Westermeier R., Naven T., H?pker H.-R. Proteomics in Practice: A Guide to Successful Experimental Design

Подождите немного. Документ загружается.

1 Electrophoretic Techniques70

1.5.2.1 Lysine Labeling (Minimal Labeling)

It must be avoided to block all lysines of the proteins with a label,

because the proteins would become very hydrophobic and would not

stay in solution. Therefore only a small portion of the proteins is

labeled. In practice only 400 pmol/L of dye is added to 50 mg of pro-

tein. In this way only 3 to 5% of the proteins will receive a tag; the

other 95–97% proteins remain unlabeled. This measure, which is

called “minimal labeling”, prevents also the appearance of multiple

labels: the occurrence of multiple labels on proteins with several

lysines will statistically be so low that these molecules are below the

limit of detection. Overlabeling with minimal dyes could be recog-

nized very easily in the false color image (see above): Two and more

dye molecules per protein would cause vertical streaking due to

increased molecular weight of a part of the molecules of the labeled

protein.

It is, however, very important that these dyes are charge and size

matched. Therefore a basic buffering group is added to each of the

dyes, thus the isoelectric points are not changing. Also in the second

dimension the proteins migrate to the same point independently of

their label, because the molecular weights added are in the range

between 434 and 464 for the three dyes Cy2, Cy3, and Cy5. This little

difference cannot be resolved by SDS electrophoresis. The 2-D pat-

tern obtained with minimal labeling does not differ from a pattern of

post-stained 2-D gels of non-labeled proteins.

The sensitivity of minimal labeling is comparable to good quality

silver staining, however exhibiting a much wider dynamic range and

linearity. It is very rare that a protein does not pick up a label or does

not contain any lysine. To be on the safe side it is recommended to

stain a gel with a sensitive fluorescent dye to detect such proteins.

In order to make the method reliable and reproducible, the labeling

has to be performed under optimum conditions and needs to be well

controlled (see Figure 1.19). The proteins are solubilized and dena-

tured with high molar urea or thiourea/urea. The optimal concentra-

tion of the protein lysate is between 5 and 10 mg/mL. But also sam-

ples containing from 1 mg/mL to 20 mg/mL have been successfully

labeled. If the sample is too diluted, labeling will be less efficient.

Lysine labeling is performed on ice and in absence of IPG buffers or

carrier ampholytes and reductants: The primary amino groups of

amphoteric buffers would be labeled as well; the reductant would

interfere with the reaction. These compounds are added after labeling

had been completed. A sample pH value above pH 8.0 is essential,

the optimum lies at pH 8.5. Therefore the sample solution should

always contain 30 mmol/L Tris-base. Samples, which have been

In downstream analysis the

non-labeled peptides are in a

great majority versus the

labeled ones, therefore only the

non-modified peptides will be

detected with mass spectro-

metry.

In the lower molecular weight

area of the SDS gel an offset of

non-labeled from labeled

protein spots can be observed.

Therefore, when spot picking is

intended, it is recommended to

stain the gel after scanning with

a fluorescent dye and match

the spots with the labeled spots

using the spot matching algo-

rithm of the evaluation soft-

ware.

Underlabeling is not an issue: if

more protein is labeled with the

same amount of dye, the

scanned image will be the

same. However, when proteins

are overlabeled, multiple labels

can become visible, particular

in the cathodal area of the 2-D

gel.

The distribution of dyes is

uniformly and quantitative

across all the different proteins.

1.5 Two-dimensional Electrophoresis 71

extracted with TCA acetone or have been cleaned up with precipita-

tion, are often very acidic. In this case it can be necessary to adjust

the pH value by adding 50 mmol/L NaOH solution. The reaction is

carried out on ice for 30 minutes and stopped by adding a 0.1% (w/v)

lysine solution.

Fig. 1.19: Simplified diagram of lysine labeling for DIGE.

No derivatization is needed, because the e-amino group of

the lysine is readily accessible.

