1130 Biophysical Journal Volume 73 August 1997 1130-1133

Stable Magnetic Field Gradient Levitation of Xenopus laevis: Toward

Low-Gravity Simulation

James M. Valles, Jr.,* Kevin Lin,* James M. Denegre,# and Kimberly L. Mowry#
*Department of Physics and #Department of Molecular Biology, Cell Biology, and Biochemistry, Brown University, Providence,
Rhode Island 02912 USA

ABSTRACT We have levitated, for the first time, living biological specimens, embryos of the frog Xenopus laevis, using a
large inhomogeneous magnetic field. The magnetic field/field gradient product required for levitation was 1430 kG2/cm,
consistent with the embryo's susceptibility being dominated by the diamagnetism of water and protein. We show that unlike
any other earth-based technique, magnetic field gradient levitation of embryos reduces the body forces and gravity-induced
stresses on them. We discuss the use of large inhomogeneous magnetic fields as a probe for gravitationally sensitive
phenomena in biological specimens.

INTRODUCTION
Large inhomogeneous magnetic fields have been employed on it will oppose the gravitational pull of the earth. Explic-
to levitate pure solids and fluids for the elaboration of their itly, the force per unit mass acting on a pure material with
physical properties in the absence of extraneous forces a specific magnetic susceptibility Xp (units of cm3/g) is
exerted by gravity or the surfaces of a container (see, for
example, Weilert et al., 1996). Recently, Beaugnon and =Xp d(B2) - (1)
Tournier established that magnetic fields produced by ex- f 2 dz g
isting high-field solenoids can exert forces that are suffi- where g is the acceleration due to gravity, B is the applied
cient to levitate many diamagnetic organic materials magnetic field in gauss, and z is the height of the object in
(Beaugnon and Tournier, 1991a,b). Consequently, it should centimeters. By adjusting d(B2)/dz, the effective gravity,
be possible to perform magnetic field gradient levitation
geff = f, can be reduced to zero and the object levitated.
(MFGL) of biological specimens. Such a capability would Because xp is a molecular or atomic property for organic
be of interest for experiments to determine the sources of materials, every molecule or atom in the object experiences
gravitational sensitivity in biological systems. It would con- the same balance of gravitational and magnetic forces. Con-
stitute an earth-based alternative to experiments done in the sequently, the value of d(B2)/dz required for levitation is
microgravity environment of space. independent of the mass of the object. And, more pertinent
Here we present the results of an investigation of MFGL to this work, all gravity-induced internal pressure gradients
as a technique for simulating low gravity for biological or stresses in the object are absent when it is levitated. This
systems. We have stably levitated embryos of the frog state truly mimics weightlessness. By contrast, levitating an
Xenopus laevis, using the force generated by a large inho- object by floating it in a fluid does not mimic the state of
mogeneous magnetic field. We are able to show, using weightlessness. The forces that buoy the object act only at
measurements of the susceptibilities and densities of frac- the object's surface and do not affect the internal pressure
tions of the embryos, that this levitation reduces the stresses gradients and stresses. For diamagnetic materials, this lev-
on the embryos induced by gravity. In particular, levitation itation can be stable, because V2B2 ' 0 (Brandt, 1989).
decreases the body forces on the embryo by an order of The levitated state of a heterogeneous material, such as a
magnitude. biological specimen, can only approximate weightlessness.
Each type of molecule within a specimen has its own Xp and
THEORY OF MAGNETIC FIELD therefore experiences its own geff. When a heterogeneous
GRADIENT LEVITATION object levitates, the effective gravity or body force acting on
the ith constituent of the object is
The idea behind MFGL is straightforward (Beaugnon and
Tournier, 1991a,b; Brandt, 1989). If a diamagnetic material
is placed above a strong magnet, the magnetic force exerted glev,i = g(l - ) (2)
where
Received for publication 27 March 1997 and in final form 13 May 1997.
Address reprint requests to Dr. James M. Valles, Department of Physics, \Xp/ = (3)
Brown University, Providence, RI 02912. Tel.: 401-863-7559; Fax: 401-
863-2024; E-mail: [email protected]. and where Xpi and mi are the specific susceptibility and mass
( 1997 by the Biophysical Society of the ith constituent, respectively. The smaller the variation
0006-3495/97/08/1130/04 $2.00 in the Xp, the better MFGL simulates weightless conditions.
Valles et al. Magnetic Levitation of Xenopus laevis 1131

