Role of the interferon gamma/interferon gamma receptor complex in signal transduction


Material Information

Role of the interferon gamma/interferon gamma receptor complex in signal transduction
Physical Description:
x, 108 leaves : ill. ; 29 cm.
Larkin, Joseph, 1973-
Publication Date:


Subjects / Keywords:
Microbiology and Cell Science thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Microbiology and Cell Science -- UF   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 2000.
Includes bibliographical references (leaves 95-107).
Statement of Responsibility:
by Joseph Larkin III.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 024895091
oclc - 45427185
System ID:

This item is only available as the following downloads:

Full Text








I would first like to thank God, the head of my life, for His guidance and

mercy. It is in Him that I find the strength to strive forward.

I would like to thank my graduate mentor, Dr. Howard M. Johnson for

accepting me into his laboratory, first as an undergraduate researcher and then as a

graduate student. My time in his laboratory has taught me a lot about the world of

scientific research. I also express my thanks to the other committee members: Dr.

Edward M. Hoffmann, Dr. Ammon B. Peck, Dr. Janet K. Yamamoto, and

especially Dr. Henry C. Aldrich for their assistance in research and for providing

me with a better understanding of what it takes to be a successful researcher. I

would like to thank Dr. Prem S. Subramaniam, with whom I have worked closely

for these past four years, for being a "good teammate" and instructor on our project.

I would also like to express gratitude to my labmates and friends including Amy,

George, Karrie, Mustafa, Scott, Faith, Martez, Tim, and Barbara; who have helped

me a great deal, not only with science, but also with life.

I would also like to express thanks to my undergraduate mentor Dr. Keith

Legg for his continued support and assistance. The last thing that he told my

parents during a meeting when we were deciding where I would attend college was

for them not to worry because "We (The University of Florida) are going to take

care of him." This sort of support and backing made the transition from home-to

college-to life much, much easier.

Finally, I would like to express both thanks and love to my family: Mom,

Dad, Dee, and Tab; who have been an undying well of support through sometimes

troubled times. I want to thank them for teaching me that I have the ability to rise

above adverse situations and press on into my destiny.


ACKNOWLEDGEMENTS ............................................................. ii


LIST OF FIGURES ....................................................................vii

A B STR A C T ........................................................................ix

CHAPTER 1 INTRODUCTION............... ........................................ 1

Discovery of Interferons .................................................. ..... 1
Discovery of IFNy .............................................................. 3
Biological Activities of IFNy .................................................. 3
Characterization and Cloning of IFNy ......................................... 6
Interferons as Therapy for Human Disease ................................. 7
Receptor Mediated Endocytosis ..................................... ........ 8
Functional Sites on the IFNy Molecule ........................................ 9
Interferon Gamma Receptor and Signal Transduction .................... 11
Nuclear Translocation of STATla ........................................ 14
Experimental Rationale ....................................................... 15

CHAPTER 2 MATERIALS AND METHODS ............................ 19

Materials .................................. ............................ 19
Preparation of Nanogold/ IFNy Complex ............................. 19
Nuclear Translocation of IFNy/ Nanogold Conjugate ................ 20
Radioiodination ........................................................ 20
Binding Assays .......................................................... 21
Binding of 2 I-IFNy to WISH cells .................................. 22
Immunofluorescence ................................................... 22
Conjugation of IFNyRa Chain Antibodies to FITC .................. 24
Preparation of Cytosolic and Nuclear Extracts ......................... 24
Immunoprecipitation Experiments ...................................... 25
Purification of Recombinant Human IFNy (1-123) ...... ......... 26
Cell Culture .......................................................... .. 28

Peptide Synthesis ....................................................... 28
Preparation of Import Substrates (APC conjugates) ................ 28
Nuclear Import Assays ................................................ 29

CHAPTER 3 RESULTS ..................................................... 31

Nuclear Translocation of IFNy ........................................ 31
Interaction of IFNy with IFNy R Cytoplasmic Domain ............. 33
Differential Nuclear Localization of the a and p Subunits of the
IFNGR Complex Following Activation by IFNy ............ 41
Human WISH cells Express a Large Number of IFNy
Receptors ............................................... 41
IFNy dependent selective nuclear translocation of IFNGRa
versus IFNGRp in human WISH cells ..................... 41
IFNGRa and STAT1a co-localize to the nucleus after
IFNy activation in a time and dose-dependent fashion... 52
Immunoprecipitation of isolated cytoplasmic and nuclear
extracts confirm the selective nuclear accumulation of
IFN GRa .............................................. 62
The C-terminal region of IFNy has biological significance 68
Coprecipitation of STAT1 a and IFNy with importin a ... 72
The carboxy terminus of HuIFNy contains a functional
NLS ................ ............................. .. 74

CHAPTER 4. DISCUSSION .............................................. 82

REFERENCE LIST ........................................................... 95

BIOGRAPHICAL SKETCH ................................................. 108


Table Page

Table I. An overview of interferons .................................. 2

Table II General biological activities of IFN ....................... 5

Table III Sequences referred to in this study .......................... 69

Table IV Putative NLS sequences found on some
JAK/STAT utilizing proteins ......................... 92


Figure Page

Figure 1. The current model of IFNy signal transduction ................... 13

Figure 2. Nuclear translocation of HuIFNy ................................. 32

Figure 3. Effect of HuIFNy and IFNy peptides on 125I HuIFNy
Binding to HuIFNyR and MuIFNyR ............................... 34

Figure 4. Dose response of HuIFNy and HuIFNy peptides on
125I HuIFNy binding to HuIFNyR and MuIFNyR ............... 36

Figure 5. Scatchard analysis of the binding of HuIFNy to the
HuIFNyR cyoplasmic domain protein and the
soluble M uIFNyR ................................................. 37

Figure 6. Effect of MuIFNyR peptide dosage on 125I HuIFNy
binding to receptors ....................................... 39

Figure 7. Effect of HuIFNyR cytoplasmic domain protein and
MuIFNyR on 125I HuIFNy binding to receptors .................. 40

Figure 8. WISH cells express a high number of IFNy receptor
m olecules ................................................. .......... 42

Figure 9. IFNy treatment of WISH cells induces the nuclear
translocation of the IFNGRa subunit but not that
of the IFN G Rp ...................................................... 45

Figure 10. IFNy treatment of cells induces the nuclear
translocation of IFNGRa along with STATa .................... 48

Figure 11. Differential nuclear localization of the IFNGRa
and IFNGRp subunits visualized within the
sam e cells ................................................ ..... ... 51

Figure 12. Time courses analysis of the nuclear translocation
of IFNGRa and STAT1 a in WISH cells treated with
IFN y .................................................................. 53

Figure 13. Dose response analysis of the nuclear translocation
of IFNGRa and STATla in WISH cells treated with IFNy ..... 59

Figure 14. Immunoprecipitation of IFNGRa but not of IFNGRp
from nuclear extracts of IFNy treated WISH cells .............. 63

Figure 15. Comparison of binding constants for IFNy and
IFNy (1-123) on W ISH cells ....................................... 71

Figure 16. Deletion of the NLS in IFNy inhibits nuclear
translocation of STAT1 a .......................................... 73

Figure 17. The formation of an Npi-1/STAT a complex
requires the IFNy NLS ............................................. 75

Figure 18. The peptide HulFNy (122-132) mediates the
nuclear import of the heterologous protein APC .............. 78

Figure 19. Nuclear import directed by HulFNy (122-132)
is strictly energy dependent. ..................................... 79

Figure 20. Nuclear import directed by HulFNy (122-132) is
sequence specific and inhibited by the SV40
T -N L S .......................................... ........ ......... 8 1

Figure 21. Proposed nuclear transport of STAT1 model ......................94

Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy




August, 2000

Chairperson: Howard M. Johnson
Major Department: Microbiology and Cell Science

Interferon gamma (IFNy) is a multipotent cytokine known for its numerous

immunomodulatory activities, including antiviral and antitumoricidal effects. In

collaboration with others, I have examined the possible importance of the

IFNy/IFNy receptor (IFNGR) complex in receptor/cytokine signaling. Expanding on

previous peptide studies that identified an IFNGR alpha chain (IFNGRa)

cytoplasmic binding site for IFNy, we have shown that HulFNy interacts with both

the cytoplasmic domain of human (Hu) IFNGRa chain and soluble marine (Mu)

IFNGRa. Through competition studies we have shown that amino acids 95-133 of

HulFNy and 95-134 of MuIFNy interact in a species nonspecific manner with the

cytoplasmic domains of both human and murine IFNy receptors. This binding was

inhibited by preincubation of 125I labeled HulFNy with peptides corresponding to

MulFNGRa (253-287) showing that this is in fact the region of interaction.

Additionally we have shown the nuclear translocation of the HulFNGRa,

but not the beta (HulFNGRP) chain. This nuclear translocation possessed kinetics

similar to those of the nuclear translocation of STATla, a protein known to

undergo nuclear translocation after activation via IFNy, but itself lacks a nuclear

localization sequence (NLS).

