t l c ~ r i u p h o i e ~ i1992. s 13. 129-132
2-DI: database o f coiiirnon human proteins
729
Dorothe Burggraf Kerstin Andersson Christoph Eckerskorn Friedrich Lottspeich
Towards a two-dimensional database of common human proteins
Max-Planck-Institut fiir Biochemie, Martinsried
A protein pattern of common human proteins was constructed by comparing the two-dimensional gel electrophoresis (2-DE) protein patterns from five cell lines of different germ layers. Total cell lysate and the isolated and purified nuclei of each cell line were separated by parallel electrophoresis runs in a multiple casting system of highest reproducibility. The computerized image analysis of the digitized 2-DE gels revealed a master protein pattern for each cell line. By comparison of all master protein patterns a 2-DE protein map of only common human proteins was constructed as a basis for a new 2-DE database. In a first step we have started characterizing a number of spots by microsequencing, amino acid composition analysis, and mass spectroscopy.
Current estimates indicate that there are no more than 5000 different expressed proteins per cell [l].To date, only a small portion of the total set of proteins from humans has been characterized. Despite the successful strategy to separate proteins by the most powerful separation method available, high resolution two-dimensional gel electrophoresis (2-DE), little is known about the protein composition of different cell types. Great efforts have been made to catalog the completely expressed proteins by 2-DE. Based on 2-DE master gels, Leftkovits et al. [2] presented a proteinpaedia correlated to a gene catalogue. The Quest system generated by Garrels e t a / . [3]is based on a network structure of matchsets. Other laboratories have started the approach of a protein database by 2-DE analysis. Celis and his co-workers [4] presented master gels from single cell lines, Neidhard has placed the names of 200 E. coli proteins on a reference gel map [5], Mc Laughlin et a/. [6] used a similar strategy to identify yeast proteins by 2-DE maps and Anderson eta/.[7] identified many proteins of the human serum on 2-DE gels. Proteins common to all human cells have not yet been described and characterized by aid of a 2-DE protein pattern. It seems necessary to map these proteins and to characterize them proteinchemically for a survey of the basic set of proteins of a human cell. These universal proteins should be responsible for important household processes, for structural functions and for regulation. A protein database of common proteins derived from high resolution 2-DE should facilitate studies of genome organization and function and complement the human genome sequencing project by identifying polypeptides. Here we describe the basis for creating a 2-DE database of common human cell proteins. Additional maps of subcellular fractions should complete this 2-DE database. A map of common nuclear proteins, to reveal those proteins responsible for the regulation and control of important cellular processes, has been added. The identification and characterization of common proteins was started by methods of microsequencing, amino acid composition analysis, and mass spectroscopy.
fetal calf serum at 37°C with 5% CO,. Cells grown at 90% confluence were rinsed with cold phosphate buffered saline (PBS) and scraped into a buffer containing 250 mM sucrose, 20 mM N(2-hydroxyethyl) piperazine-N'-(2-ethanesulfonic acid) (HEPES), 1 mM MgCl,, 50 mM P-mercapto-
-
IEF
i
i
The cell lines, from American Type Culture Collection, were grown in DMEM or MEM supplemented with 10% a s s i s t a n t gels
Correspondence: Dorothe Burggraf, Max-Planck-Institut fur Biochemie, Am Klopferspitz 18a, DW-8033 Martinsried, Germany Abbreviation: 2-DE, two-dimensional gel electrophoresis
0VCH Verlagsgesellschaft mbH, D-6940 Weinheim,
1992
