Cell. Signal. Vol. 11, No. 5, pp.

331–335, 1999 ISSN 0898-6568/99 $–see front matter

Copyright  1999 Elsevier Science Inc. PII S0898-6568(98)00049-7

Role of STAT5 in Interferon-a

Signal Transduction in Ba/F3 Cells
Robert Jaster,* Edda Tschirch, Thomas Bittorf and Josef Brock
Institute of Medical Biochemistry, Medical Faculty
of the University Rostock, Schillingallee 70, 18057 Rostock, Germany

ABSTRACT. In interferon-a (IFN-a) signalling, the essential role of the transcription factors STAT1 and
STAT2 is well established. In contrast, the involvement of other STAT proteins, including STAT5, is much
less well understood. Here we show that, in IFN-a-responsive Ba/F3 cells, this cytokine stimulates the DNA-
binding of STAT5A and B but that IL-3 is a much more potent activator of both STAT5 isoforms. A stably
expressed dominant-negative mutant of JAK2 suppressed the IL-3- but not the IFN-a-dependent DNA binding
of STAT5, suggesting independent mechanisms of its activation. Northern blots revealed that IL-3 strongly in-
duced the expression of two STAT5-regulated genes, pim-1 and oncostatin-M, whereas IFN-a had a weak stimula-
tory effect on pim-1 expression only. In summary, our results suggest that, despite the capability of IFN-a to stim-
ulate DNA binding of STAT5, this transcription factor does not play a pivotal role in IFN-a signalling in Ba/
F3 cells. cell signal 11;5:331–335, 1999.  1999 Elsevier Science Inc.

KEY WORDS. Interferon-a, Signal transduction, STAT5, Ba/F3 cells

INTRODUCTION STAT1 plays an essential and specific role in the induction

of IFN-dependent biological responses [5].
Cytokines regulate survival, growth and differentiation of
their target cells through binding to specific cell surface re- The attractive model of a JAK/STAT pathway-mediated
specificity of cytokine action, however, has become more
ceptors and the activation of intracellular signalling cas-
and more complex with the discovery that many STATs are
cades [1]. Recent data from many laboratories indicate that
activated in response to a much broader range of cytokines
intracellular tyrosine kinases of the Janus kinase (JAK) fam-
and growth factors than originally expected. For example,
ily and their key effectors, the signal transducer and activa-
IFN-a has also been shown to activate STAT3 and STAT5
tor of transcription (STAT) proteins, are essential in signal-
in some cell types [6–9]. The biological consequences of
ling by the interferons (IFNs) and other cytokines [1–4].
STAT5 activation by interferons are only partly under-
The specific set of genes induced by individual cytokine re-
stood. Some data suggest that STAT5 might be part of the
ceptors is determined, at least in part, by the specific combi-
IFN-induced differentiation response of myeloid cells [9].
nation of the JAKs and STAT proteins that are activated
The specific role of the two STAT5 isoforms A and B is cur-
[1–4]. In IFN-a signal transduction, for example, ligand
rently unclear.
binding to the receptor induces the activation of the Janus
Most cytokine receptors that are known to activate
kinases JAK1 and TYK2, which stimulate tyrosine phos-
STAT5, including, for example, the receptors for erythro-
phorylation and activation of the transcription factors
poietin, granulocyte-macrophage colony-stimulating factor
STAT1 and 2 [1–3]. Subsequently, STAT1-STAT2 hetero-
and interleukin-3 (IL-3) [10, 11], stimulate both growth and
dimers can form and associate with p48, an IFN regulatory
differentiation of haematopoietic progenitor cells [12, 13]. In
factor family protein. The resulting transcription factor
contrast, interferons are inhibitors of cell proliferation [14].
complex termed ISGF3 then binds to DNA sequences des-
How IFN-a activates STAT5 has not been addressed so
ignated the IFN-stimulated response element (ISRE) in the
far. According to the current model of cytokine receptor
promoter of target genes. Alternatively, STAT1 homodi-
signalling, STAT5 activation is mediated by JAK2, but the
mers can bind to IFN-g activation site (GAS) sequences. A
Janus kinases activated by IFN-a are JAK1 and Tyk2 [1, 2].
targeted disruption of the STAT1 gene revealed that Although there is evidence for a principal interchangeabil-
ity of different JAKs in STAT activation [4], which Janus
*Author to whom all correspondence should be addressed. E-mail: jaster@ kinase is the physiological mediator of the IFN-a-depen-
med.uni-rostock.de
Abbreviations: JAK–Janus kinase; STAT–signal transducer and activa- dent STAT5 activation still remains to be shown. Further-
tor of transcription; IFN–interferon; ISRE–interferon-stimulated response more, the STAT5-binding site(s) of the IFN-a receptor
element; GAS–interferon-g activation site; IL-3–interleukin-3; DN-JAK2– have not been identified so far.
dominant-negative Janus kinase 2; EMSA–electrophoretic mobility shift
assay; WT–wild type. In this study, we analysed the role of STAT5 in IFN-a
Received 5 May 1998; and accepted 1 June 1998. receptor signal transduction in Ba/F3 cells. The prolifera-
332 R. Jaster et al.

