Dilution-Free Protein Concentration Measurement for High Protein Concentration Samples

A widely used method for determining the concentration of purified proteins in solution is by absorption at 280 nm using commercially available ultraviolet/visible (UV/VIS) spectrophotometers. The concentration is determined according to the Beer-Lambert Law (1) calculation adapted for proteins of known absorptivity as shown below:

The absorbance at 320 nm is subtracted to account for light scattering by the test sample due to aggregation and should be low relative to the absorbance at 280 nm. According to the equation, the measured concentration is inversely proportional to the path length of the absorption cell (cuvette) used for the absorption measurement. As an example, for bovine serum albumin (BSA) with an absorptivity value of 0.667 mL mg^-1 cm^-1, a concentration range of 0.3 to 1.8 mg/mL can be measured using a standard spectrophotometer with a 1-cm path length cell, assuming the maximum absorbance of approximately 1.2 in the spectrophotometer. The absorption limits or linear dynamic range of a spectrophotometer is limited by noise and stray light and is influenced by the type of detector in use. It should be noted that absorbance by the cuvette and buffer solution used for blanking the spectrophotometer is part of the total absorption and must be considered.

For the BSA described, if the path length of the cell is decreased by a factor of five to 0.2 cm, the measurable range can be increased by a factor of five to 1.5 to 9.0 mg/mL. If the path length is decreased by a factor of 100 to 0.01 cm, then the measurable range will be 30 to 180 mg/mL. A standard spectrophotometer can, therefore, be used with commercially available shorter path length cuvettes to measure higher protein concentrations without dilution. Spectrophotometers and spectrophotometer accessories (such as fiber optic probes) tailored for high protein concentration measurement are also available and are based on the same principle of reduced path lengths.

The biopharmaceutical industry manufactures recombinant proteins at various protein concentration levels for therapeutic applications. Using the previously described BSA concept, for example, on purified recombinant transforming growth factor beta superfamily proteins with an absorptivity of 1.36 mL mg^-1 cm^-1or a monoclonal antibody (mAb) with a similar absorptivity, the measureable range is approximately 0.15 to 0.9 mg/mL with a 1 cm path length cell. However, if the cuvette path length is decreased by a factor of ten to 0.1 cm, the protein concentration range changes to approximately 1.5 to 9 mg/mL. For a cuvette with a path length 100 times shorter, 0.01 cm, the protein concentration range is approximately 15 to 90 mg/mL.

Therefore, according to the calculated results shown in Table I, it is possible to measure a wide range of protein concentrations without performing any dilution of the samples. Protein concentrations for biopharmaceutical drug substances and drug products can be determined directly without prior dilution by selecting the appropriate path length absorption of the cuvette according to the absorptivity of the protein and the expected concentration.

Table I: Measureable concentration ranges by cuvette path length.