Figure /. Strategy for creating a master image with a basic image and 4 assistant gels representing different electrophoretic runs, sample preparations and cell culture conditions. These are matched into the basic image. O173-0835/92/09 10.0729 $3.50+.25/0
730
Elrcfr-ophorr\rs 1992. 13. 729-732
D. Burggraf ~t o /
ethanol and protease inhibitors. Crude nuclei were obtained by pressure homogenization [8].The suspension was layered onto a 1.1 M sucrose cushion containing 20 mM H E P E S and 1 mM MgC1, and centrifuged at 100 000 g for 1 h . The nuclei thus obtained were morphologically satisfactory by phase-contrast microscopic analysis. Plasma membrane contaniination was measured by testing the enzyme activity of the S'nucleotidase [9]. It is lower than 8 %. Samples of the total cellular proteins were prepared from cultured cells that had been rinsed thoroughly with PBS and pelleted. The pellet was lysed in 9 M urea and 2"/0 Nonidet P-40 (NP-40) [ lO].The same procedure was used forthe nuclear pellet. After sonification (10 X 15 s) and centrifugation, P-mercaptoethanol (5 O/o) and Ampholine carrier anipholytes (0.80/~)were added and the samples applied to isoelectric focusing (IEF), which was done according to Gorg et al. [ l l ] in immobilized pH gradients (IPG). The individual I P G strips were equilibrated in 50 mM Tris-HC1 buffer, pH 6.8, containing 30% glycerol, 6 M urea, 3% sodium dodecyl sulfate (SDS) and 50 mM dithiothreitol (DTT).The gel strips were transferred onto 12% SDS polyacrylamide gels, prepared in a multicasting and parallel running system (Investigator, Millipore) for separation in the second dimension [12].The dimension ofthe gels is 11 X 20 cm. Typically 20 pL (100 pg) of total cell lysates and 100 pL (400 pg) of nuclear proteins were applied to each gel of the pH interval 4-7. The gels were silver-stained according to [ 131. Image proce'ssing is performed by Bioimage software (Millipore). Scanning is done with a 1024 X 1024 pixel C C D camera (Kodak). In this report we describe 2-DE maps of common human proteins from 6 different tissues and of the 3 germ layers (Table 1).
Table 1. Cell lines for the map of common human proteins Tissues
Germ layer
Cell line
Blood Muscle Connective tissue Nervous tissue Epithelia Organ suecific (liver)
Mesoderm Mesoderm Mesoderm Ectoderni Endoderm Endoderm
Jurkat HISM Ht1080 1MR32 HeLa HeuG2
The real key towards a master protein pattern which compares complete gels (with ca. 1400 spots) is the realization of a standardized and reproducible 2-DE technique, so that complex patterns can be evaluated qualitativelyand quantitatively. This includes (i) a precisely defined protocol, apparatus, and chemicals, (ii) standardized sample preparation and application and (iii) defined running conditions and staining. Computerized image analysis of 2-DE patterns, a prerequisite for protein databasing, is only feasible if polypeptide patterns are consistent from experiment to experiment. Having established a standard protocol we now obtain highly reproducible 2-DE protein patterns. The molecular weights of polypeptides differed by an average of 4.5 %, which corresponds to approximately a 1.8 kDa difference between gels. The p l differed by an average of 2%, which corresponds to approximately 0.06 pH unit difference between gels. This represents a high level of reproducibility. Figure 1 shows the strategy used to create a master protein pattern of each single cell type. Five gels from different electrophoresis runs were chosen to eliminate minimal experimental variations in cell culture conditions, sample preparation, and electrophoretic conditions. Anyone of these 5 is elected as the basic gel and the 4 others, the assistant gels, are matched with it. The achieved master image is the
IEF
IEF
7P 97-
5p
.
6,0
70
97-
45-
31-
21-1
2ld
14-
14-
master gel of one cell line (Ht1080)
map of common
human proteins
(from 5 cell lines)
Figure2. Masier prolein pattern o f one cell line (Ht1080) leading to the map of common human proteins in the pH interval from 4-7 with molecular masses between 15 and 100 kDa.
