156
N. R. Vaidya et al.
Narendra R. Vaidya Balwant P. Gothoskar Ashutosh P. Banerji Biological Chemistry Division, Cancer Parel, Bombay Research
Electrophoresis 1990,11, 156-161
Column isoelectric focusing in natural pH gradients generated by biological buffers Using 2 or 3 simple Good zwitterionic buffers at a 16 or 18 mmol/L final column concentration of the mixture, natural p H gradients of 4 to 8 and 3 to 9.5, respectively, were generated in a liquid LKB column. The pH gradients, stabilized by an anticonvective sucrose gradient, were linear, reproducible and stable in the electric field up to 5h. The p H gradients were used for isoelectric focusing of a number of impure proteins such as human hemoglobin, bovine serum albumin and chicken egg white lysozyme. The protein components could be well separated in the gradient, wereeasily recovered and appeared to be quite pure when analyzed by sodium dodecyl sulfategel electrophoresis. Furthermore, the p H gradient 4-8 was effectively used to isolate one of the acidic isozyme (PI 5.6) components of mouse liver alcohol dehydrogenase (EC 1.1.1.1) in an enzymatically active state, suggesting that the procedure does not denature proteins. The low cost, the ease with which the pH gradients are formed, their linearity, stability for a sufficient period to allow proteins to reach equilibrium and their subsequent recovery from buffer eluates should make the procedure interesting for electrofocusing of proteins.
1 Introduction Isoelectric focusing (IEF) of proteins, using wide or narrow ranges of p H gradients, generated by mixtures of synthetic carrier ampholytes (SCAMS), has come to be accepted as a standard procedure to characterize or separate proteins from their mixtures. Carried out largely on polyacrylamide gel systems, these carrier ampholytes, which generate a linear and stable pH gradient, bring about efficient separation of proteins. However, the cost of the carrier ampholytes, the time involved and the efforts needed to separate polypeptides from the carrier ampholytes put severe restrictions on their use. Earlier, mixtures of amino acids [ 1I, or nonamphoteric buffers were used toobtain pH gradientsin anelectric field which were at best quasi-stable. The credit for putting I E F with buffer mixtures or buffer electrofocusing (BEF) on a stable footing goes to Chrambach's group 121. Nguyen and Chrambach described methods for producing natural p H gradients from 10 nonamphoteric and amphoteric buffers 131 but the gradients suffered from nonlinearity and/or narrowness of range. Availability of amphoteric buffers with wide ranges of pK values, and their zwitterionic properties of migration in either direction in an electric field, upheld the belief that for effective BEF, pH gradients can be tailor-made when using these buffers. T o suit one's need, one could thus scan their pK values and then add or substract constituents in the mixture [ 2 ] . Such pH gradients, made from 9-14 amphoteric/nonamphoteric buffers in the range of 3-7 on polyacrylamide gels [41 and 4-8 on Sephadex gels I51 or narrow acidic range on Sephadex gels [61, were found to be useful. However, the most successful effort in establishing a wide p H gradient of 3-10 on polyacrylamide slab gels, stable for 18h, needed 100 trials as Correspondence:Dr. N. R. Vaidya, Biological Chernistry Division, Cancer Research Institute, Parel, Bombay - 400 012, India
well as 47 buffers in the final analysis [71. Even computational approaches to simulation of pH gradients have not fulfilled the promise of BEF as an inexpensive and practical protein separation tool [21. With this background we present here a simple liquid column BEF technique using 2-3 amphoteric buffers which could be useful for analytical as well as preparative separation of proteins.
2 Materials and methods 2.1 Chemicals and proteins Unless specifically mentioned, all the chemicals were from Sigma (St. Louis, MO). The following proteins (BDH, Poole, Dorset, England) were used: crystalline bovine serum albumin (BSA), human hemoglobin (Hb) type IV, chicken egg ovalbumin and egg white lysozyme. 2.2 Buffers The following Good buffers were used for developing the pH gradients: 2 (N-morpho1ino)ethanesulfonic acid (MES) with a pK value of 6.1 at 25 "C and useful p H range of 5.5 to 6.7, N-2hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES) with pK 7.5 at 25 OC and useful pH range of 6.8 to 8.2, N,Nbis(2-hydroxyethyl/glycine (BICINE) with pK 8.3 at 25 "C and useful p H range of 7.6 to 9.0, 3-(cyclohexylamino)-1propanesulfonic acid (CAPS) with pK 10.4 at 25 OC and useful p H range of 9.7 to 11.1.
2.3 Buffer mixtures To generate a gradient with a pH range of 4-8, I E F was car-
ried out with a mixture of HEPES and BICINE buffers, Abbreviations: Bicine, (N,N-bis[2-hydroxyethyllglycine);BEF, buffer whereas for obtaining p H gradient in the range of 3-9.5, the electrofocusing; BSA, bovine serum albumin; %C, relative percentage of cross linker in a polyacrylamide gel; CAPS, (3-lcyclohexylamino~-1 -pro- mixture comprised CAPS, BICINE and MES buffers. panesulfonic acid): Hb, hemoglobin; HEPES, (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonicacid); IEF, isoelectric focusing; kDa, kilodalton; 2.4 IEF apparatus MES, (2"-morpholinol ethanesulfonic acid); ML-ADH, mouse liver alcohol dehydrogenase; SDS-PAGE, sodium dodecyl sulfate-polyacrylI E F was carried out in a 110 mL-LKB 8101 column (LKB, amide gel electrophoresis; %T, total monomer concentrations in polyacrylamide gel Bromma, Sweden). Accessory equipment comprised a D C 0VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990
0173-0835/90/0202-0156 %2.50/0
Buffer isoelectric focusing
Eleclrophoresis 1990,lI. 156-161
157
power supply, a gradient mixer and a peristaltic pump from LKB.
2.9 Purification of mouse liver alcohol dehydrogenase (ML-ADH) by salt fractionation and IEF
2.5 Procedure for preparative IEF of proteins
Livers (5 gm) of ICRC mice (20-24 g body wt.) were homogenized with cold 5 m~-Tris-HC1-0.5mM NAD-0.7 mM dithiothreitol (DTT)-0.25 M sucrose buffer, pH 8.5, in accordance with the procedure of Hasabe et al. [ 111. The clear supernatant obtained at 40 000 x g for lh at 2 "C was subjected to ammonium sulfate fractionation between 35-75 % saturation as suggested by Dalvi [ 121. The precipitate obtained by centrifugation at 15 000 rpm was dialyzed against 5 mmol/L HEPES buffer. The dialyzate (about 5mg protein) was subjected to IEF in pH 4-8 gradient as above. All the operations were carried out at 4 "C. The crude liver extract, the dialyzate of 35-75 % ammonium sulfate precipitate and the protein peak focused at p15.6 were assessed for ML-ADH activity and specific activity in accordance with the method of Dalvi [ 121. The protein eluted from the column after dialysis and concentration by vacuum dialysis in cold, was measured by the method of Lowry et al. (131.