Absorbance at 280 nm

Cuvette path length

1.0 cm

0.2 cm

0.1 cm

0.02 cm

0.01 cm

Bovine serum albumin (BSA) with absorptivity 0.667 mL mg-1 cm-1

1.2

1.80

9.00

17.99

89.96

179.91

1.0

1.50

7.50

14.99

74.96

149.93

0.7

1.05

5.25

10.49

52.47

104.95

0.5

0.75

3.75

7.50

37.48

74.96

0.3

0.45

2.25

4.50

22.49

44.98

0.2

0.30

1.50

3.00

14.99

29.99

Protein with absorptivity 1.36 mL mg-1 cm-1

1.2

0.88

4.41

8.82

44.12

88.24

1.0

0.74

3.68

7.35

36.76

73.53

0.7

0.51

2.57

5.15

25.74

51.47

0.5

0.37

1.84

3.68

18.38

36.76

0.3

0.22

1.10

2.21

11.03

22.06

0.2

0.15

0.74

1.47

7.35

14.71

MATERIALS AND METHODS
UV absorption of protein solutions was measured at 280 and 320 nm using an Agilent 8453 UV/VIS spectrophotometer, with various 1 cm path length quartz absorption cells. Quartz cuvettes with path lengths of 0.01, 0.1, and 0.2 cm from Agilent were used for high protein concentration measurements. The dimensional tolerances were ± 0.001 cm (0.1%) for the 1-cm path length cell, 0.001 cm (1%) for the 0.1-cm cell, 0.001 cm (0.5%) for the 0.2-cm cell, and 0.0005 cm (5%) for the 0.01-cm cell (2). For comparison, the acceptable systematic plus random error for mechanical pipettes are 1.4% for 10 μL, 1.0% for 100 μL, and 0.8% for 1000 μL. In addition, the viscosity of the solution, the goodness of fit of the tips used, and the technique of the operator can introduce error. Thus the dilution error introduced by one dilution can be larger than the dimensional error for all but the shortest path length cuvette (0.01 cm) and for multiple dilutions the error increases.

The NanoDrop 8000 (Thermo Fisher Scientific), a spectrophotometer with fixed light paths of 0.1 and 0.02 cm for which a cuvette is not needed, was also used for comparison. Each sample was measured in triplicate. The assessment of measurement repeatability is used to ensure that there are no problems with sample carry over or surface cleanliness.

For implementation of the use of short path length cells on a routine basis, all operations such as filling the cuvette, ensuring adequate rinsing between samples, and cleaning and drying the cuvette should be proceduralized with specific instructions and validated to ensure that the procedure is robust with multiple analysts.

RESULTS
Accuracy and precision for protein concentration determinations using shorter path length cells with no sample dilution was assessed using BSA, a mAb, and a transforming growth factor beta. BSA concentrations of approximately 10 mg/mL and 70 mg/mL were used in the study. The purified recombinant transforming growth factor beta superfamily proteins and mAb proteins used in the studies have absorptivities of 1.36 mL mg-1 cm-1 and 1.38 mL mg-1 cm-1. The concentrations examined were 6, 8, 20, and 60 mg/mL. The results are presented in the following sections.

6.0 mg/mL protein concentration. Accuracy of the procedure for determining protein concentration of highly concentrated protein samples using the shorter path length cuvette was tested using recombinant protein at about 6 mg/mL concentration with absorptivity of 1.36 mL mg-1 cm-1. The concentration was measured directly (without dilution) using a 0.1-cm cuvette and the results are shown in Table II. The result of 6.5 mg/mL is in excellent agreement (99% of the value) with the result of 6.6 mg/mL obtained using a 1-cm cuvette by a product-specific validated method using multiple dilutions at various concentrations. The data presented in Table II also indicate that the precision of this method using the 0.1-cm cuvette is very good with only about 0.1% RSD (relative standard deviation) for six measurements.

Table II: Protein concentration results for 6 mg/mL purified protein with an 0.1-cm path length cuvette (absorptivity 1.36 mL mg-1 cm-1) compared with the accepted value obtained using a 1-cm path length cuvette.

6 mg/mL sample¹

Absorbance280 nm0.1 cm

Absorbance²320 nm0.1 cm

Concentrationusing 0.1-cm cuvette(mg/mL)

Accepted³concentrationusing 1-cm cuvette(mg/mL)

Accuracy⁴

6 mg/mL_a

0.90626

0.01880

6.525

6.751

6 mg/mL_b

0.90585

0.01765

6.531

6.623

6 mg/mL_c

0.90559

0.01749

6.530

6.603

6 mg/mL_d

0.90527

0.01754

6.527

6.519

6 mg/mL_e

0.90623

0.01804

6.531

6.456

6 mg/mL_f

0.90685

0.01778

6.537

N/A5

average

6.530

6.590

99%

standard deviation

0.004

0.112

% relative standard deviation

0.1

1.7

¹Six independent measurements of a homogeneous sample were made.  The cuvette was removed from the holder, emptied, rinsed, and then filled with sample for each measurement.  
²A₃₂₀ values were subtracted.  ³The accepted value was obtained using a 1-cm path length cuvette and five different dilutions within the absorbance range of the spectrophotometer.
⁴% accuracy = (observed/accepted) x 100. ⁵Only five replicates consisting of different dilutions were performed for the 1-cm path length cuvette measurements.

