CURRENT MICROBIOLOGY Vol. 42 (2001), pp. 89 –95
DOI: 10.1007/s002840010184
Current
Microbiology
An International Journal
© Springer-Verlag New York Inc. 2001
Influence of Physico-Chemical Factors on the Oligomerization and
Biological Activity of Bacteriocin AS-48
Hikmate Abriouel,1 Eva Valdivia,1 Antonio Gálvez,2 Mercedes Maqueda1
1
Dpto. Microbiología, Facultad de Ciencias, Universidad de Granada, Fuentenueva s/n, E-18071-Granada, Spain
Area de Microbiología, Facultad de Ciencias Experimentales, Universidad de Jaén, E-23071 Jaén, Spain
2
Received: 3 July 2000 / Accepted: 11 August 2000
Abstract. Bacteriocin AS-48 forms a mixture of monomers and oligomers in aqueous solutions. Such
oligomers can be clearly differentiated by SDS-PAGE after formaldehyde crosslinking, and we have
verified that these associates are stable to acid treatment after fixation. In addition, they show antimicrobial activity and are recognized by anti-AS-48 antibodies. AS-48 oligomers can be dissociated by the
detergents SDS and Triton X-100. The degree of oligomerization of AS-48 depends on the pH of the
solution and the protein concentration. At pH below 5, AS-48 is in the monomeric state at protein
concentrations below 0.55.mg ml21, but it also forms dimers above this protein concentration. This
bacteriocin forms oligomers at pH values above 5, in agreement with the observation that it is also more
hydrophobic at neutral pH. AS-48 is stable to mild heat treatments irrespectively of pH. At 120°C it is
more heat resistant under acidic conditions, but it inactivates at neutral pH. Activity of AS-48 against E.
faecalis is highest at neutral pH, but it is highest at pH 4 for E. coli. The influence of pH on bacteriocin
activity could be owing to changes in the conformation/oligomerization of the bacteriocin peptide as well
as to changes in the surface charge of the target bacteria.
The bacteriocin AS-48 produced by Enterococcus faecalis subsp. liquefaciens S-48 is unique in its cyclic structure and its broad antimicrobial spectrum [1, 4, 6, 7, 15,
21]. AS-48 is a strongly basic molecule (pI close to
10.5), and it contains a high proportion (49%) of hydrophobic amino acids (Ala, Pro, Val, Met, Ile, Leu, and
Phe) and uncharged hydrophilic residues (Ser, Gly, Thr,
and Tyr). Therefore, the combination of a net positive
charge with a large proportion of hydrophobic residues
suggests an amphipathic structure for AS-48 [4, 5]. This
bacteriocin exerts bactericidal and bacteriolytic effects
against most of the Gram-positive bacteria and some
Gram-negative bacteria [4, 6]. Its bactericidal effect on
Escherichia coli and Salmonella is well documented [1,
7, 8], although the bacteriocin concentrations required
are much higher, probably because of the protective
effect of the bacterial outer membrane.
Lactic acid bacteria produce strongly cationic peptides with potent antimicrobial activity like nisin [9, 10],
Correspondence to: M. Maqueda; email: [email protected]
lacticin 481 [18], carnocin UI49 [22], lactocin S [17],
pep-5 [20], or bacteriocin C3603 [11], among others. All
of them are similar in molecular sizes (3– 6 kDa), isoelectric points (ca. 10), and biological activities [12].
Together, these inhibitory substances may represent the
conservation in the course of evolution of a general
mechanism of antibiosis by means of basic peptides.
Bacteriocin AS-48 shows striking similarities to these
cationic antimicrobial peptides, but the lack of lanthionine and its capacity to inhibit several Gram-negative
bacteria represent two major differences. Because of its
stability and solubility over a wide pH range and its
broad antimicrobial spectrum, AS-48 is a good candidate
to be used as a natural food preservative. In this respect,
its activity against several food-borne bacteria such as E.
coli, Salmonella, and Listeria monocytogenes is noteworthy [16]. In this context, it is important to gain
knowledge into the effect of several factors (especially
those that may be encountered in food systems) on the
physico-chemical behavior, stability, and biological activity of this bacteriocin, which was the aim of this work.