The procedure followed was that described in the LKB IEF instruction and operation manual. The catholyte was 0.4 % ethanolamine (Merck, Darmstadt, FRG) in 60 % sucrose (BDH) solution at the bottom of the column while the anolyte was 0.1 M NaCl solution adjusted to pH 2.7 with HCl in accordance with the procedure of Rao and Duraiswami IS]. For focusing proteins in a pH 4-8 gradient, stabilized with alinear 20-50 % sucrose density gradient, 3-4 mg of protein (Hb, BSA, etc.) dissolved in 16 mmol/L BICINE was added to 50 mL of the same BICINE solution, containing 50 % sucrose in one vessel of the gradient mixer and an equal volume of 20 % sucrose in 16 mmol/L HEPES solution in the other vessel. The concentration of each buffer in thecolumn was 8 mmol/L. For BEF of a basic protein like chicken egg white lysozyme, in pH 3.0 to 9.5 gradient, the solutions in the gradient mixer were 50 % sucrose, containing 16 mmol/L CAPS and 20 % sucrose with 16 mmol/L BICINE plus 4 mmol/L MES. The column concentrations of each buffer were8,8 and 2 mmol/L, respectively. All solutions were prepared in freshly deionized water with an electrical conductivity less than 0.06 pS/cm at 25 "C. Focusing was initiated with a current of 4.8 mA. The electrical potential applied to the column was then gradually raised to 1100 V, so that the current did not exceed 4.8 mA with a power of 3 W. Later, the current dropped to a constant minimum of less than 0.4 mA and the proteins focused as sharp zones. The power was then switched off and the column eluted (1.5 mL/fraction), with aflow rate of 2 mL/min. The experiments were conducted at 4 "C.
2.6 pH Measurement Immediately following collection, the pH of each fraction was determined on Pye Radiometer Model pH meter.
2.7 Protein measurement The absorbance at 280 nm (A 280 nm) of each fraction was recorded and plotted against fraction number. For human Hb, A560 nm was also determined for the fractions.
2.8 Polyacrylamide slab gel electrophoresis The fractions corresponding to high A280 nm values were pooled and dialyzed in the cold against distilled water. Following lyophilization, the powder was dissolved in the sample buffer (2.3 % SDS, 0.0625 M Tris-HC1, pH 6.8, 11 % glycerol- 1 % 0-mercaptoethanol) and heated at 100 "C for 3 min, and then subjected to SDS-polyacrylamide slab gel electrophoresis (SDS-PAGE) by the method of Laemmli 191. The gels contained 15 % acrylamide and 0.04 % N,N'-methylenebisacrylamide (Bis). The proteins (50-180 pg) were loaded onto 5 % T, 2.6 % C stacking gels and electrophoresed in an 0.025 M Tris, 0.19 M glycine buffer, pH 8.3, containing 0.1 % SDS. Bromophenol Blue was the tracking dye. Next, the gels were fixed in 40 % methanol in 10 % acetic acid solvent mix and stained with0.25 % Coomassie Brilliant Blue R-250in the above solvent, followed by destaining. For silver staining, the gels were fixed and stained by the procedure described by Blum et al. 1101.
3 Results Natural pH gradients can be formed from simple and chemically defined biological Good buffers such as BICINE, HEPES, CAPS and MES [ 141. Preliminary experiments to establish an effective broad range of pH gradient carried out with final concentrations of 4-7 mmol/L of each of the 2 buffers in the column, such as BICINE and HEPES, proved unsatisfactory because the pH gradient was not sufficiently broad (pH 3-6)and also nonuniform. Raising the final column concentration of each buffer to 8 mmol/L brought about a satisfactory linear pH gradient within 90 min of electrofocusing. 3.1 Time course ofthe developmentoflinear pH 4-8 gradient with a BICINE-HEPES buffer mixture BICINE and HEPES at a concentration of 8 mmol/L each were used. After initiation offocusing, the current begins to fall sharply with time so that by 150 min the current drops to less than 0.4 mA with voltage stabilized at 1100 V (Fig. IA). Before application of the electric field, i. e. at 0 time, the pH gradient is essentially flat as the pH is almost uniform (ca. 7.2) along the column (Fig. 1B). At 30 min, the pH gradient starts to manifest itself and with the drop in current at 50 min, it tends to be linear until such a state is attained at 90 min. The conductance, which drops sharply with time, later reaches a constant negligible value with the establishment of the pH gradient (Fig. 1A inset). The residual constant conductance implied that buffers had attained their p1 positions.
3.2 Stability and reproducibilityof the pH gradient and the effect ofthe nature of anolyte solutionon the formationof the gradient This natural pH gradient developed in the electric field was stable up to 5h (Fig. lB, C) at a potential of 1100 V beyond which time the gradient tended to be unstable and later deteriorated. The linearity and stability were reproducible for over several experiments using these buffer mixtures (Figs. 1C and 2). It remained to be seen whether the gradient was stable
158
N. R. Vaidya et al,
Electrophoresis 1990,11, 156-161
5
I
I
4
E I
25;
"
w
z
Q
w .Y
5
"
.Y
>
3
"\
3.3 Resolution ofcrude proteinmixtures by IEF in a pH 4-8 gradient generated by the HEPES-BICINEbuffer system
B 0 Min.
long enough to allow proteins to reach equilibrium and thus enable their effective separation 151. Moreover, as observed earlier in column BEF using 2 amphoteric buffers [Sl, we also noted that the choice ofthe right anolyte, rather than catholyte solution, had significant influence on the nature and stability of the pH gradient. When 0.5 % H,SO, was used as an anolyte, the gradient deteriorated, possibly due to the generation of protons by electrolysis of this solution with resultant drop in the pH along the anodic end (Fig. 1C) and disturbance in the gradient 181. However, use of acidified 0.1 M NaCI, pH 2.7, produced no acidic drag at the anode. It appears that electrolysis of NaCl causes an Na' flux into the column, neutralizing the effect of protons on pH gradient formation during focusing as the ratio of sodium ions to protons is nearly 1 at pH 2.7 1161.
30 Min
The ability ofthe pH gradient produced by the buffer system to satisfactorily resolve crude protein mixtures was then assessed. Human Hb type IV and crystalline BSA were subjected to IEF by the procedure described above. By about 2.5 h, when H b focused as a fine sharp colored zone or BSA as a white ring in the column, the current was switched off and elution initiated. The H b protein peak (fraction No. 43-44, Fig. 2), focused at pH 6.6 (p1 of Hb), and the peak component, eluting at pH 4.8 (fraction No. 67-70), which corresponds to the plvalue of BSA [ 171, were collected. They were subjected to SDS-PAGE analysis, after dialysis, lyophilization etc. to assess their purity and appeared to be homogeneous (Fig. 3a, 3b). Thus, the procedure of IEF detailed above generates a reproducible, linear and stable pH gradient adequate for an analysis of proteins.
50 Min.
3.4 Purificationof partiallypurified ML-ADH by IEF with the HEPES-BICINE buffer system
12
24
36
48
60
72
12
0
FRACTION
C
36
24
48
60
72
NO
BEF TIME ( M i n ) o
150 - o 5 % 150
0 200
I
I
,
20
40 FRICTION
10
IM
+so4 NoCl p H
AS ANOLYTE 2 7 AS
ANOLYTE
1 __--I_
60 NO
80
This was carried out to achieve the following objectives: (i) to see whether the procedure evolved can improve the degree of purification after its extraction from liver tissue and subsequent salt fractionation, (ii) to isolate the major (PI5.3-6.4) isozyme of ML-ADH [ 1 I], (iii) to ascertain that the procedure does not denature proteins during the process offocusing. IEF of the 35-75 % ammonium sulfate precipitate, containing appreciable alcohol dehydrogenase activity [ 121 of mouse liver extract (Table l), could be resolved into several protein peaks (Fig. 2). One major peak (highest A 280 nm value) focused at p15.6 (fraction No. 61,62), the known p1range of the major isozyme component of ML-ADH [ l 11. This peak showed significant ML-ADH activity and also migrated as a single band on SDS-PAGE (Fig. 3c). The purification achieved was about 25-fold (Table 1). 4Figur-e 1. (A) Time course of change in current with stabilization of voltage during establishment ofpH 4-8 gradient, using theHEPES-BICINE buffer mixture, stabilized by a sucrose gradient in LKB-I 10 glass column. The final column concentrations of BICINE and HEPES were 8 mM each. Insert shows the plot of conductance against time. Anolyte: 0.1 M NaCI, pH 2.7. Catholyte: 0.4 % ethanolamine. Duration of IEF was 3.5 h. (B) Generation of alinear pH 4-8 gradient with the HEPES-BICINE buffer mixtureas afunctionoftime. (C)ThestabilityofthepHgradientwith0.5 % H,SO, and 0.1 M NaC1, pH 2.7 as anolytes for BEF with the BICINEHEPES mixture. Note the proton effect in the pH gradient at the anodic end with 0.5 % H,SO, as the electrolyte.