20 mg/mL protein concentration. Next, a study was performed using an 0.01-cm path length cuvette for a protein with absorptivity of 1.38 mL mg-1 cm-1. Based on cuvette path length and absorptivity, the expected measurable protein concentration range is approximately 15 to 90 mg/mL. In this study, a recombinant protein of approximately 20 mg/mL concentration was analyzed with the 0.01-cm path length cuvette in the Agilent 8453 UV/Vis spectrophotometer. Samples from six batches targeted to the same concentration of approximately 20 mg/mL were used and each batch was analyzed in triplicate. The results of the absorbance values and the protein concentrations measured with the 0.01-cm path length cuvette are shown in Table III. The protein concentrations of the 20 mg/mL samples determined using the 1-cm cuvette by a specific validated method for this product are also shown in Table III. This value (i.e., the accepted value) was used to calculate the accuracy compared to that of the value obtained from the 0.01-cm path length cuvette.

Table III: Protein concentration results for 20 mg/mL purified protein with an 0.01-cm path length cuvette
(absorptivity of 1.38 mL mg-1 cm-1) compared with the accepted value obtained using a 1-cm path length cuvette.

Sample¹

Absorbance280 nm0.01 cm

Absorbance 320 nm0.01 cm

Concentration²(mg/mL)for 0.01 cm

Average,standard deviation,% RSD³

Accepted⁴concentrationusing 1 cm(mg/mL)

Accuracy⁵

1a

0.28071

0.00161

20.224

20.084
0.2386
1.2

19.0

106%

1b

0.27434

-0.00469

20.219

1c

0.27703

0.00367

19.809

2a

0.28773

0.00182

20.718

20.979
0.2387
1.1

20.3

103%

2b

0.28621

-0.00406

21.034

2c

0.29580

0.00344

21.186

3a

0.28483

-0.00120

20.727

20.769
0.0410
0.2

20.0

104%

3b

0.29036

0.00321

20.808

3c

0.29054

0.00387

20.773

4a

0.28894

-0.00098

21.009

21.030
0.0191
0.1

19.7

107%

4b

0.28991

-0.00041

21.038

4c

0.29056

0.00015

21.044

5a

0.28688

0.00073

20.736

20.780
0.0653
0.3

20.0

104%

5b

0.29164

0.00384

20.855

5c

0.29078

0.00445

20.749

6a

0.29180

0.00576

20.727

20.994
0.4054
1.9

20.0

105%

6b

0.30345

0.00730

21.460

6c

0.29302

0.00607

20.793

¹Three independent measurements were made for each sample.  For example, 1a, 1b, and 1c are three measurements for Sample 1. ²A320 values were subtracted. ³RSD is relative standard deviation.
⁴The accepted value was previously obtained using a product specific validated method for a 1 cm path length cell. ⁵Accuracy = (observed/accepted) x 100.

Based on these results, the protein concentration values obtained for the six batches using 0.01-cm path length cell without any sample dilution are within 3 to 7% of the accepted value. The RSD value was slightly more than 1% or less for most of the measurements, indicating the method has good precision. The 1% RSD is less than expected for multiple volumetric dilutions, which can be up to 5% depending on the protein concentration. For example, multiple dilutions for the 6 mg/mL data in Table III by volumetric dilutions had RSD 1.7. The precision for the accepted value for which the dilutions were performed gravimetrically and not volumetrically (precision data not shown) was up to 1% and is comparable to the results with the 0.01-cm path length. The results have a small positive bias compared with the 1-cm cuvette results but the bias is consistent towards the positive side and not randomly distributed.