We reexamined the importance of the C-terminal sequence of IFNy. We

showed that a truncated HulFNy (1-123), although capable of binding to the IFNy

receptor with an affinity similar to that of intact gamma, has less than 1% of the

biological activity of intact IFNy as determined by antiviral assay.

HulFNy (1-123) lacked the ability to facilitate STATla nuclear translocation as

efficiently as intact HulFNy and had reduced ability to bind to importin a. In

addition, we showed that this region of human IFNy, HulFNy 122-132 contains a

functional NLS sequence. These findings directly support a hypothesis stating that

the NLS of IFNy mediates the nuclear translocation of STATIa through their

common interaction with the IFNGRa.



Discovery of Interferons

In 1957 Isaacs and Lindermann discovered that the treatment of chick

chorio-allantoic membrane fragments with inactivated influenza viruses

"interfered" with the ability of fresh influenza virus to replicate within these

samples (Isaacs and Lindenmann, 1957). This research led to the identification of a

soluble protein (produced in response to viral challenge) which has been termed

interferon (IFN). This factor was found to be a member of a large family of

proteins characterized by their induction of biological responses and immunological

reactivity. Collectively known as interferons, original nomenclature identified these

proteins based on their antibody reactivity and the cellular source where they were

produced. Currently, IFN nomenclature is based on an agreed convention whereby

IFNs are named according to Greek letters. IFNs are currently divided into two

main types, type I and type II. Included among type I Interferons are interferons

alpha (a) and omega (co) (produced by leukocytes), beta (P) (produced by

fibroblasts), and tau (T) (produced by trophoblast cells of the conceptss. Currently,

the only known type II IFN is gamma (y), produced by natural killer cells and T

cells. Presented in Table 1 is an overview of interferons.

Table I. An overview of interferons

Cellular source





Biological effects

antiviral, antiproliferative

antiviral, antiproliferative

antiviral, antiproliferative
Pregnancy recognition

antiviral, antiproliferative

Type I

a and 0o

Type II


a Also produced by macrophages, virally infected cells and epithelial cells
(Reviewed in Baron et al. 1991)

Discovery of IFNy

Wheelock first described IFNy as an IFN-like inhibitor of the cytopathic

effects of the Syndbis virus (1965). IFNy was first induced in cultures of human

leukocytes by phytohemagglutinin (PHA) (Wheelock, 1965). The sole type II

interferon, IFNy, differs from the interferon described by Issacs and Linderman by

its instability at pH =2 and its cellular source (generally leukocytes as opposed to

non-immune tissues). Electrophoretic mobility studies of IFN derived from

leukocytes revealed two distinct peaks of IFN activity of different molecular masses

(Stewart and Desmyter, 1975). It was also demonstrated numerous times that

antisera raised against "well defined" type I interferons was unable to neutralize the

activity of this interferon derived from leukocytes (Younger and Salvin, 1973;

Harvell et al., 1975; Valle et al., 1975).

Biological Activity of IFNy

Interferon gamma possesses antiviral activities similar to that of type I

interferons (a,p,m,r), however the antiviral activity is substantially lower than that

of the interferons a and p (Stuart and Desmyter, 1975). In addition, IFNy has been

shown to act as a potent immunomodulatory molecule. IFNy has been shown to

upregulate the surface expression of the major histocomapatibility complex (MHC)

class I and class II antigens on numerous cell types (Baron et al., 1991, Sen and

Lengyel, 1992). MHC class I molecules are necessary for the general function of

cytotoxic CD 8+ T cells, whereas MHC class II molecules help to present foreign

particles to CD4+ helper T cells. Expression of surface Fc receptors on

macrophages (Friedman et al., 1980; Itoh et al, 1980; Weyenbergh et al., 1998) and

expression of interleukin 2 receptors on lymphocytes can both be increased by

IFNy (Johnson and Farrar, 1983).

The expression of cell surface receptors is not the only immunomodulatory

activity of IFNy. IFNy plays a crucial role in the induction of macrophages to a

tumoridicidal state (Kleinschmidt and Shultz, 1982; Pace, et al., 1983) and the

enhancement of natural killer cell activity. IFNy plays an important role in antibody

(Ab) production. IFNy acts to suppress Ab production before B cell clonal

expansion (Gisler et al., 1974; Sonnenfeld et al., 1977. Activated B cells react to

IFNy by differentiating terminally and by secreting antibodies (Sidman et al., 1984;

Leibson et al., 1984).

In addition to the above attributes, IFNy also has the ability to combat both

the formation of tumors and the replication of various intracellular pathogens. IFNy

is capable of inducing numerous genes, including nitric oxide synthase and

indoleamine 2,3 dioxygenase which has been shown to mediate IFNy activity

(Struehr et al., 1989; Lyons et al., 1992; Xie et al., 1992). It has also been

demonstrated that IFNy has the ability to inhibit cellular proliferation as relating to

tumors (Lee et al., 2000; Kominsky et al., 1998; Hobeika et al., 1998; Knupfer et

al., 2000). Some of the biological activities of IFNy are outlined in Table II.


Table II. General biological activities of IFNy

B cell antibody regulation
Induction of tumor necrosis factor in macrophages
Regulation of natural killer cell activity
Regulation of nitric oxide synthetase production
Induction of indolamine 2,3 dioxygenase
Upregulation of MHC class I and Class II surface expression
Upregulation of cell surface receptors

Characterization and Cloning of IFNy

IFNy is a glycoprotein produced by natural killer cells and T cells (both CD4+ and

CD8+) of approximately 17 kilodaltons (kd). IFNy expression is closely linked to

the process of T cell activation. In order to closely examine the properties of IFNy,

work was conducted to obtain a purified protein and/or to clone its genes. In 1979

it was discovered that IFNy would adhere to pore glass beads (Laybord et al., 1979).

Utilizing this discovery, a purification scheme was established whereby IFNy was

first adsorbed to glass beads, bound to Concanavalin-A sepharose, gel filtered, and

finally HPLC purified (Rinderknecht et al., 1984).

In the early 1980s the cDNA for murine (Gray et al., 1983) and human

IFNy (Gray et al., 1982) was first cloned and expressed. There is a single copy of

genes for IFNy for mice and humans (chromosomes 10 and 21 respectively). The

cleavage of a 23 amino acid signal peptide (Devos et al., 1982; Goeddel et al., 1980;

Gray et al., 1982b) yields a 143 residue mature protein (purification of natural

human IFNy has revealed some C terminal heterogeneity among IFNy molecules

probably due to end peptidases). Mature murine IFNy protein consists of 133

amino acids and is similar to human IFNy in that it has two potential sites for N-

linked glycosylation (Gray et al., 1983). Glycosylation does not, however, appear

to be a prerequisite for interferon function in that recombinant IFNy produced in E

coli is not glycosylated and retains functionality. In nature, IFNy exists as an anti-

parallel dimer where the N-terminus of one molecule interacts with the C-teminus

of another (Ealick et al., 1991).

Interferons as Therapy for Human Disease

The use of IFNy and other IFNs as treatment for disease has shown promise

in recent years. Although somewhat controversial, clinical trials using IFNs in the

treatment of various diseases has often found success. Currently, the recommended

treatment for chronic hepatitis C virus (HCV) infection is three million international

units (mIU) of IFNa three times a week for one year (Ahmed and Keeffe, 1999).

Clinical trials are currently underway to determine the efficacy of using IFNa as a

treatment for multiple sclerosis, in particular the exacerbation-remission form

(Cabrera-Gomez and Lopez-Saura, 1999), in addition to IFNp (Giovannoni and

Miller, 1999; Blumhardt, 1999; Arason, 1999). Further, It has been shown that

IFNa is one of the most effective treatments for chronic myeloid leukemia

(CML)(Silver et al, 1999) and is still under clinical trials for the treatment of non-

Hodgkin's lymphoma (Haas-Statz and Smalley, 1999).

It has been shown that IFNy is an effective inducer of CD20 on malignant

plasma cells in multiple myeloma (MM) at physiological doses (Treon et al, 2000).

Studies are currently underway to ascertain the use of this marker for antibody

mediated immunotherapy as a possible treatment for MM. It has recently been

shown that IFNy has the ability to modulate the activity of 5-fluorouracil (5-FU),

making it more effective in the treatment of colorectal cancer (Makower and

Wadler, 1999). IFNy has also been approved for treatment of chronic granulomatus

disease (CGD), a rare inherited immunodeficiency disease (Meischl and Roos,

1998) and is the recommended treatment in the prevention of disease relapse (Ahlin

et al., 1997; Kume and Dinauer, 2000). IFNy is thought to help restore the ability of

phagocytes to produce sufficient amounts of reactive oxygen intermediates to

combat this chronic infection (Meischl and Roos, 1998; Weening et al., 1996) and

increase the level of nitric oxide production in polymorphonuclear neutrophils

(Ahlin et al., 1999). It has been shown that IFNy is an effective treatment for

leishmaniasis (Murray and Delph-Etienne, 2000), a condition whereby ulcers are

caused by a parasite of the genus Leishmania. It has been suggested that adjuvant-

containing antigens that stimulate IFNy may serve as an effective vaccine against

leishmaniasis (Aebischer et al., 2000). IFNy with and without other anti-parasitic

compounds has been shown to control/cure the syndrome through the action of

nitric oxide production that kills the intracellular parasite in infected macrophages.