-
Elcwrophoresis 1992, 13, 129-132
I ) IEF
2-DE database of common human proteins
731
IEF
laster gel of nuclear p t e i m
~p
of c o m n human nuclear proteins (from 5 cell lines)
(Ht1080)
F f g u r p 3. Master protein pattern of nuclear proteins of one cell line (Ht1080) leading to the map of common nuclear proteins.
same when another one of the 5 gels served as basic image. The long-term reproducibility has been proved by matching gels made half a year later. Here we present our initial data, the common protein pattern of human cells in the pH interval from 4-7 (Fig. 2). Completing this effort in developing a comprehensive database for the human cell, we are nowfollowing two additional strategies: (i) using the method of the overlapping pH ranges, the map of the common proteins will be expanded to pH intervals in the acidic and basic regions [14]. There will be a composite master image of the whole cell proteins. (ii) Further fractionation of the cell would make it possible to associate the proteins with cell organelles.The first step is a master image of common human nuclear proteins (Fig. 3). Comparing the map of common proteins, with 450 spots, to the single master gels of each cell line (HT1080: 923 spots, HepG2: 1012 spots, IMR 32: 897 spots, etc.), only 45-50% of the proteins have the same gel coordinates ( p l and Mr). We are able to separate ca. 520 nuclear proteins from different cells; about 280 are found in all master images ofthe nuclear proteins. Note that with this strategy only a subset of the cellular proteins (pH 4-7) is registered. In addition, based on the algorithm chosen to generate the map of common proteins, all those proteins with slightly differing pH, (because of posttranslational modifications or amino acid exchanges) are neither included in the spot list nor detected as a common protein.The universality of the proteins ofthe spot list from the map of common proteins needs to beverified protein-chemically. The o-phthaldialdehyde (OPA) amino acid composition analysis from blotted and Coomassie Blue-stained protein spots [15]confirms the universality of the proteins. For2 protein samples the result of the OPA analysis is given in Fig. 4. It is shown that the ratio of 5
SDOt
6
normed to Ala
1.50
0.90
0.60
0.30
0.00 Val
1
HepG2
lie
0 HtlOBO
Phe
Leu
Phe
Leu
IMR32
SDOt
8
normed to Ala
1.50
0.90 0.60 0.30
0.00
Val
Ile
Figure 4. Two examples for the verification of common proteins by
0-phthaldialdehyde amino acid composition analysis. The columns represent the ratios of the amino acids Val, Ile, Phe, Leu to Ala in the different cell lines.
732
H. Kiier and F. Lottspeich
C i e c r r o p h o r e m 1992, 13, 732-735
representative amino acids is the same in all these cells. We began to characterize the common proteins by amino acid sequence from siliconized glass fiber membranes, amino acid composition and molecular mass (published elsewhere).
In this report we have shown our initial data aimed at developing a common human protein database. With recent interest in the sequencing of the human genome, 2-DE in conjunction with protein microsequencing should play a vital role by confirming the expression of predicted gene products. Received August 10 1992
References [ l ] Celis, J. E., Leffers. H., Rasmussen, H. H., Madsen, P., Honore, B., Gessen, B., Dejgaard. K., Olsen, E., Retz, G . P., Lauridsen. J . B., B a s e , B., Andersen, A . H., Nelbum, E., Brandstrup, B., Celis, A,, Puype, M.. Danne. J. andvandekerckehove, J., Electrophoresis 1991. 12. 765-801.