FIGURE 1. Induction of b-casein

oligonucleotide probe-binding ac-
tivities by IL-3 and IFN-a. Ba/F3
cells were IL-3-deprived for 8 h
and stimulated with IL-3 (100
pM) or IFN-a (1000 U/mL) for
the indicated periods of time.
Nuclear extracts from 2 3 105
cells per sample were subjected
to gel mobility shift analysis by
using a labelled b-casein oligo-
nucleotide probe.

tion of this murine pro-B cell line depends on the presence
of IL-3 and is inhibited by IFN-a [15]. We provide evidence
that IFN-a, like IL-3, is capable of stimulating the DNA-
binding of both STAT5A and STAT5B and that this effect
is not mediated by JAK2. Our data, however, also suggest
that the physiological relevance of STAT5 activation by
IFN-a, at least in BA/F3 cells, is questionable.
cells. Stable expression of DN-JAK2 was confirmed by im-
munoblot analysis. These cells were cultured in complete
medium supplemented with G418 at 1 mg/mL. For induc-
tion studies, Ba/F3 cells were depleted of IL-3 for 8 h. Re-
combinant murine IL-3 (R&D Systems, Minneapolis, MN,
USA) and IFN-a (Life Technologies, Eggenstein, Germany)
were used to stimulate the cells where indicated.

MATERIAL AND METHODS

Cell Culture
Ba/F3 is a murine IL-3-dependent pro-B cell line [16]. Cells
were maintained in RPMI 1640 (Life Technologies, Eggen-
stein, Germany) medium supplemented with 10% foetal
calf serum and 10% conditioned medium from WEHI-3B
cells as a source of murine IL-3. Generation and characteris-
tics of Ba/F3 subclones stably expressing a dominant-nega-
tive form of JAK2 (DN-JAK2) truncated at amino acid 829
have been described before [15]. Briefly, wild-type cells were
co-transfected with DN-JAK2 in pBOS along with pSVneo
by electroporation, followed by G418 selection of resistant
Preparation of Nuclear Extracts and EMSAs
Nuclear extracts from Ba/F3 were prepared essentially as de-
scribed previously [17]. Electrophoretic mobility shift assays
(EMSAs) with the use of a double-stranded, 32P-labelled oli-
gonucleotide derived from the bovine b-casein promoter
(59-AGA TTT CTA GGA ATT CAA ATC-39) as a
STAT5-binding probe have been described before [18].
Binding reaction mixtures (20 mL) contained nuclear ex-
tracts from 2 3 105 cells, 10 mM Tris-HCl (pH 7.5), 50 mM
potassium chloride, 0.1 mM EDTA, 1 mM dithiothreitol, 1
mg/mL BSA, 5% glycerol, 0.1% NP40, 1 mM Pefabloc

FIGURE 2. Effects of DN-JAK2

on the IL-3- and IFN-a-depen-
dent formation of b-casein oligo-
nucleotide probe-binding com-
plexes. The indicated BA/F3
subclones, depleted of cytokine
for 8 h, were treated with (A)
IL-3 (100 pM) or (B) IFN-a
(1000 U/mL) for various times,
and b-casein oligonucleotide
probe-binding activities in nu-
clear extracts from 2 3 105 cells
per sample were analysed by
EMSA. Note that the intensi-
ties of band-shift signals in (A)
and (B) are not comparable, ow-
ing to different times of X-ray
exposure.
Role of STAT5 in Interferon-a Signalling 333

(Boehringer Mannheim, Germany), 1 mg of poly(dI-dC)

and 16 fmol of end-labelled oligonucleotide. Binding was
for 30 min on ice. For supershift analysis, 0.5 mg of the indi-
cated STAT-antibody (Santa Cruz Biotechnology, CA,
USA) was added afterwards, and the incubation continued
for an additional 20 min.
Complexes were analysed by electrophoretic separation
on a 6% polyacrylamide gel in 0.25 3 TBE buffer.

