Spatial Distribution of
Spermine/Spermidine Content and
K?-Current Rectification in Frog
Retinal Glial (Mu ¨ller) Cells
SERGUEI N. SKATCHKOV,1MISTY J. EATON,1JAN KRUSˇEK,2
RU¨DIGER W. VEH,3BERND BIEDERMANN,5ANDREAS BRINGMANN,5
THOMAS PANNICKE,5RICHARD K. ORKAND,4
1CMBN, Department of Biochemistry, School of Medicine, Universidad Central del Caribe,
Bayamon, Puerto Rico
2Institute of Physiology, Czech Academy of Science, Prague, Czech Republic
3Institute of Anatomy, Humboldt University, Berlin, Germany
4Institute of Neurobiology, University of Puerto Rico, MSC, San Juan, Puerto Rico
5Institute of Brain Research, Leipzig University, Leipzig, Germany
AND ANDREAS REICHENBACH5*
glia; retina; potassium channels; polyamines; immunocytochemistry;
the dominant membrane currents are mediated by K?inward-rectifier (Kir) channels
(Newman and Reichenbach, Trends Neurosci 19:307–312, 1996), and (2) rectification of
these Kir channels is due largely to a block of outward currents by endogenous poly-
amines such as spermine/spermidine (SPM/SPD) (Lopatin et al., Nature 372:366–369,
1994). In frog Mu ¨ller cells, the degree of rectification of Kir-mediated currents is
significantly higher in the endfoot than in the somatic membrane (Skatchkov et al., Glia
27:171–181, 1999). This article shows that in these cells there is a topographical
correlation between the local cytoplasmic SPM/SPD immunoreactivity and the ratio of
inward to outward K?currents through the surrounding membrane area. Throughout
the retina, Mu ¨ller cell endfeet display a high SPM/SPD immunolabel (assessed by
densitometry) and a large inward rectification of K?currents, as measured by the ratio
of inward to outward current produced by step changes in [K?]o. In the retinal periphery,
Mu ¨ller cell somata are characterized by roughly one-half of the SPM/SPD immunoreac-
tivity and K?-current rectification as the corresponding endfeet. In the retinal center,
Mu ¨ller cell somata are virtually devoid of both SPM/SPD immunolabel and K?-current
inward rectification. Comparing one region of the retina with another, we find an
exponential correlation between the local K?rectification and the local SPM/SPD con-
tent. This finding suggests that the degree of inward rectification in a given membrane
area is determined by the local cytoplasmic polyamine concentration. GLIA 31:84–90,
Previous studies in retinal glial (Mu ¨ller) cells have suggested that (1)
© 2000 Wiley-Liss, Inc.
Inwardly rectifying potassium (Kir) channels as well
as glutamate receptors (NMDA and AMPA types) and
nicotinic acetylcholine receptors are regulated by the
endogenous polyamines, spermine, spermidine (SPM/
SPD) and putrescine (reviewed by Nichols and Lopatin,
1997; Williams, 1997). Neuroglial cells express Kir
channels (Sontheimer, 1994) and glutamate receptors
(Steinha ¨user and Gallo, 1996) and contain polyamines
(Laube and Veh, 1997; Biedermann et al., 1998; Gilad
et al., 1999). In the retina of several mammalian spe-
cies, endogenous SPM/SPD immunoreactivity was lo-
Grant sponsor: National Institute of Health; Grant number: RCMI-RR03035
Projects A and B; Grant sponsor: NINDS; Grant sponsor: FIRCA; Grant sponsor:
M-RISP; Grant sponsor: National Science Foundation; Grant sponsor: EPSCoR;
Grant sponsor: RIMI; Grant sponsor: German Bundesministerium fu ¨r Bildung,
Forschung und Technologie (BMB?F); Grant number: 01KS9504 Project 5; Grant
sponsor: Deutsche Forschungsgemeinschaft; Grant number: DFG-Ve-187-1-2;
Grant sponsor: Czech Ministry of Education; Grant number: VS 97099.
