems are limited to counting particles no smaller than 1 µm in size, and being able to differentiate shape information for particles no smaller than 2–3 µm in size (4). To image particles smaller than these limits requires electron microscopy, where the sample size and number of particles that can be imaged are extremely limited, and therefore statistically significant numbers of particles cannot be achieved. Other synthetic image techniques, in which an image is produced by sampling another characteristic (e.g., Brownian motion, or atomic force microscopy), can also be used for smaller particles, but these are not direct optical images. 

The second factor to consider when using dynamic imaging particle analysis is the effect of thresholding. Not only are digital images quantized in the spatial domain in terms of a limited number of pixels, but each pixel is quantized in terms of gray-scale (or color in the case of color imaging) resolution, also. In most typical systems, the gray-scale (i.e., intensity) value of each pixel is limited to 256 levels or eight bits. In a color system, there are eight bits each of red, green, and blue. To make rapid measurement calculations on the image data in particle image analysis, the gray-scale intensity is reduced to a single bit (i.e., on or off) through a thresholding process. In imaging particle analysis, this thresholding is performed by comparing each pixel of an incoming image that may contain particles with the same pixel of a background image taken when no particles were present in the system. Because most systems are "bright-field" or backlit, a particle in the optical path will reduce the amount of light passing through to the camera sensor, therefore the incoming pixel intensity will be darker (i.e., a smaller number) than the background calibrated value for the same pixel. For this reason, most imaging particle analysis systems will define a threshold as either a delta value or percent value darker than the background. 

For opaque particles, thresholding in this fashion (i.e., in which pixels are darker than the background) works quite well. However, as previously mentioned, protein aggregates are semitransparent and amorphous. In fact, because of the way light is bent through the structure of the aggregate, many of the pixels within the aggregate will be brighter than the background. Therefore, if only a dark threshold is used, the result will be that a single aggregate will be chopped up into multiple small particles by the thresholding process. This technique results in a severe overcounting of small particles and undercounting of large particles (5). 

To avoid this problem, the imaging particle analysis system should allow for thresholding based on pixels that are either darker or lighter than the background. While this solution is certainly helpful, it still may allow the thresholding process to form some image artifacts. Fortunately, common image-processing algorithms are available to overcome this problem by using neighborhood analysis to group disparate clusters of thresholded pixels into logical whole images. Figure 4 shows the differences obtained for a thresholded image of a single protein aggregate image using a dark-only pixel threshold, a dark-and-light pixel threshold, and finally, a dark-and-light pixel threshold plus neighborhood analysis. 

Figure 4: The effect of varying threshold on the binarization of a protein-aggregate image. Red pixels are particles, based on the threshold. The green boxes show enclosed particles found after thresholding. Image 1 is the original protein aggregate image. Image 2 shows the result of a threshold of 25 darker than background. Image 3 shows the result of a threshold of 15 darker. Image 4 shows the result of a threshold of 15 lighter. Image 5 shows the result of thresholding 15 lighter and darker. Image 6 is the same as Image 5, but with the addition of neighborhood analysis.			

The third factor to consider when using dynamic imaging particle analysis is the effect of image quality. Although the overall topic of image quality is quite broad and well beyond the scope of this article, a basic tenet of the subject is that image sharpness is the most critical measurement of image quality (6). Indeed, in imaging particle analysis, the sharpness of the particle images is directly proportional to the accuracy of the measurements obtained.  

Figure 5: Thresholded images of National Institute of Standards and Technology traceable 10-Âµm spheres. Sphere images in sharp focus are on the left, and less sharp images are at right. The table shows variance in measured diameter based on different threshold values.			

Figure 5 demonstrates this by showing images and measurements obtained in an imaging particle analysis system for National Institute of Standards and Technology traceable size bead standards. The images at left are beads in sharp focus, and the images at right are beads in less sharp focus. The variation in size and shape caused when the bead images are less sharp is easy to see, and Table II shows the variation in size measurements that would be obtained by the system with different thresholds. For 10-µm calibrated spheres, the variation in measurement for the blurry images is more than 12 µm for a difference of 100 in threshold value, whereas the variation in measurement for the sharp images is only 1.67 µm over the same threshold range. It is clear that image sharpness greatly affects the accuracy and precision of particle measurements. 

Table II: Equivalent spherical diameters (ESD).			

Because the filtering operations that automatically separate various particle types in imaging particle analysis are based on the image measurements, it seems to follow that less-than-sharp images with inaccurate measurements would not produce good-quality particle characterization. To demonstrate this idea with protein aggregates, the author analyzed a sample protein-based therapeutic with the FlowCAM using a deeper-than-normal flow cell to yield particle images of varying sharpness. Two image libraries were created for use with a statistical pattern-recognition algorithm. One contained eight aggregate images in sharp focus, and the other contained eight aggregate images in less sharp focus. Each library was then used to perform a statistical pattern match on the entire run of aggregate particles.  

The results are shown in Figure 6. Sixty percent fewer aggregates were found using the less focused library images. In statistical filtering, this occurs because the library particles form a looser, more ambiguous cluster in the 

-dimensional pattern-recognition space because of the higher variance in measurements (3). This cluster, in turn, yields lower statistical confidence in the identity of the library particles, and therefore a lower match rate. 

Not only will the resulting match rate be lower for less sharp particle libraries, but the inaccuracy of the measurements can also lead to false positives. To summarize this discussion of image quality: Fuzzy Images = Fuzzy Measurements = Fuzzy Classifications. 

Figure 6: Results of statistical pattern recognition with two libraries of various image quality. Results on left (152 particles/mL) used the sharp library particles, and results on the right (91 particles/mL) used the less sharp library particles.			

For the characterization of protein aggregates and other particulates in biologics, volumetric techniques such as light obscuration, laser diffraction, and electrozone (Coulter) sensing have limited effectiveness because of the fact that each must assume that particles are spherical in shape. This assumption means that the techniques cannot distinguish between protein aggregates and other particulates, such as silicone droplets, found in biologics. 

Imaging particle analysis can overcome the limitations of these volumetric techniques by measuring shape and gray-scale parameters of particles. It also has the benefit of detecting transparent particles, such as protein aggregates, where techniques such as light obscuration fail or mischaracterize the particles. However, as with any other method of particle analysis, it is important to understand the limitations and factors that affect particle measurements using this method. This article has shown that three factors must be kept in mind when evaluating the efficacy of an imaging particle analysis solution for characterizing biologics: resolution, thresholding, and image quality. Once these factors are understood, one can have higher confidence in one's interpretation of results based on this technique. 

 is the director of marketing at Fluid Imaging Technologies, 65 Forest Falls Dr., Yarmouth, ME 04096, 

 1. S. Aldrich, US Pharmacopeia Workshop on Particle Size: Particle Detection and Measurement (Rockville, MD, 2010). 

 3. L. Brown, "Particle Image Understanding–A Primer," 

www.fluidimaging.com/resource-center-whitepapers.htm

 4. L.Brown, "Imaging Particle Analysis: Resolution and Sampling Considerations," 

 5. J.S. Pedersen, Bio-Process International European Conference and Exposition (Nice, France, 2011). 

 6. L. Brown, "Imaging Particle Analysis: The Importance of Image Quality," 

Overcoming limitations of volumetric techniques and detecting transparent particles.

BioPharm International-08-02-2011

Analysis of PEGylated Protein by Tetra Detection Size Exclusion Chromatography

Structural Characterization of Monoclonal Antibodies

Characterizing Biologics Using Dynamic Imaging Particle Analysis

BioPharm International, August 2011 Supplement