Electrophoresis 1990, 11, 156- 16 1
4
BOVINE SERUM ALBUYIN
-
1.00
;
0.15
159
Buffer isoelectric focusing
12.0
. :
n 0.50
I
E
0.25
P
FRACTION NO
FRACTION NO
ALCOHOL DEMYDROGENASE L Y S O Z Y YE
'.
12.0
10.0
r-5
5.0
IL "
0
20
60
40 FRICTION
NO.
z
2.5
I
80 20
0
j -
I
I
1
40
60
110
F R A C T I O N NO.
Figuve2. Absorptioncurves at 280 nmofeffluentfractionscollectedfromLKB-I 1Ocolumn,superimposedonpHvaluesoffractions(at 27 T ) a f t e r I E F o f human hemoglobin type IV, bovine serum albumin, partially purified mouse liver alcohol dehydrogenase in a p H 4-8 gradient, generated by the HEPESBICINE buffer system, egg whitelysozymein a p H 3-9.5 gradient,generatedbyusing theBICINE-MES-CAPS buffer mixture. For Hb, A 560 nm was also measured for the fractions. Electrolytes were as in Fig. 1 A.
Table 1. Purification of ML-ADH Fraction
Liver extract 35-75 % saturated (NH,)SO, IEF
Total Protein Activitya) Specific Purification volume mg/mL nmoles/mL activity factor mL min nmoles min/mg protein 12 3 2
31.7 40 0.45
200 616 71.1
6.3 15.4
1 2.4
158.0
25.1
a) Unit of enzyme activity was defined as the amount of enzyme that catalyzes formation of 1000 nmoles of NADH/min at 340 nm.
4 Figure 3. SDS-PAGE patterns of (a) Human hemoglobin (lane 1) and that of the protein (lane 2) recovered from the peak focused at p16.6 (Fig. 2), (b) bovine serum albumin (lane 3) and ofprotein (lane 4) electrofocused at p l 4.8 (Fig. 2), (c) M, markers ovalbumin (42 000, lane 5) and hemoglobin (16 000,lane6),protein(lane7)eluting atpI5.6(Fig.2)afterIEFofpartially purified mouse liver alcohol dehydrogenase. IEF was carried out in a4-8 p H gradient, (d) egg white lysozyme (lane 8) and protein peak focused at pZ 9.5 (lane 9) in pH gradient 3-9.5. Slab gels (a), (b), (d) were stained with Coomassie Brilliant Blue R-250, while gel (c) was stainedwith silver. Arrow indicates origin.
160
N. R. Vaidya el al.
3.5 Establishmentofa broadrange pH 3-9.5 gradientusing a mixture of 3 buffers MES, BICINE and CAPS, at different concentrations, were subjected to IEF. The final concentrations of the buffers in the column to obtain a linear pH 3-9.5 gradient (Fig. 2) was found to be 2 mmol/L of MES and 8 mmol/L, each of CAPS and BICINE. When a protein such as egg whitelysozyme with a p l value greater than 11 was focused in this gradient, the crude mixture was separated into 6 protein peaks. Since the ultimate limit of p H gradient formed by this buffer system is pH 9.5, the protein peak (fraction No. 19-20), eluting first at the plvalue of pH 9.5, was analyzed by SDS-gel electrophoresis. It appeared to be a homogeneous protein (Fig. 3d).
Electrophoresis 1990.11, 156-161
(Figs. lB, l C , 2). The conductance, approaching anear-zero value after 2.5 h during focusing of crude protein mixtures, implied that the carrier and protein constituents had reached their equilibrium pZ positions by this time. Further, the constancy of elution patterns [ 181of Hb, BSA andlysozyme(data not shown) on extension of their I E F time from 2.5-5h gave credence to steady-state protein IEF in the gradients generated by the buffer systems. Thus, though the stability of the pH gradients that can be formed by our systems were not quite comparable to those reported by the buffer system of Cuono and Chapo 171 or in systems using SCAMS, they were nonetheless quite efficient for the separation of the proteins studied.
Theoretically, the success of steady-state IEF of proteins is determined by the properties of the carrier ampholytes such as 3.6 Stability of pH gradients suficient to allow proteins in their high buffering capacity and conductivity at their pZ[2,7, a mixture to reach equilibrium state during IEF 191. Though the buffering capacities of mixtures of the two buffer systems used in this study were not assessed, it may be After I E F is initiated, there is a sharp dlrop in conductance expected from their low M , and known relationship between reaching a negligible value with the establishment of the pH buffering capacity and conductance [ 2 11 to be less than that of gradients by 90 min (Fig. lA, inset). The conductance con- SCAMS [ 71. The conductivity of buffer mixtures at low molar tinues to be uniformly low until the proteins focus as zones concentrations has already been reported to be l/lOth that after 2.5 h. This implies that after the initial attainment of reported for SCAMS (71. In spite of these adverse points, the isoelectric position by the carrier constituents, the proteins separation of protein mixtures by the method presented here gradually attain their pzpoints after approximately 3h. Exten- appears quite satisfactory as indicated by homogeneity on sionofIEF timeofHb,BSA,orlysozymefrom2Shto5hgave SDS-PAGE analysis of proteins separated by IEF. This is superimposable protein elution profiles of the respective pro- because buffering capacity and higher conductivity of carrier teins with the pZ values of the proteins remaining the same constituents are not the critical factors - not as in preparative (data not shown). This constancy of pattern 1181 and the IEF - for the analytical resolution of proteins [7, 21, 221. above observation indicate that the two p H gradients, narrow or wide, generated in the liquid column were sufficiently stable An argument which is often preferred to explain why BEF has so that the proteins reached a steady state during IEF. not been the most widely used procedure for protein analysis, unlike I E F with SCAMS, is the nonavailability of a wide range of amphoteric buffers with small pl-pK difference and therefore their inability to buffer protein ampholytes at the 4 Discussion isoelectric state 12, 19, 231. However, in practice it appears Since the advent of SCAMS, which can formlinear, stable and that neither the pl-pK values nor the total number of such bufwide pH gradients, IEF using amphoteric: buffers has taken a fers used essentially determine the buffering capacity or back seat mainly due to the nonavailabiliity of buffers with pZ stability of the pH gradient. This can be inferred from the fact covering the entire pH scale [2, 191. Thus, buffer mixtures, that all 13 buffers, consisting largely of Good buffers [61, and used to generate natural pH gradients mclstly on solid media, 43 out of 47 similar buffers [71 used to generate pH gradients, were not ideal for protein IEF because the gradients were had pZ-pK differences > 2 p H units. Perhaps factors such as either narrow or nonlinear [2l. The BEF procedure which can buffer concentration and the nature or concentration of media produce a more stable, reproducible and linear p H gradient, have some role to play. uses a gel system and a carefully weighed mixture of 47 amphoteric buffers 171. According to Almgren 1201, using car- Righetti 191has reviewed how pH gradients have been known rier ampholytes with widely separated PIvalues reduces the to