60 mg/mL protein concentration. Similar to the 20 mg/mL protein sample, a 60 mg/mL protein sample was evaluated using the 0.01-cm cuvette because this cuvette’s range is predicted to be up to 90 mg/mL. The results obtained for the 60 mg/mL protein samples from six batches are presented in Table IV. The accepted value for these batches by the product specific validated method is also shown in the table. Based on these results, the protein concentration values obtained for the six batches using the 0.01-cm path length cell without any sample dilution is within 3 to 7% of the accepted value. Also the RSD value for the replicates is less than 1%, indicating that the method has good precision, similar to the data shown in the previous studies. The results obtained in this study also have a small positive bias compared with the 1-cm cuvette results, similar to the 20 mg/mL results.

Table IV: Protein concentration results for 60 mg/mL purified protein with an 0.01-cm path length cuvette
(absorptivity of 1.38 mL mg-1 cm-1) compared with the accepted value obtained using a 1-cm path length cuvette.

Sample¹

Absorbance280 nm0.01 cm

Absorbance320 nm0.01cm

Concentrationmg/mL²for 0.01 cm

Average,standard deviation,%RSD³

Accepted⁴concentrationusing 1 cm(mg/mL)

%Accuracy⁵

1a

0.84307

0.00344

60.843

60.818
0.1166
0.2

57.4

106%

1b

0.83702

-0.00126

60.745

1c

0.84022

0.00027

60.866

2a

0.69401

-0.00053

50.291

49.951
0.3783
0.8

46.7

107%

2b

0.69028

-0.00121

50.020

2c

0.68370

-0.00033

49.543

3a

0.77884

0.00080

56.380

56.623
0.2632
0.5

53.4

106%

3b

0.78401

0.00310

56.588

3c

0.79184

0.00658

56.903

4a

0.79391

-0.00053

57.530

57.853
0.5023
0.9

53.9

107%

4b

0.79485

-0.00104

57.598

4c

0.80708

0.00072

58.432

5a

0.78717

-0.00327

57.041

57.036
0.4979
0.9

53.1

107%

5b

0.78018

-0.00190

56.535

5c

0.79392

-0.00109

57.530

6a

0.79559

-0.00712

57.651

57.905
0.2484
0.4

56.1

103%

6b

0.80244

-0.00329

58.148

6c

0.79925

-0.00400

57.917

¹Three independent measurements were made for each sample. For example, 1a, 1b, and 1c are three measurements for Sample 1. ²A320 values were subtracted. ³RSD is relative standard deviation.
⁴The accepted value was previously obtained using a product specific validated method for a 1 cm path length cell. ⁵Accuracy = (observed/accepted) x 100.

70 mg/mL protein concentration. The accuracy of the procedure for determining protein concentration of high protein samples using the shorter path length cells with no sample dilution was also tested using BSA. Two batches of BSA samples of approximately 70 mg/mL concentration were obtained from National Institute of Standards and Technology (NIST) (lot # SRM 927c and SRM 927d) and analyzed using the cuvette with a path length of 0.01 cm. The official protein concentrations of the NIST BSA samples are 72.112 mg/mL and 71.57, both ± 0.74 mg/mL, respectively. Each lot was analyzed in triplicate and the results are shown in Table V.

Table V: Protein concentration results for bovine serum albumin (BSA) (7% or 70 mg/mL) with a 0.01-cm path length cuvette (absorptivity of 0.667 mL mg-1 cm-1) compared with the accepted value. RSD is relative standard deviation.