*Correspondence to: Andreas Reichenbach, Department of Neurophysiology,
Paul Flechsig Institute for Brain Research, Leipzig University, Jahnallee-59,
D-04109, Leipzig, Germany. E-mail: email@example.com
Received 2 February 2000; Accepted 29 March 2000
GLIA 31:84–90 (2000)
© 2000 Wiley-Liss, Inc.
cated exclusively in Mu ¨ller cells (Biedermann et al.,
1998). In the salamander, Valentino et al. (1996) de-
scribe SPM/SPD immunolabeling in the inner retinal
layer(s), which they ascribe mainly to retinal ganglion
cells. The present study was aimed at studying possible
correlations between polyamine content and inward
rectification of Kir-mediated currents.
Frog Mu ¨ller cells were chosen because their mem-
brane currents are carried virtually exclusively by Kir
channels (Skatchkov et al., 1995). Kir channels are
dominant in Mu ¨ller cells from all species studied so far
(Newman and Reichenbach, 1996) but, particularly in
mammalian cells, there is a variable contribution of
other types of K?channels (Chao et al., 1994; Bring-
mann et al., 1997; Kusaka et al., 1998). Mu ¨ller cell Kir
channels are strongly rectifying when studied by cell-
attached patch-clamp recordings (i.e., at symmetrical
high [K?]), but they do permit significant K?outward
currents at physiological extracellular K?(Brew et al.,
1986; Newman, 1993; Skatchkov et al., 1995). In am-
phibian Mu ¨ller cells, the membrane of the endfoot (fac-
ing the vitreous body) expresses a much higher density
of Kir channels than that of the soma (Brew et al.,
1986; Newman, 1993; Skatchkov et al., 1995; Rojas and
Orkand, 1999). The degree of inward rectification dif-
fers between endfoot and soma membranes (Skatchkov
et al., 1999). In particular, endfoot membranes display
stronger inward rectification of K?currents than so-
matic cell membranes when tested with step increases
or decrease of [K?]o. There is also evidence that this
endfoot-soma difference in inward rectification is
larger in cells from the central frog retina than in the
retinal periphery (Skatchkov et al., 1999). Thus, frog
Mu ¨ller cells display significant intracellular (endfoot
vs. soma) and intercellular (retinal center vs. periph-
ery) heterogeneity in Kir current rectification. The cur-
rent study addresses the question of whether similar
heterogeneity exists in the cytoplasmic distribution of
SPM/SPD. A preliminary report of the data has been
presented in abstract form (Skatchkov et al., 1998).
MATERIALS AND METHODS
Immunocytochemical Detection of SPM/SPD
Adult frogs (Rana pipiens) were anesthetized on ice
and decapitated; their eyes were enucleated and fixed
in a solution of 4% paraformaldehyde, 0.05% glutaral-
dehyde and 0.2% picric acid in 0.1 M phosphate buffer,
pH 7.4 for 16 hours. The fixation was followed by im-
mersion in 0.15 M sucrose in 0.1 M phosphate buffer,
pH 7.4 (for 24 h), 0.5 M sucrose (for 48 h) and freezing
in a dry ice/ethanol bath at ?80°C; 25-?m cryostat
sections of retinae were pretreated with 1% sodium
borohydride in phosphate-buffered saline (PBS) for 15
min, and subsequently permeabilized with 0.3% Triton
X-100 for 30 min. After incubation with the primary
antibody (affinity-purified anti-SPM/SPD antibody, cf.
Laube and Veh, 1997) for 36 h at ?4°C, freely floating
sections were treated with the secondary antibody (bi-
otinylated goat anti-rabbit IgG, 1:2,000, Vector Labo-
ratory) for 18 h and with an ABC complex (Vectastain
Elite, 1:1000, Vector Laboratory) for 6 h. Peroxidase
activity was demonstrated with 1.4 mM 3,3?-diamino-
benzidine, 10 mM imidazole, 0.3% nickel ammonium
sulfate and 0.015% H2O2in 50 mM Tris-HCl, pH 7.6 for
3 min at room temperature. Controls were obtained by
omitting the primary antibody.