be formed in an electric field. Briefly, they can be generated number of carrier ampholytes required to give a good pH (i) artificially by diffusion of buffers of different pH, (ii) nongradient. We were therefore keen to develop a system where isoelectrically, by steady state or isotachophoretic stacking the use of few amphoteric buffers could give linear, reproduc- of nonamphoteric buffers, (iii) isoelectrically, by sequential ible and stable gradients to achieve satisfactory separation of alignment of ampholytes in increasing order of plvalues from proteins in a liquid column from which proteins could easily be the anode to cathode, (iv) electrolysis of a mixture of acids and recovered. ampholytes, (v) alignment of amphoteric buffers by moving boundarymechanisms [2],and(vi)by pHvariationsofbuffers Earlier, a mixture of HEPES and MES buffers was used (81to due to variations in dielectric constant in different concentragenerate a p H gradient, stabilized by a sucrose gradient, which tions of solvents such as gradients of glycerol or sucrose. It has - being somewhat nonlinear and unstable - affected protein also been remarked that p H gradients are formed by passeparation. Using both above buffers, as well as BICINE and sage of current through gradients of sucrose or acrylamide, CAPS, it was possible to generate p H 4-8 and 3-9.5 gra- (vii) computer simulation by using ampholytes of known elecdients, stabilized by a sucrose gradient in the column. The gra- trochemical parameters such as pK, pZ, mobility coefficient, dients were as linear as those generated by SCAMS and the etc. [2,191. Themechanism(s)underlying the generationofthe duration of stability quite similar to that reported previously pH 4-8 and 3-9.5 gradients by the procedure detailed here for two Good buffers 181. Furthermore, the procedures could cannot be ascertained, but it is possible that some of these facgenerate pH gradients with reproducible slope and linearity tors are operative. Whatever the mechanism(s) involved, the
Buffer isoelectric focusing
EIertrophoresis 1990,11, 156-161
slope of the two gradients and their shallowness indicate that they can be used for satisfactory separation of proteins. Some of the advantages of BEF over I E F with SCAMS have already been recognized [2,5,7,24,251. Thefocusingpatterns obtained with carrier ampholytes are known to vary not only with manufacturing source [261 but also from lot to lot from a given source [ 181. Further, in I E F with carrier ampholytes on polyacrylamide gels, locations of focused proteins, handling of the gels, and subsequent recovery of the protein from gels/ carrier ampholytes become a formidable task, whereas by BEF in liquid column, proteins can be obtained easily by simple dialysis or ultrafiltration. Unlike carrier ampholytecontaining proteins, those eluted from the column after BEF can be directly introduced into biological systems where they are well tolerated. The procedure described here takes only 3h as against 18h o r more required when carrier ampholytes are used in the LKB column. Furthermore, one can use large protein loads for preparative studies and also analyze large M , proteins which do not enter polyacrylamide gels. However, one cannot overlook the drawback of I E F in a liquid column. A possible remixing of focused protein zones by diffusion or during elution, particularly when the protein mixture focused has a number of components with very close plvalues, cannot be ruled out. This may give rise to contaminants in the eluted protein which necessitates refocusing. The best application of the BEF technique could be in the field of purification of enzymes because the possibility of denaturation of proteins by heat is considerably reduced as the liquid column is maintained at 4 "C. This is amply borne out by the ability to purify about 25-fold (Table 1) the major acidic isozyme component of ML-ADH by subjecting a partially purified ML-ADH preparation to BEF in the pH gradient 4-8. The activity focused at p15.6, with an M,of 40 000, appeared to be homogeneous on SDS-PAGE (Fig. 3c). Another area of research where this technique can find application is for the separation of cells. In fact, we have successfully used biological buffers such as MES-HEPES to generate a linear p H gradient of 3 to 6 which satisfactorily resolves heterogeneous populations of peripheral blood leukocytes in normal and leukemic patients.
W e thank Shrikrishna Athavale for his excellent technical assistance. Received February 28, 1989; in revised form September 29, 1989
161
5 References Caspers,M. L. amd Chrambach, A., Anal. Biochem. 1977,8/, 28-39. Chrambach, A,, in: Neuhoff, V. (Ed.), Electrophoresis '84, Verlag Chemie, Weinheim 1984, pp. 3-28. Nguyen, N. Y. and Chrambach, A,, Anal. Biochem. 1976, 74, 145-153.
Nguyen,N.Y. andchrambach, A.,Electrophoresis 1980,1,14-22. Prestidge, R. L. and Hearn, M. T. W., Anal. Biochem. 1979, 97, 95-102. Auzan, C. and Michoud, A,, in: Peltre, G. (Ed.)Electrophorese-par is^ 8 2 , Institut Pasteur, Paris 1982, p.1. Cuono. C. B. and Chapo, G. A., Electrophoresis 1982,3, 65-15. Rao, K. V. and Duraiswami, S., Indian J. Exptl. Biol. 1978, 16, 1221-1225. Laemmli, U. K., Nature 1970,227, 680-685. Blum,H.,Beier,H. andGross,H.J.,Electrophoresis1987,8,93-99. Haseba,T.,Hirakawa,K.,Tomita,Y. and Watanabe,T., in: Hirai, H. (Ed.), Electrophoresis '83, Walter de Gruyter, Berlin 1984, pp. 393-400. Dalvi, R. R., Indian, J. Biochern. Biophys. 1987,24,248-25 1. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J., J. Biol. Chem. 1951,193,265-275. Good, N. E., Winget, G. D., Winter, W., Connolly, T. N., Izawa, S. and Singh, R. M. N., Biochemistry 1966,5,467-477. Sherins, R. J., Vaitukaitis, J . L. and Chrambach, A., Endocrinology 1974,92, 1135-1141. Hjerten, S., Liao, H. L. and Yao, K., J . Chromatogr. 1987, 387, 127-138. Carpenter, H. C., Skerritt, J. H., Wrigley, C. W. and Margolis, J., Electrophoresis 1986, 7,221-226. Finlayson, G. R. and Chrambach, A., Anal. Biochem. 1971, 40, 292-3 1 1. Righetti, P. G.,in: Work, T. S. and Burdon, R. H. (Eds.), Laboratory Techniques, Elsevier Biomedical Press, New York 1987, pp. 1-86. Almgren, M., Chem. Scripta 1971, I , 69-75. Gelsema, W. J., DeLigny, C. 1 . and Van Der Veen, N. G., J. Chromatogr. 1979,173,33-4 1. Charlionet, R., Morcamp, C., Sesboue, R. and Martin, J. P., J. Chromatogr. 1981,205, 355-366. Rilbe, H. in: Catsimpoolas, N. (Ed.), lsoelectric Focusing, Academic Press, New York 1976, pp. 14-52. Nguyen, N. Y. and Chrambach, A., Anal. Biochem. 1977, 79, 462-469. Chrambach, A., andNguyen, N. Y.in: Radola,B. J. andGraesslin,D. (Eds.), Electrofocusing and Zsotachophoresis, Walter de Gruyeter, Berlin 1977, pp. 51-58. Allen, R. C., Christopher, J., Lorlincz, L., Allen, R. C. Jr. and Liu, P. in: Radola, B. J. (Ed.), ElectrophoreseForum '80,Technical University, Munich 1980, pp. 117-125.