90
CURRENT MICROBIOLOGY Vol. 42 (2001)
Fig. 1. Electrophoretic separation of formaldehyde-crosslinked AS-48 in aqueous solution. A) Coomassie blue-stained SDS-PAGE gel showing the
different multimers. AS-48 has a molecular mass of 7.1 kDa (lane 2). Bands corresponding to various mono- and multimeric forms of AS-48 after
formaldehyde crosslinking are shown in lane 3. Lane 1: standard markers (in kDa). B) Antimicrobial activity against E. faecalis S-47 of
formaldehyde-crosslinked AS-48. C) Immunological detection of AS-48 in Western blotting by using specific anti-AS-48 antibodies. Lane 1:
formaldehyde-crosslinked AS-48. Lane 2: untreated AS-48.
Preliminary data on AS-48 purification indicated that this
bacteriocin showed an abnormal behavior on size-exclusion chromatography, since multiple forms seemed to be
present (unpublished results), and the bacteriocin eluted
as a peak of low-molecular mass only under acidic pH
[4]. In this work we present data about formation of
oligomers, its biological activity, and the effect of different physico-chemical factors on these features.
Materials and Methods
Strains and growth conditions. Enterococcus faecalis A-48-32 was
used to produce AS-48 E. faecalis S-47 and Escherichia coli U-9 from
our collection were used as indicator strains. The bacteria were propagated in brain heart infusion broth (BHI, Oxoid).
AS-48 preparation and activity assay. Bacteriocin preparations were
obtained as described elsewhere [5]. Protein concentration of stock
solution (2.2 mg.ml21) was determined by the method of Bradford [3].
The inhibitory activity of AS-48 was determined by the agar-well
diffusion method [4]. For the assay of antimicrobial activity at different
pH values, exponential-phase cells of the bacteria were used as described by Abriouel et al. [1]. The logarithmic reduction factor (LRF)
was calculated as log10 CFU/ml in controls minus log10 CFU/ml in
treated samples.
Effects of heat treatments. Bacteriocin solutions (50 mg.ml21) in
different buffers were held at different temperatures (60°C for 30 min;
70°C for 10 min; 80°C for 5 min; 120°C for 15 min; and 140°C for 4 s).
After heat treatments, samples were cooled rapidly and tested for
remaining antimicrobial activity. Positive controls were held at room
temperature for similar periods of time at the desired pH values. Buffer
solutions without bacteriocin were used as negative controls.
Buffers. Sodium citrate (pH 3.0 and 4.0), sodium acetate (pH 5 and
5.6), and sodium phosphate (pH 6 – 8) buffers were used at 0.1 M
concentration.
Electrophoretic techniques. AS-48 samples were separated by SDSPAGE on 10% slab gels as described by Laemmli [13] and transferred
to a nitrocellulose (NC) membrane according to Towbin et al. [23].
Immunoblotting was carried out with specific anti-AS-48 antibodies
[14]. The antimicrobial activity of proteins separated by SDS-PAGE
was carried out according to the method reported by Bhunia et al. [2].
Chemical cross-linking. Bacteriocin solutions in different buffers
were cross-linked by incubation with 1% formaldehyde (from paraformaldehyde powder, Taap, Altermaston, England) for 1 h at room
temperature before they were separated by SDS-PAGE. Alternatively,
samples crosslinked by formaldehyde were incubated at room temperature for 2 h with 10% trichloroacetic acid. The precipitated proteins
were collected by centrifugation (12,000 g for 10 min) in a microfuge,
redissolved in sample buffer, and separated by SDS-PAGE.
Peptide hydrophobocity. The hydrophobicity of AS-48 was evaluated
by phase-partition between the aqueous phase and n-octanol (Sigma)
according to Puyal et al. [19].