Subsequently the samples are combined and either submitted

directly to 2-D electrophoresis or stored in a deep freezer until use.

1.5.2.2 Cysteine Labeling (Saturation Labeling)

In contrast to lysine labeling here all available cysteines will be

tagged. Labeling the cysteines is pH neutral and does not increase

the hydrophobicity. However the resulting 2-D electrophoresis pat-

tern looks different from those originating from minimal labeled or

non-labeled proteins: proteins with many cysteines will pick up many

labels, which results in slower migration in the SDS gel and a much

stronger light emission signal. The increase of molecular weight is

650 Da per label. Because of multiple labels, here it is particularly

important that the dyes are size matched. Not all proteins contain

cysteine. It is estimated that in human, animal, and plant proteomes

about 95% of the proteins possess at least one cysteine, and can there-

fore be detected with this method.

The sensitivity of this method is very high. The application of down

to 1,000 human cells (equal to 2.5 mg protein) provides still very good

2-D gels, as shown in a paper by Sitek et al. (2005).

Cysteine labeling requires more work than lysine labeling. First the

disulfide bridges are cleaved with the reductant TCEP, thereafter the

However, there are bacteria

species, which do not contain a

single cysteine. Needless to say,

the method cannot be applied

for those.

Sitek B, Lttges J, Marcus K,

Klçppel G, Schmiegel W, Meyer

HE, Hahn SA, Sthler K.

Proteomics 5 (2005) 2665–

2679

1 Electrophoretic Techniques72

dye with the maleimide reaction group is added (see Figure 1.20).

The temperature is higher and the reaction time is longer than for

minimal labeling in order to achieve complete labeling of all

cysteines. To avoid bleaching of the dyes these steps are therefore per-

formed in the dark. Furthermore, some optimization for certain sam-

ple types may be necessary, because the quantitative amount of

cysteines in the protein mixture cannot be predicted. If more dye had

been applied than cysteines were available, non-specific labeling of

some lysines can be observed, indicated by a pI shift of some spots

towards the acidic side. If there was not enough dye for all the

cysteines, some proteins will partly have lower molecular sizes and

will migrate a little faster, resulting in vertically elongated spots or

vertical streaks. The optimal concentrations can be assessed by a

same–same experiment with different label concentrations (see also

Sitek et al. 2005). The applied ratio of reductant to dye is constant.

For this method only the two dyes Cy3 and Cy5 can be used,

because Cy2 does not work properly for this application.

Fig. 1.20: Simplified diagram of cysteine labeling for DIGE.

The more cysteinyls are available after reduction, the more dye

molecules will become attached to a protein molecule.

This technique is mainly

applied for very scarce samples.

Here is one of the major differ-

ences between minimal

labeling and cysteine labeling:

overlabeling causes vertical

streaks in minimal labeling, but

horizontal displacement in

cysteine labeling.

Cy2 shows too strong self-

quenching effects, when dye

molecules are in close proximity

to each other.

1.5 Two-dimensional Electrophoresis 73

For downstream analysis the spots could in principle be picked

directly from the image file (see spot picking below). However,

because the method is so sensitive, there will not be enough protein

material in a spot for mass spectrometry analysis. Therefore a pre-

parative 2-D gel of a reference proteome loaded containing about

300 mg protein has to be run. Also for this gel the cysteines must

carry a label; otherwise the spots to be excised cannot be matched

with the spots in the analytical gel. The selection of the optimal refer-

ence proteome, which should show the most similarities with the

analytical sample, is dependent on the type of sample. Labeling of the

preparative sample proteins is performed with Cy3.

1.5.2.3 The Internal Standard

The multiplex feature of DIGE offers the unique possibility to reserve

one channel for an internal standard. Some papers are describing the

application of a mix of labeled commercial standard proteins. How-

ever, by far the greatest effect is obtained, when the internal standard

is a pool of aliquots taken from each sample of an experiment, and

labeling this mixture with one of the dyes. This concept is shown in a

simplified schematic diagram in Figure 1.21. Of course, “control” and

“treated” can be replaced by “healthy” and “diseased” or by “wild type”

and “mutant”.