Semiquantitative Analysis of SPM/SPD
Immunolabeled retinae (a total of 18 retina sections;
each 3 sections from the 6 eyes of the 3 animals used)
were analyzed by using a microscope (Zeiss, Germany,
?50 oil immersion lens) with an interactive imaging
analysis system (SIS, Mu ¨nster, Germany). Labeled
Mu ¨ller cells (or parts of them) were identified by their
location and shape (for details, see Newman and
Reichenbach, 1996; Biedermann et al., 1998). Briefly,
Mu ¨ller cell endfeet form funnel-shaped compartments
directly abutting the vitreous body and continue as
thin fibers into the inner plexiform layer; at the high
magnification used, they were easily distinguished
from the spherical ganglion cells which are located
distant from the vitreous. The elongated Mu ¨ller cell
somata are located in the inner third of the inner
nuclear layer and extend their two opposite stem pro-
cesses into both inner and outer plexiform layer,
whereas the neighboring bipolar and amacrine cells
have spherical somata. Light absorption was measured
along lines crossing either the endfeet or the somata
and a 5-?m long part of the surrounding unstained
tissue. In order to find such unstained neighboring
areas (generally in the adjacent inner plexiform layer),
the analyzed lines had to be placed at various oblique
angles. The difference in absorption between Mu ¨ller
cell (compartment) and background was used as a mea-
sure of labeling intensity. Three independent measure-
ments were performed for every labeled compartment
(endfoot or soma; there was negligible variation within
such series of measurements); at least five independent
neighboring compartments were measured at every
retinal region studied (represented by the various so-
ma-endfoot distances at the analyzed regions in Fig.
2A). Maximal optical density was taken as 100% to
normalize the data. All data were calculated and plot-
ted using SigmaPlot, mean values with standard errors
are given in the diagrams.
Dissociation of Living Cells for
Mu ¨ller cells were isolated from the retina of adult
Rana pipiens by a modification (Skatchkov et al., 1999)
of the technique used by Chao et al. (1994). Briefly, the
animals were anesthetized on ice in the dark for two
hours and then decapitated. The eyes were excised, and
K?CHANNEL RECTIFICATION IN RETINAL GLIAL CELLS
the retinal periphery was separated from the central
retina. Retinal pieces from both sources were put in 0.1
mg/ml Nagarse (subtilisin, E.C.188.8.131.52 Protease type
XXVII, Sigma Chemical Co., Deisenhofen, Germany) in
Ca-Mg-free PBS for 1 h, at 23 °C. After washing in
Leibovitz medium (L-15, Gibco) diluted by 27% with
H2O, the slices were rinsed 3 times in diluted L-15
containing 0.02 mg/ml DNase-I (Sigma, D-4263 from
bovine pancreas), and washed in DNase-free L-15 me-
dium. Dissociated Mu ¨ller cells were transferred to a
small (0.03-ml) recording chamber on the stage of an
inverted microscope (Nikon Diaphot-TMD, Nikon, Ja-
pan) and used for electrophysiological recording in the
extracellular solution (ECS) containing mM NaCl 115
mM, KCl 3 mM, CaCl22 mM, MgCl21 mM, and Hepes
5 mM; the pH was adjusted with NaOH to 7.4.
Whole Cell Electrophysiology
Membrane currents were measured with the single
electrode patch-clamp technique using an Axopatch-1B
or Axopatch-200 patch amplifier with a CV-3 headstage
or CV-203BU (Axon Instruments); high frequencies
(?10 kHz) were cut-off. P-Clamp software (Axon In-
struments; versions 5.5, 6, and 7), and a Dell-425s or
GP6-266 computer was used to acquire, store, and an-
alyze the data. Signals were digitized by an interface
(TL-1 DMA or Digidata 1200, Axon Instruments).
