Journal of Immunological Methods 258 2001.

5563
www.elsevier.comrlocaterjim

Two-step negative enrichment of CD4q and CD8q T cells from

murine spleen via nylon wool adherence and an optimized
antibody cocktail
Matthias Gunzer a,) , Carsten Weishaupt a , Lourdes Planelles b, Stephan Grabbe a
a

Department of Dermatology, Institute of Cell Biology, Uniersity of Muenster, Von Esmarch Strae 56, D-48149 Munster,
Germany

b
Departamento de Biologia Molecular, Instituto de Parasitologia y Biomedicina A Lopez Neyra,B CSIC, Calle Ventanilla 11,
E-18001 Granada, Spain
Received 3 April 2001; received in revised form 28 June 2001; accepted 11 July 2001

Abstract
We developed a method to highly purify CD4q and CD8q T cells from murine spleen by negative enrichment strategy.
Single-cell suspensions of spleen cells were depleted from erythrocytes by ammonium chloride-mediated lysis. The obtained
cell suspension contained approximately 28% CD4q cells and 14% CD8q cells. Passing of these cells over a nylon wool
column removed up to 75% of all cells, leading to a suspension containing approximately 50% CD4q and 23% CD8q cells.
These cells were further purified by a single immunomagnetic depletion step using a panel of eight antibodies in
combination with MACS magnetic beads and an autoMACS machine. After purification, cells were viable and mostly
non-activated based on the expression of activation markers and did not or only minimally respond to polyclonal stimuli
such as soluble anti-CD3 antibodies or Concanavalin A. With this method, 1938% of all CD4 cells and 1029% of all CD8
cells in a spleen cell suspension were recovered at the mentioned purity. The whole procedure is fast - 4 h of preparation.,
simple and cost effective, as all antibodies and the magnetic beads have been titrated to the minimal concentration needed
for purification. The method is highly reproducible, routinely leading to CD4q cells with ) 97% purity range 97.499%.
and CD8q cells with ) 96% purity range 95.696.7%.. The described protocol should facilitate studies aiming at the
physiology of AuntouchedB murine T cells. q 2001 Elsevier Science B.V. All rights reserved.
Keywords: MACS; Negative enrichment; CD4; CD8; Spleen; Mouse

1. Introduction
The analysis of T-cell function requires the enrichment of CD4 or CD8 cells to as high a purity as
technically possible in order to obtain clear-cut sig-

)
Corresponding author. Tel.: q49-251-83-52749; fax: q49251-83-58579.
E-mail address: mgunzer@uni-muenster.de M. Gunzer..

nals. In principle, there are two strategies on how to

isolate a certain cell type from a mixture of contaminating other cells. The first approach is the marking
of the wanted cell type, e.g. with a specific monoclonal antibody mAb. which is either fluorescently
labeled or carries paramagnetic particles positive
enrichment. and the purification of the marked cells
by a cell sorter or a magnet, respectively. The second
approach tries to deplete all non-wanted cells from a
preparation the technical methods being the same as

0022-1759r01r$ - see front matter q 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 0 2 2 - 1 7 5 9 0 1 . 0 0 4 6 6 - 5

56

M. Gunzer et al.r Journal of Immunological Methods 258 (2001) 5563

for positive selection., leaving the wanted cell type

behind as a purified fraction negative selection..
This approach is particularly useful for cells to which
no known mAb exists. Furthermore, it is also desirable when cells are required to remain unaltered by
the purification method since the binding of an antibody to a surface structure of a cell cannot be
expected to have no influence on the physiology of
this cell. In fact, numerous reports show that antibody binding to T cells induces multiple effects such
as cytokine production Campbell et al., 2000., gene
activation Milia et al., 1997. or even apoptosis
Wang et al., 1994; Berndt et al., 1998.. Therefore,
the gold standard method for the purification of a
cell type should be the negative enrichment, which
leaves the target cell Auntouched.B Technically, immunomagnetic depletion is superior to flow cytometric cell sorting, as it is faster, treats cells much more
gently and does not rely on expensive laboratory
machinery.
Since we intended to analyze the activation of
unaltered murine CD4q and CD8q T cells, we
developed a method, which allows the fast, easy and
reproducible negative enrichment of CD4q as well
as CD8q T cells from murine spleen by a combination of nylon wool adherence and a complex antibody cocktail followed by immunomagnetic depletion via the MACS system. With this method, we
reproducibly obtained CD4q T cells of up to 99%
and CD8 cells of ) 96% purity after less than 4 h of
preparation. Cell yields were 2.911.6 = 10 6 CD4
and 1.63.6 = 10 6 CD8 cells per mouse spleen. The
remaining contamination did not contain B cells,
dendritic cells or other MHC II positive cells as
detected by FACS-staining and did not or only minimally respond with proliferation to polyclonal stimuli such as soluble anti-CD3 mAb or Concanavalin
A.

