CURRENT MICROBIOLOGY Vol. 36 (1998), pp.

131–135

An International Journal
R Springer-Verlag New York Inc. 1998

Comparison of Different Methods of Cell Lysis and Protein

Measurements in Clostridium perfringens: Application
to the Cell Volume Determination
Patrick Guerlava, Véronique Izac, Jean-Luc Tholozan
Laboratoire de Génie des Procédés et de Technologie Alimentaires, Institut National de la Recherche Agronomique, 369 rue J. Guesde, BP 39,
F-59651 Villeneuve d’Ascq Cedex, France

Received: 11 April 1997 / Accepted: 6 August 1997

Abstract. Four cell lysis methods (NaOH-SDS solubilization, French press treatment, sonication,
mutanolysin treatment) and three methods of protein assays (Lowry, Bradford, Pierce) were studied for
their applicability to determination of cell volume in Clostridium perfringens NCTC 8798 cell suspen-
sions. Protein contents were higher after a mechanical disruption of the cells than with the other
techniques of lysis. The lowest concentrations of protein were obtained with the Bradford procedure. With
each of the three protein assay methods, Clostridium perfringens NCTC 8798 protein cell contents were
45% to 58% of protein. Other factors possibly involved in variations of the intracellular volume
measurements were examined. A control of the level of protein concentration in the test sample and the
type of silicone oil used for the centrifugation were of prime importance during sample preparation. Under
our conditions, an intracellular volume of 4 µl/(mg of protein) was routinely found for Clostridium
perfringens NCTC 8798.

Study of the main physiological characteristics of micro-
organisms reveals their ability to cope with a stress. The
cell response to stress involves the internal ions, K1 or
Na1 concentration [3], the concentration of some solutes
[10], the internal pH regulation [17], and membrane
potential maintenance [28]. An accurate measurement of
the cell volume is thus needed, because results are often
given as ion or probe concentrations. In addition, measure-
ments of all these parameters require an accurate measure
of biomass, since results are currently expressed in the
literature per mg of protein or per mg of dry weight.
Protein amounts of microorganism are in the most
case assayed by the Lowry technique [21] or the Bradford
technique [5]. The techniques for cellular lysis currently
used in the literature are French press, sonication, NaOH
treatment in the presence of SDS, and even biological
lysis, i.e., mutanolysin treatment [16]. The internal vol-
ume of microorganisms is usually calculated after Rotten-
berg [27], using the silicone oil technique with [14C]car-
Correspondence to: P. Guerlava
boxyl-dextran and [3H]water; a cell volume of 2–4 µl/(mg
of protein) is usually found [22, 27].
We focused our attention on the factors influencing
biomass and cell size measurement in Clostridium perfrin-
gens NCTC 8798, which was chosen as a model strain.
Indeed, Clostridium perfringens type A is a wide-spread,
anaerobic, ubiquitous micro-organism, causing food spoil-
age outbreaks and food poisoning through the excretion
of an enterotoxin [20]. The precise measurement of the
above-described cell parameters is thus of utmost impor-
tance in understanding the cell response to preservation
processes of the food industry.
Materials and Methods
Chemicals. Tritium-labeled water, [14C]carboxyl-dextran, [14C]sorbitol
were from Amersham Life Science (Les Ulis, France). Mutanolysin
came from Sigma (Saint-Quentin Fallavier, France). The Bradford
reagent was purchased from Bio-Rad (Ivry-sur-Seine, France), and the
Pierce reagent from Interchim (Montluçon, France). The Folin-
Ciocalteu reagent was from Merck (Nogent-sur-Marne, France). Sili-
cone oils were Rhodorsil 508v70 (d 5 1.03) and Rhodorsil 47v50
(d 5 0.96) from Rhône Poulenc (Courbevoie, France).
132 CURRENT MICROBIOLOGY Vol. 36 (1998)

Microorganism and culture procedures. Clostridium perfringens

type A strain NCTC 8798 was kindly provided by G. Daube (University
of Liège, Belgium). Culture conditions were derived from Miller and
Wolin [23]. Carbonated buffered medium (pH 7.2) was from Samain et
al. [30]. Growth was performed at 37°C under N2 in 1.2-L culture vials
with 11 mM glucose as substrate. Growth was routinely followed by
medium absorbance measurements and fermentation products accumu-
lation, as described previously [36].