It was found that IFNy has the ability to reduce intracellular bacterial load in murine

macrophages, a result that could possibly have future therapeutic applications

(Mahon and Mills, 1999). In addition, IFNy has been shown to enhance the

resistance of macrophages to Chlamydia pneumoniae infection (Airenne et al.,


Receptor-Mediated Endocytosis

Receptor mediated endocytosis (RME) is one process by which extracellular

proteins are internalized by cells. In RME proteins first bind to the extracellular

domain of their specific receptor. Ligand binding then generally induces

dimerization and activation of receptor subunits usually by phosphorylation (Vieira

et al., 1996). The binding of the ligand to its respective receptor then induces a

conformational change that facilitates the binding of clathrin to the cytoplasmic face

of the cell membrane surrounding the receptor (Goldstein et al., 1985). Dynamin, a

GTPase recruited to the area of clathrin concentration, is thought to induce the

invagination of the plasma membrane. Although GTP is required for dynamin

action, GTP hydrolysis is not required for the production of this invagination

(Schmid and Damke, 1995). Both GTP and ATP, however are required for the

budding of this vesicle into the cell. This internalized vesicle then fuses with other

vesicles (including lysosomal vesicles) and travels to their desired area of the cell.

It is then thought that the ligand is degraded and that the receptor is likewise

degraded and/or recycled (Goldstein et al., 1985).

Functional Sites on IFNy Molecule

Monoclonal antibody studies have indicated that both the amino terminal

and carboxy terminal epitopes of IFNy are important for function (Schreiber et al.,

1985; Russell et al., 1986). Both murine and human IFNy were inhibited from

binding to their respective cell surface receptors through the use of monoclonal

antibodies against the N terminal regions of these proteins, inhibiting both their

antiviral activity on fibroblasts and their induction of macrophage cytotoxic

activities (Johnson et al., 1982; Russell et al., 1986; Farrar and Schreiber, 1993).

The importance of the first 39 N-terminal amino acids of both human and murine

IFNy was displayed when synthetic peptides constructed corresponding to these

regions were capable of inhibiting IFNy binding to receptor proteins (Magazine et

al., 1988). It was also shown that the removal of the first 9 amino acids of murine

IFNy resulted in a protein without any observable antiviral activity (Zavodny et al.,


The importance of the C terminus has also been examined. Although there

is no significant homology between IFNy proteins of different species, a strictly

conserved region of 4-5 basic amino acids at the C-terminal region is observed

between species (Ealick et al., 1991). It has been shown that monoclonal antibodies

directed against this region of murine IFNy resulted in the abrogation of both the

antiviral and macrophage priming activities of IFNy (Schreiber et al., 1985). In the

same respect, monoclonal antibodies against the basic residues or the substitution of

these amino acids with non basic amino acids resulted in a severely diminished

induction of biological response (Arakawa et al., 1986; Leinikki et al., 1987; Wetzel

et al., 1990; Lundell et al., 1991). Interestingly it was shown using synthetic

peptides that these C-terminal amino acids are involved in receptor binding in a

region that is different from the region where the amino terminal end binds (Griggs

et al., 1992).

The Interferon Gamma Receptor and Signal Transduction

The type II interferon (IFN) IFNy is a multipotent cytokine secreted by

activated T cells and natural killer cells that is responsible for the modulation of

many facets of the immune response. IFNy exerts its effects on the surface of target

cells through interaction with a specific receptor complex. The IFNy receptor

complex (IFNGR) is a heterodimeric complex consisting of an a subunit,

IFNGRa (90 kDa), and a P subunit (60 kDa), IFNGRP (Bach et al., 1997; Pestka

et al., 1997). The IFNGRa subunit binds IFNy with high affinity, whereas the

IFNGRp, although contributing to ligand binding and required for signaling (Farrar

and Schreiber, 1993; Hemmi et al, 1994; Soh et al., 1994), attaches to IFNy with a

significantly lower affinity (Bach et al., 1997; Pestka et al, 1997). The active

ligand-receptor complex consists of a dimer of IFNy bound by two molecules each

of IFNGRa and IFNGRp. After ligand binding, two tyrosine kinases are activated

at the cytoplasmic domains of these subunits: JAK1 by constitutive association with

IFNGRa (Farrar et al., 1991; Greenlund et al., 1994; Kaplan et al., 1994) and JAK2

by constitutive association with IFNGRp (Bach et al., 1996). The cytoplasmic

domain of the IFNGRa contains additional signaling sites that include a membrane

proximal dileucine motif domain required for receptor internalization and a domain

for binding the transcription factor STATI a, tyrosine 440 (Bach et al., 1997; Pestka

et al, 1997). This STATla binding site is activated on the cytoplasmic domain by

phosphorylation mediated by JAK1 and JAK2. Subsequently, STATla is tyrosine

phosphorylated and forms a dimer that is translocated to the nucleus by as yet

unknown mechanisms. STATlca is ultimately thought to drive IFNy-specific gene

regulation. In light of the events at the IFNGRa cytoplasmic domain, the role of

IFNGRp appears to be limited to bringing JAK2 in proximity to the IFNGRa to

initiate or participate in the signaling events (Bach et al., 1997; Pestka et al., 1997).

Figure 1 shows the current model for IFNy signal transduction.

Previous studies have shown that besides these signaling events leading to

translocation of the transcription factor STATla to the nucleus, the ligand, IFNy,

when bound to the receptor complex is itself rapidly transported to the nucleus via

receptor-mediated endocytosis (Bader et al., 1994; Macdonald et al., 1986). We

have recently shown that nuclear import of IFNy is driven by a specific nuclear

localization sequence in the C-terminus of IFNy (Subramaniam et al., 1998).

Further, studies have suggested that internalization of the IFNy receptor may be

linked to signal transduction (Farrar et al., 1991). These observations question the

simple assumption that receptor endocytosis is strictly a recycling mechanism for

either receptor and/or ligand. Also, given the multimeric nature of the IFNy

receptor complex, the individual fates of the IFNGRa and IFNGRp subunits are

unknown. This question becomes particularly relevant in light of the fact that the

only route for intracellular delivery of ligand required for subsequent nuclear

transport of the ligand is via the process of receptor-mediated endocytosis.

SF- I. 4 IFNy homodimer

*-- IFNyRB

Plasma membrane

UAK2* m m -----
EU Cytoplasm


[ IFN? Inducible Genes

Nuclear Membrane

Figure 1. The current model of IFNy signal transduction. IFNy binds the IFNGR and induces a cascade of intracellular
events. JAK2 transfers from the IFNGRp to the IFNGRa chain. Both JAKI and JAK2 are phosphorylated and then
phosphorylate tyrosine 440 of the IFNGRa chain. STAT then docks on tyrosine 440, becomes phosphorylated,
homodimerizes, and translocates to the nucleus where it activates the transcription of gamma inducible genes.


Nuclear Translocation of STATla

STATI a is translocated to the nucleus through the use of the Ran/importin

pathway (Sekimoto et al 1996 and Sekimoto et al 1997). In this pathway, proteins

destined to travel to the nucleus interact with a heterodimeric protein known as

importin which consists of an a and p subunit. Nuclear proteins interact with the a

subunit via what is known as a nuclear localization sequence (NLS). Nuclear

localization sequences utilized by the Ran/importin pathway generally consist of a

cluster of basic amino acids, or two short clusters of basic amino acids separated by

a spacer of variable length (Gorlich and Mattaj, 1996). One of the best known

simple polybasic NLS sequences is that of the SV40 T-antigen, KKKRK. The p

subunit of importin mediates the binding of the NLS bearing protein/importin a

complex to Ran GTPase, which is present at the nuclear pore complex. Ran

GTPase then mediates the transport of the protein-containing complex through the

nuclear pore requiring the hydrolysis of both GTP and ATP to provide the energy

necessary for this process to take place (Gorlich and Mattaj, 1996).

This NLS region, however has not been identified on STAT1. Mutational

analysis of STAT1 failed to reveal a conventional NLS capable of binding to NPI-1,

which is a homologue of Importin a (Sekimoto et al., 1997). The absence of this

polybasic NLS in STAT1 gives credence to the hypothesis that another molecule is

supplying the necessary NLS, thus assisting in the nuclear transport of STATI

(Johnson et al., 1998).