sruhe, Germany.. FCS, penicillinrstreptomycin

penrstrep. and L-glutamine were from PAA Linz,
Austria.. Non-essential amino acids NEAA. and
sodium pyruvate were from Seromed Berlin,
Germany.. NH 4 Cl, KHCO 3 , Concanavalin A and
NaEDTA were from Sigma Deisenhofen, Germany..
Mouse serum was from Caltag Hamburg, Germany..
Cell strainers of 100 mm were from Falcon Catalog
No. 2360, Le pont de claix, France.. Four-way stopcocks were from PVB Medizintechnik Kirchseeon,
Germany.. Nylon wool combed, scrubbed, MKN50. was from Kisker Muhlhausen,
Germany,

www.kisker-biotech.com.. Syringes 20 and 5 ml.

were from AMEFA Limburg, Germany.. Goat antirat, anti-DX-5 pan-NK cells. and streptavidin microbeads were from Miltenyi Bergisch Glabach,
Germany.. Anti-CD2 RM2-5., -CD62-L MEL14.,
-CD3 145-2C11. and -CD45RB 16A. and appropriate isotype controls as well as all depletion antibodies see Table 1. were from Pharmingen Heidelberg, Germany..
2.2. Animals
Female Balbrc mice between 8 and 14 weeks of
age were purchased from Harlan Winkelmann Borchen, Germany. and housed under SPF conditions.
2.3. Media
T cell medium Tmed.: RPMI 1640, 1% NEAA,
5% FCS, 2 mM L-glutamine, 10 mM HEPES, 1 mM
sodium pyruvate, 500 mM b-mercaptoethanol, 100
Urml penrstrep. Lysis buffer: 4.15 g NH 4 Cl, 0.5 g
KHCO 3 , 1.85 mg NaEDTA, ad 500 ml H 2 O. Labeling buffer: PBS q 1% FCS. MACS buffer: PBS, 1%
FCS, 2 mM EDTA.
2.4. Preparation of spleen leukocytes

2. Materials and methods

2.1. Materials
Phosphate-buffered saline without calcium and
magnesium PBS., RPMI-1640 and HEPES were
purchased from Gibco Life Technologies, Karl-

Two to four spleens were converted into singlecell suspensions by squeezing through a cell strainer
with the rough end of a 5-ml syringe plunger. Cells
were spun down in a 50-ml Falcon tube 5 min,
300 = g . and the supernatant was discarded. The
pellet was vigorously resuspended three to five times
in erythrocyte lysis buffer 5 mlrspleen. and then

M. Gunzer et al.r Journal of Immunological Methods 258 (2001) 5563

57

Table 1
Antibodies and beads used for isolation of CD4q and CD8q cells
The shown amount of each antibody, which was the minimal quantity necessary for optimal depletion of labeled cells, was titered for each
antibody individually using a nylon wool purified sample at saturating doses of magnetic beads. After autoMACS depletion, samples were
tested for the degree of depletion by flow cytometry, and the minimal dose of antibody giving optimal depletion was used. With this amount
of antibody, the minimal quantity of magnetic beads was established analogously.
Target

Clone

Isotype

Label

mg Antibodyr10 8 cells after nylon

CD4 sort

CD4
H 129.19
CD8
53-6.7
CD11b
M1r70
CD16r32
2.4G2
CD24
M1r69
CD45rB220
RA3-6B2
GR-1rLy-6G
RB6-8C5
gd-T cells
GL3
NKT cells
U5A2-13
Goat anti rat microbeads ml.
Streptavidin microbeads ml.
Anti DX-5 NK cells. microbeads ml.