Hannelore Klier Friedrich Lottspeich Max-Planck-Institute for Biochemistry, Martinsried
[2] Lefkovits, J., Kettmann, J . R. and Coleclough, C., / / ~ m u / r o E/ . ~ d q 1990, I ! , 157-162. [3] Garrels, J. I.. J . Biol. Chem. 1989, 264, 5269-5282. [4] Bravo, R. and Celis, J. E., Clin. Chem. 1982, 28, 766-781. IS] Neidhard, F. C.,Vaughn,V., Phillips,T. A. and Bloch, P. L., Microhio[. Rev. 1983, 47, 23 1-284. L6] Ludwig.J. R.,Foy,J.J.,Elliott,S.G.and McLaughlin,C. S.,Mol. Cell. Biol. 1982, 2, 117-126. 171 Anderson,L.andAnderson,N. G.. Proc. Nar/.Acud.Sci. USA 1977.24, 5421-5425. 181 Dowben, R. M., Gaffey, T. A. and Lynch, P. M.. FEBS letrers 1968,2, 1-8. 191 Kai, M., White, G . 1.and Hawthorne, J. N., Biorhem. J. 1966, 101, 320-332. (101 O’Farrell. P. H., J . Biol. Chem. 1975, 250, 4007-4021. [ l l ] Bjellqvist,B.,Righetti.P. G., Gianazza,E., Gorg,A., Westermeier, R. and Poslel,W., J. Biochem. Biophys. Methods 1982, 6, 317-339. F., [12] Patton,W.F.,Pluskal,M.G.,Skea,W.M.,Buecker,J.L.,Lopez,M. Zimmermann, R., Belange, L. M . and Hatch, P. D., BioTechniques, 1990, 8, 518-527. [I31 Heukeshoven, J. and Dernick, R., Elecrrophoresis 1988, 9, 53 1-546. [14] Kehl, M., Andersson, K., and Lottspeich, F., in: Radola, B. J., (Ed.), Electrophorese Forum Technische Universitit, Miinchen 1991, pp. 367-372. [15] Eckerskorn. C. and Lottspeich, F., Electrophoresis 198X, Y, 518-527.
Detection of the hypusine-containing protein (HP = eIF-SA) in crude yeast extracts by two-dimensional Western blots The hypusine-containing protein (HP) with its unique modification of a specific lysine residue resulting in the amino acid hypusine is highly conserved among all eukaryotes and is also found in Archaebacteria. Studies of the protein function in translational processes showed a stimulatory effect in the methionyl puromycin assay, but not in in vitro translation of native mRNA. It was therefore also designated as eJF-SA. To further investigate the role of H P in cellular metabolism, we purified the protein from Saccharoinyces cerevisiae and raised polyclonal antibodies in chicken. Immunoglobulin preparations from the eggs of the immunized hens were used for Western blot analysis of H P in crude yeast extracts. For those studies, the soluble protein fraction of the yeast was resolved on two-dimensional gels (first dimension: isoelectric focusing using an immobilized pH gradient (JPG), pH 4-7; second dimension: sodium dodecyl sulfate-polyacrylamide gel electrophoresis, 12 O/o T) and subsequently blotted onto Fluorotrans membrane. Anodic versus cathodic application o f t h e extracts of the JPG strips was compared.
The hypusine-containing protein (HP) is the only protein known to be covalently modified by polyamines.This modification of a specific lysine residue is formed in a two-step process resulting in K-(4-amino-2-hydroxybutyl)-lysine (= hypusine) [l]. Despite the broad knowlege of the bio-
chemical features of HP, little is known about its function. I n vitro assays utilizing the artificial methionyl puromycin
Corresondence: Hannelore Klier. Max-Planck-Institute for Biochemistry, Am Klopferspitz 18A. D-8033 Martinsried, Germany
system for measuring the first peptide bond formation show a two- to threefold stimulation upon addition of H P [2].This observation,plus the fact that the protein was copurified with rabbit reticulocyte factors [3], led to the designation eukaryotic initiation factor 5A (eIF-SA) [4]. Nevertheless, the function ofthis protein in intact cells remains to be elucidated.
