140

K. Altland

Electrophoresis 1990,II, 140-147

I201 West,K. A.andCrabb,J. W.,in:T.E. Hugli,(Ed.), Techniquesinfrotein Chemistry. Academic Press, SanDiego, CA 1989,pp. 295-304. [21) Hunkapillar, M. W. and Lujan, E., in: Shively, J. E., (Ed.) Microcharacterization of Polypeptides: A Practic,al Manual, Humana Press, Clifton NJ 1986, pp. 89-101. [221 Crabb, J. W., Johnson, C. M., Carr, S. A., Armes, L. G. and Saari. J. C.,J. B i d . Chem. 1988,263, 18678-18687. 1231 Jennissen, H. P., Peterson-VonGehr, J. K. H. and Botzet, G., Eur.J. Biochem. 1985, f47,619-630.

Klaus Altland University of Giessen

[241 Matsudaira, P., 2. B i d . Chem. 1987,262, 10035-10038. I251 Eckerskorn, C., Memes, W., Goretski, H. and Lottspeich, F., Eur. J . Biochem. 1988,176,509-519. [261 Aebersold, R. H., Leavitt, J., Saavedra, R. A., Hood, L. E. and Kent, S. B., Proc. Natl. Acad. Sci. USA 1987,84, 6970-6974. [271 Scott, M. G., Crimmins, D. L., McCourt, D. W., Tarrand,J. J.,Eyerman, M. C. and Nahm, M. H., Biochem. Biophys. Res. Commun. 1988, f55,1353-1359. 1281 Wilson, K. J., Trends Biochem. Sci. 1989,14,252-255.

IPGMAKER: A program for IBM-compatible personal computers to create and test recipes for immobilized pH gradients The program “IPGMAKER’ is a computational aid for creating and testing recipes for near-linear immobilized p H gradients. It was written for fast IBM personal cornpiiters (with aType 80386 processor and 80387 coprocessor) and compatibles equipped with a VGA, E G A or Hercules (mono) graphics card. The program is limited to thie use of up to 10 acids and/or bases, and to ranges spanning between pH 2 and 12. The resulting recipes are presented either as final concentrations in the 2 chambers of a mixing device for linear gradients or as volumes from 0.2 moles/L stock solutions adjusted to a user-defined average buffering power. One of the subroutines determines the pH, gradient slope and buffering capacity at any location of the gradient and includes a facility to estimate the p l of proteins from the composition of their primary structure.

1 Introduction The access to full flexibility of immobilized pH gradients (IPG) requires the availability of Immobiliiies with dissociation constants (pKa) covering the p l scale of proteins, as well as recipes and technical facilities to generate appropriate IPGs in a gel matrix. The commercially available Immobilines of pKa 3.58,4.51,6.21,7.06,8.50 and 9.59 111 make it possible to generate appropriate recipes for gradients spanning the ranges between pH 4 and 10. The ranges can be extended to pH 2.5- 1 1.5 by including two other Immobiilines - of pKa 0.8 and 10.3 - which, hopefully, will soon be available to everyone. Many recipes for narrow and wide pH ranges have been made available for linear gradient mixers by Gianazza et al. [21 after the development of an adequate computer program, and Altland and Altland [ 3,4] have developed computerized mixing devices to use these recipes for the creation of any linear or nonlinear gradient with high reproducibility. While the user has free access to the chemicals and the devices to pour gels from published recipes, the flexibility of recipes remains limited as the program for their generation has not been made available. Furthermore, linearity of published recipes is Correspondence: Prof. Dr. Klaus Altland, Justus-Lielbig-University,Institute of Human Genetics, Schlangenzahl 14, D-6300 Giessen, Federal Republic of Germany Abbreviations:CV, coefficient ofvariation; IPG, immobilized pH gradient; PC, personal computer; SD, standard deviation 0VCH Verlagsgesellschaft mbH, D-6940 Weinheim, I990

expected to get lost if they are not corrected for the effects of additives like urea on pKa [51. The profile of buffering power along the pH range has not been presented for any of the recipes although it varies considerably. In the extreme pH ranges immobilized buffering power is considerably superimposed by that of water ions. Their relative quantitative contribution should be known but cannot be read from numbers merely giving an average level over the pH range of a given recipe. To eliminate these shortcomings a program for generally available personal computers (PCs) is presented which creates new recipes, tests existing recipes and provides information such as pH, slope of gradient and immobilized as well as free buffering power at any site of linear or nonlinear pH gradients generated by devices for linear gradients.