Experimental Rationale

Although the associated biological activities of IFNy occur after it binds to

the high-affinity extracellular domain of the receptor (Pestka et al., 1987), at least

three independent lines of experimentation have identified an intracellular role for

IFNy in cell activation. These include the observations that 1) HulFNy delivered

by a liposome vector was able to activate murine macrophages to a tumoricidal state

(Fidler et al., 1985), 2) secretion-defective HulFNy expressed in murine fibroblasts

triggered antiviral activity (Sanceau et al., 1987), and 3) microinjected HulFNy-

induced Ia expression in murine macrophages (Smith et al., 1990). The activity of

HulFNy in these murine cells defies the well known species specificity of

exogenously applied HulFNy which has no activity on murine cells. Thus, the

IFNy molecule most likely interacts with some intracellular elements) to induce a

biological response. Recently, through the use of lipopeptides, additional evidence

was produced supporting an intracellular role of interferon gamma. IFNy C-

terminal peptides, which when exogenously applied to cells do not confer

detectable biological activity, when conjugated to a lipid moiety (which allows

penetration through the cell membrane) were able to upregulate the expression of

MHC class II molecules and confer viral protection to the treated cells (Thiam et

al., 1998).

Previously, IFNy binding sites were identified on the cytoplasmic domain of

the soluble MulFNGRa chain subunit using synthetic peptides (Szente et al., 1994;

Szente and Johnson, 1994). Briefly, C-terminal peptides MulFNy (95-133) and

hulFNy (95-134) bound to the cytoplasmic domain of the MulFNGRa chain at

residues 253-287. Binding to this membrane proximal region was demonstrated

with peptides to the cytoplasmic domain of the receptor as well as with

fixed/permeabilized mouse L cells where site-specific antibodies to residues 253-

287 specifically blocked binding. The C-terminus of IFNy contains a polycationic

sequence that is required for intracellular binding (Szente et al., 1994; Szente et al.,

1996). Functionally, the C-terminal peptide induced an antiviral state and up-

regulation of MHC class II molecules when taken up by pinocytosis by a

macrophage cell line (Szente et al., 1994). Thus, intracellular binding of IFNy C-

terminus to the receptor cytoplasmic domain is associated with biological activity.

Adjacent to this receptor cytoplasmic domain region is a binding site for the

tyrosine kinase JAK2 (Igarashi et al., 1994). JAK2 binding to this site is enhanced

by IFNy C-terminus (Szente et al., 1995).

Because the above binding experiments involved only peptides to the

MulFNGR cytoplasmic domain, we have subcloned and expressed the HulFNGR

cytoplasmic domain to evaluate its binding site specificity and relative affinity for

HulFNy. Through direct and competitive binding experiments, we show that the

newly expressed HulFNGR cytoplasmic domain protein can specifically bind both

HulFNy and MulFNy via their C-terminal regions.

We mentioned previously that IFNy has been shown to rapidly translocate to

the nucleus through indirect means (Bader and Wietzerbin, 1994; MacDonald et al.,

1986). Two putative NLS sequences have been identified previously within HulFNy

(Bader and Weitzerbin, 1994) and a functional polycationic NLS sequence has been

identified within MulFNy (Subramaniam et al., 1998). It has therefore been

established that the carboxy region of IFNy plays a significant role in IFNy signal

transduction, which is probably not due to interaction with the extracellular domain

of IFNGR. In addition to the intracellular roles of IFNy mentioned earlier, and the

interaction of the C-terminus of IFNy with the cytoplasmic domain of the

IFNGRa chain, crystallographic data showing the interaction of IFNy with the

extracellular domain of the IFNGRa failed to show any interaction between the C-

terminus of IFNy and the extracellular domain of the receptor (Walter et al., 1995).

It therefore seems likely that a major portion of this signal transduction modulation

occurs after the cellular internalization of IFNy. For this reason we have studied

those cellular events that surround cellular activation via IFNy.

Although IFNy nuclear translocation has been shown previously, the

actuality and functionality of this event remains controversial. Insulin, like many

other Jak/STAT utilizing cytokines has been shown to translocate to the nucleus,

(Shah et al., 1995) through the use of direct labeling with nanogold. In order to

shed light on the significance of nuclear translocated IFNy we labeled HulFNy with

nanogold and observed its nuclear translocation.

IFNy is translocated to the nucleus, has intracellular biological activity and

interacts with the cytoplasmic domain of the IFNy receptor. Many of IFNy's

biological activities are attributed to the nuclear translocation of STAT1, a protein

that binds to tyrosine 440 of the cytoplasmic domain of the IFNyRa chain after

cellular stimulation by IFNy. Although STAT1 is nuclear translocated via the

ran/importin pathway (Sekimoto et al 1996, and Sekimoto et al., 1997) as we have

previously noted, it lacks the NLS necessary for nuclear translocation. We

therefore propose a chaperone model whereby STAT1 undergoes nuclear

translocation via a complex with another NLS containing protein. Because IFNy

contains an NLS and STAT1 is known to interact with the IFNyRa, it is therefore

logical that we examine the IFNyRa after IFNy stimulation. We show that after

IFNy binds to its receptor the IFNGRa subunit of the receptor, like the ligand, is

rapidly internalized and translocated to the nucleus. In marked contrast, the

IFNGRp chain does not undergo endocytosis and nuclear translocation.

It has been shown that under certain conditions the C-terminal region of

IFNy is sufficient for biological activity (Szente et al., 1994). We have attempted to

answer the question of whether or not IFNy possess biological activity in the

absence of this region. We have expressed a HuIFNy devoid of the polycationic C-

terminal region and discovered that although capable of binding to the extracellular

domain of the receptor, it has both severely reduced antiviral activity and ability to

nuclear translocate STAT1. Finally, we have tested to see if HuIFNy contains a

functional NLS and found that this was indeed the case.



Recombinant human IFNy (HulFNy) was purchased from BioSource

International. The antibodies used in these experiments were as follows: rabbit

polyclonal IgG raised against a peptide corresponding to amino acids 466-485 of the

HulFNGRa subunit (Santa Cruz Biotechnology), rabbit polyclonal IgG raised against

a peptide corresponding to amino acids 318-337 of the HulFNGRP chain (Santa Cruz

Biotechnology), goat polyclonal IgG raised against a peptide corresponding to amino

acids 702-739 of human STATla (R & D Systems). Polyclonal phospho-STATI

(Tyr701) antibody was purchased from New England BioLabs. Na1251 was purchased

from Amersham. B-PERM Bacterial Protein Extraction Reagent for purification of

IFNy (1-123) was from Pierce (Rockford, IL). Fluorescein-5-isothiocyanate was

obtained from Molecular Probes.

Preparation ofNanogold/ IFNy Complex

Interferon gamma (0.02 mg) was added to sufficient SHS nanogold

conjugate (Nanoprobes Inc.) to label 1.2 nMoles of amine sites. The reaction

mixture was placed at 40C overnight. The reaction was quenched by adding 5 pl of

1M tris-HCl. Uncoupled HulFNy was separated from the IFNy/nanogold complex

via gel filtration using sephadex 75 (Amersham).

Nuclear Translocation of IFNy/Nanogold Conjugate

Human WISH cells were plated at 2x106 cells on a 60x15 mm tissue

culture dish and allowed to adhere overnight. The culture dishes were allowed to

cool for 30 minutes at 40C and then incubated with 10 ml of 12.25 nM

HulFNy/nanogold conjugate or BSA/nanogold conjugate in the cold room. The

tissue dishes were then transferred to a 370C incubator for 15 minutes. The treated

cells were then rinsed 3x with ice cold EMEM + 10% FBS followed by rinsing with

0.1 M cacodylate buffer (3x) to remove unbound protein. Partial fixation of

samples was accomplished by incubation of samples in 0.2 M cacodylate buffer

containing 1% paraformaldehyde and 0.5% tween for 15 minutes at 4C. The cells

followed by washing 3x with nanopure water. Gold particles conjugated to IFNy

within cells were then silver intensified for 6 minutes according to manufacturer's

instructions (Nanoprobes, Inc.) for visualization via electron microscopy followed

by vigorous washing with nanopure water (5x). Fixation was then completed by

subjecting samples to 0.2 M cacodylate buffer containing 2% glutaraldehyde and

2% paraformaldehyde for 2 hours at 40C. The cells were then dehydrated through a

series of graded ethanols followed by acetone. The samples were then embedded in

Spur's plastic and polymerized at 600C. The plastic was then ultrathin sectioned on

an ultramicrotome and observed via transmission electron microscopy using a Zeiss

EM 10CA microscope.