Northern Blot Analysis

Total RNA from Ba/F3 cells was isolated by using TRIzol-
reagent (Life Technologies, Eggenstein, Germany) according
to the manufacturer’s instructions. RNA samples (20 mg) were
electrophoresed on denaturating formaldehyde gels, blotted
onto nylon membranes (Hybond-N, Amersham, UK) and im-
mobilised by ultraviolet cross-linking. The indicated cDNA
probes were labelled with [a-32P]dCTP (Amersham, UK) by
using a Prime Kit (Boehringer Mannheim, Germany) and
hybridised as described previously [15]. To test for loading
equivalence, blots were rehybridised with a glyceraldehyde-3-
phosphate dehydrogenase cDNA probe. The cDNA probes for
murine oncostatin-M and pim-1 were kindly provided by J.
Ihle (Memphis, TN, USA).

RESULTS
To investigate whether IFN-a stimulates the binding of
protein complexes to a consensus DNA-binding sequence
of STAT5, we performed EMSAs by using nuclear extracts
from IFN-a-stimulated Ba/F3 cells and an oligonucleotide
derived from the bovine b-casein promoter. The effect of
IFN-a was compared with the effect of IL-3, a known acti-
vator of the JAK2/STAT5 pathway [1, 11] and essential
growth factor for Ba/F3 cells [16]. The concentrations of cyto-
kine applied in the experiments corresponded to a maximum
stimulation (IL-3) or inhibition (IFN-a) of cell growth (data
not shown).
Figure 1 indicates that IL-3 strongly induced the rapid
and transient formation of protein complexes binding to
the b-casein promoter probe. The maximum stimulation of
DNA-binding activity was observed after 20–60 min. IFN-a
also induced protein complex binding to the b-casein pro-
moter probe. Both the kinetics of stimulation of DNA-binding
activity and the position of the band shifts on the gel were the
same as those after IL-3 stimulation. The IL-3 effect, however,
was much more pronounced.
Previous experiments have shown that, in Ba/F3 cells,
IFN-a, unlike IL-3, does not induce the tyrosine phosphor-
ylation of JAK2 [15]. To study more directly the role of this
Janus kinase in the mediation of the IFN-a- and IL-3-
dependent induction of b-casein probe-binding activity, we
compared the effects of these cytokines in wild-type (WT)
Ba/F3 cells with the effects in Ba/F3 cells stably expressing
a dominant-negative deletion mutant of JAK2, which lacks
the kinase domain of the enzyme [19]. In the presence of
DN-JAK2, activation of WT-JAK2 by IL-3 is strongly sup-
FIGURE 3. Supershift analysis of IL-3- and IFN-a-induced
b-casein oligonucleotide probe-binding activities (A) Parental
Ba/F3 cells or (B) DN-JAK2 expressing Ba/F3 were starved for
8 h and stimulated with cytokine for 60 min. Nuclear extracts
were prepared, and supershift analysis was performed by incu-
bating the EMSA binding reaction mixtures with the indicated
STAT antibodies as described.
pressed [15]. Figure 2A shows that the IL-3-dependent in-
duction of protein binding to the b-casein oligonucleotide
is much weaker in the DN-JAK2-expressing clone than in
WT-Ba/F3 cells. In contrast, Figure 2B indicates that DN-
JAK2 does not suppress the IFN-a-induced complex bind-
ing. Moreover, the IFN-a effect is even enhanced in Ba/F3
cells stably transfected with the DN-JAK2 cDNA. This
phenomenon may be, at least in part, due to the fact that
this mutant cell line expresses higher protein levels of both
STAT5A and STAT5B than WT-Ba/F3 cells do (R. Jaster,
unpublished data).
The presence of STAT5 in the b-casein promoter probe-
binding complexes was examined by supershift analysis with
the use of specific antibodies (Fig. 3). On long-run gels (see
also Fig. 1), the DNA–protein complexes separate into at
least two distinguishable components. Figure 3A shows that
IL-3 stimulated the DNA binding of both STAT5A and
STAT5B because two different supershift bands (indicated
by arrows) were observed if the binding reaction was incu-
bated with an antibody detecting both STAT5 isoforms
334
(lane 2), but only one was observed if a STAT5A-specific
antibody was applied (lane 3). Furthermore, the faster-