Sample preparation and dry weight measurements for protein
assays. A late log-phase culture was centrifuged anaerobically. The
pellets were then washed and resuspended in 50 mM MOPS (3-[N-
morpholino])propane sulfonic acid. Pellets of the second anaerobic
centrifugation were stored at 220°C for further cell lysis and protein
concentration characterization.
Cell dry weight of culture was determined after pellet resuspen-
sion in water and drying at 105°C to a constant weight [2].
Cell lysis and protein assays. Four methods of cell lysis were
compared. One milliliter of NaOH-SDS (0.2 M, 2.5%) was added to the
pellets above and placed in a water bath at 100°C for 5 min. Pellets were
also resuspended in water before sonication (Sonics Materials, Dan-
bury, CT, USA) or French press treatment at 140 MPa (Bioritech,
Janville-sur-Juine, France). The effect of mutanolysin on cell lysis
(100–200 IU/ml) was studied as described by Kämpfer [16] for
Gram-positive bacteria.
Protein amounts were determined by the Lowry method [21], the
Bradford method [5], and the Pierce method [32] with bovine serum
albumin as the standard. Absorbances were measured with an UVIKON
810 spectrophotometer (Kontron Instruments, Montigny-le-Breton-
neux, France).
Intracellular volume measurements. Dense resting-cell suspensions Fig. 1. Influence of mutanolysin treatment (A) and sonication time (B)
were obtained from late log-phase, glucose-grown cells. The cultures on the protein concentration of samples of Clostridium perfringens
were centrifuged anaerobically and suspended in culture medium suspensions determined with the Bradford assay. In (A) the mutanolysin
without substrate. The protein concentrations were varied from 2 to 12 concentrations were as follows: (M) 100 IU/ml; (N) 200 IU/ml.
mg per ml with the Bradford protocol and cell lysis with the French
press.
Table 1. Comparison of different methods of cell lysis and of methods
74 kBq of [3H]-water and 35.1 kBq of [14C]carboxyl-dextran were
of protein concentration measurement (mg of protein/ml) on the
added to 2 ml of resting cells. The sample was mixed and left for 10 min
protein contents of samples of Clostridium perfringens
at room temperature. 4 3 300 µl of the cell suspension was then placed
in eppendorf tubes containing a 300-µl layer of silicone oil Rhodorsil
Protein concentration (mg of protein/ml) with
508v70 (d 5 1.03) over 100 µl of HClO4 (1 M) at the bottom of the tube.
Cells were centrifuged at 11,000 g for 5 min, and the amount of Cell lysis method Lowry Bradford Pierce
radioactivity in supernatants and pellets was measured in a Betamatic II
liquid scintillation counter (Kontron), as described previously [17]. The No .7 1.0 6 0.1 2.8 6 0.1
internal cell volume (Vi) was calculated according to Rottenberg [27]. NaOH/SDS/100°C 3.2 6 0.2 1.9 6 0.4 2.2 6 0.1
French Press 3.3 6 0.3 2.6 6 0.2 3.1 6 0.2
Sonication 3.4 6 0.1 2.5 6 0.2 3.0 6 0.2
Results Mutanolysin nd 1.6 6 0.6 2.0 6 0.2

Influence of the cell disruption mode on the protein
content of cell suspension. For all experiments, pellets
from a single culture were used. All the protein assay
methods were thus performed on identical samples.
Increasing the mutanolysin concentrations and longer
incubation times increased the protein concentration of
the pellet determined with the Bradford technique (Fig.
1A). However, compared with other disruption methods,
lysis was not complete. In identical samples, a 3- to 5-min
sonication was enough for complete lysis of the cells
(Fig. 1B). Results were similar after one or more French
press treatments of the sample (data not shown). French
Data points represent mean value 6SD (n 5 3). nd: not determined.
press results and sonication results were close (Table 1)
and were higher than those obtained with mutanolysin or
NaOH/SDS/boiling lysis.