Sample¹

Absorbance280 nm0.01 cm

Absorbance320 nm0.01 cm

Concentrationmg/mL²for 0.01 cm

Average, standard deviation,%RSD³

Accepted⁴concentrationmg/mL

%Accuracy⁵

927c _a

0.48508

0.00404

72.119

72.112
0.033
0.0

71.57
± 0.74

101 %

927c _b

0.48403

0.00328

72.077

927c _c

0.48518

0.00400

72.141

927d _a

0.48333

0.00786

71.285

71.316
0.070
0.1

70.10
± 0.74

102%

927d _b

0.48075

0.00541

71.266

927d _c

0.48250

0.00629

71.396

¹Samples are BSA and are National Institute of Standards and Technology Standard Reference Materials (NIST SRM). “927” is the designation for 7% BSA and “c” and “d” refer to specific lots.  Both lots were measured in triplicate. Sample name 927c_ a is the first measurement for lot 927c. ²A320 values were subtracted. ³RSD is relative standard deviation.
⁴The accepted value is the NIST “certified” value for SRM 927c and the “reference” value for 927d.  The 927d reference value is used for consistency with values historically reported by NIST per the NIST certificate of analyses for these materials. ⁵% Accuracy = (observed/accepted) x 100.

The % accuracy in protein concentration calculated for the 0.01-cm cuvette values from the NIST official protein values are 101% and 102%, respectively, indicating that the results for non-diluted sample procedure using the 0.01-cm cuvette is not significantly different from the actual value. The precision of the method with these samples is also good with an RSD value of less than 1%, indicating that even a single measurement is sufficient using the 0.01 cm cuvette procedure.

METHOD COMPARISON
The NanoDrop 8000 is one of a series of instruments that can be used to measure protein concentration without any sample dilution. In this instrument, a small volume of sample (1 or 2 µL) is applied to a pedestal in the optical path and the surface tension of the solution is relied upon to form a column between two optical surfaces. A sample cell is, therefore, not required. The NanoDrop 8000 has two available path lengths, 0.1 cm and 0.02 cm, and switches automatically between the two lengths depending on the concentration of the solution (other models are available with different path lengths.) For a protein with absorptivity of 1.38 mL mg-1 cm-1, the measurable range of protein concentrations is approximately 9 mg/mL to 45 mg/mL. Samples with two protein concentrations, 10 mg/mL BSA and 20 mg/mL purified recombinant protein, were used to compare the protein concentration values obtained from the spectrophotometer with 0.01- and 0.1-cm path length cuvettes and the protein concentration values obtained by the NanoDrop 8000. The results are shown in Table VI for the 10 mg/mL BSA results for both the undiluted sample method with a 0.1-cm path length cuvette and the NanoDrop 8000 method. The results obtained for the NanoDrop 8000 are shown in Table VII and the corresponding results for the 0.01-cm path length cuvette are shown in Table III. The data are also shown side-by-side in Table VIII.

Table VI:  Protein concentration results for 10 mg/mL bovine serum albumin (BSA) using an 0.1-cm path length cuvette (absorptivity of 0.667 mg^-1 cm^-1) and NanoDrop 8000 compared with the accepted value using a 1-cm path length cuvette.

10 mg/mL BSA¹

Absorbance280 nm0.1 cmcuvette

Absorbance320 nm0.1 cmcuvette

Concentrationmg/mL²0. 1 cmcuvette

Concentrationmg/mL²byNanoDrop

Acceptedconcentrationmg/mLfor 1 cm cuvettes³

1a

0.692320

0.004130

10.318

9.790

10.320

1b

0.692040

0.004036

10.315

9.865

10.445

1c

0.693060

0.004729

10.320

9.970

10.492

1d

0.692420

0.004344

10.316

10.120

10.569

1e

0.692940

0.004191

10.326

10.000

10.492

1f

0.693360

0.004887

10.322

9.955

10.473

average

10.319

9.950

10.465

standard deviation

0.004

0.114

0.082

% relative standard deviation

0.0

1.1

0.788

% Accuracy⁴

99%

95%

N/A⁵

¹Six independent measurements of a homogeneous sample were made for the 0.1 cm path length cuvette. ²A320 values were subtracted.  ³The accepted value was obtained using a 1 cm path length cuvette and six different dilutions within the absorbance range of the spectrophotometer. ⁴% Accuracy = (observed/accepted) x 100. ⁵The accuracy for the other two methods is calculated relative to the accepted value.