N. R. Vaidya et al.
Narendra R. Vaidya Balwant P. Gothoskar Ashutosh P. Banerji Biological Chemistry Division, Cancer Parel, Bombay Research
Electrophoresis 1990,11, 156-161
Column isoelectric focusing in natural pH gradients generated by biological buffers Using 2 or 3 simple Good zwitterionic buffers at a 16 or 18 mmol/L final column concentration of the mixture, natural p H gradients of 4 to 8 and 3 to 9.5, respectively, were generated in a liquid LKB column. The pH gradients, stabilized by an anticonvective sucrose gradient, were linear, reproducible and stable in the electric field up to 5h. The p H gradients were used for isoelectric focusing of a number of impure proteins such as human hemoglobin, bovine serum albumin and chicken egg white lysozyme. The protein components could be well separated in the gradient, wereeasily recovered and appeared to be quite pure when analyzed by sodium dodecyl sulfategel electrophoresis. Furthermore, the p H gradient 4-8 was effectively used to isolate one of the acidic isozyme (PI 5.6) components of mouse liver alcohol dehydrogenase (EC 1.1.1.1) in an enzymatically active state, suggesting that the procedure does not denature proteins. The low cost, the ease with which the pH gradients are formed, their linearity, stability for a sufficient period to allow proteins to reach equilibrium and their subsequent recovery from buffer eluates should make the procedure interesting for electrofocusing of proteins.
1 Introduction Isoelectric focusing (IEF) of proteins, using wide or narrow ranges of p H gradients, generated by mixtures of synthetic carrier ampholytes (SCAMS), has come to be accepted as a standard procedure to characterize or separate proteins from their mixtures. Carried out largely on polyacrylamide gel systems, these carrier ampholytes, which generate a linear and stable pH gradient, bring about efficient separation of proteins. However, the cost of the carrier ampholytes, the time involved and the efforts needed to separate polypeptides from the carrier ampholytes put severe restrictions on their use. Earlier, mixtures of amino acids [ 1I, or nonamphoteric buffers were used toobtain pH gradientsin anelectric field which were at best quasi-stable. The credit for putting I E F with buffer mixtures or buffer electrofocusing (BEF) on a stable footing goes to Chrambach's group 121. Nguyen and Chrambach described methods for producing natural p H gradients from 10 nonamphoteric and amphoteric buffers 131 but the gradients suffered from nonlinearity and/or narrowness of range. Availability of amphoteric buffers with wide ranges of pK values, and their zwitterionic properties of migration in either direction in an electric field, upheld the belief that for effective BEF, pH gradients can be tailor-made when using these buffers. T o suit one's need, one could thus scan their pK values and then add or substract constituents in the mixture [ 2 ] . Such pH gradients, made from 9-14 amphoteric/nonamphoteric buffers in the range of 3-7 on polyacrylamide gels [41 and 4-8 on Sephadex gels I51 or narrow acidic range on Sephadex gels [61, were found to be useful. However, the most successful effort in establishing a wide p H gradient of 3-10 on polyacrylamide slab gels, stable for 18h, needed 100 trials as Correspondence:Dr. N. R. Vaidya, Biological Chernistry Division, Cancer Research Institute, Parel, Bombay - 400 012, India
well as 47 buffers in the final analysis [71. Even computational approaches to simulation of pH gradients have not fulfilled the promise of BEF as an inexpensive and practical protein separation tool [21. With this background we present here a simple liquid column BEF technique using 2-3 amphoteric buffers which could be useful for analytical as well as preparative separation of proteins.
2 Materials and methods 2.1 Chemicals and proteins Unless specifically mentioned, all the chemicals were from Sigma (St. Louis, MO). The following proteins (BDH, Poole, Dorset, England) were used: crystalline bovine serum albumin (BSA), human hemoglobin (Hb) type IV, chicken egg ovalbumin and egg white lysozyme. 2.2 Buffers The following Good buffers were used for developing the pH gradients: 2 (N-morpho1ino)ethanesulfonic acid (MES) with a pK value of 6.1 at 25 "C and useful p H range of 5.5 to 6.7, N-2hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES) with pK 7.5 at 25 OC and useful pH range of 6.8 to 8.2, N,Nbis(2-hydroxyethyl/glycine (BICINE) with pK 8.3 at 25 "C and useful p H range of 7.6 to 9.0, 3-(cyclohexylamino)-1propanesulfonic acid (CAPS) with pK 10.4 at 25 OC and useful p H range of 9.7 to 11.1.
2.3 Buffer mixtures To generate a gradient with a pH range of 4-8, I E F was car-
ried out with a mixture of HEPES and BICINE buffers, Abbreviations: Bicine, (N,N-bis[2-hydroxyethyllglycine);BEF, buffer whereas for obtaining p H gradient in the range of 3-9.5, the electrofocusing; BSA, bovine serum albumin; %C, relative percentage of cross linker in a polyacrylamide gel; CAPS, (3-lcyclohexylamino~-1 -pro- mixture comprised CAPS, BICINE and MES buffers. panesulfonic acid): Hb, hemoglobin; HEPES, (N-2-hydroxyethylpiperazine-N'-2-ethanesulfonicacid); IEF, isoelectric focusing; kDa, kilodalton; 2.4 IEF apparatus MES, (2"-morpholinol ethanesulfonic acid); ML-ADH, mouse liver alcohol dehydrogenase; SDS-PAGE, sodium dodecyl sulfate-polyacrylI E F was carried out in a 110 mL-LKB 8101 column (LKB, amide gel electrophoresis; %T, total monomer concentrations in polyacrylamide gel Bromma, Sweden). Accessory equipment comprised a D C 0VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990
0173-0835/90/0202-0156 %2.50/0
Buffer isoelectric focusing
Eleclrophoresis 1990,lI. 156-161
157
power supply, a gradient mixer and a peristaltic pump from LKB.
2.9 Purification of mouse liver alcohol dehydrogenase (ML-ADH) by salt fractionation and IEF
2.5 Procedure for preparative IEF of proteins
Livers (5 gm) of ICRC mice (20-24 g body wt.) were homogenized with cold 5 m~-Tris-HC1-0.5mM NAD-0.7 mM dithiothreitol (DTT)-0.25 M sucrose buffer, pH 8.5, in accordance with the procedure of Hasabe et al. [ 111. The clear supernatant obtained at 40 000 x g for lh at 2 "C was subjected to ammonium sulfate fractionation between 35-75 % saturation as suggested by Dalvi [ 121. The precipitate obtained by centrifugation at 15 000 rpm was dialyzed against 5 mmol/L HEPES buffer. The dialyzate (about 5mg protein) was subjected to IEF in pH 4-8 gradient as above. All the operations were carried out at 4 "C. The crude liver extract, the dialyzate of 35-75 % ammonium sulfate precipitate and the protein peak focused at p15.6 were assessed for ML-ADH activity and specific activity in accordance with the method of Dalvi [ 121. The protein eluted from the column after dialysis and concentration by vacuum dialysis in cold, was measured by the method of Lowry et al. (131.