2 Operation 2.1 Getting started The program was written in Pascal and should be run on an IBM PC or PC compatible with a Type 80386 microprocessor, a Type 80387 coprocessor and a VGA-, EGA- or HERCULES (Mono) graphics card with a minimum resolution of 650 x 350pixel. Hard copies ofnongraphic screens can be made with any attached printer by using the [PRINT] function key. Copies of graphic screens as those shown in this paper can be achieved with an NEC P6 or compatible 24needle matrix printer by pressing key [PI. The following 10 01 73-0835/90/002-0140 $2.50/0

Electrophoresis 1990.11, 140-141

IPGMAKER: a program to create and test recipes for IPGs

141

IPGMAKER: Create a Recipe, User Data.

Number of Immobilines: 10 Low pH limit (LpH) : 3.00 High pH limit (HpH) : 11.00 No.

pKa :

Acid/Base

1 2

0.80 3.00

3 4 5 6

3.58 4.51 6.21 7.06 8.50 9.59 10.30

Acid Acid Acid Acid Base Base Base Base Base Base

1

8

9 10

14.00

: 10 Testpoint Number Buffering power : 3.00 meq/pH/L Max. Irnrnob. Conc. : 16.00 mmoles/L Max.incr.0.5^n*max.n=: 3 chamber 2 Code chamber 1 (low pH) (high pH) 1.000 1.000 T 1.000 1.000 T 1.000 1.000 T 1.000 1.000 B 1.000 1.000 B 1.000 1.000 B 1.000 1.000 B 1 .ooo 1.000 B 1,000 1.000 B 1.000 1.000 T

Are these default data correct [y]/[n]?

figures and comments will demonstrate by example the development of a recipe for the p H range 3.5-6.5. Thediscette contains the program file 1PGMAKER.EXE and 2 graphics drivers EGAVGA.BG1 and HERC.BG1. The program is called by its name (IPGMAKER) and - after a nonvisible hardware check until no problems are encountered - offers on request some hopefully useful notes on the first screen. The second screen asks whether a new recipe is to be created or an already existing recipe to be tested. Upon activating the first choice the screen shown by Fig. 1 is presented.

2.2 Entering data All default numbers and characters can be changed within reasonable limits. Some of them require a few comments: The pH range between the low pH limit (LpH) and the high p H limit (HpH) is divided into a number of equal fractions. The test point number (TN) is the number of these fractions and may vary between 1 and 20. A test for buffering capacity or pH is made at any site with pH = LpH

+ n/TN * (HpH - LpH)

(1) for 0 I n 5 TN. A low TNreduces the number oftests and may miss undesirable deviations of the resulting gradient from expectation but increases the speed of calculation. A high TN provides a safer result but will cost more computational time. Narrow gradients will require only a small TNrelative to wide gradients at a given level ofprecision. One may start with alow TN and try again with a high TNto improve the final result.

Figure 1. First screen with default values to create an IPG recipe.

Whenever the program does not find an improvement with a given increment (decrement) it tries again with IZ + 1 for 1 increasing from 1 to 13 - n. The actual size of the resulting increment (decrement) is presented on the graphic screens. The individual Immobilines must be labelled with a character code, which can be B, T or F. Immobilines labelled with B are selectively used to smooth the buffering capacity at the given average while those with a T are used for titrating to the expected p H levels. The label F for "Fine tuning" will be explained later. For a trial to create a recipe of a gradient of p H 3.5-6.5 with commercially available Immobilines, changes were made to result in the screen givenin Fig. 2 . AllImmobilines with significant buffering capacity within the pH range were labelled with B and one extra Immobiline with almost no buffering power in the range 3.5-6.5 was included and labelled with T. The data were accepted as correct and the user was given the choice to start the algorithm for smoothing the buffering capacity (beta) or that for adjusting p H levels. This screen can always be recalled by interrupting the optimization algorithms. The user will learn that at this stage it is wise to select (b) for automatic smoothing of the buffering capacity at the entered level of the average.