Comparison of results of different methods for protein
measurements. With the same lysis method, the three
protein assay methods gave different protein concentra-
tions (Table 1). The Bradford technique resulted in lower
protein concentrations whatever the lysis method. Aver-
age measurements were 17% lower than with the Pierce
P. Guerlava et al.: Clostridium perfringens Cell Volume
Table 2. Influence of the protein concentration of two samples (a and b
from two different cultures) on internal cell volume measurements of
Clostridium perfringens cell suspensions. Protein concentrations were
determined according to Bradford [5]. For each sample, measurements
of the internal volume were performed in duplicate, first on the
concentrated sample and then on the same sample diluted
Protein Internal
concentration cell volume
Sample (mg of protein/ml) (µg/mg of protein)
Concentrated (a) 8.3 4.8 6 0.3
Diluted (a) 2.0 9.0 6 0.5
Concentrated (b) 7.1 4.8 6 0.3
Diluted (b) 1.8 8.5 6 0.5
method and were 23% lower than with the Lowry method
for the same sample. With the three methods of lysis, the
results were close with the Lowry assay (Table 1).
However, this method used without cell lysis gave much
higher protein concentrations than the dry weight content
in samples, which was measured as 5.7 mg/ml. Whether
the Bradford protocol, the Pierce technique, or the Lowry
method was used, Clostridium perfringens contained up
to 45%, 54%, 58% of protein per mg of dry weight,
respectively.
Effect of the cell density in samples on cellular volume
measurements. The calculated cell volume depended on
the protein concentration of resting-cell suspensions.
Different cultures of Clostridium perfringens leading to a
protein concentration of cell suspension from 1 to 12 mg
protein/ml were used. An increase of the protein concen-
tration in cell suspensions resulted in a decrease of the
calculated cell volume. An average volume of 4 6 0.5 µl
was routinely calculated between 6 and 10 mg protein/ml.
Moreover, for the same sample, the calculated cell
volume was different if it was measured on concentrated
or diluted cell suspensions: the volume ratio of diluted
sample to concentrated sample was approximately 1.8
(Table 2). This was confirmed by all assays made on
different cultures of Clostridium perfringens (Fig. 2).
Influence of the oil layer on cellular volume measure-
ments. Different oil volumes (100 µl, 200 µl, 300 µl or 1
ml) of silicone oil were utilized for pellet and supernatant
separation without modifying the calculated value of
internal volume. In contrast, the oil density was of prime
importance to avoid cross-contamination and to allow a
good separation of pellets and supernatants. With a
70/30% mixture of Rhodorsil oils (508v70 d 5 1.03 and
47v50 d 5 0.96, respectively), the internal volume was
15% higher than with Rhodorsil 508v70 oil applied
alone. A complete separation between supernatant and
cells settled in the perchloric acid, at the bottom of the
eppendorf tube, was obtained with Rhodorsil oil 508V70.
133
Fig. 2. Influence of the resting cell protein concentration (mg protein/
ml) in the test sample on the calculated internal cell volume (µl/mg
protein). Each point represents measurements carried on a different
culture.
Influence of the cell disruption method on cellular
volume measurements. Radioactivity counts in the
pellets were not affected by the amount of perchloric acid
used for cell lysis during volume measurements. An
increase from 100 µl to 1 ml of perchloric acid demon-
strated 100 µl as a sufficient volume for complete cell
lysis. Nonspecific adsorption of radioactive probes led to
a maximum calculated internal cell volume of 0.1 µl/mg
of protein after a French press treatment of the samples.
Cell volume calculated with different extracellular
probes. The specific probes routinely used for cell
volume and periplasmic volume measurements in Gram-
negative bacteria have been compared. Dextran and
sorbitol were not consumed by the dense cell suspen-
sions. The internal volume of the same sample estimated
with [14C]carboxyl-dextran and with [14C]sorbitol as
external probe was 4.7 6 0.5 µl/mg protein and 2.3 6 0.3
µl/mg protein, respectively.