Two Narishige hydraulic micromanipulators (Nar-
ishige, MMW-203, Japan) were used for (1) voltage-
clamp recording, and (2) positioning a theta double-
barrel micropipette, 10-?m tip diameter, containing 1
mM and 10 mM [K?] ECS test solutions for local K?
application. A third micromanipulator (Narishige,
MP-2, Japan) was used to position the tube for total
(whole-cell) perfusion. The cells with the pipette tip
placed at the endfoot or at the soma, were put in front
of the total outlet tube (200-?m diameter, flow 1.7 ?l/s)
at a distance of less than 150 ?m, to provide continuous
superfusion by 3 mM [K?] ECS solution. In order to
apply test solutions from the outlet of the theta tube,
the 3-mM [K?] total perfusion was interrupted without
delay. A five valve system for fast drug application (MS
Concentration Clamp; VS-2001, Vibraspec, PA), con-
trolled by a second computer, was connected to the
Electrodes were pulled from 1-mm-diameter soft-
glass tubes (Fisher Scientific, Cat. No 02-668-68),
and filled with intracellular solution containing KCl
90 (or rarely 120 mM), MgCl20.8 mM, CaCl21 mM,
EGTA 5, Hepes 10 mM, and Na2ATP 3 mM; pH was
adjusted to 7.1 with N-methyl-D-glucamine. After
filling, their resistances were 4-6 M?, and after
membrane rupture, the access resistance was 10-15
M?. The liquid-junction potentials (?2 to ?5 mV)
were compensated to zero after the pipette touched
the bath solution; as they were small compared with
the ?[K?]e-induced Emchanges used in our study,
they were not further considered. Only cells with
membrane potentials close to EKwere accepted for
this study; these cells were kept at holding potentials
equal to their native membrane potentials. Through-
out the study, currents were recorded from freshly
penetrated cells only, i. e., within the first two min-
utes after establishment of the whole-cell configura-
tion (see Discussion, for details). Capacity and series
resistances were compensated to the maximal possi-
ble extent, usually not exceeding 30%.
Kir Current Stimulation by Local K?
In order to assess the inward rectification of the K?cur-
rents, application of solutions with enhanced (10 mM for
inward currents; Iin) or lowered (1 mM for outward cur-
rents; Iout) potassium (K?steps) was used (Skatchkov et
and from 3 mM to 1 mM (i.e., well within the physiolog-
ical range) produce comparable changes in the driving
for the outward current) but evoke Kir currents with
clear inward rectification in Mu ¨ller cells (Skatchkov et
al., 1999). Furthermore, the technique of K?steps al-
lowed rapid (50-ms) changes of [K?]oaround the cell
compartment being studied (soma or endfoot) without
participation of other cell areas.
Immunostaining for Spermine/Spermidine
Figure 1 shows that the SPM/SPD immunoreactivity
varied along sections of the adult frog retina. Intense
immunoreactivity was visible in many cells at the pe-
ripheral retinal margin where cells are generated and
differentiate life-long in amphibians (Fig. 1A), whereas
the central retina was virtually devoid of label, with
the exception of the innermost layer(s) (GCL in Fig.
1E). Throughout the retina, strongly labeled spherical
cell somata were found in the ganglion cell layer (GCL);
this corresponds to the observations of Valentino et al.
(1996), who found SPM/SPD immunoreactivity in the
ganglion cells of salamander retinae. Furthermore,
faint immunoreactivity was displayed by numerous
spherical cell somata in the inner nuclear layer (INL),
probably representing a subpopulation of bipolar cells.
Somata in the outer layers of the retina (ONL) showed
distinct label only in the far periphery (Fig. 1A), where
virtually all immature cells displayed strong SPM/SPD
immunolabel (left margin of Fig. 1A). The dark color of
retinal pigment epithelial cells (RPE) was due to their
melanin granula, and was visible also in the control
sections (Fig. 1A?, bottom); no immunolabel was
present in these cells.
