Assessment of the purification of a protein by ion exchange and gel permeation chromatography



A 3-week group laboratory project experiment that involves the partial purification of myoglobin (Mb) from bovine hamburger is described. The experiment compares alternate purification methods (gel filtration and ion exchange chromatography) as to both overall yield and enhancement in relative purity. The purity and molecular weight of purified Mb is established by SDS-PAGE. Group reports are directed toward having students put together all the information that is collected in the experiment. We discuss what we have learned and continue to learn from experiments done in the cooperative learning mode.

In recent years the content of biochemistry laboratories has shifted to a greater emphasis on molecular biology techniques, and several of the newer laboratory manuals have even reflected this change in their titles [1, 2, 5]. Though these techniques are extremely important for the training of a student, the emergence of proteomics as an important subdiscipline requires that students also understand the underlying principles of protein purification. In many biochemistry laboratory manuals [3, 4] gel filtration (molecular exclusion or gel permeation) chromatography and ion exchange chromatography are demonstrated with different proteins. In other laboratory manuals [1, 5, 6] only a single method is demonstrated. Boyer [7] uses two methods (gel filtration and affinity chromatography) to purify milk α-lactalbumin with an emphasis only on the A280 chromatography profile without the corresponding α-lactalbumin profile. Hardin et al. [2] use both forms of chromatography (in addition to affinity chromatography) to purify Caenorhabditis elegans β-galactosidase-myo-3 fusion protein. However the focus is only on those fractions that contain β-galactosidase activity, and protein measurements are not made for the entire separation scheme, i.e. the A280 profile. Farrell and Ranallo [8] do a sequential purification of lactate dehydrogenase with ion exchange, affinity, and gel filtration chromatography with an emphasis on finding the active lactate dehydrogenase fractions. We feel that for the students to gain a greater appreciation and complete understanding for protein purification, side-by-side comparisons of two methods on the same protein with both activity and protein measurements is warranted. We have accomplished this by extending the work of Bylkas and Andersson [9], which describes the redox properties of myoglobin (Mb). The two purifications methods are evaluated by comparing qualitative evidence (visual inspection of SDS-PAGE gels) to quantitative evidence (Mb and protein determinations).

This lab exercise, as indeed all lab exercises in our lab curriculum, involves cooperative learning. Students work in groups, learn to manage multiple tasks, and have joint responsibility for a group report.


At the beginning of this 3-week experiment a pre-lab discussion covers essentially all of the various aspects of the experiment. We use an in-house laboratory manual; like most published laboratory manuals it contains separate sections on spectroscopy, chromatography, and electrophoresis. In addition to requiring the students to review these sections, the students are reminded to review the protein purification and characterization sections found in their textbook.

Groups of two or three students perform the following set of experiments during three 4-h laboratory periods.

  • Period 1, isolation of crude Mb and investigation of its redox properties.

  • Period 2, partial purification of crude Mb by ion exchange and gel permeation chromatography.

  • Period 3, SDS-PAGE of crude and partially purified samples.

Period 1—

In the first period the initial isolation of Mb is accomplished by the method of Bylkas and Anderson [9] in which 10.0 g of freshly ground lean hamburger (preferably top round, with its lower fat content) is suspended in 20.0 ml of 20 mm, pH 5.6, potassium phosphate (KPi) 11 buffer. This suspension is centrifuged at 20.0 °C for 20 min at 10,000 × g (15,300 rpm with a Beckman JA-14 rotor). The supernatant is then filtered through glass wool, and the resulting filtrate is passed through a 0.45-μm syringe filter. Volume is measured to the nearest 0.1 ml, and total mg of protein content is determined by using the extinction coefficient at 595 nm generated from the reaction of bovine serum albumin with the Bradford reagent [10]. The Bradford reagent is prepared by dissolving 100 mg of Coomassie Brilliant Blue G in 100 ml of 95% ethanol, adding 100 ml of 85% (w/v) phosphoric acid, diluting to 1.0 liter with water, and filtering. The standard curve is prepared as shown in Table I.