Radioiodinations were performed as previously described (Szente et al.,

1994). HuIFNy (5 ug) was radioiodinated by combining 10 Ml with 5 /l (500 4Ci)

Na125I (16.9 mCi//ig, Amersham) in the presence of 25 ul of 0.15 M potassium

buffer, pH 7.4, and 10 /l Chloramine-T (5 mg/ml) for 2 minutes. After

neutralization of the reaction with 10 gl of sodium metabisulfite (10 mg/ml),

potassium iodide (70 mg/ml), and BSA (20 mg/ml), the reaction mixture was

chromatographed over a 10-ml Sephadex G-10 column equilibrated with a

Tris/NaCl/BSA buffer (10 mM Tris-HCI, pH 7.4, 0.15 M NaCI, and 0.33 mg/ml

BSA). Fractions of 500 to 600 /l were collected, and the fraction containing the

greatest activity was used for receptor binding studies. The specific activity of 1251-

IFNy was 90-120 /Ci/tg of protein.

Binding Assays

Binding assays were performed as previously described (Szente et al.,

1994). EIA/RIA plates (Costar, Cambridge, MA) were seeded with HuIFNyR

cytoplasmic domain protein or soluble MuIFNGR (10 ng) or MuIFNGR peptide

(253-287) in 0.1 M carbonate/bicarbonate buffer, pH 9.6, and allowed to incubate

for 18 h at 40C. The remainder of the assay was performed at room temperature.

Plates were washed three times with wash buffer (0.15 M NaCI, 0.05% Tween) then

blocked for 1 h in 5% Carnation powdered milk dissolved in PBS. Following this,

125I-HuIFNy was added to the wells at a final concentration of 5 nM for 1.5 h. For

competition assays, competitors were incubated in the wells for 1 h prior to the

addition of 125I-HuIFNy. In the case of saturation assays, various concentrations of

HulFNy were added to the wells for 1 h prior to the addition of 125I-HuIFNy (10

nM). When HuIFNGR cytoplasmic domain protein or soluble MuIFNGR were

used as competitors, 125I-HuIFNy was preincubated with the competing receptor in

a microcentrifuge tube for 1 h and then added to the plate wells for 1.5 h. The wells

were then removed and the radioactivity quantified in a scintillation counter

(Beckman instruments, Irvine, CA). Saturation and competition data were analyzed

with the Equilibrium Binding Data Analysis (EBDA) program for "cold" saturation

binding and ligand displacement analysis (McPherson, 1983; McPherson and

Summers, 1983; McPherson, 1985; McPherson, 1985).

Binding of 125I-IFNy to WISH cells

The binding of 125I-IFNy to confluent WISH cells was performed at 4C

using a standard "cold saturation" procedure as previously described for protein

binding studies above. Data were analyzed using the LIGAND computer program

for determination of binding constants (McPherson, 1985). The Scatchard curve in

Figure 8 was replotted from values from the program output.


WISH cells were grown on tissue culture treated slides (Falcon) at 3 x 105

cells per slide. Just before use, the cells were washed with ice-cold culture medium

and brought to 40C. Cells were then incubated at 40C with IFNy (20,000 units/ml) in

ice-cold culture medium for 1.5 hr. Cellular events were initiated by transferring

cells to 370C and incubation for the indicated periods of time. After the appropriate

times, the cells were immediately fixed in methanol (-200C), and then

permeabilized using 0.5% Triton X in 100 mM Tris-HC1, 0.9% NaCl (TBS) for 10

minutes. Slides were washed in TBS and non-specific sites were blocked with TBS

containing 5% dried nonfat milk containing 0.1% Triton X, and the cells incubated

in the same blocking solution with antibodies against IFNGRa, IFNGRp or

STATI a, either alone or in combination as indicated. After 1.5 hours of incubation

at room temperature cells were washed with TBS containing 0.1% Triton-X,

followed by incubation with FITC conjugated donkey anti-goat IgG and/or Texas

red conjugated donkey anti-rabbit IgG, as the case may be. The slides were again

washed repeatedly and stained with a solution of DAPI according to manufacturer's

recommendations (Molecular Probes). Following washing, cells were mounted in

Prolong antifade solution (Molecular Probes), covered with a coverslip, and sealed

with nail varnish.

The images observed were obtained through the use of an Olympus 1X70

deconvolution microscope under oil immersion 60x objective and an auxiliary 1.5x

magnification. Subsequent to the acquisition of these images, they were then

further deconvolved through the use of Applied Precision's Delta Vision

deconvolution algorithm [reviewed in Hiraoka et al., 1991].

In the case of experiments using double-staining for the IFNGRa chain and

the IFNGRp chain (Figure 11) within the same cells, the cells were first stained for

IFNGRp. Following washing of the secondary antibody, the cells were incubated

with FITC-conjugated IFNGRa chain antibodies for 1 hour. All other procedures

are identical to that of the other immunofluorescence experiments.

Quantitation of fluorescence was performed on images using the NIH Image

software. The mean fluorescence (f) intensity from approximately equal areas in

the cytoplasm (fc) and the nucleus (fn) from each cell within a field was measured.

The areas were chosen arbitrarily within cells and across fields. This was designed

to give truly average values. The ratio fn/fc for each cell from treated samples was

subtracted against the average fn/fc ratio from measurements on untreated cells, and

the resulting values, Fn/Fc, were averaged for each field. Averaged Fn/Fc ratios

were then plotted against time or dose of IFN treatment.

Conjugation of IFNGRa Chain Antibodies with FITC:

For IFNGRa antibody labeling, purified rabbit anti-human IFNGRa chain

antibodies were obtained from Santa Cruz Biotechnology. Fluorescein-5-

isothiocyanate was obtained from Molecular Probes. Ten mgs FITC was dissolved

in anhydrous DMSO. The FITC solution was then added to purified IFNGRa chain

antibodies at a concentration of 80 ptg per mg of antibody. This mixture was then

incubated at room temperature for 1 hour. The unreacted FITC was removed via

gel filtration.

Preparation of Cvtosolic and Nuclear Extracts

IFNy (20,000 units/ml) in ice-cold culture medium was added to previously

cooled (4C) flasks containing confluent WISH cells, and the cells incubated for 1

/2 hours at 40C. For initiation of cellular events, the flasks were incubated at 370C

for the indicated time periods. After these time periods, the cells were removed by

scraping, collected by centrifugation at 4C for 4 min, and cell pellets were

immediately flash-frozen in liquid nitrogen. The cells were lysed in buffer

containing 10 mM Hepes (pH 7.9), 40 mM KC1, 50 mM NaF, 3 mM MgCI2, 20 mM

p-glyceroylphosphate, 1 mM DTT, 5% glycerol, 0.2% NP-40, 1 mM PMSF, 2 mM

sodium orthovanadate, and 1 jtg/ml each of pepstatin, aprotinin, and leupeptin for

10 minutes on ice. Samples were centrifuged for 10 minutes at 1,000 rpm at 40C,

and supernatants containing the cellular cytosolic fraction were then collected and

saved for further experimentation (Greenburg and Bender, 1994).

The pellets containing isolated nuclei were carefully resuspended in lysis

buffer and washed twice with lysis buffer. The presence of isolated nuclei was

verified by trypan blue staining, and nuclei counted. Nuclear extracts were

prepared from the intact nuclei by lysis, for 30 minutes on ice, using a high salt

buffer consisting of 20 mM Hepes (pH 7.9), 420 mM KC1, 1.5 mM MgCI2, 50 mM

NaF, 0.2mM EDTA, 0.5 mM DTT, 2 mM sodium orthovanadate, 25% glycerol, 0.5

mM PMSF, 1 pg/ml each of pepstatin, aprotinin, and leupeptin, and centrifugation

for 15 minutes at 14,000 rpm at 40C to remove the insoluble fraction. The nuclear

extracts were assessed for contamination with cytosolic proteins by measuring the

activity of the cytosolic enzyme lactate dehydrogenase. Nuclear extracts contained

less than 1% of the cytosolic specific activity of this marker enzyme.

Immunoprecipitation Experiments

Equal amounts of protein, within a given experiment, for each of the

cytosolic extracts and nuclear extracts were incubated at 40C with the indicated

antibodies for 2 hr. Immunoprecipitation and immunoblotting of immune

complexes were performed as previously described (Subramaniam and Johnson,


For co-immunoprecipitations using 125I-labeled IFNy, human WISH cells

(see legends) were incubated at 4C with 0.33 gig/ml (3300 units/ml) of 125I-labeled

human IFNs for 1 hour. Cells were then shifted to 370C for the appropriate time

periods. Control cells were maintained at 40C. Following incubation at 370C, cells

were washed three times with ice-cold growth medium, once with ice-cold PBS.

Control cells were processed last. Cells were dislodged by scraping into cold PBS,

and flash-frozen in liquid N2, for storage at -800C till use. Cells were lysed at 4C

into lysis buffer (see above), to provide cytoplasmic extracts.