In addition to immature retinal cells and the
above-mentioned types of retinal neurons, Mu ¨ller
(glial) cells were strongly immunoreactive for SPM/
SPD. From the very peripheral margin up to the
SKATCHKOV ET AL.
retinal center, virtually all Mu ¨ller cell endfeet were
intensely labeled (black arrowheads in Fig.1A–E).
Also, the adjacent vitread stem processes of Mu ¨ller
cells showed more or less distinct label throughout
the retina (white arrows in Fig. 1A–E). The elon-
gated somata of Mu ¨ller cells were strongly labeled in
the retinal periphery (white arrowheads in Fig.
1A–C) but, approaching the more central areas,
gradually became less intensely labeled. They were
faintly labeled near the retinal center and devoid of
label in the very center (white arrowheads in Fig.
1D,E). The labeling intensity of Mu ¨ller cell somata
and endfeet was assessed semiquantitatively by den-
sitometry (see Materials and Methods). The results
are given in Figure 2A, where the distance between
endfoot and soma of the cells is used to indicate their
area of origin (maximal distance represents retinal
center, short distance indicates periphery; 100%
?110 ?m). The labeling intensity of the endfeet de-
creases only slightly toward the retinal center,
whereas the somata are devoid of visible label when
the soma-endfoot distance is about 80 ?m or more. In
the midperiphery (soma-endfoot distances of 60–70
?m), the large standard errors are caused by the
intense labeling of some somata, whereas others are
The results of the electrophysiological experiments
are shown in Figure 2B,C. When high or low K?steps
were applied, inward and outward currents were
evoked through Kir channels, respectively (for details,
see Skatchkov et al., 1999). Cells were clamped at
either soma or endfoot, where local K?changes were
applied. In peripheral cells (soma-endfoot distance of
40 ?m in the example illustrated), the amplitude of the
inward currents (Iin) always exceeded that of the out-
ward currents (Iout), both in the endfoot (Iin/Iout? mean
?SD 5.98 ? 0.79, n ? 7) and in the soma (Iin/Iout?
3.74 ? 0.19; n ? 16) (upper row of traces in Fig. 2B). A
similar relationship (Iin/Iout? 4.50 ? 0.19; n ? 9) was
observed in endfoot membranes of cells from the cen-
tral retina (soma-endfoot distance of 90 ?m in the
example illustrated; lower traces in Fig. 2B). However,
the soma membrane of such long cells from retinal
center showed almost no rectification (Iin/Iout? 1.20 ?
0.08, n ? 9).
Detailed data from cells of various lengths (soma-
endfoot distances; 100% ?120 ?m) are summarized in
Figure 2C. Endfoot membranes displayed strong in-
ward rectification in cells from all retinal regions; only
in very long cells from retinal center there was a slight
but significant (P ? 0.0001) decrease of the Iin/Ioutratio
(open circles in Fig. 2C). By contrast, the soma mem-
branes showed less inward rectification even in periph-
eral cells, and the degree of inward rectification de-
creased with increasing cell lengths down to Iin/Iout
ratios of close to 1 in cells from retinal center (i.e., with
soma-endfoot distances of more than about 80 ?m;
filled circles in Fig. 2C).
Fig. 1. Spermine/spermidine (SPM/SPD) immunoreactivity in reti-
nal sections from adult Rana pipiens. The insert (top) indicates the
retinal regions A: Far periphery. B–E: Retinal center, from which
retinal microphotographs are shown. A?: Negative control (omission of
first antibody) from retinal periphery. Black arrowheads, Mu ¨ller cell
endfeet; white arrows, vitread Mu ¨ller cell stalks; white arrowheads,
Mu ¨ller cell somata. GCL, ganglion cell layer; IPL, inner plexiform
layer; INL, inner nuclear layer; OPL, outer plexiform layer; ONL,
outer nuclear layer; RPE, retinal pigment epithelium.
K?CHANNEL RECTIFICATION IN RETINAL GLIAL CELLS
The major findings of this study are the following.
