Interdiscip Sci Comput Life Sci (2009) 1: 81–90
Electromagnetic Signals Are Produced by Aqueous Nanostructures
Derived from Bacterial DNA Sequences
1(Nanectis Biotechnologies, S.A. 98 rue Albert Calmette, F78350 Jouy en Josas, France)
2(Vironix LLC, L. Montagnier 40 Central Park South, New York, NY 10019, USA)
Jamal A¨ISSA1,St´ ephane FERRIS1,
Recevied 3 January 2009 / Revised 5 January 2009 / Accepted 6 January 2009
Abstract: A novel property of DNA is described: the capacity of some bacterial DNA sequences to induce
electromagnetic waves at high aqueous dilutions. It appears to be a resonance phenomenon triggered by the
ambient electromagnetic background of very low frequency waves. The genomic DNA of most pathogenic bacteria
contains sequences which are able to generate such signals. This opens the way to the development of highly
sensitive detection system for chronic bacterial infections in human and animal diseases.
Key words: DNA, electromagnetic signals, bacteria.
Pathogenic microorganisms in this day of age are not
only submitted to high selective pressure by the im-
mune defenses of their hosts but also have to survive un-
der highly active antiviral or antibiotic treatments. Not
surprisingly, they have evolved in finding many ways to
escape these hostile conditions, such as mutations of re-
sistance, hypervariability of surface antigens, protective
biofilms, latency inside cells and tissues.
We initially observed (Montagnier and Lavallee, per-
sonal communication) that some filtration procedures
aimed at sterilizing biological fluids can yield under
some defined conditions the infectious microorganism
which was present before the filtration step. Thus, fil-
tration of a culture supernatant of human lymphocytes
infected with Mycoplasma pirum, a microorganism of
about 300 nM in size, through filters of 100 nM or
20 nM porosities, yielded apparently sterile fluid. The
latter however was able to regenerate the original my-
coplasma when incubated with a mycoplasma negative
culture of human lymphocytes within 2 to 3 weeks.
Similarly, a 20 nM filtration did not retain a minor in-
fective fraction of HIV, the causal agent of AIDS, whose
viral particles have a diameter averaging 100-120 nM.
In the course of investigating the nature of such filter-
ing infectious forms, we found another property of the
filtrates, which may or may not be related to the former:
their capacity to produce some electromagnetic waves
of low frequency in a reproducible manner after appro-
priate dilutions in water. The emission of such waves is
likely to represent a resonance phenomenon depending
on excitation by the ambient electromagnetic noise. It
is associated with the presence in the aqueous dilutions
of polymeric nanostructures of defined size. The super-
natant of uninfected eukaryotic cells used as controls
did not exhibit this property.
In this paper we provide a first characterization of the
electromagnetic signals (EMS) and of their underlying
nanostructures produced by some purified bacteria.
In addition to M. pirum, a more classical bacterium,
E. Coli, was utilized for the purpose of the analysis.
The nanostructures produced by HIV will be the sub-
ject of another paper.
M. pirum is a peer-shaped small bacterial cell,
ressembling M. pneumoniae, which can be grown in syn-
thetic enriched medium (SP4) (Tully et al., 1977) but
also mutiplies at the surface of human T lymphocytes.
The strain (Ber) used in our experiments was isolated
from a T lymphocyte culture derived from the blood
of an apparently healthy subject (Grau et al., 1993).
The strong mycoplasma adherence to lymphocytes is
mediated by a specific adhesin, whose gene had been
previously cloned and sequenced by the authors (Tham
et al., 1994).
We used as primary source of the mycoplasma, super-
natants of infected human T lymphocyte cultures or of
cultures of the CEM tumor T cell line. All cell cultures
were first tested for the lack of M. pirum contamination
by polymerase chain reaction (PCR) and nested PCR,
before starting the experiments. Titers of 106-107infec-
82Interdiscip Sci Comput Life Sci (2009) 1: 81–90
tious Units/ml of M. pirum were readily achieved after
5-6 days of incubation following deliberate infection of
both types of cultures.