The procedure followed was that described in the LKB IEF instruction and operation manual. The catholyte was 0.4 % ethanolamine (Merck, Darmstadt, FRG) in 60 % sucrose (BDH) solution at the bottom of the column while the anolyte was 0.1 M NaCl solution adjusted to pH 2.7 with HCl in accordance with the procedure of Rao and Duraiswami IS]. For focusing proteins in a pH 4-8 gradient, stabilized with alinear 20-50 % sucrose density gradient, 3-4 mg of protein (Hb, BSA, etc.) dissolved in 16 mmol/L BICINE was added to 50 mL of the same BICINE solution, containing 50 % sucrose in one vessel of the gradient mixer and an equal volume of 20 % sucrose in 16 mmol/L HEPES solution in the other vessel. The concentration of each buffer in thecolumn was 8 mmol/L. For BEF of a basic protein like chicken egg white lysozyme, in pH 3.0 to 9.5 gradient, the solutions in the gradient mixer were 50 % sucrose, containing 16 mmol/L CAPS and 20 % sucrose with 16 mmol/L BICINE plus 4 mmol/L MES. The column concentrations of each buffer were8,8 and 2 mmol/L, respectively. All solutions were prepared in freshly deionized water with an electrical conductivity less than 0.06 pS/cm at 25 "C. Focusing was initiated with a current of 4.8 mA. The electrical potential applied to the column was then gradually raised to 1100 V, so that the current did not exceed 4.8 mA with a power of 3 W. Later, the current dropped to a constant minimum of less than 0.4 mA and the proteins focused as sharp zones. The power was then switched off and the column eluted (1.5 mL/fraction), with aflow rate of 2 mL/min. The experiments were conducted at 4 "C.
2.6 pH Measurement Immediately following collection, the pH of each fraction was determined on Pye Radiometer Model pH meter.
2.7 Protein measurement The absorbance at 280 nm (A 280 nm) of each fraction was recorded and plotted against fraction number. For human Hb, A560 nm was also determined for the fractions.
2.8 Polyacrylamide slab gel electrophoresis The fractions corresponding to high A280 nm values were pooled and dialyzed in the cold against distilled water. Following lyophilization, the powder was dissolved in the sample buffer (2.3 % SDS, 0.0625 M Tris-HC1, pH 6.8, 11 % glycerol- 1 % 0-mercaptoethanol) and heated at 100 "C for 3 min, and then subjected to SDS-polyacrylamide slab gel electrophoresis (SDS-PAGE) by the method of Laemmli 191. The gels contained 15 % acrylamide and 0.04 % N,N'-methylenebisacrylamide (Bis). The proteins (50-180 pg) were loaded onto 5 % T, 2.6 % C stacking gels and electrophoresed in an 0.025 M Tris, 0.19 M glycine buffer, pH 8.3, containing 0.1 % SDS. Bromophenol Blue was the tracking dye. Next, the gels were fixed in 40 % methanol in 10 % acetic acid solvent mix and stained with0.25 % Coomassie Brilliant Blue R-250in the above solvent, followed by destaining. For silver staining, the gels were fixed and stained by the procedure described by Blum et al. 1101.
3 Results Natural pH gradients can be formed from simple and chemically defined biological Good buffers such as BICINE, HEPES, CAPS and MES [ 141. Preliminary experiments to establish an effective broad range of pH gradient carried out with final concentrations of 4-7 mmol/L of each of the 2 buffers in the column, such as BICINE and HEPES, proved unsatisfactory because the pH gradient was not sufficiently broad (pH 3-6)and also nonuniform. Raising the final column concentration of each buffer to 8 mmol/L brought about a satisfactory linear pH gradient within 90 min of electrofocusing. 3.1 Time course ofthe developmentoflinear pH 4-8 gradient with a BICINE-HEPES buffer mixture BICINE and HEPES at a concentration of 8 mmol/L each were used. After initiation offocusing, the current begins to fall sharply with time so that by 150 min the current drops to less than 0.4 mA with voltage stabilized at 1100 V (Fig. IA). Before application of the electric field, i. e. at 0 time, the pH gradient is essentially flat as the pH is almost uniform (ca. 7.2) along the column (Fig. 1B). At 30 min, the pH gradient starts to manifest itself and with the drop in current at 50 min, it tends to be linear until such a state is attained at 90 min. The conductance, which drops sharply with time, later reaches a constant negligible value with the establishment of the pH gradient (Fig. 1A inset). The residual constant conductance implied that buffers had attained their p1 positions.
3.2 Stability and reproducibilityof the pH gradient and the effect ofthe nature of anolyte solutionon the formationof the gradient This natural pH gradient developed in the electric field was stable up to 5h (Fig. lB, C) at a potential of 1100 V beyond which time the gradient tended to be unstable and later deteriorated. The linearity and stability were reproducible for over several experiments using these buffer mixtures (Figs. 1C and 2). It remained to be seen whether the gradient was stable
158
N. R. Vaidya et al,
Electrophoresis 1990,11, 156-161
5
I
I
4
E I
25;
"
w
z
Q
w .Y
5
"
.Y
>
3
"\
3.3 Resolution ofcrude proteinmixtures by IEF in a pH 4-8 gradient generated by the HEPES-BICINEbuffer system
B 0 Min.
long enough to allow proteins to reach equilibrium and thus enable their effective separation 151. Moreover, as observed earlier in column BEF using 2 amphoteric buffers [Sl, we also noted that the choice ofthe right anolyte, rather than catholyte solution, had significant influence on the nature and stability of the pH gradient. When 0.5 % H,SO, was used as an anolyte, the gradient deteriorated, possibly due to the generation of protons by electrolysis of this solution with resultant drop in the pH along the anodic end (Fig. 1C) and disturbance in the gradient 181. However, use of acidified 0.1 M NaCI, pH 2.7, produced no acidic drag at the anode. It appears that electrolysis of NaCl causes an Na' flux into the column, neutralizing the effect of protons on pH gradient formation during focusing as the ratio of sodium ions to protons is nearly 1 at pH 2.7 1161.
30 Min
The ability ofthe pH gradient produced by the buffer system to satisfactorily resolve crude protein mixtures was then assessed. Human Hb type IV and crystalline BSA were subjected to IEF by the procedure described above. By about 2.5 h, when H b focused as a fine sharp colored zone or BSA as a white ring in the column, the current was switched off and elution initiated. The H b protein peak (fraction No. 43-44, Fig. 2), focused at pH 6.6 (p1 of Hb), and the peak component, eluting at pH 4.8 (fraction No. 67-70), which corresponds to the plvalue of BSA [ 171, were collected. They were subjected to SDS-PAGE analysis, after dialysis, lyophilization etc. to assess their purity and appeared to be homogeneous (Fig. 3a, 3b). Thus, the procedure of IEF detailed above generates a reproducible, linear and stable pH gradient adequate for an analysis of proteins.
50 Min.
3.4 Purificationof partiallypurified ML-ADH by IEF with the HEPES-BICINE buffer system
12
24
36
48
60
72
12
0
FRACTION
C
36
24
48
60
72
NO
BEF TIME ( M i n ) o
150 - o 5 % 150
0 200
I
I
,
20
40 FRICTION
10
IM
+so4 NoCl p H
AS ANOLYTE 2 7 AS
ANOLYTE
1 __--I_
60 NO
80
This was carried out to achieve the following objectives: (i) to see whether the procedure evolved can improve the degree of purification after its extraction from liver tissue and subsequent salt fractionation, (ii) to isolate the major (PI5.3-6.4) isozyme of ML-ADH [ 1 I], (iii) to ascertain that the procedure does not denature proteins during the process offocusing. IEF of the 35-75 % ammonium sulfate precipitate, containing appreciable alcohol dehydrogenase activity [ 121 of mouse liver extract (Table l), could be resolved into several protein peaks (Fig. 2). One major peak (highest A 280 nm value) focused at p15.6 (fraction No. 61,62), the known p1range of the major isozyme component of ML-ADH [ l 11. This peak showed significant ML-ADH activity and also migrated as a single band on SDS-PAGE (Fig. 3c). The purification achieved was about 25-fold (Table 1). 4Figur-e 1. (A) Time course of change in current with stabilization of voltage during establishment ofpH 4-8 gradient, using theHEPES-BICINE buffer mixture, stabilized by a sucrose gradient in LKB-I 10 glass column. The final column concentrations of BICINE and HEPES were 8 mM each. Insert shows the plot of conductance against time. Anolyte: 0.1 M NaCI, pH 2.7. Catholyte: 0.4 % ethanolamine. Duration of IEF was 3.5 h. (B) Generation of alinear pH 4-8 gradient with the HEPES-BICINE buffer mixtureas afunctionoftime. (C)ThestabilityofthepHgradientwith0.5 % H,SO, and 0.1 M NaC1, pH 2.7 as anolytes for BEF with the BICINEHEPES mixture. Note the proton effect in the pH gradient at the anodic end with 0.5 % H,SO, as the electrolyte.