Fig. 1.21: Simplified schematic representation showing how to

create a pooled internal standard in difference gel electrophoresis

(DIGE). The sample proteins are labeled with Cy3 and Cy5, the

internal standard with Cy2. From Westermeier and Scheibe (2007).

Because in this case all

cysteine-containing proteins are

modified, the molecular weight

added to the cysteines needs to

be entered into the database

search definitions, analogous to

alkylation of cysteines.

Westermeier R, Scheibe B. In

Posch A. Ed. Sample Prepara-

tion for 2D PAGE. Methods in

Molecular Biology, Humana

Press, Totowa, NJ (2007) in

press.

1 Electrophoretic Techniques74

In reality it makes no sense to run only two samples in an experi-

ment. As already mentioned in the introduction, in order to find sig-

nificant protein changes with acceptable statistical confidence, at least

three biological replicates are required. This means that the internal

standard will be composed of aliquots from six samples: three con-

trols plus three treated. It is important to create enough pooled inter-

nal standard to be able to apply it on every gel. With minimal labeling

each gel contains two samples and one standard, with saturation

labeling only one sample and the standard are run together.

Because the proteins are co-migrating in the gel, no spot matching

is needed within the gel. The sample spot volumes can directly be

compared to the spot volumes of the standard, which is the mixture

of all samples. The dedicated software, DeCyder, does this automati-

cally, employing a co-detection algorithm. The result is expressed in

spot volume ratios. For the comparison of several gels the spot match-

ing procedure is highly facilitated, because the images of the standard

are matched, which originate from the identical protein composition.

In this way it is possible to perform a fully automated evaluation of a

high number of images. Furthermore image analysis can be carried

out fully automatically with a batch processor, which reduces the user

bias to a minimum.

& Note: the result issued by DeCyder consists of

spot volume ratios to the internal standard

instead of absolute spot volumes. In this way

artifactual spot volume variations caused by gel-

to-gel variations are eliminated.

The use of the pooled internal standard makes the evaluation of the

results much easier and considerably more reliable:

Every protein in the population must appear on

the standard image in each gel, otherwise it will

not be included in the result.

The sample spot volumes are normalized against

the internal standard.

Variations of spot volumes due to gel-to-gel varia-

tions are eliminated.

Each sample is compared internally with the

same standard.

It enables fully automated accurate spot statis-

tics.

Gel-to-gel matching becomes much easier.

When minimal labeling is

employed, the standard is

always labeled with Cy2, for

saturation labeling it is Cy3.

All this will be described in

more detail in the chapter on

image analysis.

Artifactual spot volume varia-

tion can be caused by slight

different conditions during

sample application.

1.5 Two-dimensional Electrophoresis 75

1.5.2.4 Experimental Design

First of all, the pooled internal standard should be labeled with one

and the same dye within an experiment. For spot matching it is

important that the conditions for the internal standard are identical.

In order to obtain statistically significant results, the distribution of

the labels among samples has to be well planned. For instance, any

preferred labeling of certain proteins by one of the dyes must be sta-

tistically excluded. If the experiment consists only of a small number

of samples, it is recommended to run the samples two times, but

with inverse labeling. For larger experiments, no replicates are

required, but a “planned” randomization of the labels and sample

application is proposed:

The two available dyes should be evenly distribu-

ted between treated sample and control.

Treated sample and control from separate origins

should be applied to each gel.

An example for randomization of sample application and inverse

labeling is shown in Table 3.1 on page 309.

To minimize methodical variation all samples of an experiment

should be run together in one instrument in the first and also the sec-

ond dimension.