migrating complex was supershifted exclusively by the anti-
body directed against both STAT5A and STAT5B but not
by the STAT5A-specific one, suggesting that this complex
contains only the B isoform of STAT5.
To study the composition of the IFN-a-induced probe-
binding complexes (Fig. 3B), we took advantage of the en-
hanced IFN-a effect in DN-JAK2 expressing Ba/F3 cells
(Fig. 2B), which facilitated the supershift analysis. Again
using antibodies detecting either both STAT5A and
STAT5B (lane 2) or STAT5A (lane 3) only, we observed
the same band-shift pattern as that described for the IL-3-
activated probe-binding complexes (compare lanes 2 and 3
in Fig. 3A and B). In conclusion, IFN-a, like IL-3, stimu-
lates the binding of both STAT5A and STAT5B to the
b-casein promoter probe. No supershift was detected with
anti-STAT1 or 3 antibodies (Fig. 3B, lanes 4 and 5), indi-
cating that the IFN-a-induced protein complex binding to
the b-casein promoter probe does not contain these STAT
proteins.
The results obtained from DN-JAK2-expressing Ba/F3
cells were confirmed by the additional analysis of nuclear
extracts from WT-Ba/F3 cells (data not shown).
The capability of IFN-a to induce DNA binding of
STAT5 raised the question of possible effects on the expres-
sion of genes known to be regulated by STAT5. Therefore,
Northern blots were performed to study the IFN-a-depen-
dent expression of two members of this gene family, pim-1
and oncostatin-M [20–22]. For comparison, RNA samples
from IL-3-stimulated Ba/F3 cells were included in the inves-
tigations. Figure 4 shows that IL-3 strongly induced the
rapid and transient expression of oncostatin-M, whereas
IFN-a had no effect. The expression of the pim-1 mRNA
R. Jaster et al.
FIGURE 4. Northern blot anal-
ysis of the effects of IL-3 and
IFN-a on oncostatin-M and
pim-1 mRNA expression. Ba/F3
cells, cultured for 8 h in the ab-
sence of IL-3, were stimulated
with IL-3 (100 pM) or IFN-a
(1000 U/mL) for the indicated
periods of time. Total RNA was
isolated, and Northern blot analy-
sis was performed as described by
using 32P-labelled probes specific
for oncostatin-M, pim-1 and
glyceraldehyde-3-phosphate de-
hydrogenase.
was stimulated by both cytokines with similar kinetics, but
the IFN-a effect was very weak compared with the response
to IL-3 stimulation.
DISCUSSION
IFNs are a family of cytokines with pleiotropic effects on
target cells, including induction of an anti-viral state, inhi-
bition of cell growth and modulation of the immune re-
sponse [23]. They can be subdivided into two distinct types:
type I IFNs (comprising IFN-a, -b and -v) and type II IFN
or IFN-g, which recognise structurally related but not iden-
tical cell surface receptors [24, 25]. After receptor binding,
IFNs induce the transcription of specific target genes
through the activation of intracellular signalling cascades
including the JAK/STAT pathway. In IFN-a signal trans-
duction, the specific involvement of the Janus kinases JAK1
and TYK2 and the STAT proteins 1 and 2 is well estab-
lished [1, 2]. In contrast, the contribution of other STATs
to the mediation of IFN-a action is poorly understood.
To study the role of STAT5 in IFN-a signal transduc-
tion, we analysed the effects of IFN-a on the activation of
both isoforms of STAT5, A and B, in murine Ba/F3 cells.
IFN-a is a potent suppressor of the IL-3-dependent growth
of these cells [15]. Our data indicate that IFN-a activates
the DNA binding of both STAT5A and STAT5B; but,
compared with the IL-3 response, the IFN-a effect is weak.
Furthermore, only a small fraction of cellular STAT5 be-
comes tyrosine phosphorylated subsequent to IFN-a stimu-
lation of the cells (R. Jaster, unpublished data).
We previously showed that, in Ba/F3 cells, IFN-a in-
duced the expression of a luciferase reporter with an ISRE
enhancer element, indicating an activation of the JAK1/
Tyk2 and STAT1/2 signalling cascade [15]. In contrast,
Role of STAT5 in Interferon-a Signalling
only IL-3, but not IFN-a, activated a luciferase reporter
construct with GAS elements as STAT5-binding sites.