rat IgG2a
rat IgG2a
rat IgG2b
rat IgG2b
rat IgG2b
rat IgG2a
rat IgG2b
hamster IgG2
rat IgG2a

Biotin

incubated for 3 min at room temperature. Then, the

mixture was filled up with labeling buffer to 50 ml,
spun down 5 min, 300 = g . and resuspended in 10
ml Tmed. Before filling the cell suspension into the
nylon wool column, the cells were filtered through a
fresh cell strainer. Cells were counted and a sample
of this suspension was stored on ice for later analysis
fraction AerylysisB .. By this method, 80125 = 10 6
cellsrspleen were obtained and used for further enrichment.
2.5. Nylon wool enrichment of T lymphocytes
A stock of nylon wool columns was prepared
before each series of experiments. The used nylon
wool is able to enrich up to 4 = 10 8 spleen cellsrg
of nylon. For the described experiments, 1.6 g of
nylon wool were weighed, teased into fine threads
with gloved hands, rolled to a loose ball and stuffed
into a fresh 20-ml syringe. With the plunger, the
nylon wool cushion was pressed to the mark of 12
ml, packed in a sterilization bag together with a
piece of aluminum foil ca. 5 = 5 cm. and steamsterilized before use. At the beginning of each T cell
preparation, a sterilized nylon wool column was put
into a support. A three-way stopcock was fixed on
the outlet, and the column was gently filled from the

CD8 sort

1.3

4.1

0.9
0.9
2.1
1.7
4.2
1.2
0.6

67

100
5
20

bottom with 20 ml Tmed, making sure that no air

bubbles remained in the column. Then, the column
was covered with the aluminum foil and incubated at
37 8C, 5% CO 2 for at least 45 min usually the time
needed for preparation of the spleens as described in
Section 2.4.. Together with the column, a 50-ml tube
filled with Tmed was warmed in the incubator.
The nylon wool column was put into a support,
and the Tmed was allowed to drip out until the
surface of the liquid reached the top of the nylon
wool. The whole 10 ml of cell suspension obtained
under Section 2.4 were filled into the column, a fresh
tube was put under the outlet and the stopcock was
opened until the first cells had run through the entire
column bed detectable from a cloudy eluate.. The
collected eluate was poured back into the column
before it was covered with the aluminum foil. Then,
the column was incubated for 45 min at 37 8C, 5%
CO 2 . After incubation, the column was fixed in the
stand, a 20-gauge needle was fixed to the outlet and
a fresh 50-ml collection tube in ice was put under the
needle. The column was filled up completely with
the pre-warmed Tmed, and the elution speed was
adjusted to approximately 1 droprs ; 1 mlrmin..
Until 40 ml had been eluted, the column was steadily
refilled with warm Tmed. Subsequently, cells were
counted and a sample was stored on ice for later
analysis fraction Anylon woolB ..

58

M. Gunzer et al.r Journal of Immunological Methods 258 (2001) 5563

Table 2
Numbers of CD4q and CD8q cells during the different steps of the purification process
Step

Erylysis
Nylon wool
MACS
CD4rCD8.

Cellsrmouse =10 6 .

81 " 19
6199.
30.9 " 8.8
20.836.5.
7r2.4
2.911.6r1.63.6.

Purity % all cells.

Yield CD4rCD8 % from step.

CD4

CD8

Erylysis

Nylon wool

28 " 2.9
24.730.3.
47 " 3.9
42.750.2.
98.1 " 0.8
97.499.

13.6 " 1.5

11.914.5.
20.8 " 2.3
18.322.7.
96.1 " 0.6
95.696.7.

27.9r21.9
1938r1029.

51.1r40.3
1675r2373.

Numbers are mean values for three independent experiments.

Numbers in brackets give the range over all experiments.