Abbreviations: eIF-SA, eukaryotic initiation factor 5A; HP, hypusine-containing protein; IPG, immobilized pH gradient; km, kilovolthours; NP-40, Nonidet P-40; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis
In the budding yeast Saccharomyces cerevisiae? two genes (HYP1 =TJF51B=ANB1 and HYP2=T1F51A) coding for H P have been identified [5,6], which are reciprocally regu-
0VCH
Verlagsgcsellsc1i.ift m b H , D-6940 Weinheim, 1992
0 I73-0835/92/0910-0732 $3.50+.25/0
2-DI: database o f coiiirnon human proteins
729
Dorothe Burggraf Kerstin Andersson Christoph Eckerskorn Friedrich Lottspeich
Towards a two-dimensional database of common human proteins
Max-Planck-Institut fiir Biochemie, Martinsried
A protein pattern of common human proteins was constructed by comparing the two-dimensional gel electrophoresis (2-DE) protein patterns from five cell lines of different germ layers. Total cell lysate and the isolated and purified nuclei of each cell line were separated by parallel electrophoresis runs in a multiple casting system of highest reproducibility. The computerized image analysis of the digitized 2-DE gels revealed a master protein pattern for each cell line. By comparison of all master protein patterns a 2-DE protein map of only common human proteins was constructed as a basis for a new 2-DE database. In a first step we have started characterizing a number of spots by microsequencing, amino acid composition analysis, and mass spectroscopy.
Current estimates indicate that there are no more than 5000 different expressed proteins per cell [l].To date, only a small portion of the total set of proteins from humans has been characterized. Despite the successful strategy to separate proteins by the most powerful separation method available, high resolution two-dimensional gel electrophoresis (2-DE), little is known about the protein composition of different cell types. Great efforts have been made to catalog the completely expressed proteins by 2-DE. Based on 2-DE master gels, Leftkovits et al. [2] presented a proteinpaedia correlated to a gene catalogue. The Quest system generated by Garrels e t a / . [3]is based on a network structure of matchsets. Other laboratories have started the approach of a protein database by 2-DE analysis. Celis and his co-workers [4] presented master gels from single cell lines, Neidhard has placed the names of 200 E. coli proteins on a reference gel map [5], Mc Laughlin et a/. [6] used a similar strategy to identify yeast proteins by 2-DE maps and Anderson eta/.[7] identified many proteins of the human serum on 2-DE gels. Proteins common to all human cells have not yet been described and characterized by aid of a 2-DE protein pattern. It seems necessary to map these proteins and to characterize them proteinchemically for a survey of the basic set of proteins of a human cell. These universal proteins should be responsible for important household processes, for structural functions and for regulation. A protein database of common proteins derived from high resolution 2-DE should facilitate studies of genome organization and function and complement the human genome sequencing project by identifying polypeptides. Here we describe the basis for creating a 2-DE database of common human cell proteins. Additional maps of subcellular fractions should complete this 2-DE database. A map of common nuclear proteins, to reveal those proteins responsible for the regulation and control of important cellular processes, has been added. The identification and characterization of common proteins was started by methods of microsequencing, amino acid composition analysis, and mass spectroscopy.
fetal calf serum at 37°C with 5% CO,. Cells grown at 90% confluence were rinsed with cold phosphate buffered saline (PBS) and scraped into a buffer containing 250 mM sucrose, 20 mM N(2-hydroxyethyl) piperazine-N'-(2-ethanesulfonic acid) (HEPES), 1 mM MgCl,, 50 mM P-mercapto-
-
IEF
i
i
The cell lines, from American Type Culture Collection, were grown in DMEM or MEM supplemented with 10% a s s i s t a n t gels
Correspondence: Dorothe Burggraf, Max-Planck-Institut fur Biochemie, Am Klopferspitz 18a, DW-8033 Martinsried, Germany Abbreviation: 2-DE, two-dimensional gel electrophoresis