When the mean protein concentrations of the 10 mg/mL BSA samples were compared to the mean accepted value, accuracies of 99% and 95% were observed for the 0.1-cm path length cuvette and the NanoDrop 8000, respectively (Table VI). Although the data obtained by the NanoDrop 8000 method are statistically different (p<0.0005 by student’s t test as well as with unequal variances by Welch’s test), the percentage difference in the mean protein concentration value is less than 5% from the 0.1-cm path length procedure as well as the theoretical (accepted) values. The impact of this difference would need to be assessed before implementing the method in a manufacturing environment or in a quality-control analytical lab; otherwise, discrepancies between different process steps or different laboratories could result in out-of-specification results. In addition, the root cause of the bias would need to be investigated, understood, and either eliminated or accounted for to have a robust method that is accurate and reproducible when used in multiple laboratories. Potential sources for differences between the dilution methods and the short path length methods are presented in the following.

When the protein concentrations of the 20 mg/mL purified protein samples were compared to the accepted values, accuracies of 103% to 107% were observed for the 0.01-cm path length cuvette (Table III) and accuracies of 95% to 100% were observed for the NanoDrop 8000 (Table VII). The data are shown side-by-side in Table VIII for direct comparison. The protein concentration value obtained by the NanoDrop 8000 method is also statistically different when compared to the 0.1-cm path length cuvette method. As noted previously, the cause of the difference in the protein concentration would need to be understood and addressed to achieve a robust method for the intended purpose for manufacturing applications or quality-control labs. Interestingly, for the 0.01-cm cuvette with the 60 mg/mL sample, the accuracies compared to the accepted method were similar to those for the 20 mg/mL samples.

Table VII: Protein concentration results for 20 mg/mL purified protein obtained with the NanoDrop 8000
(absorptivity of 1.38 mL mg^-1 cm^-1) compared with the accepted value using a 1-cm path length cuvette.

Sample¹

Absorbance²280 nm

Absorbance³320 nm

Concentration mg/mL

Average,standard deviation% RSD⁴

Acceptedconcentrationmg/mLfor 1 cm⁵

%Accuracy⁶

1a

2.633

0.017

18.957

18.998
0.0906
0.5

19.0

100%

1b

2.609

-0.004

18.935

1c

2.566

-0.070

19.101

2a

2.616

-0.090

19.609

19.476
0.1341
0.7

20.3

96%

2b

2.671

0.002

19.341

2c

2.714

0.026

19.478

3a

2.666

0.021

19.167

19.070
0.0840
0.4

20.0

95%

3b

2.658

0.032

19.029

3c

2.653

0.029

19.014

4a

2.714

0.051

19.297

19.319
0.0261
0.1

19.7

98%

4b

2.679

0.009

19.348

4c

2.704

0.039

19.312

5a

2.673

0.051

19.000

19.036
0.0959
0.5

20.0

95%

5b

2.649

0.032

18.964

5c

2.666

0.024

19.145

6a

2.653

0.031

19.000

18.942
0.0664
0.4

20.0

95%

6b

2.649

0.033

18.957

6c

2.649

0.045

18.870

¹ Three independent measurements were made for each sample using the NanoDrop 8000.  For example, 1a, 1b, and 1c are three measurements for Sample 1.
²Absorbance values reported are normalized to a 0.1 cm path length. ³A320 values were all subtracted. ⁴ RSD is relative standard deviation.
⁵The accepted value was previously obtained using a product specific validated method for a 1 cm path length cell. ⁶Accuracy = (observed/accepted) x 100.

Table VIII: Comparison of protein concentration results obtained with an 0.01-cm path length cell and
with the NanoDrop 8000 for 20 mg/mL purified protein (absorptivity of 1.38 mL mg^-1 cm^-1).