Results and Discussion
The bacteriocin AS-48 shows several interesting features
that make it an attractive candidate for food application,
and also to study protein-protein interactions, especially
when these are accompanied by changes in the biological
activity of the molecule. In this study, we present data
confirming earlier observations about oligomerization of
H. Abriouel et al.: Oligomerization and Biological Activity of Bacteriocin AS-48
91
Fig. 2. Effect of detergents on AS-48 association. A) Aqueous solutions of AS-48
(0.54 mgzml21) were added of Triton
X-100 (1%) or SDS (1%) before or after
formaldehyde fixation. Lane 1: Control
(untreated AS-48); 2: formaldehyde-fixed
AS-48; 3: AS-48 treated with Triton
X-100 alone; 4: AS-48 treated with Triton
X-100 before fixation with formaldehyde;
5: AS-48 fixed before treatment with Triton X-100; 6: AS-48 treated with SDS
before fixation; 7:AS-48 fixed before addition of SDS. B) Effect of SDS on AS-48
association. Aqueous solutions of AS-48
(0.51 mgzml21) were added of SDS at a
final concentration of 0% (lane 2), 0.1%
(lane 3), 0.2% (lane 4), 0.4% (lane 5),
0.8% (lane 6) or 1% (lane 7) before formaldehyde fixation. Lane 1: AS-48 without
formaldehyde fixation (control). M: Standard markers (in kDa).
bacteriocin AS-48, and suggesting how this phenomenon
affects AS-48 stability and biological activity.
Oligomerization of AS-48 in aqueous solutions. The
degree of oligomerization of AS-48 in aqueous solution
(1.3 mg.ml21) was studied by SDS-PAGE after crosslinking with formaldehyde. The results obtained revealed
that AS-48 forms a mixture of monomers and different
multimers (associates of 13, 8, 5, 4, 3, and 2 molecules of
AS-48), with M1 ranging from 91 to 7 kDa (Fig. 1A).
Non-fixed AS-48 solutions yielded a single monomeric
band when separated under the same conditions.
Gels prepared as above were treated to remove SDS
and then incubated with a soft agar overlay seeded with
the indicator strain E. faecalis S-47. After incubation, the
agar overlay showed zones of inhibited growth that corresponded to the different bacteriocin protein bands in
the gel (Fig. 1B). The sizes of the inhibition zones
correlated with the intensity of the protein bands. According to the amino acid composition of AS-48, the
residues involved in methylene bridge formation after
fixation with formaldehyde could be two residues of
arginine, one residue of asparagine, and another one of
tyrosine, and these bridges remained stable after acid
treatment. To verify this fact, crosslinked bacteriocin
samples were treated with 10% trichloroacetic acid, and
the acid-precipitated proteins were collected by centrifugation, dissolved in buffer, and separated by SDSPAGE. In these conditions, we still got the same types of
oligomers as in crosslinked samples without acid treatment (data not shown). These results clearly indicate
that associates crossslinking by formaldehyde were
stable to acidic conditions. Therefore, the antimicrobial activity detected on gels should correspond to the
oligomers and not to dissociated forms originated by
acid treatment.
Immunological studies show that AS-48 monomer
and also oligomers and multimers were recognized by
anti-AS-48 antibodies after being transferred to a NC
membrane (Fig. 1C). This behavior is comprehensive
because rabbits were immunized with concentrated bacteriocin solutions containing the different oligomeric
forms [14], and reactive sera probably contain a mixture
92
CURRENT MICROBIOLOGY Vol. 42 (2001)
Fig. 3. Effect of pH on AS-48 association.
Aliquots of AS-48 (0.54 mgzml21) buffered at pH values of 3 (lane 2), 4 (lane 3),
5 (lane 4), 6 (lane 5), 7 (lane 6), and 8
(lane 7) fixed with 1% formaldehyde and
separated by SDS-PAGE. Lane 1: AS-48
without formaldehyde fixation. M: Standard markers (in kDa).
of antibodies that can recognize both the bacteriocin
monomer and the oligomers.
neutral, and the degree of association increased markedly
as pH rose from 6 to 8.
Effect of detergents on bacteriocin association. Bacteriocin oligomerization in aqueous solutions is a reversible process, and it probably arises from the attraction
between hydrophobic groups. In order to investigate the
effect of detergents on AS-48 association, aqueous bacteriocin solutions (0.54 mg.ml21) at neutral pH were
treated with SDS or Triton X-100 (1% final concentration) and then fixed with formaldehyde before separation
by SDS-PAGE. In samples treated with SDS, the bacteriocin dissociated completely into monomers (Fig. 2A).
The minimal concentration of SDS required for complete
dissociation of AS-48 oligomers (0.8%; Fig. 2B) was
much higher than the critical micellar concentration of
this detergent (ca. 0.27%), suggesting that the hydrophobic interactions between AS-48 molecules must be quite
strong. As expected, neither of the detergents tested was
able to dissociate AS-48 oligomers if they were previously fixed with formaldehyde.