Filtration of the clarified supernatant was first per-
formed on 0.45 µM (450 nM) Millipore filters to remove
debris, and subsequently on 0.1 µM (100 nM) Milli-
pore filters or on 0.02 µM (20 nM) Whatman filters,
to remove mycoplasma cells. Indeed, the two 100 nM
and 20 nM filtrates were confirmed sterile when aliquots
were incubated for several weeks in SP4 medium. Re-
peated search for traces of mycoplasma DNA by PCR
and nested PCR using specific primers for the adhesin
gene or for the 16S ribosomal gene was consistently neg-
However when the filtrates were incubated for two
weeks (100 nM filtrate) or three weeks (20 nM filtrate)
with a culture of human activated T lymphocytes, the
mycoplasma was recovered in the medium with all its
original characteristics as previously observed.
The same filtrates were analyzed just after filtra-
tion for production of electromagnetic waves of low fre-
quency. For this purpose we used a device previously
designed by Benveniste and Coll (1996; 2003) for the
detection of signals produced by isolated molecules en-
dowed with biological activity. The principle of this
technology is shown in Fig. 1.
Fig. 1 Device for the capture and analysis of electromag-
netic signals (EMS): (1) Coil: a bobbin of copper
wire, impedance 300 Ohms; (2) Plastic stoppered
tube containing 1 mL of the solution to be analyzed;
(3) Amplifier; (4) Computer with softwares.
Briefly, the 100 nM or 20 nM filtrates are serially di-
luted 1 in 10 (0,1 +0,9 in sterile water (medical grade).
The first 2 dilutions (1/10 and 1/100) are done in
serum-free RPMI medium, in order to avoid eventual
protein precipitation in deionized water.
Each dilution is done in 1.5 mL Eppendorf plastic
tubes, which are then tightly stoppered and strongly
agitated on a Vortex apparatus for 15 seconds . This
step has been found critical for the generation of signals.
After all dilutions have been made (generally 15-20
decimal dilutions), the stoppered tubes are read one by
one on an electromagnetic coil, connected to a Sound
Blaster Card itself connected to a laptop computer,
preferentially powered by its 12 volt battery.
emission is recorded twice for 6 seconds, amplified 500
times and processed with different softwares for vizual-
ization of the signals on the computer’s screen (Fig. 1).
The main harmonics of the complex signals were an-
alyzed by utilizing several softwares of Fourier transfor-
In each experiment, the internal noise generated by
the different pieces of the reading system was first
recorded (coil alone, coil with a tube filled with water).
Fourier analysis shows (Fig. 2(c, d)) that the noise was
predominantly composed of very low frequencies, prob-
ably generated at least in part by the 50/60 Hz ambi-
ent electric current. The use of the 12 V battery for the
computer power supply did reduce, but not abolish this
noise, which was found to be necessary for the induc-
tion of the resonance signals from the specific nanos-
When dilutions of the M. pirum filtrate were recorded
for wave emission, the first obvious phenomenon ob-
served was an increase of the overall amplitude of
the signals at certain dilutions over the background
noise (Fig. 2(a)) and also an increase in frequencies
(Fig. 2(b)). This change was abolished if the tube to be
analyzed was placed inside a box sheltered with sheets
of copper and mumetal (David, 1998).
Fourier analysis of the M. pirum signals showed a
shift towards higher frequencies close to 1000 Hz and
multiples of it. Profiles were identical for all the dilu-
tions showing an increase in amplitude (Fig. 2(c) and
The first low dilutions were usually negative, showing
the background noise only. Positive signals were usu-
ally obtained at dilutions ranging from 10−5to 10−8or
10−12. Higher dilutions were again negative (Fig. 3).
The positive dilutions varied according to the type of
filtration, the 20 nM filtrate being generally positive at
dilutions higher than those of the 100 nM filtrate.