0VCH Verlagsgesellschaft mbH, D-6940 Weinheim,
1992
Figure /. Strategy for creating a master image with a basic image and 4 assistant gels representing different electrophoretic runs, sample preparations and cell culture conditions. These are matched into the basic image. O173-0835/92/09 10.0729 $3.50+.25/0
730
Elrcfr-ophorr\rs 1992. 13. 729-732
D. Burggraf ~t o /
ethanol and protease inhibitors. Crude nuclei were obtained by pressure homogenization [8].The suspension was layered onto a 1.1 M sucrose cushion containing 20 mM H E P E S and 1 mM MgC1, and centrifuged at 100 000 g for 1 h . The nuclei thus obtained were morphologically satisfactory by phase-contrast microscopic analysis. Plasma membrane contaniination was measured by testing the enzyme activity of the S'nucleotidase [9]. It is lower than 8 %. Samples of the total cellular proteins were prepared from cultured cells that had been rinsed thoroughly with PBS and pelleted. The pellet was lysed in 9 M urea and 2"/0 Nonidet P-40 (NP-40) [ lO].The same procedure was used forthe nuclear pellet. After sonification (10 X 15 s) and centrifugation, P-mercaptoethanol (5 O/o) and Ampholine carrier anipholytes (0.80/~)were added and the samples applied to isoelectric focusing (IEF), which was done according to Gorg et al. [ l l ] in immobilized pH gradients (IPG). The individual I P G strips were equilibrated in 50 mM Tris-HC1 buffer, pH 6.8, containing 30% glycerol, 6 M urea, 3% sodium dodecyl sulfate (SDS) and 50 mM dithiothreitol (DTT).The gel strips were transferred onto 12% SDS polyacrylamide gels, prepared in a multicasting and parallel running system (Investigator, Millipore) for separation in the second dimension [12].The dimension ofthe gels is 11 X 20 cm. Typically 20 pL (100 pg) of total cell lysates and 100 pL (400 pg) of nuclear proteins were applied to each gel of the pH interval 4-7. The gels were silver-stained according to [ 131. Image proce'ssing is performed by Bioimage software (Millipore). Scanning is done with a 1024 X 1024 pixel C C D camera (Kodak). In this report we describe 2-DE maps of common human proteins from 6 different tissues and of the 3 germ layers (Table 1).
Table 1. Cell lines for the map of common human proteins Tissues
Germ layer
Cell line
Blood Muscle Connective tissue Nervous tissue Epithelia Organ suecific (liver)
Mesoderm Mesoderm Mesoderm Ectoderni Endoderm Endoderm
Jurkat HISM Ht1080 1MR32 HeLa HeuG2
The real key towards a master protein pattern which compares complete gels (with ca. 1400 spots) is the realization of a standardized and reproducible 2-DE technique, so that complex patterns can be evaluated qualitativelyand quantitatively. This includes (i) a precisely defined protocol, apparatus, and chemicals, (ii) standardized sample preparation and application and (iii) defined running conditions and staining. Computerized image analysis of 2-DE patterns, a prerequisite for protein databasing, is only feasible if polypeptide patterns are consistent from experiment to experiment. Having established a standard protocol we now obtain highly reproducible 2-DE protein patterns. The molecular weights of polypeptides differed by an average of 4.5 %, which corresponds to approximately a 1.8 kDa difference between gels. The p l differed by an average of 2%, which corresponds to approximately 0.06 pH unit difference between gels. This represents a high level of reproducibility. Figure 1 shows the strategy used to create a master protein pattern of each single cell type. Five gels from different electrophoresis runs were chosen to eliminate minimal experimental variations in cell culture conditions, sample preparation, and electrophoretic conditions. Anyone of these 5 is elected as the basic gel and the 4 others, the assistant gels, are matched with it. The achieved master image is the
IEF
IEF
7P 97-
5p
.
6,0
70
97-
45-
31-
21-1
2ld
14-
14-
master gel of one cell line (Ht1080)
map of common
human proteins
(from 5 cell lines)
Figure2. Masier prolein pattern o f one cell line (Ht1080) leading to the map of common human proteins in the pH interval from 4-7 with molecular masses between 15 and 100 kDa.