The total mg of Mb is determined at 417 nm using an extinction coefficient of K = 7.57 ml·mg−1·cm−1 (from calculations based on Ref. 9). Normally the term specific activity (total units of enzymatic activity/mg of total protein) is used for following the purification of an enzyme. As Mb is not an enzyme, we have adopted the term “relative purity.” The relative purity (R.P.) of the crude Mb solution (and subsequent purified fractions) is defined as the total mg of Mb divided by the total mg of protein. During the remainder of the period, the students examine the redox properties of Mb with the reducing agent sodium dithionite and the oxidizing reagent potassium ferricyanide that has been extensively described by Bylkas and Anderson [9]. Basically this involves adding to separate 1.5-ml aliquots of the crude Mb supernatant a few small crystals of either sodium dithionite or potassium ferricyanide. The reduced Mb solution is then diluted 1 to 8 and scanned from 300 to 700 nm. This procedure differs from that of Bylkas and Anderson [9] in that the desalting step is eliminated as the reducing agent does not interfere with the subsequent spectra. As the potassium ferricyanide does interfere with the spectral scans, the oxidized Mb solution is subjected to a desalting step in which the 1.5-ml solution is added to a desalting column (10.0 ml of de-gassed Sigma Sephadex G-25 equilibrated with the 20 mm, pH 5.6, KPi buffer in a 12-cm × 1.5-cm inner diameter Bio-Rad Econo-Pac column). A 0.5-ml aliquot of the leading yellowish-colored Mb band is collected, diluted 1 to 6 with the KPi buffer, and scanned from 300 to 700 nm. Leftover crude Mb solutions are labeled and frozen until the next laboratory period.

Period 2—

Frozen crude Mb solutions are thawed, and 5.0 ml is filtered through a 0.45-μm syringe filter; 2.0 ml is to be used in the chromatography separations, with the remainder frozen for use in Period 3. One group member further equilibrates a pre-poured, pre-equilibrated carboxymethylcellulose (CMC) column (10.0 ml of de-fined Sigma CMC in a 12-cm × 1.5-cm inner diameter Bio-Rad Econo-Pac column) with a 20 mm, pH 5.6, KPi buffer. During this equilibration step, the number of drops to collect 2.0 ml is determined. A second group member equilibrates a pre-poured, de-gassed, pre-equilibrated Sephadex G-75 column (∼38.0 ml of degassed Sigma G-75–120 in a Bio-Rad 50.0-cm × 1.0-cm inner diameter Econo-Column to a height of ∼48.0 cm) with a 20 mm KPi, 0.10 m KCl, pH 5.6, buffer. The head height of this buffer is kept constant at 2.0 cm above the G-75 column material so that the flow rate can be determined. In addition, the number of drops to collect 1.0 ml is determined.

Crude Mb (1.0 ml) is pipetted onto the CMC and allowed to enter the CMC matrix. One column volume (∼10.0 ml) of the 20 mm, pH 5.6, KPi buffer is then added to the CMC column, and 2.0 ml fractions are collected. After all of the KPi buffer has entered the column, a 20 mm Tris, pH 7.5, buffer is added, and 2.0 ml fractions are collected until all of the red color (Mb) moving through the column has been eluted. To each collected fraction 2.0 ml of distilled or molecular biology grade water is added. Absorbance at 417 and 280 nm is measured for each fraction; three consecutive fractions with the highest A417/A280 ratio are pooled, and the combined volume for these pooled fractions (labeled CMC fraction) is determined. The total Mb and total protein in the pooled CMC fraction is determined as described above. In addition to determining the R.P., a double Y plot of A417 and A280versus fraction number is constructed. The pooled CMC fraction is frozen for the following period.

After determining the flow rate (typically 0.3–0.7 ml/min), 1.0 ml of the crude Mb is pipetted onto the top of the G-75 column and allowed to enter the column matrix; 20 mm KPi, 0.10 m KCl, pH 5.6, buffer is added to the column to a level 2.0 cm above the gel matrix. While maintaining a constant buffer head height, 1.0-ml fractions are collected until all of the red color (Mb) has been eluted. To each collected fraction 1.5 ml of distilled or molecular biology grade water is added. Absorbance at 417 and 280 nm is measured for each fraction, and the three consecutive fractions with the highest A417/A280 ratio are pooled, and the combined volume for these pooled fractions (labeled G-75 fraction) is determined. The total Mb and total protein in the pooled G-75 fraction is determined as described above. In addition to determining the R.P., a double Y plot of A417 and A280versus fraction number is constructed. The pooled G-75 fraction is frozen for the following period.