Discussion
In comparison with mechanical disruption, mutanolysin
or NaOH/SDS/boiling lysis underestimated the quantity
of protein in our samples whatever the protein assay
chosen. Since pioneering work on colorimetric measure-
ment of protein concentrations [5, 13, 21], numerous
studies on measurement interferences were proposed [6,
8, 11, 31]. In addition to the interferences described
above, the high protein level measured by the Lowry
technique in whole-cell suspensions of Clostridium per-
fringens could also be due to the formation of cell-reagent
precipitates, as described by Peterson [25]. In a similar
way, the lower results obtained with the Bradford tech-
nique could also be due partly to the strong binding of the
dye with the bovine serum albumin used as the standard;
134
actually, a major problem with this assay is the variation
in response to different proteins [33]. Stoschek increased
sensitivity and uniformity in absorbance produced by
different types of proteins with an addition of NaOH to
the dye reagent [33]. In our experiments, the use of a
0.05% SDS concentration is generally considered as
compatible with the protein assay methods and excluded
the presence of additional interferences [34]. This under-
estimation of protein concentrations with a chemical
method of lysis could also be due to the formation of an
insoluble fraction of protein during cell lysis. Indeed, the
requirement of a prior treatment adapted to the cells was
illustrated by Saleemuddin et al. and Gogstad and Krutnes
[12, 29]. A platelet suspension was ultrasonicated and
analyzed with the Bradford protocol in the absence of
solubilizing agents (Triton X-100 or NaOH): the color
development was considerably lower than with intact
platelets, and the calculated corresponding level of pro-
tein without solubilizing agents was much lower. This
was due to proteins which did not bind the dye without a
prior treatment for cell lysis, such as the use of solubiliz-
ing agents or a mechanical disruption. This amount of
thermally denatured, insoluble protein fraction has been
sometimes estimated by linking the Bradford assay to a
resolubilization step [14]. Fanger [9] also developed a
method which allowed membrane-bound proteins to
become and remain solubilized during the Bradford assay
by using glucopyranoside detergents. Up to 20% higher
protein concentrations were measured with this special
sample treatment. Compared with mechanical lysis meth-
ods, our results demonstrated a 20–30% lower amount of
protein in cell suspensions after using mutanolysin or
NaOH/SDS/boiling cell lysis. An important part of this
difference could thus come from the sample preparation.
Although a complete separation of cells from aque-
ous environment could not be obtained by this technique,
the silicon oil method for separation allows bacterial
separation without altering the internal aqueous environ-
ment of the cells [15]. Kashket et al. [18] determined a Vi
of 2.11 µl/mg of dry cells for Streptococcus lactis,
although they found 2.64 µl/mg of dry cells without
silicon oil [17]. In addition, this method has been
recognized as the only method that prevents losses of
NH41 K1 from the cell pellet after the centrifugation step
[3]. The amount of medium trapped in the pellet is about
10-fold less than that obtained by membrane filtration
methods [17, 27]. Previous results demonstrated that
some medium is nevertheless always dragged along with
the cells into the sediment through the silicon oil [26].
These nonselective adsorptions in Clostridium perfrin-
gens cell suspension accounted for a 2–5% increase in the
calculated volume. Variations in nonspecific adsorption
of the probes with the cell concentration in the pellet have
CURRENT MICROBIOLOGY Vol. 36 (1998)
been also carefully checked. No quenching of radioactiv-
ity measurement was demonstrated from 1 to 15 mg
protein/ml (data not shown). Similarly, varying the time
of contact with radioactive probes did not change cell
volume calculations.