2.6. Magnetic enrichment of CD4 q or CD8 q cells

For magnetic enrichment, cells were spun down
5 min, 300 = g ., resuspended at 10 8 cellsrml in
labeling buffer and filled into a 1.5-ml Eppendorf
tube. Then, antibodies were added according to Table
1. The suspension was resuspended and incubated at
4 8C on a rolling shaker for 10 min. Cells were
washed twice centrifugation at 450 = g for 3 min.
with labeling buffer and then resuspended in 400 ml
MACS buffer. MACS microbeads were added according to Table 1. The suspension was filled up to
2 = 10 8 cellsrml with MACS buffer and carefully
resuspended before incubation at 7 8C fridge. for 15
min. Then, cells were washed twice in MACS buffer
as described above and finally filled to 500 ml
MACS buffer. Cells were counted and CD4 sortings
were adjusted to 2 = 10 8 cellsrml with MACS buffer
minimum 500 ml., while CD8 sortings were adjusted to 10 8 cellsrml with MACS buffer minimum
500 ml.. Cells were stored on ice until processing in
an autoMACS machine Miltenyi.. On the autoMACS, the program ADEPLETE SB was used and
CD4q or CD8q cells were collected at port ANEG.B
The antibody-labeled eluate was collected from port
APOS 1B and saved for yield assessment. Cells were
stored on ice until further use fraction AMACSB ..

2.7. Assessment of cell purity, iability and actiation status

Cell purity was tested in two ways. First, cells
were analyzed for the expression of surface markers
by FACS. Second, the impact of several polyclonal
stimuli on the purified cells was tested.
All FACS-labeling procedures were performed in
96-well round-bottom plates. Before addition of specific antibodies, all cells were incubated in PBS q
12.5% mouse serum for 20 min on ice. Then, cells
were stained with FITC-labeled antibodies against
CD3, CD4, CD8, CD45RB and CD62-L. All cells
were also labeled simultaneously with a PE-labeled
anti-CD2 antibody. One sample was stained with
anti-CD25-PE and anti-CD3-FITC. Cells were incubated on ice for 20 min, washed once, and analyzed
on an EPICS MCL flow cytometer Coulter, Krefeld,
Germany.. All measurements were performed in
PBS q 1 mM PI for exclusion of dead cells with
20,000 cells collected for each analysis.
For polyclonal stimulation of T cells, three different methods were applied. First, cells were stimulated by addition of an anti-CD3 mAb 2C11, 10
mgrml. with or without addition of an anti-CD28
mAb 37.51, 2 mgrml. to the culture. Second, a
culture was supplemented with 10 mgrml anti-CD3

Fig. 1. The enrichment of CD4q and CD8q cells over a two-step protocol from murine spleens yields highly purified preparations, which
show no sign of pre-activation. Spleens were converted into a single-cell suspension and erythrocytes were lysed Erylysis.. These cells
were incubated within a nylon wool column and non-adherent cells were eluted Nylon.. Nylon wool cells were split and one part was
negatively enriched for CD4q, the other part for CD8q cells by antibody cocktails followed by immunomagnetic depletion of labeled cells.
After this procedure, all cells were stained for expression of lineage as well as activation markers. Numbers indicate the percentage of cells
in the upper right quadrant positive for both markers.. R1 depicts the region used for morphological gating of cells for the analysis of
surface receptors.