Sample1

0.01 cm path length mg/mL

0.01 cm path length averagemg/mL

0.01 cm path length SD%RSD

NanoDrop mg/mL

NanoDropaverage

NanoDropSD%RSD

1

20.224

20.084

0.2386
1.2

18.957

18.998

0.0906
0.5

20.219

18.935

19.809

19.101

2

20.718

20.979

0.2387
1.1

19.609

19.476

0.1341
0.7

21.034

19.341

21.186

19.478

3

20.727

20.769

0.0410
0.2

19.167

19.070

0.0840
0.4

20.808

19.029

20.773

19.014

4

21.009

21.030

0.0191
0.1

19.297

19.319

0.0261
0.1

21.038

19.348

21.044

19.312

5

20.736

20.780

0.0653
0.3

19.000

19.036

0.0959
0.5

20.855

18.964

20.749

19.145

6

20.727

20.994

0.4054
1.9

19.000

18.942

0.0664
0.4

21.460

18.957

20.793

18.870

¹Six independent measurements were made for each method for each sample.

A difference in protein concentration results of up to 7% is observed between the short path lengths and the five dilution method using 1-cm cuvettes. Potential sources of errors can be from both methods and they include the dimensional tolerances of the shortest path length cells, loss of protein by adsorption to tube walls during sequential dilutions, viscosity effect on sample pipetting, and tolerances of pipettes. Application of an orthogonal method such as amino-acid analysis could provide additional information regarding the true protein concentration value and this may reduce the apparent variability of the measured protein concentration. Also, assessing cuvettes from different manufacturers for smaller dimensional tolerances in optical path length may reduce some error. In summary, it is reasonable to conclude that acceptable results can be obtained using the 0.01-cm, 0.1-cm, and 0.2-cm cuvettes with a standard spectrophotometer. Shorter path lengths are also available and could be assessed.

DISCUSSION
The data presented in this study demonstrate that it is possible to measure a wide range of protein concentrations without diluting the samples using standard spectrophotometric equipment that are already available and qualified for use in most biopharmaceuticals analytical laboratories. This represents a significant advantage over the implementation of new technology in terms of time and resources. An alternative to dilutions for protein concentration measurements is readily available for most laboratories by purchasing new accessories, the shorter path length cuvettes, and developing the operating parameters for the protein of interest.

As noted in the experiments described, operations related to the handling of protein solutions and the absorbance measurement should be specific. For example, excessive vortexing or vigorous pipetting of the solution may cause aggregation. This may result in high background at 320 nm, resulting in erroneously low concentration values. Upper limits for allowable background should, therefore, be specified in the operating procedure and would need to be specific for each of the path length cuvettes used in the analysis.

CONCLUSION
Data have been presented to demonstrate that protein concentration measurements by the A280 method using commercially available short path length cuvettes is a viable alternative to the traditionally widely used 1-cm path length cuvettes for determination of higher protein concentrations. Comparability to existing methods should be assessed where applicable and an understanding of the cause of any differences should be understood and addressed to ensure the method is robust. Use of the shorter path length cuvettes can be of significant advantage by eliminating the need for multiple sample dilutions and reducing sample preparation and analysis time. In addition, implementation of the method does not require qualification of new instrumentation given that UV/VIS spectrophotometers are already in use in most laboratories. The appropriate path length of the cuvette for a given application will need to be selected based on the concentration and the absorptivity of the target protein.

ACKNOWLEDGEMENT
The authors would like to thank Denny Maratea, Quality Science and Technology, Pfizer, Andover for comments on this article.

REFERENCES
1. A. Aitken and M.P. Learmonth, “Protein Determination by UV Absorption,” in The Protein Protocols Handbook, J.M. Walker, Ed. (Humana Press, a part of Springer Science+Business Media, New York, NY, 3rd ed., 2009), pp. 3-6.
2. Agilent Technologies, Accessories and Supplies for UV-Visible Spectroscopy, Publication Number 5988-5967EN (2005).  

About the AuthorLyn Watson, PhD, is a senior scientist, lyn.r.watson@pfizer.com, and Kannappan Veeraragavan, PhD, is associate director, both at Quality Science and Technology, Pfizer Global Supply, 1 Burtt Road, Andover, MA 01810 USA.