Cell volume measurements of bacteria have been
often reported in the literature. However, the authors use
different units for cell quantitations, rendering compari-
sons between the proposed values difficult to evaluate. In
addition, some variations were obtained during measure-
ments: measurements in E. coli cells with [G-14C]hydroxy-
methylinulin and [3H]inulin resulted in a Vi of 2.68 µl/mg
of cells (dry weight) in samples containing 3 mg/ml (dry
weight) [4], and in a Vi of 1.5 µl/mg of cells (dry weight)
at a concentration of 4 mg/ml of dry cells [1]. In
Clostridia, measurements of the intracellular volume of
Clostridium pasteurianum were determined during growth
of the microorganism with [3H]water and [14C]dextran: a
decrease from 4 to 2.1 µl/mg of cells (dry weight) was
obtained when comparing growing and late log phase
cells [26]. These authors also observed that cells of the
early growth phase generally appeared to be much longer
than cells of the late growth phase. In a very similarly
shaped Clostridium, Clostridium sporogenes, an internal
volume of 2.32 µl/mg of cells was determined by using
[3H]water and [14C]inuline with a protein concentration
of 0.6 to 1 mg of cells (dry weight)/ml [24]. Assuming a
50% protein content of the cell dry weight, this would
lead to 4.6 µl/mg protein, as calculated for Clostridium
perfringens (this study). The intracellular volume measure-
ment in another Gram-positive bacterium, Streptococcus
cremoris, demonstrated a constant volume value at 3.8
µl/mg cell protein during growth [35]. These authors used
[3H]water and [14C]taurine for their measurements. How-
ever, it is important to note that [14C]dextran is excluded
by the cell wall, while [14C]taurine as well as [14C]sorbi-
tol can penetrate up to the plasma membrane, allowing
measurement of the periplasmic space in Gram-negative
bacteria [17]. These properties, and the eventual use of
sorbitol as growth substrate for some C. perfringens
strains [7], though it was not the case for our strain, could
explain the differences obtained in our volume measure-
ments when comparing dextran and sorbitol. In a similar
way, the 20% decrease of volume measured after hours of
storage of the cell suspensions at room temperature could
be due to potassium leaks, as movements of this ion are
routinely involved in the regulatory volume changes in
animal and bacterial cells [1, 19].
ACKNOWLEDGMENT
This research program was supported by a grant from the Region
Nord-Pas de Calais, France.
P. Guerlava et al.: Clostridium perfringens Cell Volume
Literature Cited
1. Ahmed S, Booth IR (1983) The use of valinomycin, nigericin and
trichlorocarbanilide in control of the protonmotive force in Esch-
erichia coli cells. J Biochem 212:105–112
2. ASM (1994) Methods for general and molecular bacteriology. In:
Gerhardt P, Murray RGE, Wood WA, Krieg NR (eds) Washington,
DC: American Society for Microbiology, p 206
3. Bakker EP (1993) Cell K1 and K1 transport systems in prokary-
otes. In: Bakker EP (ed) Alkali cation transport systems in
prokaryotes. Boca Raton: CRC Press, pp 205–229
4. Booth IR, Mitchell WJ, Hamilton WA (1979) Quantitative analysis
of proton-linked transport systems. The lactose permease of
Escherichia coli. J Biochem 182:687–696
5. Bradford MM (1976) A rapid and sensitive method for the
quantitation of microgram quantities of protein utilizing the
principle of protein-dye binding. Anal Biochem 72:248–254
6. Brown RE, Jarvis KL, Hyland KJ (1989) Protein measurement
using bicinchoninic acid: elimination of interfering substances.
Anal Biochem 180:136–139
7. Cato EP, George WL, Finegold SM (1986) Genus Clostridium. In:
Sneath PHA, Mair NS, Sharpe ME, Holt JG (eds) Bergey’s manual
of systematic bacteriology, volume 2. Baltimore: Williams and
Wilkins, pp 1141–1200
8. Compton SJ, Jones CG (1985) Mechanism of dye response and
interference in the Bradford protein assay. Anal Biochem 151:369–
374
9. Fanger BO (1987) Adaptation of the Bradford protein assay to
membrane-bound proteins by solubilizing in glucopyranoside deter-
gents. Anal Biochem 162:11–17