M. Gunzer et al.r Journal of Immunological Methods 258 (2001) 5563

59

60

M. Gunzer et al.r Journal of Immunological Methods 258 (2001) 5563

2C11. and a defined AcontaminationB of externally

produced dendritic cells bmDC generated by an
8-day culture in GM-CSFq IL-4 and the addition of
1 mgrml CD40-L for the final 2 days as described
Labeur et al., 1999... Finally, cells were stimulated
with different doses of Concanavalin A, a plant
lectin which is known to be dependent on the presence of antigen presenting cells for its full function
Habu and Raff, 1977.. All cultures were performed
in triplicate in 96-well plates with 2 = 10 5 cells in
200 ml Tmedq 10%FCSrwell. After 48 h of culture, 50% of the medium was exchanged and the
cultures were then pulsed with 1 mCi w3 Hx Thymidinerwell for the final 16 h. Cell bound radioactivity
was assessed by liquid scintillation counting.
3. Results
3.1. Cell recoeryr loss during separation
Cell recovery was assessed in three experiments
Table 2.. The erylysis step yielded 81 " 19 = 10 6
cellsrmouse spleen. These cells were 28 " 2.9% for
CD4q and 13.6 " 1.5% for CD8q Fig. 1.. The
nylon wool adherence removed 60 " 18% of all
cells, thereby increasing the purity to 47 " 3.9% for
CD4q and 20.8 " 2.3% for CD8q Fig. 1 and Table
2.. After the final autoMACS preparation, the cells
were 98.1 " 0.8% for CD4q and 96.1 " 0.6% for
CD8q, respectively Fig. 1.. After magnetic bead
depletion, we recovered 27.9 " 9.7% of all CD4q
cells and 21.9 " 10% of all CD8q cells which were
present in the original erylysis fraction Table 2..
The enrichment was from 28% to 98% for CD4q
3.5-fold. and from 14% to 96% for CD8q 6.9-fold..
Cells were viable over the entire enrichment protocol
as judged by trypan blue exclusion during all cell
countings as well as the absence of PI staining in
flow cytometry analyses.
3.2. Flow cytometric analysis of the enriched cells
CD4q and CD8q T cells prepared after the described protocol were not pre-activated as judged by
high expression of CD45RB Ernst et al., 1993. in
the majority of CD4q and almost all CD8q cells as
well as high CD62-L expression Jung et al., 1988.
in most CD4q and CD8q cells. The activation marker
CD25 Lowenthal et al., 1985. was also missing in

most cells Fig. 1.. The residual contamination of the

preparation was F 1% for the markers CD11c, CD19
and MHC II, therefore, virtually free of dendritic
cells, B cells or other APC not shown..
3.3. Polyclonal stimulation of the enriched cells
The analysis by flow cytometry only gives a
limited information on the purity of a given cell
preparation, as already a small contamination of T
cells with APC - 1%. which is hardly detectable,
may well influence the behavior of the purified cells.
To scrutinize the purity of our cell preparations, we
therefore chose to test the proliferative response of
the cells against several polyclonal stimuli such as
soluble anti-CD3 with or without anti-CD28 or the
plant lectin Concanavalin A Kruisbeek and Shevach, 1991; Habu and Raff, 1977.. Additionally, we
introduced a defined AcontaminationB of the T cells
with mature autologous dendritic cells, the most
effective APC Banchereau and Steinman, 1998.. A
really pure preparation of both CD4 and CD8 cells is
not able to respond with proliferation to a polyclonal
stimulus such as soluble anti-CD3 mAbs or lectins
since these agents depend on the presence of trace
amounts of antigen presenting cells in order to
cross-link the relevant surface receptors on the T
cells, which ultimately produces the signal necessary
for the induction of T cell proliferation. Thus, only
the absence of proliferation to a polyclonal stimulus
is a proof of the absence of even small amounts of
APC, which are not detectable by FACS.
Under these conditions, we found that our cell
preparations were very pure. A high concentration
10 mgrml. of soluble anti-CD3, either alone or in
combination with anti-CD28 or a low concentration
of Concanavalin A 1 mgrml., was not able to
induce any proliferative signal above background
Fig. 2.. High amounts of Concanavalin A 2.5
mgrml. only introduced a minor proliferative response in these cells, which was in the range of
background proliferation seen by others Habu and
Raff, 1977.. However, the introduction of as little as
0.1% mature DC that is 200 cells in 200,000
T cells. to the cells together with anti-CD3 antibodies already induced a significant response in both
CD4 and CD8 cells, which was dramatically up-regulated, when we increased the contamination to 1%.
Fig. 2.. Likewise, both the erylysis as well as the

M. Gunzer et al.r Journal of Immunological Methods 258 (2001) 5563

61

Fig. 2. CD4q and CD8q cells purified with the protocol described in this paper do not respond to polyclonal stimuli. 2 = 10 5 CD4 or CD8
cells were incubated in Tmed q 10% FCS in the presence of soluble anti-CD3 mAb with or without soluble anti-CD28 mAb or with
deliberate addition of different doses of externally produced mature dendritic cells DC.. In different cultures, cells were stimulated by
different doses of soluble Concanavalin A or by 1% externally added DC alone. Cells from the erylysis or nylon wool fraction were also
stimulated by soluble anti-CD3q soluble anti-CD28 mAbs. After 48 h, cells were pulsed with w3 Hx Thymidine and cell bound radioactivity
was assessed 16 h later by liquid scintillation counting. Data represent the means of triplicate cultures each.