2.3 Smoothing and adjusting the buffering capacity The idea of adjusting the buffering capacity to a constant level as a first step of creating a p H gradient was taken from Gianazza et al. [ 61. The results of this trial are shown in Fig. 3. The 4 curves at the bottom of the graph represent the distribution of the buffering capacity along the pH gradient range of the 4 buffering Immobilines with pKa 3.6,4.5,6.2 and 7.1, while the resulting total is represented by the wavy line near the upper frame of the graph. The total buffering capacity Pt @ = dC/ dpH with C being the amount of strong base or acid in moles to change pH of 1 L sample solution by one pH unit)is calculated by the formula:

The maximum Immobiline concentration is limited to 5 20 mmoles/L assuming that final gel concentrations > 20 are not useful. Entering small numbers may force the program to find a recipe with a more even distribution of Immobiline concentrations but may also block program execution if the entered levels of average buffering power or p H cannot be achieved. The basic concept for a good result is to change the Immobiline concentrations in chambers 1 and 2 of a mixing device for linear gradients by increments and decrements of decreasing size. The maximum increment (decrement) is entered by the number n with 0 5 n 5 13 and calculated from the formula

Pt = ln(10) * ([OH-] + [H'I + sum (Ci * Kai* IH'I / (Kai + [H'l)^2))

increment (decrement) = 0.5 * n * maximum Immobiline concentration (2)

with i = 1 to n, where C i s the concentration of the individual Immobiline, K a its dissociation constant and n their total

(3)

142

K. Altland

Electrophoresis 199O,II, 140-147

IPGMAKER: C r e a t e a Recipe, User Data.

Number of Immobilines: 5 : 3.50 Low pH l i m i t (LpH) High pH l i m i t (HpH) : 6.50 No.

pKa :

Acid/Base

1 2 3

3.58 4.51 6.21 7.06 9.59

Acid Acid Base Base Base

4 5

T e s t p o i n t Number : 10 B u f f e r i n g power : 3.00 meq/pH/L Max. Immob. Conc. : 16.00 mmoles/L Max.incr.0.5*nxmax,n=: 3 chamber 1 chamber 2 Code (low pH) ( h i g h pH) 1.000 B 1.000 1.000 1.000 B 1.000 1.000 B 1.000 1.000 B 1.000 1.000 T Figure 2. First screen after presumably appropriate changes to create a recipe for an IPG of pH 3.5-6.5.

S t a r t w i t h smoothing [ b l e t a o r w i t h a d j u s t i n g [plH ?

IPGHf4KER: Snooth Beta beta This job i s done ! 3 .O 2.7

[ E l x i t ; I[Ulser interaction pKa

Ch . # I

Ch .#2

3. 6

3.884

7.841

II

4. 5

0.000

7.293

6.2

12.569

1 ill

/ I

7.1

3.259

5.732

9.6

1 .ooo

I .ooo

..............................

1

11

1.8

I

1.5 1.2

0.9 0.6 0.3 0 .o 3 .5

3.8

4 . 1 4.4

4.7

............. nean beta

Best Results

cu

1x3 : MaxJeuiation: at pH :

5.0

5 . 3 5 . 6 5 . 9 6 . 2 6.5 pH actual f i t of beta

Clctual Result 0.970 CU [%I : -0.052 MaxJeuiation: 5.300 at pH :

Increnent(decrenent): 0.970 pKa : -0.052 Ch.111: 5.300 Ch.112:

7.060 3.258 5.731

number. The individual betai contribution of an Immobiline corresponds to the partial-function

pi = ln(10) * C * Ka * [H’l/

(Ka

+ [H+])^2

(4)

and that of water pwwith

pw= In(1O) * (lOH-1 + [H’]) On the upper right of the screen the actual composition of the recipe is shown. As expected there is no change with Immobiline of pKa 9.6 labelled with a “T”. At the bottom of the screen there are 3 columns of data: The left column gives the best achieved fit of smoothing and adjusting of beta with the expected level represented by the coefficient of variation (CV) and the maximum deviation. The second column contains the result of testing the actual change which cam be read by comparing the data of the 3rd column with the data in the actual recipe (i.e. a change in concentration).

0.002

Figure 3. Screen with the achieved best result of smoothing the buffering capacity at 3 meqv/pH/L.

2.4 Adjusting pH to the expected gradient By pressing the key [El for “Exit” the program activates the algorithm for adjusting the pH to the expectedlevels while the screen is as shown in Fig. 4. The algorithm tries to improve adjustment of pH by concentration changes of Immobiline pKa 9.6 which was labelled with “T”. Calculation of pH is performed by solving Eq. (6)

l@(-pH)+sum(Ci/(l+ l@(pH-pKa,))) = 10“(pH- 14)+sum(CJ( 1+ 10“(pKarpH)))

(6)

with Ci and pKai being the concentrations and dissociation constants of the Immobilines at a given site of the gradient. From the actually achieved fit of pH gradient to the expected levels and the actual Immobiline concentrations in chambers 1 and 2 it is quite obvious that the pH level in chamber 1 cannot be adjusted to the expected level of 3.5 without support of