Period 3—

Calculations for the crude Mb, the CMC fraction, and the G-75 fractions are performed to determine the volume of sample needed to give 10 μg of total protein (without exceeding a volume of 25 μl). Typically, the crude Mb fraction has to be diluted 1:10 (with ∼ 3.3 μg added to the SDS-PAGE gel), and the CMC and G-75 fractions are used without dilution as 10 μg cannot be obtained, and the actual amount of protein in the 25 μl is determined. All samples are made to 25 μl and are added to 1.5-ml microfuge tubes; 50 μl of Bio-Rad Laemmli sample dilutor (containing the tracking dye bromphenol blue) is added to each sample. A Bio-Rad low MW SDS standard containing six proteins, a pure horse heart myoglobin standard, and a water blank are also prepared in the same manner. The blank, standards, and samples are then placed in a boiling water bath for 5 min. Bio-Rad Ready Gels (10 well 4–15% T gradient SDS gels) and the electrophoresis apparatus are prepared for use. Use of pre-cast Bio-Rad gels prevents exposure of the students to acrylamide, which in the unpolymerized form, is a neurotoxin. Blank, crude Mb fraction, SDS MW standard, and Mb standard (25 μl each) and CMC and G-75 fractions (45 μl each) are added to the gel; two groups can place their samples on one gel with one set of standards. The electrophoresis is completed in ∼30 min at a constant 200 V. After electrophoresis, the gel is removed and stained in Coomassie Blue G-250 (0.04% (w/v) and 3.5% (w/v) perchloric acid) for 30 min and further de-stained overnight in 5.0% (v/v) acetic acid. Appropriate measurements of the protein bands on pictures of the gels captured with a Photodyne Visionary photodocumentation system are taken, and a log MW versus relative mobility (Rm) standard curve is constructed.


Students are required to write a group laboratory report in the style of a biochemistry journal article. After comparing and discussing the spectra (data not shown) for the OxyMb and the MetMb, the remainder of the report is dedicated to determining which of the two purification methods yields the better separation. Students are asked to make a judgment based upon the two double y plots that are constructed from their data. Typically, the students report that the carboxymethylcellulose ion exchange chromatography (Fig. 1) provides a better separation than the Sephadex G-75 molecular exclusion chromatography (Fig. 2). Their rationale appears to be a straightforward interpretation of the two graphs. The CMC plot shows two widely separated protein peaks (A280) with an Mb peak (A417) under the second protein peak. The Sephadex G-75 plot, on the other hand, shows two protein peaks that are not completely separated with the Mb peak under the second protein peak. Students apparently interpret the complete separation of protein peaks as the better separation method. The more insightful and correct answer would of course be that this evidence by itself is not sufficient to determine which is the better method.

Students also are asked to evaluate the two chromatography methods based on a results table that includes protein concentration, Mb concentration, R.P., and percent recovery for the three fractions, crude Mb, CMC fraction, and G-75 fraction. A results table from a recent semester is given in Table II.

Students should see these results conflicting with their naïve predictions based on the double y plots; the Sephadex G-75 column gives a relative purity value approximately twice that of the CMC column. The students also have evidence that suggests that the CMC separation yields a somewhat better recovery than the Sephadex G-75 column. How can this conflict be resolved, and which purification method is better?

The third period provides the means for answering this question. Fig. 3 shows a picture of an SDS-PAGE gel obtained from this experiment. This gel clearly shows that there are fewer bands in the G-75 fractions than the CMC fractions. Thus, the student has visual proof that correlates to the relative purity results. A standard curve of relative mobilities (Rm) versus log of known MW standards is shown in Fig. 4. From the gel it is very easy for the student to determine which band in the crude, CMC, and G-75 fractions corresponds to the Mb band. From the results generated by the six groups in Table II, the average molecular mass was determined to be 17.0 kDa (± 0.1 kDa S.E.) with a range of 16.3 to 18.0 kDa. This compares favorably with the reported molecular mass of 16.9 kDa for bovine Mb [11].


As there may be limitations to available equipment it should be noted that we have performed this experiment successfully for several years without the use of fraction collectors. One student can easily perform the CMC separation whereas two students are required to perform the Sephadex G-75 separation, one to count drops and collect fractions, the other to maintain the constant buffer head height. If one wanted to extend this experiment, different parameters could be altered to illustrate that resolution is affected by other factors. We believe that the variation of flow rate (0.3–0.7 ml/min) from one group to another is responsible for the greater variance of results for the Sephadex G-75 separation compared with the CMC separation.

The CMC chromatography also lends itself to modifications to illustrate the various aspects of ion exchange chromatography. Instead of eluting Mb with a change in pH to its isoelectric point, one could perform stepwise salt gradients (or a linear salt gradient if one has the equipment) while maintaining the same pH, or one could perform pH gradient elutions.

We are able to perform the separations as described in a 4-h period; however students must be ready and organized to accomplish all the tasks involved. We are fortunate that each group has access to a Genesys 2 or 5 UV-visible spectrophotometer thus preventing delays in reading the samples. Having a fraction collector for use in the G-75 separation is extremely helpful as it allows the third member of the group to start adding the water to the fractions (both G-75 and CMC) and to begin reading the absorbances.

This experiment has evolved over a number of years to its current configuration. The present experiment incorporates elements of three previous experiments, a gel filtration experiment with different colored biological polymers (blue dextran, hemoglobin, and cytochrome c) and the dye bromphenol blue, an ion exchange experiment separating the proteins alkaline phosphatase and cytochrome c, and the study of the properties of Mb [9].