The failure of IFN-a to stimulate STAT5-dependent en-
hancer transactivation [15], along with its lacking or poor
effect on the expression of STAT5-dependent genes ob-
served in this study, raises the question of the physiological
relevance of the weak but significant activation of this tran-
scription factor by the cytokine. The role of STAT5 in the
regulation of cell growth is not completely understood, but
recent studies suggest that STAT5 is required for a maximal
cell proliferation in response to haematopoietic growth fac-
tors [22, 26, 27]. IFN-a, however, acts as a suppressor of cel-
lular growth.
Further studies in other IFN-a-responsive cell lines will
be necessary to investigate whether there is a cell-type-
specific role of STAT5 and its isoforms A and B in IFN-a
signalling, possibly related to the induction of cell differen-
tiation, as recently suggested [9].
The IFN-a-induced STAT5 activation seems to be medi-
ated by a signalling cascade different from the JAK2-depen-
dent one used by many other cytokine receptors [1, 2]: IFN-a
does not activate JAK2, and DN-JAK2 does not suppress
the IFN-a-stimulated DNA-binding of STAT5. The exact
molecular mechanism of the IFN-a-dependent STAT5 ac-
tivation remains to be elucidated.
This work was supported by a grant from the Deutsche Forschungsge-
meinschaft.
References
1. Ihle J. N. (1995) Nature 377, 591–594.
2. Ihle J. N. (1996) Cell 84, 331–334.
3. Darnell J. E. Jr. (1997) Science 277, 1630–1635.
4. Pellegrini S. and Dusanter-Fourt I. (1997) Eur. J. Biochem.
248, 615–633.
5. Meraz M. A., White J. M., Sheehan K. C. F., Bach E. A.,
Rodig S. J., Dighe A. S., Kaplan D. H., Riley J. K., Greenlund
A. C., Campbell D., Carver-Moore K., DuBois R. N., Clark R.,
Aguet M. and Schreiber R. D. (1996) Cell 84, 431–442.
335
6. Raz R., Durbin J. E. and Levy D. E. (1994) J. Biol. Chem. 269,
24391–24395.
7. Yang C.-H., Shi W., Basu L., Murti A., Constantinescu S. N.,
Blatt L., Croze E., Mullersman J. E. and Pfeffer L. M. (1996)
J. Biol. Chem. 271, 8057–8061.
8. Pfeffer L. M., Mullersman J. E., Pfeffer S. R., Murti A., Shi W.
and Yang C. H. (1997) Science 276, 1418–1420.
9. Meinke A., Barahmand-Pour F., Wöhrl S., Stoiber D. and
Decker T. (1996) Mol. Cell. Biol. 16, 6937–6944.
10. Gouilleux F., Pallard C., Dusanter-Fourt I., Wakao H., Haldo-
sen L. A., Norstedt G., Levy D. and Groner B. (1995) EMBO
J. 14, 2005–2013.
11. Mui A. L., Wakao H., O’Farrell A. M., Harada N. and Miya-
jima A. (1995) EMBO J. 14, 1166–1175.
12. Krantz S. B. (1991) Blood 77, 419–434.
13. Whetton A. D. and Dexter T. M. (1989) Biochim. Biophys.
Acta 989, 111–132.
14. Pestka S., Langer J. A., Zoon K. C. and Samuel C. E. (1987)
Annu. Rev. Biochem. 56, 727–777.
15. Jaster R., Zhu Y., Pless M., Bhattacharya S., Mathey-Prevot B.
and D’Andrea A. D. (1997) Mol. Cell. Biol. 17, 3364–3372.
16. Palacios R. and Steinmetz M. (1985) Cell 41, 727–734.
17. Zhu Y., Pless M., Inhorn R., Mathey-Prevot B. and D’Andrea
A. D. (1996) Mol. Cell. Biol. 16, 4808–4817.
18. Bittorf T., Jaster R., Lüdtke B., Kamper B. and Brock J. (1997)
Cell. Signal. 9, 85–89.
19. Zhuang H., Patel S. V., He T.-C., Niu Z. and Wojchowski
D. M. (1994) J. Biol. Chem. 269, 21411–21414.
20. Selten G., Cuypers H. T., Boelens W., Robanus-Maandag E.,
Verbeek J., Domen J., Van Beveren C. and Berns A. (1986)
Cell 46, 603–611.
21. Yoshimura A., Ichihara M., Kinjyo I., Moriyama M., Cope-
land N. G., Gilbert D. J., Jenkins N. A., Hara T. and Miya-
jima A. (1996) EMBO J. 15, 1055–1063.
22. Mui A. L.-F., Wakao H., Kinoshita T., Kitamura T. and Miya-
jima A. (1996) EMBO J. 15, 2425–2433.
23. Borden E. C. (1992) N. Engl. J. Med. 326, 1491–1493.
24. Uze G., Lutfalla G. and Morgensen K. E. (1995) J. Interferon
Cytokine Res. 15, 3–26.
25. Farrar M. A. and Schreiber R. D. (1992) Annu. Rev. Immunol.
11, 571–611.
26. Damen J. E., Sakao, H., Miyajima A., Krosl J., Humphries R. K.,
Cutler R. L. and Krystal G. (1995) EMBO J. 14, 5557–5568.
27. Friedmann M., Migone T., Russell S. and Leonard W. (1995)
Proc. Natl. Acad. Sci. USA 93, 2077–2082.