Electrophoresis 1990, 11, 156- 16 1
4
BOVINE SERUM ALBUYIN
-
1.00
;
0.15
159
Buffer isoelectric focusing
12.0
. :
n 0.50
I
E
0.25
P
FRACTION NO
FRACTION NO
ALCOHOL DEMYDROGENASE L Y S O Z Y YE
'.
12.0
10.0
r-5
5.0
IL "
0
20
60
40 FRICTION
NO.
z
2.5
I
80 20
0
j -
I
I
1
40
60
110
F R A C T I O N NO.
Figuve2. Absorptioncurves at 280 nmofeffluentfractionscollectedfromLKB-I 1Ocolumn,superimposedonpHvaluesoffractions(at 27 T ) a f t e r I E F o f human hemoglobin type IV, bovine serum albumin, partially purified mouse liver alcohol dehydrogenase in a p H 4-8 gradient, generated by the HEPESBICINE buffer system, egg whitelysozymein a p H 3-9.5 gradient,generatedbyusing theBICINE-MES-CAPS buffer mixture. For Hb, A 560 nm was also measured for the fractions. Electrolytes were as in Fig. 1 A.
Table 1. Purification of ML-ADH Fraction
Liver extract 35-75 % saturated (NH,)SO, IEF
Total Protein Activitya) Specific Purification volume mg/mL nmoles/mL activity factor mL min nmoles min/mg protein 12 3 2
31.7 40 0.45
200 616 71.1
6.3 15.4
1 2.4
158.0
25.1
a) Unit of enzyme activity was defined as the amount of enzyme that catalyzes formation of 1000 nmoles of NADH/min at 340 nm.
4 Figure 3. SDS-PAGE patterns of (a) Human hemoglobin (lane 1) and that of the protein (lane 2) recovered from the peak focused at p16.6 (Fig. 2), (b) bovine serum albumin (lane 3) and ofprotein (lane 4) electrofocused at p l 4.8 (Fig. 2), (c) M, markers ovalbumin (42 000, lane 5) and hemoglobin (16 000,lane6),protein(lane7)eluting atpI5.6(Fig.2)afterIEFofpartially purified mouse liver alcohol dehydrogenase. IEF was carried out in a4-8 p H gradient, (d) egg white lysozyme (lane 8) and protein peak focused at pZ 9.5 (lane 9) in pH gradient 3-9.5. Slab gels (a), (b), (d) were stained with Coomassie Brilliant Blue R-250, while gel (c) was stainedwith silver. Arrow indicates origin.
160
N. R. Vaidya el al.
3.5 Establishmentofa broadrange pH 3-9.5 gradientusing a mixture of 3 buffers MES, BICINE and CAPS, at different concentrations, were subjected to IEF. The final concentrations of the buffers in the column to obtain a linear pH 3-9.5 gradient (Fig. 2) was found to be 2 mmol/L of MES and 8 mmol/L, each of CAPS and BICINE. When a protein such as egg whitelysozyme with a p l value greater than 11 was focused in this gradient, the crude mixture was separated into 6 protein peaks. Since the ultimate limit of p H gradient formed by this buffer system is pH 9.5, the protein peak (fraction No. 19-20), eluting first at the plvalue of pH 9.5, was analyzed by SDS-gel electrophoresis. It appeared to be a homogeneous protein (Fig. 3d).
Electrophoresis 1990.11, 156-161
(Figs. lB, l C , 2). The conductance, approaching anear-zero value after 2.5 h during focusing of crude protein mixtures, implied that the carrier and protein constituents had reached their equilibrium pZ positions by this time. Further, the constancy of elution patterns [ 181of Hb, BSA andlysozyme(data not shown) on extension of their I E F time from 2.5-5h gave credence to steady-state protein IEF in the gradients generated by the buffer systems. Thus, though the stability of the pH gradients that can be formed by our systems were not quite comparable to those reported by the buffer system of Cuono and Chapo 171 or in systems using SCAMS, they were nonetheless quite efficient for the separation of the proteins studied.
Theoretically, the success of steady-state IEF of proteins is determined by the properties of the carrier ampholytes such as 3.6 Stability of pH gradients suficient to allow proteins in their high buffering capacity and conductivity at their pZ[2,7, a mixture to reach equilibrium state during IEF 191. Though the buffering capacities of mixtures of the two buffer systems used in this study were not assessed, it may be After I E F is initiated, there is a sharp dlrop in conductance expected from their low M , and known relationship between reaching a negligible value with the establishment of the pH buffering capacity and conductance [ 2 11 to be less than that of gradients by 90 min (Fig. lA, inset). The conductance con- SCAMS [ 71. The conductivity of buffer mixtures at low molar tinues to be uniformly low until the proteins focus as zones concentrations has already been reported to be l/lOth that after 2.5 h. This implies that after the initial attainment of reported for SCAMS (71. In spite of these adverse points, the isoelectric position by the carrier constituents, the proteins separation of protein mixtures by the method presented here gradually attain their pzpoints after approximately 3h. Exten- appears quite satisfactory as indicated by homogeneity on sionofIEF timeofHb,BSA,orlysozymefrom2Shto5hgave SDS-PAGE analysis of proteins separated by IEF. This is superimposable protein elution profiles of the respective pro- because buffering capacity and higher conductivity of carrier teins with the pZ values of the proteins remaining the same constituents are not the critical factors - not as in preparative (data not shown). This constancy of pattern 1181 and the IEF - for the analytical resolution of proteins [7, 21, 221. above observation indicate that the two p H gradients, narrow or wide, generated in the liquid column were sufficiently stable An argument which is often preferred to explain why BEF has so that the proteins reached a steady state during IEF. not been the most widely used procedure for protein analysis, unlike I E F with SCAMS, is the nonavailability of a wide range of amphoteric buffers with small pl-pK difference and therefore their inability to buffer protein ampholytes at the 4 Discussion isoelectric state 12, 19, 231. However, in practice it appears Since the advent of SCAMS, which can formlinear, stable and that neither the pl-pK values nor the total number of such bufwide pH gradients, IEF using amphoteric: buffers has taken a fers used essentially determine the buffering capacity or back seat mainly due to the nonavailabiliity of buffers with pZ stability of the pH gradient. This can be inferred from the fact covering the entire pH scale [2, 191. Thus, buffer mixtures, that all 13 buffers, consisting largely of Good buffers [61, and used to generate natural pH gradients mclstly on solid media, 43 out of 47 similar buffers [71 used to generate pH gradients, were not ideal for protein IEF because the gradients were had pZ-pK differences > 2 p H units. Perhaps factors such as either narrow or nonlinear [2l. The BEF procedure which can buffer concentration and the nature or concentration of media produce a more stable, reproducible and linear p H gradient, have some role to play. uses a gel system and a carefully weighed mixture of 47 amphoteric buffers 171. According to Almgren 1201, using car- Righetti 191has reviewed how pH gradients have been known rier ampholytes with widely separated PIvalues reduces the to