In developing this experiment, the CMC cation exchange column was chosen, because Mb (pI = 7.0 [11]) at pH 5.6 is positively charged and therefore binds to the column matrix. As the CMC is relatively inexpensive we have not attempted to recycle it after the experiment is completed. Sephadex G-75–120 and G-100–120 (both with 40–120-μm dry bead diameters) were evaluated in the developmental stages of this experiment. Though both have acceptable fractionation ranges, 3–80 kDa for Sephadex G-75 and 4–100 kDa for Sephadex G-100 [12], we found that the Sephadex G-75 gave an overall better separation with reasonable flow rates for completing the experiment in the allotted time. After the experiment is completed the Sephadex G-75 is thoroughly washed with an aqueous 0.1% (w/v) sodium azide solution (to prevent bacterial growth) and stored at 4 °C until needed.

With respect to safety issues we have not encountered any majors concerns because of the relatively innocuous aqueous conditions of this experiment. The only minor concerns we have noted is adequate ventilation to deal with the stench of the 2-mercatoethanol in the SDS sample dilutor and a warning to the students not to get the Coomassie reagents (Bradford and SDS staining solutions) on their skin.

It was our intention in revising our biochemistry lab curriculum to go toward a group project, cooperative learning setting [1316] that we believed allows students to gain a clearer understanding of the underpinning of each experiment, in this case the various aspects of protein purification. In addition, we have been concerned over the years about how uncritically students evaluated the information that is collected in the lab. Though we have not formally evaluated student reaction to student collaboration as others have [17], we believe that overall the collaborative effort is a positive experience. We detail below some of our findings over the last four years.

The three big benefits from group work that we have noticed are as follows: 1) students are more engaged in what they are doing; 2) students are more resourceful in terms of managing their time; 3) we are able to accomplish much more experimentally in the lab.

There is noticeably more noise in the lab with group work; students are talking about what to do and why it is being done instead of silently reading directions and going through the motions. After some rocky starts students make noticeable gains in time management and efficiency. We deliberately create enough tasks such that only by properly coordinating and timing do students finish early, and we ramp up the number of tasks as the semester progresses. Though there are more sets of hands working on the same effort, we have found that those hands are doing more tasks; more pipetting, more spectral measurements, more loading of samples. The major role of the instructor is to make certain that each student is involved in all tasks in each experiment. We have not rigorously evaluated whether students learn more or better. We do know that we present students with more to learn.

On the negative side, a major problem every semester is student responsibility vis a vis his/her group. There is on occasion significant resentment of the student who does not pull his/her weight in the group, especially in terms of working on joint reports. Also the student(s) doing most of the work believe that his/her/their larger contribution is not recognized by the instructor. On the flip side the standard student excuse for not being involved in the writing is not being able to make connections with one's group. The quality of group reports is thus very spotty and “committee” efforts are sometimes disastrous with one report section having no connection to another. We have attempted different models for constructing groups but have not come up with one that is noticeably better than another; each have their unique deficiencies. We do have a self- and peer evaluation at the end of the semester that is coupled with an instructor evaluation (10% of total lab grade) that has had some effect in forcing students to be more conscientious in their group work. Additionally, the self- and peer evaluations allow the students to gain some experience that will be useful when they are in a supervisory position later in their careers.

Figure FIGURE 1..

Carboxymethylcellulose chromatography of crude myoglobin.

Figure FIGURE 2..

Sephadex G-75 chromatography of crude myoglobin.

Figure FIGURE 3..

Bio-Rad SDS 4–15% gradient Ready Gel.Lanes 1 and 10, blank; lanes 2 and 7, crude Mb fractions; lanes 3 and 8, CMC fractions; lanes 4 and 9, Sephadex G-75 fractions; lane 5, Bio-Rad low MW SDS standards; lane 6, horse heart Mb standard.

Figure FIGURE 4..

4–15% SDS-PAGE standard curve.

Table 1. Preparation of a Bradford standard curve
1.0-mg/ml bovine serum albumin standard0.0000.0100.0250.0400.050
20 mm, pH 5.6, KPi buffer0.1000.0900.0750.0600.050
Bradford reagent3.903.903.903.903.90
Table 2. Purification of myoglobin by carboxymethyl cellulose and Sephadex G-75
GroupCrude proteinCrude MbR.P.CMC proteinCMC MbR.P.% RecoveryG-75 proteinG-75 MbR.P.% Recovery
  • a

    S.E., standard error of the mean; n = 6.

 mg/mlmg/ml mg/mlmg/ml  mg/mlmg/ml  


  1. 1

    The abbreviations used are: KPi, potassium phosphate; R.P., relative purity; CMC, carboxymethylcellulose; MW, molecular weight.