Typically the amount of three times 50 mg protein is applied, which

means 150 mg total sample load. Because the detection is very sensi-

tive, it is possible that a small spot does not contain enough protein

for mass spectrometry analysis. If more protein is required in a gel, it

is suggested to spike non-labeled pooled standard into the 150 mg

labeled sample to the desired protein amount.

It should also be mentioned that the DIGE technique is very useful

for checking the quality of sample preparation:

First the same sample is labeled with two differ-

ent dyes and separated in one gel;

In a second step three sample preparation repli-

cates of the same sample source labeled with

three different dyes are run in a gel.

1.5.2.5 Major Benefits of 2-D DIGE

In short the advantages of the DIGE approach over conventional 2-D

electrophoresis are ranked in the order of their importance to prac-

tice:

It saves time: differentially expressed proteins

are detected very rapidly.

For 3-color experiments the

standard is labeled with Cy, for

2-color experiments with Cy3.

The statistics algorithms of the

DeCyder software take these

measures in calculation.

Experiments with more than 12

gels at a time are more difficult

to evaluate.

The original suggestion to run a

separate preparative gel is less

practical.

Some samples contain protein

individuals with peculiar beha-

vior. Those are detected with

this strategy.

These points have been ranked

according the comments of an

adequate number of DIGE

users.

No gel replicates, automated

image analysis.

1 Electrophoretic Techniques76

It saves work: a smaller number of gels has to be

run, because samples are combined on gels, and

no gel replicates are necessary.

It eases handling of gels: the gels are analyzed

while they are still in the cassettes.

It simplifies the evaluation: image analysis with

DeCyder software runs automatically and is sim-

ple to carry out.

It improves the quality of the results: gel-to-gel

variations are eliminated from the result because

it measures spot volume ratios relative to the

pooled internal standard, which is included in

each gel.

It delivers quantitative values: the internal stan-

dard and the wide linear dynamic range of the

fluorescent dyes allow the quantitative measure-

ment of minor changes of expression levels with

high statistical confidence.

It affords very sensitive detection: with cysteine

labeling samples containing less than 1 mg total

protein can be analyzed with 2-D electrophor-

esis.

The DIGE concept is not limited to high resolution 2-D electrophor-

esis with immobilized pH gradients: it has also been successfully

applied on 2-D gels based on carrier ampholytes IEF (Sitek et al.

2006), blue native electrophoresis (Perales et al. 2005), and acidic elec-

trophoresis with cationic detergent (Helling et al. 2006).

1.5.2.6 Specific Labeling of Cell Surface Proteins

Usually the proteins of cell or tissue lysates are labeled. It is, however,

possible to label the proteins on the surface of the intact cell (Mayr-

hofer et al. 2006). In a GE Healthcare Application Note (2005) it is

shown that after cell lysis and 2-D electrophoresis only the specifically

labeled surface proteins are visualized. The proteins inside the cell do

not become labeled, because of unfavorable conditions: the pH value

in the cytosol of an intact cell is usually below pH 7.4; the labeling

time is short and performed on ice. This approach is much easier to

carry out than using the standard protocol of biotin labeling. As

shown in Figure 1.22, the differential comparison of cell surface pro-

teins and whole cell lysate labeled with the second dye in the same

gel, is a very straightforward way to identify outer membrane proteins

against the background of the total proteins.

This reduces also costs for

consumables, equipment, and

labor.

Gels do not swell, shrink, or

break.

Image analysis is the bottle

neck in conventional 2-D elec-

trophoresis

In contrast replicate conven-

tional gels do not eliminate

these variations, they just

average them.

Note that those are still relative

quantitative results.

Sitek B, Scheibe B, Jung K,

Schramm A, Sthler K. In:

Proteomics in Drug Research

(M Hamacher et al. Eds.)

Wiley-VCH, Weinheim (2006)

pp 33–55.

Perales M, Eubel H, Heine-

meyer J, Colaneri A, Zabaleta

E, Braun H-P. J Mol Biol 350

(2005) 263–277.