Effect of the protein concentration on AS-48 association-dissociation at different pH values. Association
of AS-48 was also influenced by bacteriocin concentration. In fact, formation of bacteriocin dimers could also
be observed under acidic conditions (pH 3–5) at protein
concentrations above 0.55 mg.ml21 (Fig. 4A). Dilution
of protein samples below this concentration under acidic
conditions resulted in dissociation to the monomeric
form. These data are consistent with the fact that dilution
can perturb subunit interactions in oligomeric proteins by
changing the association-dissociation equilibrium. However, at pH 7 all of the samples showed oligomers
regardless of bacteriocin concentration (Fig. 4B), although the proportion of oligomers was lower as the
concentration of AS-48 decreased.
Effect of pH. Bacteriocin solutions at different pHs
(3– 8) were crosslinked with formaldehyde and separated
by SDS-PAGE to study the influence of pH on bacteriocin association. Samples cross-linked at acidic pH (3–5)
showed a single protein band on the gels corresponding
to momoners (Fig. 3). Since AS-48 is highly cationic (pI
10.5), the molecule should be highly positively charged
under acidic conditions, and therefore the hydrophobic
interactions should be weaker than they are at higher
values of pH. There was a strong correlation between
bacteriocin oligomerization and pH change from acid to
Peptide hydrophobicity. Hydrophobic interactions
should play a key role in AS-48 oligomerization. Therefore, we studied the hydrophobicity of AS-48 as a function of pH by phase partition into n-octanol (Fig. 5). At
pH 3 and 4, most of the bacteriocin (63%) remained in
the aqueous phase. At higher pH, larger amounts of
AS-48 migrated to the octanolic phase. Migration was
highest at pH 7 and 8 (at which only 9% and 11% of the
initial amount of protein remained in the aqueous phase,
respectively). These results indicate that AS-48 is highly
hydrophilic at pH 4 or below, while it becomes gradually
hydrophobic from pH 5 to 8. Accordingly, bacteriocin
oligomerization may provide a mechanism to increase its
solubility in aqueous systems under conditions in which
H. Abriouel et al.: Oligomerization and Biological Activity of Bacteriocin AS-48
93
Fig. 4. Concentration-dependent associationdissociation equilibrium of AS-48 at different
pHs. Different dilutions of buffered solutions
of AS-48 (2.2 mg.ml21) fixed with formaldehyde before SDS-PAGE. Lanes 1–5: solutions
containing different AS-48 concentrations:
2.2, 1.1, 0.73, 0.55, and 0.44 mg.ml21 respectively; (A) fixed at pH 3, (B) at pH 7. M:
Standard markers (in kDa).
certain regions of the molecule become more hydrophobic.
Effect of pH on the antimicrobial activity of AS-48.
To test whether the changes in pH could have any
influence on the antimicrobial activity of AS-48 (in addition to their influence on bacteriocin hydrophobicity
and antimicrobial activity, as described above), bacterial
cell suspensions at different pHs were incubated with
AS-48 for 30 min, and then the number of survivors was
determined. The results obtained indicated that the antimicrobial activity of AS-48 was influenced markedly by
the pH of the solution. Bacteriocin activity against E.
faecalis S-47 was higher in the range of pH from 5 to 7
(Table 1). The highest activity was obtained at pH 6, and
the lowest at pH 4. Some decrease in antimicrobial
activity was also detected at alkaline pH. The results
obtained by using E. coli U-9 as indicator strain were
much different. In this case, some antimicrobial activity
was detected at pH 6 (followed by that of pH 7 and 8) by
using tenfold higher bacteriocin concentrations. Nevertheless, the activity of AS-48 against this bacterium
increased markedly by lowering the pH from 5 to 4
(Table 1). Exposure to pH 4 alone did not have much
effect on the viability of E. coli cells (LRF 5 0.41; data
not shown). These results indicate that AS-48 is more
active against Gram-positive bacteria when the bacteriocin forms oligomers, while it is more active on E. coli
at acidic pH, when it is in the monomeric state. In
addition to this, we should take into account that pH also
modifies the surface charge of the cell wall/membrane of
target cells, and that changes in sensitivity of Grampositive and Gram-negative bacteria over a pH range
may also be owing to the great differences in their cell
wall composition. This phenomenon may be of great
importance in food preservation, since the antimicrobial
activity of AS-48 against Gram-negative bacteria could
94
CURRENT MICROBIOLOGY Vol. 42 (2001)
Fig. 5. Phase partition of AS-48 into n-octanol in function of pH.