Elecfrophoresis 1990.11, 140-141

pH

IPGMAKER: a program to create and test recipes for IPGs

IPGMIKER: Adjust pH, Titration [Elxit; IUlser interaction; [Ilutomatic; ISltepwise; [Bleta readjustment; CPlrint this screen (NECP6) beta pKa

143

Ch,Bl Ch,ft2

7. 50 3. 6

3,884

7.841

4.5

0.000

7.293

6 . 2 12.569

1.111

7.1

3.259

5.731

9.6

0,000

5.500

7 -00 6.50

6 .OO 5.50 5 .OO 4 -50 4 .OO

3.50 3 -00 2.50 20 30 40 50 60 70 80 90 100 X Partial Uolune (PU) pH expected -aH actual fit -------betasfor exp. pH Best Results Clctual Result Increment(decrement): 0.002 SDCpH-units1 : 1.9184SD : 1.9184 aKa : 9.590 MaxgH-diff : 3.430 MaxgH-diff : 3.430 Ch.#l: 0,000 at PUCXI : 0 : 0 Ch.12: 5.502 at PU

0

10

.............

PH 7. 50

7 .OO 6.50

IPGMRKER: Adjust pH, Titration [Elxit; [Ulser interaction; [Ilutonatic; [Sltepwise; This job is done ! [Plrint this screen (NEW61 beta

-_- - _ _ _ _ - - - .

I

.-.- - -.- - - -.

*

-

’I

--___-

p ~ a Ch,ftl

Figure 4 . Screen after the first automatic trial to adjust pH to the expected gradient with Immobiline of pKa 9.6. It is obvious that further adjustment at the acidic end requires addition of acid, either with more Immobiline of pKa 3.6 and/or 4.5 at the expense of drastic changes of buffering capacity, or by adding a new Immobiline, i.e. that of pKa 0.8 (see Fig. 5).

Ch.12

3.1 0 . 8 14.969

5.703

3,884

7.841

4.5 0.000

7.293

6. 2 12.569

1,111

7.1

3.259

5.731

9.6

0.008 16.000

2.8

I---

3.6

6 .OO 5.50 5.00

4 .SO 4 -00 3.50 3-00 2.50

20 30 40 50 60 70 80 90 100 X Partial Uolune (PU) pH expected -pH actual fit ------.betas for exp. pH Best Results flctual Result Increnent(decrenent): 0.002 SDtaH-units1 : 0.0731SD : 0.0733 pKa : 9.590 MaxgH-diff : -0.181 MaxgH-diff : -0.182 Ch.#l: 0,006 at PUCXI : 0 at PU : 0 Ch.12: 16.000 0 10 .............

more acid. If one or more of the available acids (i.e. Immobiline pKa 3.6 and 4.5) were relabelled with “T” the result of smoothing the buffering capacity would be lost. Another alternative would be to enter a new acid, i.e. Immobiline pKa 0.8 by [Ulser interaction into the first data screen together with a T code, and reactivate the algorithm for adjusting pH. The best fit achieved under these conditions is shown in Fig. 5. The results given in the graph and the data in the lower left corner in terms of standard deviation (SD) and maximum pH difference between achieved and expected gradient are not good. It becomes obvious that smoothing the buffering capacity - if it does not result in something much better than

Figure 5 . Screen after adding Imrnobiline of pKa 0.8 and automatic titration. The achieved gradient is near expected levels but far from linear. Notice that there are 5.703 mmol/L of Immobiline pKa 0.8 and 16 mmol/L of Immobiline pKa 9.6 in chamber 2, respectively. By user interaction these concentrations can bechanged to 0 for Immobiline pKa 0.8 and to 16 -5.7 = 10.3 mmol/L without changing the gradient. Hand-made changes of this type are often useful to avoid unnecessary loss of Immobilines into a recipe.

achieved in this example - combined with pH adjustment with nonbuffering Immobilines, is an unsatisfactory strategy if not complemented by something else. The decision to be made is to renounce the actual fit of buffering capacity and to include Immobilines of pKa 3.6 to 7.1 in trials to improve pH adjustment. This is done by going back to the data screen and changing codes “B” into “T” or “F”, followed by reactivation of pH adjustment. Now the differential effects of codes T and F become relevant. When less than two Immobilines are labelled with “F”, the active algorithm selects Immobilines labelled with “T” one by one and checks whether