The original unfiltered suspension was negative at all
dilutions, a phenomenon observed for all the microor-
Size and density of the structures producing the
signals in the aqueous dilutions:
An aliquot of the 20 nM filtrate was layered on the
top of a 5-20% (w/v) sucrose gradient in water and
centrifuged for 2 hours at 35,000 rpm in a swinging
bucket rotor.These conditions had previously been
used to obtain the density equilibrium of the intact my-
coplasma cells wich formed a sharp bound at 1,21 den-
sity. Fractions were collected from the bottom of the
tubes, pooled 2 by 2 and assayed for signal emission.
Fig. 4 shows that the signal emitting structures were
distributed in a large range of densities from 1.15 to
1.25 and also had a high sedimentation coefficient.
Interdiscip Sci Comput Life Sci (2009) 1: 81–9083
Fig. 2Detection of EMS from a suspension of Mycoplasma pirum: Left: background noise (from an unfiltered suspension or
a negative low dilution). Right: positive signal (from a high dilution D-7 (10-7)). (a) actual recording (2 seconds from
a 6 second recording) after WaveLab (Steinberg) treatment; (b) detailed analysis of the signal (scale in millisecondes);
(c) Matlab 3D Fourier transform analyzis (abcissa: 0-20 kHz, ordinate: relative intensity, 3D dimension: recording at
different times); Frequencies are visualized in different colors; (d) Sigview Fourier transform: note the new harmonics
in the range of 1000-3000 Hz.
Fig. 3 A typical recording of signals from aqueous dilutions of M. Pirum (Matlab software): note the positive signals from
D-7 to D-12 dilutions.
84 Interdiscip Sci Comput Life Sci (2009) 1: 81–90
123456789 10 11 12 13
Fig. 4 Sucrose density centrifugation (35000 rpm, 2 Hr)
of a 0.02µ filtrate of Mycoplasma pirum suspension.
The collected fractions were pooled 2 by 2 and di-
luted up to D-15 and tested for EMS. The bars in-
dicate the fractions positive for EMS.
We then turned to a more classical bacterium, E.
Coli, using the laboratory strain K1.
A culture of E. Coli in agitated (oxygenated) con-
ditions, yielded 109bacterial units/mL, measured by
spectrometry. The suspension was then centrifuged at
10,000 rpm for 15 minutes, the supernatant was fil-
tered on 450 nM filter and the resulting filtrate was
filtered again on a 100 nM filter. The final filtrate was
found sterile, when plated on nutrient agar medium and
was analyzed for electromagnetic wave emission, as de-
scribed above for M. pirum. Signal producing dilutions
usually range from 10−8to 10−12, with profiles upon
Fourier transformation, similar to those of M. Pirum
(Fig. 5). In one experiment, some very high dilutions
were found positive, ranging from 10−9to 10−18. An
aliquot of the unfiltered supernatant did not show any
signals above background up to the 10−38dilution, in-
dicating again the critical importance of the filtration
step for the generation of specific signals.
The only difference with M. pirum was that no sig-
nal appeared after filtration on 20 nM filters, suggesting
that the structures associated with the signals were re-
tained by these filters and, therefore, had a size greater
than 20 nM and lower than 100 nM.
We then asked why the lower dilutions, which
logically should contain a larger number of signal-
producing structures, were “silent”. When we added
0.1 mL of a negative low dilution (e.g. 10−3) to 0.4
mL or 0.9 mL of a positive dilution (10−8), the latter
became negative. This indicate that the ”silent” low
dilutions are self-inhibitory, probably by interference of
the multiple sources emitting in the same wave length
or slightly out of phase, like a radio jamming. Alterna-
tively, the abundance of nanostructures can form a gel
in water and therefore are prevented to vibrate.
-Evidence for homologous “cross talk” between
We then wonder whether or not it was possible to
generate new signal-emitting structures from tube to
tube by using wave transfer. The following experiment
Interdiscip Sci Comput Life Sci (2009) 1: 81–90 85
Fig. 5EMS from E. Coli 0.1 µ filtrate. EMS positive from dilution D-8 to D-11: (a) Actual recording; (b) millisecond
analysis; (c) Fourier transform analysis Matlab; (d) Fourier transform analysis SigView. NF: not filtered.