First, frog Mu ¨ller cell endfeet contain more SPM/SPD
immunoreactivity than found in the somata. Second,
somatic SPM/SPD immunoreactivity is moderately
strong in peripheral cells but gradually and completely
disappears toward the retinal center. Third, the distri-
bution of SPM/SPD is mirrored by local membrane
inward rectification, which is stronger in endfeet than
in somata, and is almost absent in somata from central
Mu ¨ller cells.
While absolute SPM/SPD concentrations in Mu ¨ller
cells are unknown to the best of our knowledge, the
optical density data (Fig. 2A) may reasonably be as-
sumed to reflect these concentrations as a result of
such factors as proportionality due to primary antibody
binding and secondary antibody amplification. On the
basis of this assumption, we graphed optical density of
SPM/SPD immunostaining versus Iin/Ioutratio (on a log
scale) for endfeet or somata (of groups of cells with the
same soma-endfoot distances) in Figure 3. This proce-
dure demonstrated a strong correlation (r2? 0.9306)
between local SPM/SPD concentration and inward rec-
tification. The data from endfeet (empty circles) and
from somata (filled circles) fall on the same line (Fig. 3).
This supports the idea that the two parameters are
related not just accidentally but causally.
Several factors (Mg2?, Ca2?, H?, ATP) in addition to
SPM/SPD may exert cytoplasmic control over Kir chan-
nel rectification in the adjacent membrane; in particu-
lar, Mg2?has significant effects (Ficker et al., 1994;
Lopatin et al., 1994; Fakler et al., 1995; Biedermann et
al., 1998). While the cytoplasmic concentration of free
Mg2?in Mu ¨ller cells is presently unknown, we present
evidence for local differences in SPM/SPD content,
which may account for the observed differences in in-
ward rectification. The strongest argument in favor of
the polyamines is that these substances (SPM/SPD)
have been shown to be more potent in producing in-
ward rectification than other tested substances by two
Fig. 2. Dependence of spermine/spermidine (SPM/SPD) immunore-
activity (A) and inward rectification of Kir channel-mediated currents
(B,C) on both cellular compartment (soma vs. endfoot) and length
(soma-endfoot distance) of frog Mu ¨ller cells. The soma-endfoot dis-
tance of Mu ¨ller cells is used as an indicator of the retinotopographic
origin of the cells (maximal distance represents retinal center, short
distance indicates periphery; for details, see Results). A: Optical den-
sitometry of SPM/SPD immunoreactivity of individual Mu ¨ller cells in
retinal sections; each point represents measurements of at least three
neighboring cells at a given retinal site. The relative absorption is
given as percentage of maximal optical density. B: Original traces
from isolated Mu ¨ller cells in response to K?steps denoted in the
bottom trace. Outward currents were evoked by lowered K?(1 mM),
inward currents by elevated K?(10 mM). The peripheral cells had
soma-endfoot distances of about 40 ?m, the central cells of ?120 ?m.
Iin/Iout? ratio of inward-to-outward current amplitudes. C: Iin/Iout
ratios of cells with different soma-endfoot distances, measured as
illustrated in B. For both retinal sections (A, 100% ? 110 ?m) and
isolated cells (C, 100% ? 120 ?m) the soma-endfoot distances were
plotted as relative values, in order to facilitate comparison despite
shrinkage artifacts in the sections, as well as to compensate for slight
inter-individual differences between the frogs used. Numbers of cells
are given for each data point.
SKATCHKOV ET AL.
to three orders of magnitude; for instance, while SPM
is effective in nanomolar/micromolar concentrations,
the concentration of Mg2?must be within the millimo-
lar (mM) range to achieve a similar change in rectifi-
cation (Ficker et al., 1994; Lopatin et al., 1994; Nichols
and Lopatin, 1997). This would also mean that the
observed local differences in inward rectification (Fig.