-
Elcwrophoresis 1992, 13, 129-132
I ) IEF
2-DE database of common human proteins
731
IEF
laster gel of nuclear p t e i m
~p
of c o m n human nuclear proteins (from 5 cell lines)
(Ht1080)
F f g u r p 3. Master protein pattern of nuclear proteins of one cell line (Ht1080) leading to the map of common nuclear proteins.
same when another one of the 5 gels served as basic image. The long-term reproducibility has been proved by matching gels made half a year later. Here we present our initial data, the common protein pattern of human cells in the pH interval from 4-7 (Fig. 2). Completing this effort in developing a comprehensive database for the human cell, we are nowfollowing two additional strategies: (i) using the method of the overlapping pH ranges, the map of the common proteins will be expanded to pH intervals in the acidic and basic regions [14]. There will be a composite master image of the whole cell proteins. (ii) Further fractionation of the cell would make it possible to associate the proteins with cell organelles.The first step is a master image of common human nuclear proteins (Fig. 3). Comparing the map of common proteins, with 450 spots, to the single master gels of each cell line (HT1080: 923 spots, HepG2: 1012 spots, IMR 32: 897 spots, etc.), only 45-50% of the proteins have the same gel coordinates ( p l and Mr). We are able to separate ca. 520 nuclear proteins from different cells; about 280 are found in all master images ofthe nuclear proteins. Note that with this strategy only a subset of the cellular proteins (pH 4-7) is registered. In addition, based on the algorithm chosen to generate the map of common proteins, all those proteins with slightly differing pH, (because of posttranslational modifications or amino acid exchanges) are neither included in the spot list nor detected as a common protein.The universality of the proteins ofthe spot list from the map of common proteins needs to beverified protein-chemically. The o-phthaldialdehyde (OPA) amino acid composition analysis from blotted and Coomassie Blue-stained protein spots [15]confirms the universality of the proteins. For2 protein samples the result of the OPA analysis is given in Fig. 4. It is shown that the ratio of 5
SDOt
6
normed to Ala
1.50
0.90
0.60
0.30
0.00 Val
1
HepG2
lie
0 HtlOBO
Phe
Leu
Phe
Leu
IMR32
SDOt
8
normed to Ala
1.50
0.90 0.60 0.30
0.00
Val
Ile
Figure 4. Two examples for the verification of common proteins by
0-phthaldialdehyde amino acid composition analysis. The columns represent the ratios of the amino acids Val, Ile, Phe, Leu to Ala in the different cell lines.
732
H. Kiier and F. Lottspeich
C i e c r r o p h o r e m 1992, 13, 732-735
representative amino acids is the same in all these cells. We began to characterize the common proteins by amino acid sequence from siliconized glass fiber membranes, amino acid composition and molecular mass (published elsewhere).
In this report we have shown our initial data aimed at developing a common human protein database. With recent interest in the sequencing of the human genome, 2-DE in conjunction with protein microsequencing should play a vital role by confirming the expression of predicted gene products. Received August 10 1992
References [ l ] Celis, J. E., Leffers. H., Rasmussen, H. H., Madsen, P., Honore, B., Gessen, B., Dejgaard. K., Olsen, E., Retz, G . P., Lauridsen. J . B., B a s e , B., Andersen, A . H., Nelbum, E., Brandstrup, B., Celis, A,, Puype, M.. Danne. J. andvandekerckehove, J., Electrophoresis 1991. 12. 765-801.