MATERIALS AND METHODS

Eggs, embryos, and fractionation side top
Frogs (Xenopus laevis) were obtained from Xenopus I, and were induced view 4\ view A
to ovulate by injection of human chorionic gonadotrophin from Sigma
Scientific. Before levitation, the eggs (see Fig. 1) were fertilized and
dejellied according to standard protocol, and the embryos were placed in a A ,1
saline solution, one-third strength MMR (33 mM NaCl, 0.67 mM KCI, 0.67
mM CaCl, 0.33 mM MgCl2, 1.67 mM HEPES, pH 7.4) (Kay and Peng,
1991; Denegre and Danilchik, 1993). The fractions were obtained by
centrifugation of homogenized embryos. The three well-separated fractions
were 1) a pellet that consisted mostly of protein in the form of yolk L -m
platelets and pieces of membrane and cortex, 2) water-rich cytosolic
supematant material, and 3) lipids.
mirror -~~~~~~
.J
embryos ruler
Susceptibility and density measurements of the a
embryos and fractions
The xp of the fractions and (Xp) of the whole embryos with only a small
amount of saline were measured using MFGL. At a fixed vertical position
z, the levitating force per unit mass is directly proportional to xp and the
square of the current, because the magnetic field and its gradient are each
proportional to the current. We obtained Xp values relative to that of
double-distilled water by measuring the current required to levitate double- ..I
distilled water at a given z and the current required to levitate the materials '9% ks...
(fractions and the embryos) at the same z, and taking the ratio of the o_:.,
L
squares of the currents. We measured this ratio for a range of z for each b
material and took an average to produce the Xp values. The height mea-
surements were made by sighting the vertical position of the center of the
material relative to a ruler placed within the bore, as shown schematically
in Fig. 2 a. A Questar telescope with a close-focus barrel was used for the
sighting. With a height resolution of 0.5 mm, we were able to measure xp
to a fraction of a percent.
The droplets with frog eggs were placed in air, in the bore, of the Bitter
solenoid at the Francis Bitter National Magnet Laboratory in a few steps.
First a droplet was levitated and embryos were then added to it. Later,
changes in the volume fraction of embryos in a levitated droplet were d
accomplished by either removing saline from the droplet with a pipette or
by adding more embryos. The current in the solenoid was adjusted "on the
fly" to maintain stable levitation. The levitation is stable in a region on the
axis of the solenoid 8-10 cm above its centerline. Below 8 cm, the
levitation is unstable to axial perturbations, and above 10 cm it is unstable
to radial perturbations. The eggs were allowed to develop through first and,
sometimes, third cleavage while levitated. Photographs were taken through
a Questar telescope with a 35-mm camera.
The densities and relative volume fractions of the fractions were mea-
sured by standard techniques. f 9
RESULTS FIGURE 2 (a) Geometry for taking photographs of saline droplets con-
taining fertilized eggs. A mirror was placed in the bore, in the region of
To demonstrate the feasibility of MFGL as a low-gravity stable levitation, to allow simultaneous viewing of the top and side of a
simulation technique, we levitated living biological speci- sample from the top of the bore. (b-g) Photographs of levitated embryos.
mens. Frog embryos were chosen for preliminary experi-
Top views (b,d,f) and side views (c,e,g) of droplets with quantities of
embryos and average susceptibilities as follows: (b,c) three embryos, (X) =
ments because previous investigations suggested that mag- -0.72 x 10-6 cm3/g; (d,e) -55% volume fraction of embryos, (X) =
-0.70 X 10-6 cm3/g; (fg) -95% volume fraction of embryos (X) =
-0.69 X 10-cm3/g. The horizontal rulings in c, e, and g are spaced by 1
A mm. The embryos are -1.2 mm in diameter and appear nonspherical
because of the lensing effects of the droplet.