nylon wool fraction were impure under these conditions, as they showed pronounced proliferative responses against anti-CD3 q anti-CD28 Fig. 2..
Therefore, we conclude that both the CD4 as well as
the CD8 cells obtained using the above described
protocol are very pure and contain less than 0.1%
potent APC.
Polyclonal stimulation by plate bound anti-CD3
q soluble anti-CD28 Abs induced the CD4 cells to
develop into highly proliferative cells, which produced largely IL-4 not shown., also upon restimulation. This might be a result of the presence in these
preparations of CD62-L low cells, which are known
to bias the development of a naive population of T
cells into a Th2-like phenotype Gollob and Coffman, 1994.. However, the use of Th1-inducing conditions anti-IL-4q IL-12. during the polyclonal activation phase effectively suppressed IL-4 production
and led to the exclusive induction of Ifng producing
cells not shown..
4. Discussion
Our goal was the development of a protocol for
the enrichment of murine CD4q and CD8q cells

from spleen to high purity without the use of direct

antibody labeling or cell sorters, which, to the best of
our knowledge, has not been described so far.
Although positive enrichment strategies are faster
and yield higher purities, a powerful negative enrichment procedure is always the method of choice since
the cells of interest remain AuntouchedB throughout
the whole procedure, i.e. no antibody is transiently
or permanently attached to the cell surface. Antibodies on the surface of the wanted cell type are not
desirable since they cannot be excluded of inducing
any kind of modification of the target cell. It has
been described for positively sorted human CD4q
and CD8q T cells that they spontaneously, and over
at least 20 h, produce IL-4, which was not seen in
negatively sorted cells Stanciu et al., 1996.. Ligation of CD4 on T cells is associated with many direct
effects such as induction of cytokine expression
Campbell et al., 2000., gene activation Milia et al.,
1997., TCR down-regulation Chuck et al., 1993.
and even apoptosis Malcomson et al., 1997; Wang
et al., 1994; Berndt et al., 1998.. Additionally, when
activation of the CD4q or CD8q cells by APC is the
ultimate goal of the experiments, the blocking of the
CD4rCD8 co-receptor by antibodies from the sort-

62

M. Gunzer et al.r Journal of Immunological Methods 258 (2001) 5563

ing procedure might inhibit correct signaling, as it

interferes with the contact of these receptors with
their respective ligands, MHCII or MHCI, respectively. In this context, it has been shown that antibodies against CD4 can convert full agonists into
partial agonists or even render cloned T cells anergic
when present during antigen presentation Madrenas
et al., 1997.. Finally, other cells than T cells as well
express the markers CD4 or CD8a , e.g. murine
splenic DC Grabbe et al., 2000.. By a positive
enrichment strategy, these cells would also be captured and then contaminate the preparation.
Our strategy in the development of this antibody
cocktail was iterative. The majority of contaminants
was easily defined or has already been described by
others CD4 or CD8, respectively, B220 B cells.,
CD24 B cells, granulocytes, macrophages., Gr-1
granulocytes., CD11b macrophages., CD16r32 B
cells, granulocytes, dendritic cells, macrophages..
Hurst et al., 1997; Delon et al., 1998.. Here, it
turned out that the use of more than one marker for
the same cell e.g. GR-1, CD24 and CD16r32 for
granulocytes. was consistently better than the use of
just one. The goal to reach a very pure preparation
) 95% by FACS, non-responsive to polyclonal
stimulation by anti-CD3 or ConA Fig. 2.. was
found to be indispensably dependent on the depletion
of several contaminating cells, which were almost
AinvisibleB in the original AerylysisB fraction, but
turned out as a major factor in the highly pure final
preparations. These were gd-T cells, NKT-cells and
conventional NK cells, the latter being detected by
the antibody DX-5. This was especially a problem in
the CD8 enrichment, which reaches only ; 80%
without the use of the three mAbs. Interestingly,
after the use of a nylon wool column, we found a
MHC II antibody to be of no further help in combination with the described cocktail. This might well
be different when depleting the erylysis fraction
directly without the use of nylon wool. However, as
already appreciable from Fig. 2, the nylon wool step
does a great amount of the purification process in
terms of depletion of AstickyB APC at comparably
little effort and cost.
In addition to avoiding the direct isolation of the
target cells, the main parameters which we intended
to optimize during the development of our protocol
were 1. speed, 2. ease of use, and 3. cost effec-