be formed in an electric field. Briefly, they can be generated number of carrier ampholytes required to give a good pH (i) artificially by diffusion of buffers of different pH, (ii) nongradient. We were therefore keen to develop a system where isoelectrically, by steady state or isotachophoretic stacking the use of few amphoteric buffers could give linear, reproduc- of nonamphoteric buffers, (iii) isoelectrically, by sequential ible and stable gradients to achieve satisfactory separation of alignment of ampholytes in increasing order of plvalues from proteins in a liquid column from which proteins could easily be the anode to cathode, (iv) electrolysis of a mixture of acids and recovered. ampholytes, (v) alignment of amphoteric buffers by moving boundarymechanisms [2],and(vi)by pHvariationsofbuffers Earlier, a mixture of HEPES and MES buffers was used (81to due to variations in dielectric constant in different concentragenerate a p H gradient, stabilized by a sucrose gradient, which tions of solvents such as gradients of glycerol or sucrose. It has - being somewhat nonlinear and unstable - affected protein also been remarked that p H gradients are formed by passeparation. Using both above buffers, as well as BICINE and sage of current through gradients of sucrose or acrylamide, CAPS, it was possible to generate p H 4-8 and 3-9.5 gra- (vii) computer simulation by using ampholytes of known elecdients, stabilized by a sucrose gradient in the column. The gra- trochemical parameters such as pK, pZ, mobility coefficient, dients were as linear as those generated by SCAMS and the etc. [2,191. Themechanism(s)underlying the generationofthe duration of stability quite similar to that reported previously pH 4-8 and 3-9.5 gradients by the procedure detailed here for two Good buffers 181. Furthermore, the procedures could cannot be ascertained, but it is possible that some of these facgenerate pH gradients with reproducible slope and linearity tors are operative. Whatever the mechanism(s) involved, the
Buffer isoelectric focusing
EIertrophoresis 1990,11, 156-161
slope of the two gradients and their shallowness indicate that they can be used for satisfactory separation of proteins. Some of the advantages of BEF over I E F with SCAMS have already been recognized [2,5,7,24,251. Thefocusingpatterns obtained with carrier ampholytes are known to vary not only with manufacturing source [261 but also from lot to lot from a given source [ 181. Further, in I E F with carrier ampholytes on polyacrylamide gels, locations of focused proteins, handling of the gels, and subsequent recovery of the protein from gels/ carrier ampholytes become a formidable task, whereas by BEF in liquid column, proteins can be obtained easily by simple dialysis or ultrafiltration. Unlike carrier ampholytecontaining proteins, those eluted from the column after BEF can be directly introduced into biological systems where they are well tolerated. The procedure described here takes only 3h as against 18h o r more required when carrier ampholytes are used in the LKB column. Furthermore, one can use large protein loads for preparative studies and also analyze large M , proteins which do not enter polyacrylamide gels. However, one cannot overlook the drawback of I E F in a liquid column. A possible remixing of focused protein zones by diffusion or during elution, particularly when the protein mixture focused has a number of components with very close plvalues, cannot be ruled out. This may give rise to contaminants in the eluted protein which necessitates refocusing. The best application of the BEF technique could be in the field of purification of enzymes because the possibility of denaturation of proteins by heat is considerably reduced as the liquid column is maintained at 4 "C. This is amply borne out by the ability to purify about 25-fold (Table 1) the major acidic isozyme component of ML-ADH by subjecting a partially purified ML-ADH preparation to BEF in the pH gradient 4-8. The activity focused at p15.6, with an M,of 40 000, appeared to be homogeneous on SDS-PAGE (Fig. 3c). Another area of research where this technique can find application is for the separation of cells. In fact, we have successfully used biological buffers such as MES-HEPES to generate a linear p H gradient of 3 to 6 which satisfactorily resolves heterogeneous populations of peripheral blood leukocytes in normal and leukemic patients.
W e thank Shrikrishna Athavale for his excellent technical assistance. Received February 28, 1989; in revised form September 29, 1989
161
5 References Caspers,M. L. amd Chrambach, A., Anal. Biochem. 1977,8/, 28-39. Chrambach, A,, in: Neuhoff, V. (Ed.), Electrophoresis '84, Verlag Chemie, Weinheim 1984, pp. 3-28. Nguyen, N. Y. and Chrambach, A,, Anal. Biochem. 1976, 74, 145-153.
Nguyen,N.Y. andchrambach, A.,Electrophoresis 1980,1,14-22. Prestidge, R. L. and Hearn, M. T. W., Anal. Biochem. 1979, 97, 95-102. Auzan, C. and Michoud, A,, in: Peltre, G. (Ed.)Electrophorese-par is^ 8 2 , Institut Pasteur, Paris 1982, p.1. Cuono. C. B. and Chapo, G. A., Electrophoresis 1982,3, 65-15. Rao, K. V. and Duraiswami, S., Indian J. Exptl. Biol. 1978, 16, 1221-1225. Laemmli, U. K., Nature 1970,227, 680-685. Blum,H.,Beier,H. andGross,H.J.,Electrophoresis1987,8,93-99. Haseba,T.,Hirakawa,K.,Tomita,Y. and Watanabe,T., in: Hirai, H. (Ed.), Electrophoresis '83, Walter de Gruyter, Berlin 1984, pp. 393-400. Dalvi, R. R., Indian, J. Biochern. Biophys. 1987,24,248-25 1. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J., J. Biol. Chem. 1951,193,265-275. Good, N. E., Winget, G. D., Winter, W., Connolly, T. N., Izawa, S. and Singh, R. M. N., Biochemistry 1966,5,467-477. Sherins, R. J., Vaitukaitis, J . L. and Chrambach, A., Endocrinology 1974,92, 1135-1141. Hjerten, S., Liao, H. L. and Yao, K., J . Chromatogr. 1987, 387, 127-138. Carpenter, H. C., Skerritt, J. H., Wrigley, C. W. and Margolis, J., Electrophoresis 1986, 7,221-226. Finlayson, G. R. and Chrambach, A., Anal. Biochem. 1971, 40, 292-3 1 1. Righetti, P. G.,in: Work, T. S. and Burdon, R. H. (Eds.), Laboratory Techniques, Elsevier Biomedical Press, New York 1987, pp. 1-86. Almgren, M., Chem. Scripta 1971, I , 69-75. Gelsema, W. J., DeLigny, C. 1 . and Van Der Veen, N. G., J. Chromatogr. 1979,173,33-4 1. Charlionet, R., Morcamp, C., Sesboue, R. and Martin, J. P., J. Chromatogr. 1981,205, 355-366. Rilbe, H. in: Catsimpoolas, N. (Ed.), lsoelectric Focusing, Academic Press, New York 1976, pp. 14-52. Nguyen, N. Y. and Chrambach, A., Anal. Biochem. 1977, 79, 462-469. Chrambach, A., andNguyen, N. Y.in: Radola,B. J. andGraesslin,D. (Eds.), Electrofocusing and Zsotachophoresis, Walter de Gruyeter, Berlin 1977, pp. 51-58. Allen, R. C., Christopher, J., Lorlincz, L., Allen, R. C. Jr. and Liu, P. in: Radola, B. J. (Ed.), ElectrophoreseForum '80,Technical University, Munich 1980, pp. 117-125.