V lyolk mass
cross section
netic fields as large as 7 Tesla do not adversely affect their
FIGURE 1 On the left is shown a sketch of a Xenopus embryo before development (Jacobs and Fraser, 1994) and because their
first cleavage, with a cross section depicted on the right. The orientation of early development exhibits well-characterized sensitivity to
the AV axis is indicated at the left. gravity (Ancel and Vintemberger, 1948; Gerhart et al.,
1132 Biophysical Journal Volume 73 August 1997

1981; Black and Gerhart, 1985; Cooke, 1986; Vincent and 1991a,b; Pollard, 1965). When levitated, the apparent
Gerhart, 1987; Neff et al., 1993). It should be noted, how- weight of the proteins in the embryo is reduced by the factor
ever, that frog eggs fertilized and raised in the weightless
environment of space develop into normal frogs (Souza et Ulev Ppgiev,p Ps glev,s
al., 1995). In addition, the frog egg has a single axis of (4)
U
(pp-Ps)g
asymmetry, the AV axis (see Fig. 1), which is normally
oriented parallel with gravity (pigmented side up). In Fig. 2 where pp and p. are the densities of the pellet and supema-
we show a schematic of the experimental setup and pictures tant material, respectively. Our data indicate that this factor
of droplets of buffered saline solution containing different is 0.7.
volume fractions of fertilized frog eggs stably levitated in The reduction in the body forces and gravitational
the bore of a Bitter solenoid at the Francis Bitter National stresses achieved with MFGL on Xenopus laevis has not
Magnet Laboratory. The droplets assume a spherical shape, been matched by any other ground-based, low-gravity sim-
indicating that gravitational stresses on them have been ulation technique. The most commonly employed tech-
reduced below the strength of surface tension effects. Lev- nique, clinostat rotation (CR), involves rotation of speci-
itation of a pure droplet occurs at a magnetic field/field mens about an axis perpendicular to the direction of the
gradient product of 1370 kGauss2/cm. The field strength in gravitational force (Neff et al., 1993). This procedure only
this region is -130 kGauss. The embryos reside at the vector averages the gravity-induced forces to zero. It does
bottom of the droplet (see Fig. 2, c and e), indicating that the not eliminate them. In fact, the centripetal acceleration of
saline experiences a net body force up and the embryos specimens in CR, when large enough, can potentially mimic
experience a net body force down. gravity, and thus work counter to the intended effect. In
Nevertheless, the body forces on the constituents of the addition, the rotation can cause movement of fluids within
embryos are reduced significantly in the levitated state. We specimens. The resulting mass transport may interfere with
demonstrate this using measurements of both the Xp and investigations of convective processes in biological systems
densities of fractions of the embryos. The results of these (Albrecht-Buehler, 1992).
measurements and the calculated values of 91evj for the Large inhomogeneous magnetic fields can also be used to
fractions are compiled in Table 1. The significant result is continuously vary gravitational stresses. For example, the
that the body forces on each of the three constituents of the stresses on a Xenopus embryo that levitates at z = +9 cm
embryo are reduced by more than a factor of 10 from their relative to the solenoid center can be approximately doubled
unlevitated values. from their normal values simply by putting the embryo at
z = -9 cm. The normal force would be twice and the
buoyant force factor would be 1.3 times their unlevitated
values. Or, in principle, the buoyant force, like the normal
DISCUSSION force, can also be reduced to zero. For embryos, the field-
This reduction in body forces alters the forces and stresses field gradient product at which Ulev/U goes to zero is higher
that may play a role in the gravitational sensitivity of frog than the value required for levitation. When Ulev/U is zero,
embryos or other biological systems. First, levitation re- all gravitationally induced concentration gradients and
duces to zero the normal forces that support the embryo. stresses have equilibrium values of zero. The normal contact
Second, buoyant forces, induced by gravity and variations force is not zero, however, and in fact, pushes down on the
in the mass densities of the materials in the embryo, are embryo.
different in the levitated and unlevitated states. The stresses An important issue that must be addressed when applying
present in the supporting structure of the embryo depend on MFGL is the effects of large magnetic fields on the biolog-
the apparent weight of the constituents of the embryo. ical systems under study. These effects can be investigated,
Moreover, gradients in the concentration of molecules in independent of the magnetic force effects, by placing sam-
solution depend on buoyant forces (Beaugnon and Toumier, ples at the center of the solenoid, where the magnetic field