For immunoprecipitation, equal protein amounts of lysates were

immunoprecipitated using STATla antibodies. Immunoprecipitates were washed

once with 50 mM Tris-HC1, pH 6.8, containing 2 mM sodium orthovanadate, and

immune complexes separated on SDS-PAGE. Following transfer of proteins to

nitrocellulose, 125I-IFNy was detected by autoradiography. Immunoprecipitated

STATI a was detected by immunoblotting with anti-STAT1 a antibodies.

Purification of Recombinant Human IFNy (1-123):

The expression vector for the C-terminal truncated mutant of human IFNy,

IFNy (1-123), was kindly provided by Dr. Mark R. Walter (University of Alabama

at Birmingham, Birmingham, AL). The protein was expressed in E. coli grown at

370C on a shaker in LB medium (1% bacto-tryptone, 0.5% Bacto-yeast, and 1%

NaCI) supplemented with 100 mg/ml ampicillin (Green et al, 1989), and IFNy (1-

123) and was purified from inclusion bodies (Haelewyn and Ley, 1995). Briefly,

protein expression was induced via the addition of isopropylthiogalactoside (IPTG)

for 14 hours. Cells were harvested and lysed using B-PER Bacterial protein

extraction reagent (Pierce), to which 0.2 mg/ml lysozyme had been added,

according to manufacturer's recommendations. The lysate was subjected to bovine

pancreatic ribonuclease A and bovine pancreatic deoxyribonuclease I at a final

concentration of l tg/ml for 30 min. at 25C. Inclusion bodies were recovered by

centrifugation at 12,000 g for 30 min. Inclusion bodies were washed extensively

with extraction buffer (B-PER reagent), Following centrifugation and washing, the

inclusion bodies were solubilized in 6 M guanidine hydrochloride in extraction

buffer for 1 hr at 370C. Insoluble material was removed by centrifugation, and the

supernatant containing the protein was dialyzed slowly at 40C with a decreasing

gradient of guanidine hydrochloride until the protein was suspended in 10 mM Tris

Buffered Saline containing 0.1 M guanidine hydrochloride. The purity of the

renatured protein was determined by a combination of SDS-PAGE analysis and

Western blotting with antibodies to the N-terminus of IFNy. In the assays

described here the final concentration of the guanidine hydrochloride did not exceed

0.1 mM, which was not found to affect cell growth or viability.

Cell Culture

Human HeLa cells (ATCC) were grown in DMEM containing

Penicillin/Streptomycin and 10% Fetal Bovine Serum. Cells were plated onto

coverslips 24 h before use. Human WISH cells (ATCC) were grown in EMEM

containing Penicillin/ Streptomycin and 10% Fetal Bovine Serum. Cells were

plated either 24 or 48 hours prior to use as indicated for each experiment.

Peptide Synthesis

Peptides used in this study were synthesized on a PerSerptive Biosystems

9050 automated peptide synthesizer using Fmoc (N-(9-fluorenyl)methoxycarbonyl)

chemistry (as detailed in Szente et al., 1996).

Preparation of Import Substrates (APC conjugates)

Allophycocyanin (APC) activated with the bifunctional cross-linker

succinimidyl 4-(N-maeimidomethyl)cyclohexane-l-carboxylate (SMCC) was

purchased from Prozyme (San Leadro ,CA) and used according to manufacturer's

suggestions. Briefly, peptides reduced with dithiotheritol (50 mM) were coupled at

a 1: 1 or 1:2 molar ratio (APC: peptide) in 5 mM MES, pH 6.0, containing 5 mM

EDTA. After the initial separation of uncoupled peptides by gel filtration through

an Econo 5DG column (Bio Rad) in 20 mM Hepes, pH 7.3, any residual peptide

was removed by repeated concentration in the same buffer through a Centricon 50

ultrafiltration unit (MWCO 50,000; Amicon, Inc., Beverly, MA) and the conjugate

stored at 4C. The coupling efficiency (peptide/ APC) and peptide removal were

established by SDS-polyacyrlamide gel electrophresis.

Nuclear Import Assays

Transport assays with human Hela cells were based on methods previously

described (Adam, 1992). Cells grown on coverslips were washed at 40C with

transport buffer: 20 mM Hepes, pH 7.3, 110 mM potassium acetate, 5 mM sodium

acetate, 2 mM magnesium acetate, 1 mM EGTA, 2 mM dithiothreitol, 10 tg/ml

each of leupeptin, pepstatin, and aprotinin. Cells were permeabilized with digitonin

(at 40 jpg/ml) in transport buffer for 5 min at 40C. After washing with transport

buffer, cells were incubated with the transport reaction mixture for 30 min at 300C.

Complete reaction mixtures (60tpl final volume) contained 20 mM HEPES, pH 7.3;

110 mM potassium acetate; 2 mM magnesium acetate; 1 mM EGTA; 2 mM

dithiothreiotol; 10 jig/ml each of leupeptin, pepstatin, and aprotinin; 0.5 mM GTP;

2.5 mM ATP; 5 mM phosphocreatine (Calbiochem); approximately 200 nM

appropriate import substrate and 20 pl of rabbit reticulocyte lysate (untreated;

Promega, Madison, WI). Coverslips were washed in transport buffer containing 1%

bovine serum albumin, mounted on slides, and observed under a fluorescence

microscope (cooled-CCD deconvolution microscopy)

For ATP depletion experiments reticulocyte lysates were first treated with a

mixture of hexokinase (-300 units/ml), glucose (8mM), and apyrase (0.2 units/mi)

at 300C for 15 min before the addition of the rest of components. For GTP

dependence, GTP was omitted from the reaction mixture and the reticulocyte lysate

was incubated at room temperature with the analog GTPyS (Calbiochem) at 5 nM

before the addition of other components.


For peptide competition experiments, unlabeled peptides were added in

excess as described in various experiments, calculated with respect to APC, in the

presence of all other components with the exception of the import substrate. After

incubation at room temperature for five minutes, the APC substrate was added and

the mixture incubated with the cells.


Nuclear Translocation of IFNy

Various techniques were utilized in order to observe of nuclear translocation

of IFNy. We originally attempted to use colloidal gold labeled IFNy, however we

were unsuccessful. It appears that the size of the gold particle changed the

properties of the internalization of the molecule. It is also possible that the colloidal

gold/IFNy conjugate, which is very dependent on pH, dissociated within the

lysosomal vesicles leaving the colloidal gold to accumulate in multivesicular bodies

while the dissociated protein continued its respective tasks. There is evidence

supporting the dissociation of the colloidal gold/protein conjugate within the low

pH lysosomal vesicles (Shah et al., 1995). We then turned to the use of nanogold,

which is a 1 nm gold particle that has the ability to covalently bind to proteins

(Shah et al., 1995). The nanogold particle/IFNy complex was then internalized as

any other protein would be internalized due to the small size of the nanogold

particle. Subsequent to the preparation of cell samples as described in Materials and

Methods, the internalization and/or nuclear translocation of the nanogold/IFNy

complex was visualized within the treated cells via electron microscopy. It can be

observed in Figure 2 that the IFNy/nanogold complex did in fact undergo nuclear

translocation within 15 minutes as illustrated by the dark uniform circles located

a.. *-


.A YJ .1

Figure 2. Nuclear translocation of HulFNy. WISH cells were treated with
HulFNy completed to nanogold as described in "Materials and Methods". The
treated cells were prepared as indicated within "Materials and Methods" such that it
was possible to observe them via electron microscopy. The dark uniform circles
denoted by arrows illustrate the location of the interferon molecules.

within the nucleus, indicated by arrows. We also subjected WISH cells to the same

treatment using uncoupled nanogold particles as a control. Unbound nanogold

particles were not observed within these cells thus ruling out the possibility that the

gold particles observed in the nucleus are the result of nonspecific activity (data not

shown). In addition, we conducted similar experiments using nanogold completed

to bovine serum albumin (BSA). Again, no particles were observed within these

samples (data not shown).

Interaction of IFNy with the IFNyR Cytoplasmic Domain

Previously, synthetic peptides representing different regions of HulFNy

and MulFNy were used to identify binding sites on the extracellular and

cytoplasmic domains of the soluble MulFNGR (Szente and Johnson, 1994; Szente

et al., 1994). Here, we wanted to determine whether the cytoplasmic domain of the

HulFNGR, expressed without the structural and functional attributes of the

extracellular domain, could function to bind IFNy with the same relative affinity

and site specificity determined for the cloned soluble MulFNGR containing the

extracellular and cytoplasmic domains (Szente and Johnson, 1994; Szente et al.,


Initially, we established the ligand specificity of the newly expressed

HulFNGR cytoplasmic domain in solid-phase competitive binding assays (Figure

3). Binding of 25I-HulFNy was effectively competed with 800 nM unlabeled




5000 -






No competitor

Unlabeled HuIFNy (800 nM)

MuIFNy(95-133)(100I M)

SMuIFNy(1-39) 100M
................ --... --... .....-. ,

5 a

Figure 3. Effect of HulFNy and IFNy peptides on '25I-HuIFNy binding to
HuIFNy R and MulFNy R. Soluble HulFNy R cytoplasmic domain protein (10
ng) (Panel A) or MuIFNyR (10 ng) (Panel B) were absorbed to the wells of
microtiter plates for 18 h at40C. Binding of '"I2-HuIFNy (5 nM) (specific activity
of 112 iCi/tg) was competed in each well with unlabeled HuIFNy (800 nM), C-
terminus MuIFNy(95-133) (100 ciM), or N-terminus MuIFN((1-39) (100 IM).
Cpm data represent the mean of triplicate wells.