144

Electrophoresis 1990, If, 140-147

K. Altland

protonated or when the level of dissociation of a tested acid equals that of protonation of the tested base. It is obvious that the effects are not neutral when at least one of the Immobilines in a tested couple has a pKa inside or near the selected pH range. The screen presented in Fig. 6 was achieved after setting all codes to “F”. A standard deviation of0.006 pH units and a maximum deviation of 0.012 pH units, whichis about 0.3 mm in a 10 cm wide IPG gel, is in relative good agreement with expectation. The profile of buffering capacity is now far from a constant value along the gradient. This is a consistent feature of most published gradients and is of great importance in quantitative terms for reasons mentioned above.

improvements can be achieved (the criterium is the standard deviation) by incremental (decremental) changes of concentration. When 2 or more Immobilines are labelled with “F”, an alternative algorithm is activated for pH adjustment which first does the same but then changes concentrations of two Immobilines at a time. If both Immobilines in a chamber ofthe mixer are acids or bases, an increment ofthe first Immobiline is combined with a decrement of the second and vice versa. If one Immobiline is an acid and the other a base, both are increased or decreased in a given chamber. The effects are only neutral where the absolute contribution to charge is identical for both Immobilines, as for instance when both bases are fully IPGHAKER: Adjust PH, Fine Tuning DH 7.50

-.-..

-3.8

7.00

kL h.

6.50

--.-

-I-.

6.00

/’

-_

-- - - -

!

.-.--/

%\

4.50

/

,.

.---.. ‘\



-/----,’

\.

/*

,-

2.50

_.__--”

.>’., /

_.I--

3.616

4.5

0.647

6.437

6.2

9.497

1.289

7.1

0.930

6.972

9.6

0,000

4.281

-1.9

-1.5

/ I

YL

-’

,)

-1.2

\\ \

’,-0.8

./

-*-.. + - * - -0 84 _-_-_ :::------ --- .----0.0 --,



5.088

-2.3

/

LL-.t /-?

*A:
.------

--, 4

.------*

\

3 -50

-.

-

- - \+ -

\~

4 .OO

8.338

/-3.1

./

.-.-. -. -

?

5.50

0.8 ,. ”-3* 5

k.

- .-

I

~

I

?

* ?I

-

/

‘*

I

I

.--

,_

-x

-i

I

Best SDCpN-units1 : 0.0060SD HaxgH-diff : -0,012 Hax-pH-diff at PU[%I : 15 at PU

I

1

1

0.0060 pKa : : -0.012 Ch.#l: : 115 Ch.#2:

:

Figure 6. Screen after a few hand-made

0.800 pKa : 8.340 C h . l l : 0.202 Ch.12:

7.060 0.932 6.974

justment. Notice that the good tit is paid by an unequal but acceptable distribution of buffering capacity along the gradient.

IPGHAKER: Test a Recipe beta pKa Ch.#l Ch,#2

PH 7 -50

0.8

8.338

0.200

3.6

5.088

3.626

4.5

0.647

6 437

6.2

9.497

1.289

7.1

0.930

6.972

9.6

0.000

4.281

7 .OO 6 -50 I

6.00 5.50

5 .OO 4 -50 4 -00

3.50 3 -00

1

p.4

230 0 10 20 ........ .... beta;

.

30

40

50

60

70

80

90

100 % P a r t i a l Volume (PU)

PH

[Elxit, [Tlest, change t8leta ?

For W

=

36.000 PH i s 4.590 slope i s 0.307 beta is 2.840

Meantbeta) Cneau/pH/LI: 3.000 2 or nn i n a 10 cn wide gel SD(pH) [units] : 0.0°6 pH-units ~ H / c M( i n a 10 cn wide gel) nequ/pH/L with 0.059 nequ/pH/L from water

Figure 7. Screen permitting an estimate of pH, gradient slope and buffering capacity at any site of the gradient.