tiveness. We feel that with the method presented

here, all theses requirements have been fulfilled.
4.1. Speed
The entire procedure, from the filling of the nylon
wool column and preparation of the spleens to the
completion of the final autoMACS step can be performed in less than 4 h.
4.2. Ease of use
All techniques described here are simple and do
not require specialized laboratory equipment. The
only exception is the use of the autoMACS, as
described here and the use of the MACS system in
general. We recommend that whenever a MACS
system and especially an autoMACS machine is
available, it should be used due to its better reproducibility and effectiveness. However, in initial experiments, we have also used the conventional MACS
columns and also the magnetic system from Dynal.
The results obtained with these approaches were also
very encouraging and could be optimized to reach a
comparable efficiency.
4.3. Cost effectieness
The majority ) 80%. of unwanted cells was
removed by the nylon column, which is a very cheap
and straightforward method. It has to be emphasized
that also ; 50% of all target cells both CD4q and
CD8q. are lost during this procedure, which lowers
the final yield. Direct isolation methods yield two to
three times more cells 7080% present in the erylysis fraction as compared to 2228% with this protocol, Miltenyi information., albeit leading to the discussed compromise in cellular function. Therefore, if
as complete as possible recovery of target cells is
needed, it should be considered to magnetically deplete the cells after lysis of erythrocytes directly.
From our experience, the largest groups of contaminants in the erylysis fraction are B220q B cells
50%. followed by CD24q 16%. and GR-1q 10%..
CD4q cells comprise 2030% Fig. 1., while CD8q
cells make up for 1218% Fig. 1.. During the
development of this protocol, we titrated each antibody to find out the minimal amount necessary and
sufficient to obtain the highest possible cell purity
without wasting expensive material see Table 1,

M. Gunzer et al.r Journal of Immunological Methods 258 (2001) 5563

legend.. Due to these efforts, the total amount of the

different antibodies for the enrichment of CD4q
cells from 10 8 nylon wool-sorted cells was only 12.9
mg, while following the recommendations of the
manufacturer, we would have used up to 100 mg
antibodies. The same holds true for the use of the
MACS microbeads. For the second step, reagents we
were able to titrate the amount of goat anti-rat beads
to ; 30% and 50%, respectively, of the recommended amount for CD4 and CD8 enrichment, when
following the manufacturers protocol. Streptavidin
beads were titrated to 1r20 and anti-DX-5 beads to
1r5 of the recommended amount. In their protocols,
manufacturers probably take safety margins to provide a method which also works under most adverse
circumstances. However, as shown here, it is possible and worthwhile to reevaluate such standard procedures.
In summary, we developed a reliable, fast and
cost effective method for the enrichment to high
purities of both CD4q and CD8q cells from murine
spleen. With this technique studies on the physiology especially of unaltered murine T cells should be
facilitated.

Acknowledgements
We thank Christoph Specht, Clinical Immunology
Muenster and Esther Wilk, Clinical Immunology
Hannover for helpful discussions and hints. Meike
Steinert is acknowledged for her expert technical
assistance.

References
Banchereau, J., Steinman, R.M., 1998. Dendritic cells and the
control of immunity. Nature 392, 245.
Berndt, C., Mopps, B., Angermuller, S., Gierschik, P., Krammer,
P.H., 1998. CXCR4 and CD4 mediate a rapid CD95-independent cell death in CD4q. T cells. Proc. Natl. Acad. Sci. U. S.
A. 95, 12556.
Campbell, S.B., Komata, T., Kelso, A., 2000. Differential effects
of CD4 and CD8 engagement on the development of cytokine
profiles of murine CD4q and CD8q T lymphocytes. Immunology 99, 394.
Chuck, R.S., Cantor, C.R., Tse, D.B., 1993. Effect of CD4
engagement on CD4-T cell receptor complexes. Cell. Immunol. 152, 211.