HulFNy and 100 AtM C-terminus MulFNy peptide, IFNy (95-133). The N-

terminal MuIFNy peptide, MuIFNy (1-39), previously shown to bind exclusively

to the extracellular domain of the receptor (VanVolkemburg et al., 1993), did not

compete for binding in these experiments. The data show that binding of HuIFNy

to intracellular regions of the HuIFNGR and MuIFNGR was mediated through

cytoplasmic domain binding to the C-terminal region of HuIFNy.

To further test the specificity of HuIFNy binding, we determined dose

responses of the inhibition of 125I-HuIFNy binding to the HuIFNGR cytoplasmic

domain protein and the cytoplasmic domain of the soluble MuIFNGR by HuIFNy

and C-terminal peptides HulFNy (95-134) and MuIFNy (95-133). The effective

concentrations of unlabeled HuIFNy needed to block half-maximal binding to the

HulFNGR (Figure 4A inset) and MuIFNGR (Figure 4B inset) cytoplasmic domains

(ECso) were similar, 1.5 x 108 M and 3 x 10-' M, respectively. The EC50 of C-

terminus HuIFNy (95-134) and MuIFNy (95-133) needed to block binding to the

HulFNGR cytoplasmic domain protein were 7 x 10-6 M and 3.5 x 10-6 M,

respectively (Figure 4A). Similar EC50 were obtained for peptide inhibition of

binding of 125I-HuIFNy to the MuIFNGR cytoplasmic domain (Figure 4B).

Binding control IFNy peptide, MuIFNy (1-39), was ineffective at blocking

binding. Thus, HuIFNy was 200 to 500 times more effective than the peptides in

blocking binding of '25I-HuIFNy to the receptor cytoplasmic domain, which

suggests that the C-terminus of the intact HuIFNy was better recognized by the

receptor than C-terminus alone.




Competitor (nM)
Competitor (nM)



6000 -


5000 -

4000 -

3000 -



Competitor (nM)



Competitor (LM)

Competitor (iM)

Figure 4. Dose response of HulFNy, and HulFNy peptides on '2I-HuIFNy binding to HuIFNyR and MulFNyR. Binding of
"I-HulFNy to solid-phase HulFNyR cytoplasmic domain protein (Panel A) or the soluble MulFNyR (Panel B) was performed
as described in Figure 2. The competitors were unlabeled HulFNy (filled circle; insert graphs), HuIFNy(95-134) (filled
diamond), MulFNy(95-133) (empty square), and MuIFNy(1-39) (empty circle). Binding in the absence of competitor was 7,000
CPM and 11,031 CPM on the soluble MulFNyR and HulFNyR cytoplasmic domain, respectively. Data represent the mean of
triplicate wells.








-IA -- ^

1 .0.2
S40 0.15
O. 0.1
4 I
rA 20- 0
040 1- 4 f !fb
Bound (nM)

1c0 In I o I
HulFNy (nM)

41 B


60- 0.125
0.1 -
40- i 0.05-
0 20-
Bound (nM)

0 0

Hu IFNy (nM)

Figure 5. Scatchard analysis of the binding of HulFNy to the HulFNyR
cytoplasmic domain protein and the soluble MulFNyR.
Scatchard plots (insets) and Kd values were obtained using the EBDA program for
"cold" saturation binding data (19-22). Saturation binding of unlabeled hulFNy to
the HulFNyR cytoplasmic domain (Panel A) and soluble MulFNyR (Panel B) were
determined by incubating increasing concentrations of unlabeled hulFNy with 10 nM
'2I-HuIFNy in microtitier plate wells absorbed with the receptor as in Figure 2.
Control binding was determined in the absence of unlabeled HulFNy. Binding
inhibition was calculated by subtracting the binding at the various concentrations of
unlabeled HuIFNy from the specific binding (total binding in the absence of
unlabeled HulFNy- nonspecific binding determined at 800 nM unlabeled HulFNy)
and expressed as percent of control binding inhibited (100%=38,0004,000 cpm for
HulFNyR, Panel A; and 30,0004,000 cpm for mulFNyR Panel B). Data
represent the mean of triplicate wells. Scatchard plots of binding data (inserts). B,
bound; F, free.

Analysis of receptor saturation was performed using a standard "cold" saturation

experiment (McPherson and Summers, 1983; McPherson, 1985) as shown in Figure

5A and 5B. Computer analysis of the binding data was used to obtain a scatchard

plot (Figure 5A and 5B insets) and Kd values (McPherson, 1983; McPherson,

1985). The Kd of HuIFNy for the HuIFNyR cytoplasmic domain protein (3.7 x 10-8

M) and the MuIFNyR cytoplasmic domain (7.2 x 10.8 M) were similar. The Kd of

MulFNy for the MuIFNyR (5.5 x 10"9 M) was used as an experimental control for

Kd determinations in solid-phase binding assays (Data not shown). This Kd was

similar to that previously determined for MuIFNy on MuIFNGR expressing cells

(Kumar et al., 1989; Gray et al., 1989; Cofano et al., 1990; Fernando et al., 1991),

and reflects high affinity binding to the extracellular domain of the receptor. The

binding affinity of HuIFNy for the cytoplasmic domain of the HuIFNGR a chain is

similar to that of simian virus 40 (SV40) NLS binding to the nuclear transporter

importin (Hubner et al, 1997). Thus, IFNy binding to the receptor cytoplasmic

domain is of sufficient affinity for possible nuclear translocation.

The site of interaction of IFNy on the HuIFNGR cytoplasmic domain

involves the receptor region adjacent to the cell membrane from the cytoplasmic

side as indicated by the specific inhibition of 125I-HuIFNy binding to the receptor

by receptor peptides HuIFNGR (252-291) and MuIFNGR (253-287) (Figure 6A).

HuIFNGR (102-130), which is involved in the extracellular binding of HuIFNy,

had no effect on cytoplasmic binding. Similar inhibition patterns were observed for

125I-huIFNy binding to MuIFNGR (Figure 6B). Unlabeled HulFNGR cytoplasmic

domain protein (Figure 7A) and MuIFNGR (Figure 7B) were similar in their ability

60- 120

HuIFNyR (102-130)
50- 0... ...,"100



o 60-
SuIFNyR (253-287)
30- MulFNyR (253-287)
S HuFNyR(252-291)
20 20 I I

.Competitor 0M
Competitor (PM) Competitor tM

Figure 6. Effect of MuIFNy R peptide dosage on "'I2HuIFNy binding to receptors. The soluble muIFNy R (10 ng) (Panel A) or the
HuIFNy R cytoplasmic domain protein (10 ng) (Panel B) were absorbed to the wells of microtiter plates for 18 h at 40C. '25IHuIFNy (5
nM) was incubated with the receptor absorbed to the plate in the absence (control) or presence of increasing concentrations of
competitors. The competitors were HuIFNy R (252-291)(square), HuIFNy R (102-130)(circle), and MuIFNy R (253-287)(diamond).
Specific binding (total binding nonspecific binding of BSA at various concentrations) is expressed as percent of control specific binding
(100%=3700 250 cpm for huIFNyR, Panel A; and 4200 300 cpm for muIFNy, Panel B). Data represent the mean of triplicate

0,. A B
180 0

70 HulFN yR MY 60

60 0 .... .....


40 20

Competitor Concentration (nM) Competitor Concentration (nM)

Figure 7. Effect HulFNyR cytoplasmic domain protein and HulFNyR on '25I-HuIFNy binding to receptors. HulFNyR
cytoplasmic domain protein (10 ng) (Panel A) or the soluble MulFNyR (10 ng) (Panel B) were absorbed to the wells of
microtiter plates for 18 h at 40C. I-HuIFNy (5 nM) was incubated with the receptor absorbed to the plate in the absence
(control) or presence of increasing concentrations of competitors. The competitors were: HulFNyR protein (filled circle), and
MulFNyR (open circle). Specific binding (total binding nonspecific binding of BSA at various concentrations) is expressed as
percent of control specific binding (100%=3700 250 cpm for HulFNyR, Panel A; and 4200 300 cpm for MulFNy, Panel B).
Data represent the mean of triplicate wells.

to inhibit 125I-HuIFNy binding to the HulFNGR cytoplasmic domain. Thus, receptor

and receptor peptide competitions show that cytoplasmic binding of IFNy occurs at

corresponding sites for HulFNGR and MulFNGR, and this binding is species non-


Differential Nuclear Localization of the a and p Subunits
of the IFNGR Complex After Activation by IFNy

Human WISH cells express a large number of IFNy receptors

The binding of IFNy to receptors on the IFNy-sensitive human WISH cell

line was characterized by studying the binding of 125I-labeled IFNy to these cells at

4C. Binding studies were performed using standard "cold competition"

experiments, and data analyzed using the LIGAND program (McPherson, 1985).