IPGMAKER: aprogram to create and test recipes for IPGs

Electrophoresis 1990, 11, 140-147

2.5 Testing the pH gradient

145

IPGMAKER: Recipe (Final g e l concentrations)

By exiting the algorithm for pH adjustment the program switches into the testing routine, presenting the screen shown in Fig. 7. By pressing key [BI the user may readjust the average buffering capacity to other values than the one set on the data screen. By pressing the key [TI he will be prompted to enter a partial volume of the gradient in % (which is equivalent to the distance in mm in a 10cm wide gel) to get the data shown in the lower left corner of the screen. These data are the p H level at that site, the slope ofthe gradient and the total local buffering capacity with the water contribution. If the user wants to test an existing Immobiline recipe, he enters the final Immobiline concentration in chambers 1 and 2 on a separate screen and will again receive the screen shown in Fig. 7. If he decides that the recipe could be improved or be useful for adistinct modification such as extension of range, he will use the sequence of routines to generate a new gradient but start with the data to be improved. The subroutine for testing a recipe can be misused for estimating the p l of a protein if the amino acid composition is known. Instead of entering the pKa’s of Immobilines the user may enter those of the charged side chains of amino acids such as glutamic acid, aspartic acid, lysine, arginine, histidine, tyrosine and cysteine and of ligands like sialic acids. He may assume a 1 mmole/L solution of his test protein to be the only constituent in chambers 1 and 2. Then the amountsofcharged amino acids in the primary structure of his protein in moles/ mole will be equivalent with their concentrations in mmoles/L in thechambers ofthe mixing device. The program supportspl determination by presenting, apart from pH, slope and buffering capacity, the p l when the user asks for the test data at apartialvolumeofOand 100 %ofthegradient.It shouldbereminded that the pH level in an IPG gel is a site where the sum of all charges including those of free ions, i.e. those from water, is zero while the p l of a protein is the pH where the sum of all charges bound to the protein structure is zero. Therefore, plof a protein and pH level in an I P G gel significantly differ at p H levels smaller than p H 5 and larger than pH 9 where the contribution ofwater ions to the sum of charges becomes relevant.

Expected pH range [ u n i t s f : 3.500 Achieved pH range [ u n i t s ] : 3.509 Standard deviation [pH-units] : Mean buffering capacity [meqv/pH/L]: ~

6.500 6.500 0.006 3.000

Recipe f o r low and h i g h pH l i m i t s o l u t i o n : Ch.#l,pH 3.51 [mmol es / L]

Ch.%2,pH 6.50

8.338 5.088

0.200 3.616 6.437 1.289 6.912 4.281

0.80 3.58 4.51 6.21 7.06 9.59

0.647 9.497 0.930 0.000

[mrnoles / L1

[ E l x i t or use [ P r i n t ] key t o g e t a hard copy.

Figure 8. Final recipe of the IPG of pH 3.5-6.5 given in final lmmobiline concentrations in the two chambers of the gradient mixing device.

2.6 Presenting the recipes Figures 8-10 present documents of recipes for use in the laboratory. The presentation in Fig. 8 is similar to that used by Gianazza et al. [2],giving final Immobiline concentrations in the two chambers ofthe gradient mixer. The recipe in Fig. 9 is that for 3 times concentrated Immobiline stock solutions to be stored in aliquots in the freezer for repeated use. On a preceding screen the user has been asked to enter the total stock volumes for chambers 1 and 2 and whether the buffer ions for adjustment to pH 6.8 should be Tris-chloride, -phosphate or -acetate. The complementary recipe for the final gel solution in chambers 1 and 2 are given by the last screen shown in Fig. 10. The data of the upper 8 lines can be modified by the user within reasonable limits. The rest is the resulting recipe, assuming that one third of the final volume comes from the buffered Immobiline stock solution. Figure 11 shows the result of a trial to generate a recipe for pH range 3- 1 1. The maximum deviation from the expected linear

IPGMAKER: Recipe (3 times concentrated stock solutions) Expected pH range [units]: 3.500 Achieved pH range [units]: 3.509 Standard deviation [pH-units] : Mean buffering capacity [meqv/pH/L]: Reagent

6.5b0 6.500 0.006 9.109

Ch. t l , pH 3.51 [mL] Ch. 112,pH 6.50 [mLl

0.2 M Immob., pKa 0.80 0 . 2 M Immob., pKa 3.58 0.2 M Immob., pKa 4.51 0 . 2 M Immob., pKa 6.21 0.2 M Immob., pKa 7.06 0.2 M Immob., pKa 9.59 1 M Tris 1 M H3P04 .1 M Tris-phosphate,pH 6.8

0.000

0.000

268.575

261.768

Total

450.t

400.000

56.282 34.344 4.367 64.105 6.278 0.000 16.050

1.200 21.696 38.622 7.734 41.832 25.686 1.462

[Print] key to get a hard copy; [Elxit; [Rlecipe for gel solution.

Figure 9. Final recipe for 3 times concentrated Immobiline stock solutions adjusted to p H 6.8.