63

Delon, J., Bercovici, N., Liblau, R., Trautmann, A., 1998. Imaging antigen recognition by naive CD4q T cells: compulsory
cytoskeletal alterations for the triggering of an intracellular
calcium response. Eur. J. Immunol. 28, 716.
Ernst, D.N., Weigle, W.O., Noonan, D.J., McQuitty, D.N., Hobbs,
M.V., 1993. The age-associated increase in IFN-gamma synthesis by mouse CD8q T cells correlates with shifts in the
frequencies of cell subsets defined by membrane CD44
CD45RB, 3G11, and MEL-14 expression. J. Immunol. 151,
575.
Gollob, K.J., Coffman, R.L., 1994. A minority subpopulation of
CD4q T cells directs the development of naive CD4q T cells
into IL-4-secreting cells. J. Immunol. 152, 5180.
Grabbe, S., Kampgen,
E., Schuler, G., 2000. Dendritic cells:

multi-lineal and multi-functional. Immunol. Today 21, 431.

Habu, S., Raff, M.C., 1977. Accessory cell dependence of lectininduced proliferation of mouse T lymphocytes. Eur. J. Immunol. 7, 451.
Hurst, S.D., Sitterding, S.M., Ji, S., Barrett, T.A., 1997. Functional differentiation of T cells in the intestine of T cell
receptor transgenic mice. Proc. Natl. Acad. Sci. U. S. A. 94,
3920.
Jung, T.M., Gallatin, W.M., Weissman, I.L., Dailey, M.O., 1988.
Down-regulation of homing receptors after T cell activation. J.
Immunol. 141, 4110.
Kruisbeek, A.M., Shevach, E., 1991. Proliferative assays for T
cell function. In: Kruisbeek, A.M., Margulies, D., Shevach, E.,
Strober, W. Eds.., Current Protocols in Immunology. Wiley,
New York, NY, p. 3.12.1.
Labeur, M.S., Roters, B., Pers, B., Mehling, A., Luger, T.A.,
Schwarz, T., Grabbe, S., 1999. Generation of tumor immunity
by bone marrow-derived dendritic cells correlates with dendritic cell maturation stage. J. Immunol. 162, 168.
Lowenthal, J.W., Zubler, R.H., Nabholz, M., MacDonald, H.R.,
1985. Similarities between interleukin-2 receptor number and
affinity on activated B and T lymphocytes. Nature 315, 669.
Madrenas, J., Chau, L.A., Smith, J., Bluestone, J.A., Germain,
R.N., 1997. The efficiency of CD4 recruitment to ligand-engaged TCR controls the agonistrpartial agonist properties of
peptide-MHC molecule ligands. J. Exp. Med. 185, 219.
Malcomson, R.D., Clarke, A.R., Peter, A., Coutts, S.B., Howie,
S.E., Harrison, D.J., 1997. Apoptosis induced by gammairradiation, but not CD4 ligation, of peripheral T lymphocytes
in vivo is p53-dependent. J. Pathol. 181, 166.
Milia, E., Di Somma, M.M., Majolini, M.B., Ulivieri, C., Somma,
F., Piccolella, E., Telford, J.L., Baldari, C.T., 1997. Gene
activating and proapoptotic potential are independent properties of different CD4 epitopes. Mol. Immunol. 34, 287.
Stanciu, L.A., Shute, J., Holgate, S.T., Djukanovic, R., 1996.
Production of IL-8 and IL-4 by positively and negatively
selected CD4q and CD8q human T cells following a four-step
cell separation method including magnetic cell sorting
MACS.. J. Immunol. Methods 189, 107.
Wang, Z.Q., Dudhane, A., Orlikowsky, T., Clarke, K., Li, X.,
Darzynkiewicz, Z., Hoffmann, M.K., 1994. CD4 engagement
induces Fas antigen-dependent apoptosis of T cells in vivo.
Eur. J. Immunol. 24, 1549.