Solutions of AS-48 in different buffers (pH 3– 8) mixed with n-octanol
as described in Materials and Methods. Percentage of protein (bars) and
the activity against E. faecalis (m) remaining in the aqueous phase.
Table 1. Effect of pH on the antimicrobial activity of bacteriocin
AS-48. Cell suspensions in different pH buffers were incubated for
30 min at 37°C with buffer alone or with bacteriocin AS-48 (final
concentration of 5 mg.ml21 for E. faecalis S-47 and 70 mg.ml21 for
E. coli U-9). The loss of viability is expressed as the logarithmic
reduction factor (LRF)
pH
Indicator strain
4
5
6
7
8
E. faecalis S-47
E. coli U-9
0.54
4.16
1.73
0.09
2.03
0.42
1.58
0.27
1.28
0.17
be enhanced by lowering the food pH without compromising its antimicrobial activity on Gram-positives to a
large extent.
Effect of pH on heat stability of AS-48. The physicochemical characteristics of AS-48 as well as its broad
antimicrobial spectrum make it a good candidate for use
in food preservation. In this context, it should be of
interest to investigate the effects of changes in pH (and
therefore in the degree of oligomerization) on bacteriocin
stability to heat, and especially to heat treatments applied
in the food industry. In order to determine the stability of
AS-48, bacteriocin solutions buffered at pH 3– 8 were
treated by heat as described above (Fig. 6). We did not
detect any loss of activity in this range of pH at temperatures of 80°C or lower. Heating at 120°C for 15 min
caused a marked loss of bacteriocin activity at pH 6 as
well as a complete loss of activity at pH 7 and 8. Since
these pH values induce the highest degree of oligomerization, these results suggest that oligomers are much
less stable than monomers. In this treatment, some loss
Fig. 6. Heat stability of AS-48 as a function of pH. Bacteriocin
solutions at different pHs were heated for different times at various
temperatures. The remaining antimicrobial activity is shown.
of activity was also observed at pH 4 (but not at pH 3 or
5). Short-time ultra-high temperature treatments caused
an almost complete inactivation of the bacteriocin at pH
4 and pH 8. For other pH values, this treatment caused
only partial inactivation, and the remaining activity of
samples ranged from 60 to 80% of controls (Fig. 6).
Altogether, these results predict that incorporation
of bacteriocin into foods should be compatible with mild
heat treatments irrespectively of the food pH. The broad
antimicrobial spectrum of this bacteriocin and its stability at moderate temperatures over a broad range of pH
make it a suitable candidate to be used as a biopreservative in pasteurized as well as in minimally processed
foods and foodstuffs. Application of more severe heat
treatment, however, should have to be studied in more
detail, because the pH of the solution may have a marked
effect on bacteriocin stability.
ACKNOWLEDGMENTS
This research was supported by a grant from the Comisión Interministerial de Ciencia y Tecnología (BIO95-0466) of the Spanish Ministry
of Education and Science.
Literature Cited
1. Abriouel H, Valdivia E, Gálvez A, Maqueda M (1998) Response
of Salmonella choleraesuis LT2 spheroplasts and permeabilized
cells to the bacteriocin AS-48. Appl Environ Microbiol 64:4623–
4626
2. Bhunia AK, Johnson MC, Ray B (1987) Direct detection of an
antimicrobial peptide of Pediococcus acidilactici in sodium dodecyl sulfate-polyacrylamide gel electrophoresis. J Ind Microbiol
2:319 –322
3. Bradford M (1976) A rapid and sensitive method for the quanti-
H. Abriouel et al.: Oligomerization and Biological Activity of Bacteriocin AS-48
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
tation of microgram quantities of protein utilizing the principle of
protein-dye binding. Anal Biochem 72:248 –254
Gálvez A, Maqueda M, Valdivia E, Quesada A, Montoya E (1986)
Characterization and partial purification of a broad spectrum antibiotic AS-48 produced by Streptococcus faecalis. Can J Microbiol 32:765–771
Gálvez A, Giménez-Gallego G, Maqueda M, Valdivia E (1989a)
Purification and amino acid composition of peptide antibiotic
AS-48 produced by Streptococcus (Enterococcus) faecalis subsp.