146

Electrophoresis 1990,11, 140-147

K. Altland

IPGMAKER: Recipe (All gel constituents) Final Conditons: Total acrylamide, T*10 [g/Ll: 50.00 BIS c [% w/w] : 3.00 Glycerol in chamber #1 [g/L]: 200.00 Glycerol in chamber #2 [g/Ll : 0.00 TEMED [mmol/Ll : 1.00 NH4-persulf ate [mmol/L] : 1.50 Volume in chamber #l [mLl : 8.00 : 7.00 Volume i n chamber 112 [mLl Chamber # 1

Reagent

Chamber # 2

2.000 Acrylamide solution [mLl (194.00 g/L Acrylamide; 6.00 g/L BIS) Glycerol (87 %) [mLl 1.495 0.001 TEMED [mL] NH4-persulfate (100 g/L) [mL] : 0.027 Water [mL] 1.810 2.667 Immobiline stock solution [mLI :

0.000 0.001 0.024 2.892 2.333

Total volume [mLl

7.000

1.750

8.000

Figure 10.Recipe for gel solutions to pour an IPG gel assuming that one thirdofthe gel volume is taken from the Immobiline stock solutions made according to the recipe shown by the preceding figure.

Use [Print] key to get a hard copy or [Elxit IPGrmKEFi: Test a Recipe beta pKa

pH

12.00 II

--< I .

Ch.#l

Ch.12

0.8 16.464

0.000

3.6

0.433

0,000

4.5

5.448

0.000

6.2

8.270

0.000

7.1

0.000

7.154

8.5

1,850

5,750

9.6

2 202

0.000

10.3

3.272

4.292

m

-2.00 4 0

Figure 1 I. Screen comparable with that of 10

20 ............. beta;

30

40

50

60

70

80

90

100 2 P a r t i a l Uolune (PU)

PH

[ E l x i t , [Tlest, change [Bleta ? Meancbeta) [nequ/pH/Ll: 3.000 For PU = 0,000 1 or nn i n a 10 en widie gel SD(pH) [units] : 0.032 pH i s 2.992 pH-units slope is 0.741 pH/cn ( i n a 10 cn wide gel) beta i s 3.118 neau/pH/L with 2.345 nesu/aH/L fron water

gradient was 0.072 pH units and the standard deviation was 0.032 pH units, corresponding to 0.9 and 0.4 mm, respectively, in a 10 cm wide gel. This appears to be acceptable and far better than what can be achieved by gradients from free carrier ampholytes. One should notice, however, from the given test data at the zero end ofthe gradient and Eq. 1(5), that most ofthe buffering capacity (i.e. 2.345 from a total of 3.118 meqv/pH/ L) is not immobilized. A similar situation is expected at the other end of the range. Further dilution of this recipe for any reason will destroy gradient stability near the zero end by lack of stability and both dilution and concentration of Immobilines will affect the pH level near both ends of the gradient due to the buffering effect of free ions from water. From the same point of view one should consider that incomplete immobiliza-

Fig. 7 after a trial to create a linear gradient of pH 3- 11. Notice the high contribution of water ions to buffering capacity at the lower end of the gradient, which is similar to the contribution at the upper end. This gradient recipe was achieved after ca. 65 min of calculation with the PC of the author.

tion of Immobilines during the polymerization reaction will result in a similar effect. Thus, readjustment of pH may become necessary whenever Immobiline concentrations are changed and the pH gradients include levels ofpH < 4.5 and > 9.5. The presented program can easily solve this problem.

3 Conclusions The presented program for creating and testing recipes for IPGs appears to be avaluable tool for all using these gradients. Acceptable new recipes can be achieved for narrow ranges within a few minutes and for wide ranges whithin less than 2 h of calculation, provided that the PC belongs to a fast genera-

Immobiline electrofocusing of acid phosphatascs

Eleclrophoresis 1990, 11, 147-151

tion. Trials to generate gradients including ranges with pH < 4 and > 10 require access to additional Immobilines such as those with pKa 0.8 and 10.3. Note: Copies ofthe program will become available from Pharmacia-LKB, Box 305, S-16126 Bromma, Sweden.