10. Farwick M, Siewe RM, Krämer R (1995) Glycine betaine uptake
after hyperosmotic shift in Corynebacterium glutamicum. J Bacte-
riol 177:4690–4695
11. Friedenauer S, Berlet HH (1989) Sensitivity and variability of the
Bradford protein assay in the presence of detergents. Anal Biochem
178:263–268
12. Gogstad GO (1982) Measurement of protein in cell suspensions
using the Coomassie brillant blue dye-binding assay. Anal Biochem
126:355–359
13. Gornall AG, Bardawill CJ, David MM (1949) Determination of
serum proteins by means of the biuret reaction. J Biol Chem
177:751–766
14. Gotham SM, Fryer PJ, Paterson WR (1988) The measurement of
insoluble proteins using a modified Bradford assay. Anal Biochem
173:353–358
15. Hurwitz C, Braun CB, Peabody RA (1965) Washing bacteria by
centrifugation through a water-immiscible layer of silicones. J
Bacteriol 90:1692–1695
16. Kämpfer P (1995) An efficient method for preparation of extracts
from Gram-positive bacteria for comparison of cellular protein
patterns. J Microbiol Methods 21:55–60
17. Kashket ER, Barker SL (1977) Effects of potassium ions on the
electrical and pH gradients across the membrane of Streptococcus
lactis cells. J Bacteriol 130:1017–1023
18. Kashket ER, Blanchard AG, Metzger WC (1980) Proton motive
force during growth of Streptococcus lactis cells. J Bacteriol
143:128–134
135
19. Katz U (1995) Cellular water content and volume regulation in
animal cells. Cell Biochem Funct 13:189–193
20. Labbe RG (1991) Clostridium perfringens. J Assoc Anal Chem
74:711–714
21. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein
measurement with the Folin phenol reagent. J Biol Chem 163:265–
275
22. Miguelez E, Gilmour DJ (1994) Regulation of cell volume in the
salt tolerant bacterium Halomonas elongata. Lett Appl Microbiol
19:363–365
23. Miller TL, Wolin MJ (1974) A serum bottle modification of the
Hungate technique for cultivating obligate anaerobes. Appl Environ
Microbiol 27:895–897
24. Okereke A, Montville TJ (1992) Nisin dissipates the proton motive
force of the obligate anaerobe Clostridium sporogenes. Appl
Environ Microbiol 58:2463–2467
25. Peterson GL (1979) Review of the Folin phenol protein quantita-
tion method of Lowry, Rosebrough, Farr and Randall. Anal
Biochem 100:201–220
26. Riebeling V, Thauer RK, Jungermann K (1975) The internal-
alkaline pH gradient, sensitive to uncoupler and ATPase inhibitor,
in growing Clostridium pasteurianum. Eur J Biochem 55:445–453
27. Rottenberg H (1979) The measurement of membrane potential and
delta pH in cells, organelles, and vesicles. Methods Enzymol
55:547–569
28. Ruhr E, Sahl H (1985) Mode of action of the peptide antibiotic nisin
and influence on the membrane potential of whole cells and on
cytoplasmic and artificial membrane vesicles. Antimicrob Agents
Chemother 27:841–845
29. Saleemuddin M, Ahmad H, Husain A (1980) A simple, rapid, and
sensitive procedure for the assay of endoprotease using Coomassie
brilliant blue G-250. Anal Biochem 105:202–206
30. Samain E, Albagnac G, Dubourguier HC, Touzel JP (1982)
Characterization of a new propionic acid bacterium that ferments
ethanol and displays a growth factor dependent association with a
gram negative homoacetogen. FEMS Microbiol Lett 15:69–74
31. Sedmak JJ, Grossberg SE (1977) A rapid, sensitive, and versatile
assay for protein using Coomassie brilliant blue G250. Anal
Biochem 79:544–552
32. Smith PK, Krohn PK, Hermanson GT, Mallia AK, Gartner FH,
Provenzano, Fujimoto EK, Goeke NM, Olson BJ, Klenk DC (1985)
Measurement of protein using bicinchoninic acid. Anal Biochem
150:76–85
33. Stoscheck CM (1990a) Increased uniformity in the response of the
Coomassie blue G protein assay to different proteins. Anal Bio-
chem 184:111–116
34. Stoscheck CM (1990b) Quantitation of protein. Methods Enzymol
182:50–68
35. Ten Brink B, Konings WN (1982) Electrochemical proton gradient
and lactate concentration gradient in Streptococcus cremoris cells
grown in batch culture. J Bacteriol 152:682–686
36. Touzel JP, Petrof D, Albagnac G (1985) Isolation and characteriza-
tion of a new thermophilic Methanosarcina, the strain CHTI 55.
Syst Appl Microbiol 6:66–71