Helling S, Schmitt E, Joppich

C, Schulenborg T, Mllner S,

Felske-Mller S, Wiebringhaus

T, Becker G, Linsenmann G,

Sitek B, Lutter P, Meyer HE,

Marcus K. Proteomics 6 (2006)

4506–4513.

Mayrhofer C, Krieger S,

Allmaier G, Kerjaschki D.

Proteomics 6 (2006) 579–585.

GE Healthcare Application

Note: Selective labeling of cell-

surface proteins using CyDye

DIGE Fluor minimal dyes.

(2005) 11-0033-92.

1.5 Two-dimensional Electrophoresis 77

Fig. 1.22: False color representation of 2-D gel images of a

CHO-K1 Cy5 cell-surface labeled sample (green spots) and a

Cy3 sample (red spots) labeled according to standard Ettan

DIGE protocol run in the same 2-D gel. From GE Healthcare

Application Note 11-0033-92.

& Note: The multiplex approach of DIGE has some

additional benefits for the conditions during

isoelectric focusing: the conductivity differences

between the strips are lower, because several

samples are mixed together. This leads to more

reproducible spot pattern across the gels.

1.5.3

First Dimension: Isoelectric Focusing in IPG Strips

IPG strips are available with many different lengths and gradients.

Optimally the resolution of the two-dimensional gel should be suffi-

cient to get every protein separated from each other. If a protein spot

contains more than one protein, the proteins can still be identified

and characterized with mass spectrometry, however quantitative

assessments are not possible. Therefore large gel formats are pre-

ferred. If the resolution of such a gel is insufficient, the sample can

be applied on several IPG strips with narrow, overlapping pH gradi-

ents. Most frequently wide gradients with pH 3 to 10 or 3 to 11 are

run. When higher resolution is required, a combination of an acidic

gradient pH 4 to 7 (or 3 to 7) and a basic gel pH 7 to 11 are employed.

IPG strips with printed serial numbers are very useful for tracing the

samples through the 2-D electrophoresis procedure.

When detection is performed

with western blotting, small

formats are preferred.

1 Electrophoretic Techniques78

For proteomics studies large formats with 18 or 24 cm strip lengths

are preferred because of several reasons:

To reach sufficient resolution.

To be able to load higher protein concentrations

for detection of as many spots as possible and

subsequent mass spectrometry analysis of spots.

To obtain sufficient evaluative gel area.

1.5.3.1 Rehydration

The dry IPG strips need to be reconstituted to their original thickness

of 0.5 mm with rehydration solution before use. There are two possi-

bilities:

Pre-rehydration of the strip without sample; the

sample is loaded either on the anodal or cathodal

end of the IPG strip with a cup or paper bridge.

Rehydration with the sample mixed into the

rehydration solution.

& Composition of the standard “rehydration

solution”:

8 mol/L urea, 0.5% (w/v) CHAPS,

0.2% (w/v) DTT, 1.25% (v/v) carrier ampholytes,

10% (v/v) glycerol, 0.002% Bromophenol blue.

Glycerol reduces electroendosmotic effects, pre-

vents drying of the gel, and urea crystallization.

It is not always necessary.

The concentrations of the additives are lower

than in the lysis buffer, in order to reduce

unwanted side effects like:

– Crystallization of urea;

– Instability of the patterns due to conflicts of

the buffering groups, caused by overloading

with carrier ampholytes;

– Formation of micelles between the different

detergents in the second dimension, which

can cause dark and dirty background with

some of the staining techniques (Gçrg et al.

1987a);

Bromophenol blue allows a good control of the

liquid distribution during rehydration.

When a sample has been extracted with urea and thiourea, the rehy-

dration solution must also contain thiourea.

Ideally one protein/spot.

Abundant proteins need more

space.

The smaller the gel, the higher

the fringe effect.

This solution is suitable for

samples without thiourea.

For electroendosmosis see

page 21 f or glossary.