liquefaciens S-48. Antimicrob Agents Chemother 33:437– 441
Gálvez A, Maqueda M, Martínez-Bueno M, Valdivia E (1989b)
Bactericidal and bacteriolytic action of peptide antibiotic AS-48
against Gram positive and Gram negative bacteria and other
organisms. Res Microbiol 140:57– 68
Gálvez A, Valdivia E, Martínez-Bueno M, Maqueda M (1989c)
Bactericidal action of peptide antibiotic AS-48 against Escherichia coli K-12. Can J Microbiol 35:318 –321
Gálvez A, Maqueda M, Martínez-Bueno M, Valdivia E (1996)
The broad spectrum antimicrobial peptide AS-48 from Enterococcus faecalis. In: Recent research developments in antimicrobial
agents and chemotherapy, vol. 1. Trivandrum, India: Pandalai,
Research Signpost, S.G, pp 135–147
Hurst A (1981) Nisin. Adv Appl Microbiol 27:85–123
Hurst A (1983) Nisin and other inhibitory substances from lactic
acid bacteria In: Branen AL Davidson P M (eds) Antimicrobials
in foods. (Series of Food Science), vol. X. New York: Marcel
Dekker, pp 327–351
Ike T, Iwanamini T, Hirasawa M, Watamba C, McGhee JR, Shiota
T (1982) Purification and certain properties of a bacteriocin from
Streptoccoccus mutans. Infect Immun 35:861– 868
Jack RW, Tagg JR, Ray B (1995) Bacteriocins of Gram-positive
bacteria. Microbiol Rev 59:171–200
Laemmli UK (1970) Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature 227:680 – 685
Maqueda M, Gálvez A, Martínez-Bueno M, Guerra I, Valdivia E
(1993) Neutralizing antibodies against the peptide antibiotic AS-
15.
16.
17.
18.
19.
20.
21.
22.
23.
95
48: immunocytological studies. Antimicrob Agents Chemother
37:148 –151
Martínez-Bueno M, Maqueda M, Gálvez A, Samyn B, Van
Beeumen J, Coyette J, Valdivia E (1994) Determination of the
gene sequence and molecular structure of the enterococcal peptide
antibiotic AS-48. J Bacteriol 176:6334 – 6339
Mendoza F, Maqueda M, Gálvez A, Martínez-Bueno M, Valdivia
E (1999) Antilisterial activity of peptide AS-48 and study of
changes induced in the cell envelope properties of an AS-48adapted strain of Listeria monocytogenes. Appl Environ Microbiol 65:618 – 625
Mortvedt CI, Nissen-Meyer J, Sletten K, Nes I.F (1991) Purification and amino acid sequence of lactocin S, a bacteriocin
produced by Lactobacillus sake L45. Appl Environ Microbiol
57:1829 –1834
Piard JC, Muriana PM, Desmazeaud MJ, Klaenhammer TR (1992)
Purification and partial characterization of Lacticin 481, a lanthionine-containing bacteriocin produced by Lactococcus lactis
subsp. lactis Cnrz 481. Appl Environ Microbiol 58:279 –284
Puyal C, Maurin L, Miquel G, Bienven??e A, Philippot J (1994)
Design of a short membrane-destabilizing peptide covalently
bound to liposome. Biochim Biophys Acta 1195:259 –266
Sahl HG, Brandis H (1981) Production, purification and chemical
properties of an antistaphylococcal agent produced by Staphylococcus epidermidis. J Gen Microbiol 127:377–384
Samyn B, Martínez-Bueno M, Devreese B, Maqueda M, Gálvez
A, Valdivia E, Coyette J, van Beeumen J (1994) The cyclic
structure of the enterococcal peptide antibiotic AS-48. FEBS Lett
352:87–90
Stoffels G, Nes IF, Gudmundsdottir A (1992) Isolation and properties of a bacteriocin producing Carnobacterium piscicola isolated from fish. J Appl Bacteriol 173:309 –316
Towbin H, Stacheling T, Gordon J (1979) Electrophoretic transfer
of proteins from polyacrylamide gels to nitrocellulose sheets:
procedure and some applications. Proc Natl Acad Sci USA 76:
4350 – 4354