2B, C) would require the generation of large concentra-
tion gradients of, e.g., Mg2?within the cytoplasm of a
cell, which may be difficult to maintain in the case of
diffusible ions. However, differences in rectification
may be produced by only relatively small concentration
gradients of polyamines and may be maintained by
locally different turnover rates.
This idea of a local cytoplasmic regulation of Kir
inward rectification is further supported by the obser-
vation that even in endfeet of peripheral Mu ¨ller cells
the strong inward rectification disappeared with a de-
lay of several minutes after establishing the whole-cell
configuration (i.e., after dialysis of the cytoplasm by the
pipette solution); the loss of inward rectification was
essentially due to increased outward current ampli-
tudes, whereas the amplitudes of the inward currents
remained unchanged (Serguei N. Skatchkkov, unpub-
lished data). Thus, throughout our study currents were
recorded only from very freshly penetrated cells (with-
in the first 2 min after rupture of the membrane).
Why frog Mu ¨ller cells need locally different inward
rectification ratios of their Kir channels remains un-
solved. Potassium homeostasis is a predominant func-
tion of glial cells (Orkand et al., 1966), particularly in
the retina (reviewed by Newman and Reichenbach,
1996). While some authors presented arguments for
the idea that inward rectification of K?channels
should enhance the efficacy of spatial buffering of K?
ions (Newman, 1993; Amedee et al., 1997), others have
argued that linear K?channels may be well suited for
some aspects of the spatial-buffering current circuit
(e.g., Nilius and Reichenbach, 1988). In the case of the
somatic membrane, the functional requirements for
inward rectification may differ according to the local
tissue geometry (Skatchkov et al., 1999), as its distance
from the vitreous body (i.e., the “sink” of spatial buff-
ering currents) varies greatly between retinal center
and periphery. By contrast, throughout the retinal to-
pography the endfoot membrane is virtually equidis-
tant from both vitreous (K?sink) and retinal ganglion
cells and their axons (K?sources); the spiking proper-
ties of the latter elements may require strong inward
rectification of the adjacent glial cell membrane (New-
man, 1993; Amedee et al., 1997) in both central and
peripheral Mu ¨ller cells.
This research was supported by grants from the Na-
tional Institutes of Health, RCMI-RR03035 Project A (to
S.N.S.) and Project B (to M.J.E.); NINDS, FIRCA, and
M-RISP (to R.K.O.); from the National Science Founda-
tion (EPSCoR, RIMI (to R.K.O.); from the German
Bundesministerium fu ¨r Bildung, Forschung und Tech-
nologie (BMB?F), 01KS9504 Project 5 (to A.R.); from the
Deutsche Forschungsgemeinschaft, DFG-Ve-187-1-2 (to
R.W.V.); and from the Czech Ministry of Education, VS
97099 (to J.K.). The authors thank Dr. Viktoria Vlachova
for installation of the quick perfusion system, Ms. Clau-
dia Heckel for performing many of the densitometric
measurements, and Mr. Jens Grosche, Ms. Heike Hei-
lmann and Dr. Mareike Wenzel for technical assistance.
Amedee T, Robert A, Coles JA. 1997. Potassium homeostasis and glial
energy metabolism. Glia 21:46–55.
Biedermann B, Skatchkov SN, Bringmann A, Pannicke T, Veh R,
Bernstein H-G, Reichenbach A. 1998. Spermine/spermidine is ex-
pressed by retinal glial (Mu ¨ller) cells, and controls distinct K?
channels of their membrane. Glia 23:209–220.
Brew H, Gray PTA, Mobbs P, Attwell A. 1986. Endfeet of retinal glial
cells have higher densities of ion channels that mediate K?buffer-
ing. Nature 324:466–468.
Bringmann A, Faude F, Reichenbach A. 1997. Mammalian retinal
glial (Mu ¨ller) cells express large-conductance Ca2?-activated K?
channels that are modulated by Mg2?and pH, and activated by
protein kinase A. Glia 19:311–323.