Hannelore Klier Friedrich Lottspeich Max-Planck-Institute for Biochemistry, Martinsried
[2] Lefkovits, J., Kettmann, J . R. and Coleclough, C., / / ~ m u / r o E/ . ~ d q 1990, I ! , 157-162. [3] Garrels, J. I.. J . Biol. Chem. 1989, 264, 5269-5282. [4] Bravo, R. and Celis, J. E., Clin. Chem. 1982, 28, 766-781. IS] Neidhard, F. C.,Vaughn,V., Phillips,T. A. and Bloch, P. L., Microhio[. Rev. 1983, 47, 23 1-284. L6] Ludwig.J. R.,Foy,J.J.,Elliott,S.G.and McLaughlin,C. S.,Mol. Cell. Biol. 1982, 2, 117-126. 171 Anderson,L.andAnderson,N. G.. Proc. Nar/.Acud.Sci. USA 1977.24, 5421-5425. 181 Dowben, R. M., Gaffey, T. A. and Lynch, P. M.. FEBS letrers 1968,2, 1-8. 191 Kai, M., White, G . 1.and Hawthorne, J. N., Biorhem. J. 1966, 101, 320-332. (101 O’Farrell. P. H., J . Biol. Chem. 1975, 250, 4007-4021. [ l l ] Bjellqvist,B.,Righetti.P. G., Gianazza,E., Gorg,A., Westermeier, R. and Poslel,W., J. Biochem. Biophys. Methods 1982, 6, 317-339. F., [12] Patton,W.F.,Pluskal,M.G.,Skea,W.M.,Buecker,J.L.,Lopez,M. Zimmermann, R., Belange, L. M . and Hatch, P. D., BioTechniques, 1990, 8, 518-527. [I31 Heukeshoven, J. and Dernick, R., Elecrrophoresis 1988, 9, 53 1-546. [14] Kehl, M., Andersson, K., and Lottspeich, F., in: Radola, B. J., (Ed.), Electrophorese Forum Technische Universitit, Miinchen 1991, pp. 367-372. [15] Eckerskorn. C. and Lottspeich, F., Electrophoresis 198X, Y, 518-527.
Detection of the hypusine-containing protein (HP = eIF-SA) in crude yeast extracts by two-dimensional Western blots The hypusine-containing protein (HP) with its unique modification of a specific lysine residue resulting in the amino acid hypusine is highly conserved among all eukaryotes and is also found in Archaebacteria. Studies of the protein function in translational processes showed a stimulatory effect in the methionyl puromycin assay, but not in in vitro translation of native mRNA. It was therefore also designated as eJF-SA. To further investigate the role of H P in cellular metabolism, we purified the protein from Saccharoinyces cerevisiae and raised polyclonal antibodies in chicken. Immunoglobulin preparations from the eggs of the immunized hens were used for Western blot analysis of H P in crude yeast extracts. For those studies, the soluble protein fraction of the yeast was resolved on two-dimensional gels (first dimension: isoelectric focusing using an immobilized pH gradient (JPG), pH 4-7; second dimension: sodium dodecyl sulfate-polyacrylamide gel electrophoresis, 12 O/o T) and subsequently blotted onto Fluorotrans membrane. Anodic versus cathodic application o f t h e extracts of the JPG strips was compared.
The hypusine-containing protein (HP) is the only protein known to be covalently modified by polyamines.This modification of a specific lysine residue is formed in a two-step process resulting in K-(4-amino-2-hydroxybutyl)-lysine (= hypusine) [l]. Despite the broad knowlege of the bio-
chemical features of HP, little is known about its function. I n vitro assays utilizing the artificial methionyl puromycin
Corresondence: Hannelore Klier. Max-Planck-Institute for Biochemistry, Am Klopferspitz 18A. D-8033 Martinsried, Germany
system for measuring the first peptide bond formation show a two- to threefold stimulation upon addition of H P [2].This observation,plus the fact that the protein was copurified with rabbit reticulocyte factors [3], led to the designation eukaryotic initiation factor 5A (eIF-SA) [4]. Nevertheless, the function ofthis protein in intact cells remains to be elucidated.
Abbreviations: eIF-SA, eukaryotic initiation factor 5A; HP, hypusine-containing protein; IPG, immobilized pH gradient; km, kilovolthours; NP-40, Nonidet P-40; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis
In the budding yeast Saccharomyces cerevisiae? two genes (HYP1 =TJF51B=ANB1 and HYP2=T1F51A) coding for H P have been identified [5,6], which are reciprocally regu-
0VCH
Verlagsgcsellsc1i.ift m b H , D-6940 Weinheim, 1992
0 I73-0835/92/0910-0732 $3.50+.25/0