TABLE 1 xp and densities of fractions of the embryos

-Xp 106 Density Volume Mass
Material (cm3/g) (g/cm3) fraction (%) fraction (%) (g1ev,/g)
Double-distilled water 0.720* 1.00
Buffered saline (MMR) 0.720 1.01
Embryos# 0.690 1.09 96 95 <0.003
Pellet fraction 0.638 1.18 29 32 0.075
Cytosolic fraction 0.703 1.03 66 64 -0.02
Lipid fraction 0.646 0.85 5 4 0.06
*This value for the susceptibility of double-distilled water was assumed and used as a standard for the other susceptibility measurements (Ohlendorf, 1981).
#Approximately 95% volume fraction of embryos; see Fig. 2, f and g.
Valles et al. Magnetic Levitation of Xenopus laevis 1133

is maximum and the field gradient is zero. In fact, we have Barone, R. P., and L. D. Caren. 1984. The immune system: effects of
found that homogeneous magnetic fields as large as those hypergravity and hypogravity. Aviat. Space Environ. Med. 55:
1063-1068.
we employed for levitation can cause some changes in the
early development of a portion of the embryos (Denegre et Beaugnon, E., and R. Toumier. 1991a. Levitation of organic materials.
Nature. 349:470.
al., manuscript in preparation). It remains to be shown,
Beaugnon, E., and R. Tournier. 1991b. Levitation of water and organic
however, whether these alterations will have a significant substances in high static magnetic fields. J. Phys. III France.
impact on the efficacy of MFGL for investigating the 1:1423-1428.
sources of gravitational sensitivity in Xenopus. Black, S., and J. C. Gerhart. 1985. Experimental control of the site of
embryonic axis formation in Xenopus laevis eggs centrifuged before first
cleavage. Dev. Biol. 108:310-324.
SUMMARY Brandt, E. H. 1989. Levitation in physics. Science. 243:349-355.
The decrease in gravitational stresses and the order-of- Cooke, J. 1986. Permanent distortion of positional system of Xenopus
magnitude reduction in the body forces that we have embryo by brief early perturbation in gravity. Nature. 319:60-63.
achieved with MFGL on a biological specimen have not Denegre, J. M., and M. V. Danilchik. 1993. Deep cytoplasmic rearrange-
ments in axis-respecified Xenopus embryos. Dev. Biol. 160:157-164.
been matched by any other ground-based technique. Be-
cause the materials composing most biological systems are Gerhart, J., G. Ubbels, S. Black, K. Hara, and M. Kirschner. 1981. A
reinvestigation of the role of the grey crescent in axis formation in
similar to those in frog embryos, we expect that levitation Xenopus. Nature. 292:511-516.
can reduce the gravitational stresses in a wide range of Jacobs, R. E., and S. E. Fraser. 1994. Magnetic resonance microscopy of
systems. These results suggest that MFGL can serve as a embryonic cell lineages and movements. Science. 263:681-684.
unique tool for investigating how biological systems trans- Kay, B., and H. G. Peng. 1991. Methods in Cell Biology, Vol. 36, Xenopus
duce gravitational information. laevis: Practical Uses in Cell and Molecular Biology. Academic Press,
San Diego.
Neff, A. W., H. Yokota, H.-M. Chung, M. Wakahara, and G. M. Malacin-
We gratefully acknowledge seminal discussions with Professor H. Maris, ski. 1993. Brief notes: early amphibian (Anuran) morphogenesis is
helpful comments from Professor C. Elbaum, and the expert assistance of sensitive to novel gravitational fields. Dev. Bio. 155:270-274.
the personnel at the Francis Bitter National Magnet Laboratory. J. Mandell
Ohlendorf, D. H. 1981. CRC Handbook of Chemistry and Physics. CRC
kindly loaned the camera. Press, Boca Raton, FL.
This work has been partially supported by National Science Foundation Pollard, E. C. 1965. Theoretical studies on living systems in the absence of
grants DMR-9122268 and DMR-9502920 and a Salomon Faculty Research mechanical stress. J. Theor. Biol. 8:113-123.
Award.
Souza, K. A., S. D. Black, and R. J. Wassersug. 1995. Amphibian devel-
opment in the virtual absence of gravity. Proc. Natl. Acad. Sci. USA.
92:1975-1978.
REFERENCES
Vincent, J. P., and J. C. Gerhart. 1987. Subcortical rotation in Xenopus
Albrecht-Buehler, G. 1992. The simulation of microgravity conditions on eggs: an early step in embryonic axis specification. Dev. Biol. 123:
the ground. ASGSB Bull. 5:3-10. 526-539.
Ancel, P., and P. Vintemberger. 1948. Recherches sur le determinisme de Weilert, M. A., D. L. Whitaker, H. J. Maris, and G. M. Seidel. 1996.
la symetrie bilaterale dans l'oeuf des amphibiens. Bull. Biol. Fr. Belg. Magnetic levitation and noncoalescence of liquid helium. Phys. Rev.
Suppl. 31:1-82. Lett. 77:4840-4843.