I appreciate useful discussions with Dr. Bengt Bjellqvist, Pharmacia-LKB, Bromma on details increasing the efficiency of the algorithms. Received November 29, 1989

Aleksandra Kubicz' Agata Szalewicz' John S. Fawcett' Andreas Chrambach* 'Institute of Biochemistry, University of Wroclaw 'Section on MacromolecularAnalysis, Laboratory of Theoretical and Physical Biology, National Institute of Child Health and Human Development, National Institutes of Health, Bethesda, MD

147

4 References I l l Bjellqvist, B., Ek, K., Righetti, P. G., Gianazza, E., Gorg, A., Westermeier,R. andPostel, W.,J.Biochern. Biophys.Methods 1982,6, 317-339. 121 Gianazza, E., Celentano, F., Dossi, G., Bjellqvist, B. and Righetti, P. G., Electrophoresis 1984,5, 88-97. 131 Altland, K. and Altland, A., Electrophoresis 1984, 5 , 143- 147. [41 Altland, K. and Altland, A,, Clin. Chern. 1984,30, 2098-2103. 151 Gianazza, E., Artoni, G. and Righetti, P. G., Electrophoresis 1983,4, 321-326. 161 Gianazza, E., Dossi, G., Celentano, F. and Righetti, P. G.,J.Biochern. Biophys. Methods 1983,8, 109-133.

Electrofocusing of acid phosphatases from frog liver, using an immobilized pH gradient Isoelectric focusing on carrier ampholyte-containing immobilized pH gradient gels was applied (i) to gels submerged in silicone oil on a Peltier cooled apparatus, (ii) to the separation of the higher molecular weight (HMW, M , 140 000) and the lower molecular weight (LMW, M , 38 000) acid phosphatases (AcPases) from frog livers. (i) Electrofocusing was conducted on gels submerged under silicone oil cooled and stirred on a Peltier-thermoregulated horizontal gel support plate. This procedure aimed at a) improving the temperature control of the gel by direct contact of coolant with the gel surface, and thus at being able to focus at the maximal field strength and consequently highest resolution; b) preventing evaporation from the gel and c) excluding atmospheric carbon dioxide. Silicone oil submersion did not abolish water loss from the gel into the electrolyte strips during isoelectric focusing, or a rippled gel surface. Absence of water exudation on the ripples noted previously by Altland 1 11 was observed. (ii)The electrofocusing of AcPases on immobilized pH gradients yielded patterns which remained stationary as a function of time, by contrast to previous analyses on carrier ampholyte generated pH gradients. The total number of enzymatically active components found in the enzyme preparations from different stages of purification and in the isolated HMW and LMW AcPases was 18. The HMW and LMW AcPases focused in characteristic pH ranges and exhibited qualitative and quantitative pattern differences. Their band patterns add up to that of a crude preparation containing both enzymes. Neither polyacrylamide gel electrophoresis (PAGE) at any nondenaturing pH, nor isoelectric focusing in carrier ampholytes with pattern changes due to the pH gradient drift were. able to yield that result.

1 Introduction The great leap forward in isoelectric focusing (IEF) provided by immobilized pH gradients [2] was that electrofocusingpatterns were stabilized and therefore significant and reproducible. Nonetheless, this method, like every other, had its problems. One of these was water redistribution and exudation Correspondence: Dr. A. Chrambach, Bldg. 10, Rm. 6C101, NIH, Bethesda MD 20892, USA Abbreviations: AcPase, acid phosohatase; HMW, higher molecular weight; ICAPG-EF, isoelectric focusing on carrier ampholyte containing immobilized pH gradients; IEF, isoelectric focusing; LMW, lower molecular weight; PAGE, polyacrylamide gel electrophoresis

0VCH Verlagsgesellschaft mbH, D-6940 Weinheim, 1990

when the method was applied in the presence of carrier ampholytes [3, 41. Visually, the water redistribution could be observed as ripples on the gel surface; quantitatively, it was determined by weight changes across the pH gradient gel [51. Another problem encountered in common with other openfaced gel slabs was progressivegel dehydration, which was not satisfactorily controlled in spite of the passage of watersaturated argon across the gel [3]. To circumvent these problems, advantage was taken of an innovation introduced by Altland 161, viz. the operation of isoelectric focusing on carrier ampholyte containing immobilized pH gradients (ICAPG-EF) on gels submerged under paraffin oil. In one of Altland's applications of that method I 11, the oil layer appeared to suppress the formation of water droplets on the ripples induced by the presence of carrier ampholytes on 01 73-0835/90/0202-0147 %2.50/0

IPGMAKER: a program for IBM-compatible personal computers to create and test recipes for immobilized pH gradients.

The program "IPGMAKER" is a computational aid for creating and testing recipes for near-linear immobilized pH gradients. It was written for fast IBM p...
695KB Sizes 0 Downloads 0 Views