Chao TI, Skachkov SN, Eberhardt W, Reichenbach A. 1994. Na?
channels of Mu ¨ller (glial) cells isolated from retinae of various
mammalian species including man. Glia 10:173–185.
Fakler B, Brandle, U, Glowatzki E, Weidemann S, Zenner HP, Rup-
persberg JP. 1995. Strong voltage-dependent inward rectification of
inward rectifier K?channels is caused by intracellular spermine.
Ficker E, Taglialatela M, Wable BA, Henley CM, Brown AM. 1994.
Spermine and spermidine as gating molecules for inward rectifying
channels. Science 266:1068–1072.
Gilad GM, Balakrishnan K, Gilad VH. 1999. The course of putrescine
immunocytochemical appearance in neurons, astroglia and micro-
glia in rat brain cultures. Neurosci Lett 268:33–36.
Kusaka S, Kapousta-Bruneau N, Green DG, Puro DG. 1998. Serum-
induced changes in the physiology of mammalian retinal glial cells:
role of lysophosphatidic acid. J Physiol (Lond) 506:445–458.
Laube G, Veh RW. 1997. Astrocytes, not neurons, show most promi-
nent staining for spermine/spermidine-like immunoreactivity in
adult rat brain. Glia 19:171–179.
Fig. 3. Semilogarithmic plot of Iin/Iout? ratios vs. spermine/sper-
midine (SPM/SPD) immunolabel optical density (percentage of max-
imum) paired from Mu ¨ller cell compartments (E, endfeet; F, somata)
originating from corresponding sites of the retina (indicated by the
same relative soma-endfoot distances). The correlation coefficient r2
was 0.9306, which is highly significant (P ? 0.001).
K?CHANNEL RECTIFICATION IN RETINAL GLIAL CELLS
Lopatin AN, Makhina EN, Nichols CG. 1994. Potassium channel Download full-text
block by cytoplasmic polyamines as the mechanism of intrinsic
rectification. Nature 372:366–369.
Newman EA. 1993. Inward-rectifying potassium channels in retinal
glial (Mu ¨ller) cells. J Neurosci 13:3333–3345.
Newman EA, Reichenbach A. 1996. The Mu ¨ller cell: a functional
element of the retina. Trends Neurosci 19:307–312.
Nichols CG, Lopatin AN. 1997. Inward rectifier potassium channels.
Ann Rev Physiol 59:171–191.
Nilius B, Reichenbach A. 1988. Efficient K?buffering by mammalian
retinal glial cells is due to cooperation of specialized ion channels.
Pflu ¨gers Arch 411:654–660.
Orkand RK, Nicholls JG, Kuffler SW. 1966. Effect of nerve impulses
on the membrane potential of glial cells in the central nervous
system of amphibia. J Neurophysiol 29:788–806.
Rojas L, Orkand RK. 1999. K?channel density increases selectively in
the endfeet of isolated Mu ¨ller cells from the frog retina. Glia 15:
Skatchkov SN, Vyklicky L, Orkand RK. 1995. Potassium currents in
endfeet of isolated Mu ¨ller glial cells from the frog retina. Glia
Skatchkov SN, Eaton MJ, Veh R, Biedermann B, Orkand RK,
Reichenbach A. 1998. Asymmetric distribution of polyamines in
glial (Mu ¨ller) cells. Soc Neurosci Abs 24:138.
Skatchkov SN, Krus ˇek J, Reichenbach A, Orkand RK. 1999. Potas-
sium buffering by Mu ¨ller cells isolated from the center and periph-
ery of the frog retina. Glia 27:171–181.
Sontheimer H. 1994. Voltage-dependent ion channels in glial cells.
Steinha ¨user C, Gallo V. 1996. News on glutamate receptors in glial
cells. Trends Neurosci 19:339–345.
Valentino TL, Lukasiewicz PD, Romano C. 1996. Immunocytochemi-
cal localization of polyamines in the tiger salamander retina. Brain
Williams K. 1997. Modulation and block of ion channels. Cell Signal
SKATCHKOV ET AL.