As shown in Figure 8, a Kd of 4.78 x 10-9 M was determined for the binding of

IFNy, confirming high-affinity receptor sites for IFNy on these cells.

Approximately 100,000 high-affinity binding sites were estimated for IFNy on these

cells. These values are very similar to that reported previously by others (Sarcar et

al., 1984). The large number of binding sites, and hence receptor molecules, on

these cells provided us with a system for easily monitoring receptor movement by

immunofluorescence and other techniques.

IFNy-dependent selective nuclear translocation of IFNGRa versus IFNGRP in
human WISH cells

We have previously shown that the full-length human IFNy is translocated to

the nucleus (Subramaniam et al, 1998), using standard nuclear import assays.

1251-y bound to receptor cpm/conc

Competitor cone (nM)

Figure 8. WISH cells express a high number of IFNy receptor molecules.
Binding of 25I-IFNy to WISH cells was assayed as described in the "Materials
and Methods". Nonspecific binding, determined as binding in the presence of a
1000-fold excess of unlabeled IFNy, was less than 18%. Data were analyzed
using the LIGAND computer program to determine the binding constants.
Values for points on the abscissa and ordinate were obtained from the above
analysis and have been replotted here.

Human IFNy utilizes components of the Ran/importin pathway similar to

murine IFNy (Subramaniam et al, 1998), and nuclear import of human IFNy can be

competed by both the NLS in murine IFNy and the prototypical NLS in the SV40 T

antigen (Subramaniam et al, 1998). The C-terminus of human IFNy, like marine

IFNy, contains a nuclear localization sequence, 128 KRKRI31, in a highly conserved

region (amino acids 95-134) in its C-terminus that overlaps, in its position, with the

NLS of murine IFNy. As has previously shown for the NLS of murine IFNy, a

peptide containing this human IFNy NLS is able to target a heterologous protein for

nuclear import, when employed in standard nuclear import assays using digitonin

permeabilized HeLa cells. The nuclear import mediated by the human IFNy NLS

occurs in an energy-dependent fashion requiring GTP, and is strictly dependent on

the addition of cytosolic factors. These data will be illustrated later within this

dissertation. Thus, human IFNy is targeted to the nucleus via a simple polybasic

NLS that is similar to that in murine IFNy.

Using human IFNy on WISH cells, which do not produce IFNy, the effects of

IFNy stimulation on the nuclear localization of IFNy receptor subunits were

determined. Cells treated with human IFNy at various times were fixed,

permeabilized and fluorescently stained using antibodies to IFNGRa, IFNGRp and

STATIa, either alone or in combination, to monitor the localization of these

proteins following receptor stimulation. Specificity of the antibodies was verified

by incubation in the presence of excess of the peptide antigens used to generate the

antibodies. Excess of the antigen was able to completely inhibit

immunofluorescence staining (data not shown).

In these experiments, cells were simultaneously stained for the individual

receptor subunits (Texas Red) and STATla (FITC) so that we could monitor both

molecules in the same cell. Further, since STATal is well characterized with

respect to its nuclear localization, double-staining experiments allowed us to verify

the competence of these cells for IFNy mediated receptor activation. As can be seen

in Figure 9, in untreated cells both the IFNGR subunits (IFNGRa: Figure 9A;

IFNGRp: Figure 9B) essentially showed a diffuse localization on the cell. As

expected, STATIa was also present throughout the cytoplasm in untreated cells. It

should be noted, however, that labeling which may appear nuclear is actually

labeling of those receptors above the nucleus. Treatment of cells for 30 min with

IFNy altered the cellular distribution profile of IFNGRa and STATIa. In both

Figures 9A and 9B, in cells treated with IFNy, STATla was localized in the

nucleus. The pattern seen for STAT1 with respect to 2D immunofluorescence

images and immunoprecipitation are very similar to what other published reports

have shown for STAT1 in IFNy treated cells (Haspel et al., 1996; Koster and

Hauser, 1999), in which immunofluorescence images show a particularly strong

nuclear presence of STAT1, compared to immunoprecipitations. In Figure 9A,

IFNy treated cells stained for the IFNGRa subunit showed that IFNGRa was also

translocated to the nucleus. In marked contrast in Figure 9B, in IFNy treated cells

the IFNGRp showed the same diffused profile as seen in untreated cells. These data

show that the while IFNGRa, like STATla, is translocated to the nucleus, IFNGRp

is not. Thus, these data show for the first time that the IFNGRa and IFNGRp



(30 min)

Figure 9. IFNy treatment of WISH cells induces the nuclear translocation of the
IFNGRa subunit, but not that of IFNGRp. (A). WISH cells were either left
untreated (No IFNy) or treated with IFNy as indicated for 30 min, and the cells
fixed and immunofluorescently stained simultaneously with antibodies to IFNGRa
(rabbit anti-IFNGRa) and STAT1a (goat anti-STATla). (B). Cells treated as in
(A) were doubly stained with antibodies to IFNGRp (rabbit anti-IFNGRp) and
STATIa (rabbit anti-Statla). In immunofluorescence experiments, antibodies
developed in rabbits were generally found to give stronger signals than other




30 min

Figure 9 continued:


subunits of the receptor complex are differentially localized within the cell

following receptor-mediated endocytosis.

To better characterize the localization of IFNGRa, we used deconvolution

microscopy to render a 3D reconstructed image of IFNGRa and STATla staining

in WISH cells before and after IFNy treatment. Multiple image sections were made

vertically through the plane of the cells and recorded, and image sections

deconvolved. Deconvolved images were then merged to give Figure 10A. Figure

10B shows the same cells, except that in this case only a section 0.2 Pma thick

through the plane of the nucleus has been shown, projected at 00 and 900 angles.

The latter figure (Figure 10B), shows that in the plane of the nucleus both

IFNGRa and STATIa are confined to the nucleus only in IFNy-treated cells,

confirming their nuclear localization following IFNy treatment.

In these experiments, we have used STATla as a marker for nuclear

translocation of the IFNy receptor subunits. Although the relative fates of the two

receptor subunits can be inferred by comparison of the two receptor subunits with

respect to STAT a translocation, a direct comparison of the two molecules within

the same cells following IFNy stimulation would be more definitive. Thus, cells

were stained simultaneously for IFNGRa and IFNGRp subunits before and after

IFNy stimulation. As can be observed in Figure 11, when examined within the

same cells, IFNGRa chain shows nuclear accumulation subsequent to

IFNy stimulation while there is no observable change in the IFNGRp chain. Thus,

these data further confirm that IFNy treatment of cells leads to the selective nuclear

translocation of IFNGRa subunit.

IFNGR a STAT1 a Overlap


3000 U/ml

Figure 10. IFNy treatment of cells induces the nuclear translocation of the
IFNGRa along with STATIa. 3D volume reconstruction of IFNGRa and
STATIa localization in WISH cells either left untreated, or treated with IFNy
(3000 units/ml). (A). After immunofluorescence images of cells were obtained
through 64 image sections on a deconvolution microscope. After deconvolution of
the stack, deconvolved sections were merged to render a 3D reconstruction of the
cells, as shown. All image processing was done using DeltaVision (Applied
Precision, Issaquah, Washington) software attached to a Silicon Graphics
workstation. (B). 3D reconstruction as in (A) except that only 2 image sections
from the stack, above and below the best focal plane through the nucleus of the
cells (a 0.2 pm displacement each along the Z-axis), were merged. The resulting
nuclear images have been projected by rotation at 0 (top panels) and 90 (bottom
panels) along the Y-axis. Note that in the plane of the nucleus in IFNy treated cells
staining is still largely confined to the nucleus, unlike in untreated cells.



Figure 10 continued:


IFNGRa STATla Overlap

Figure 10 continued:



IFNy Treated

Figure 11. Differential nuclear localization of IFNGRa and IFNGRp subunits as
visualized within the same cells. Cells were treated as in Figure 9 with the
exception of the antibody labeling procedure. Cells were first stained for the
IFNyRp chain antibodies The cells were then stained with FITC-conjugated
IFNyRa antibodies.


IFNGRa and STATla co-localize to the nucleus after IFNy activation in a time- and
dose-dependent fashion.

Time-course experiments of IFNGRa and STATla translocation were next

performed in IFNy-treated cells that were simultaneously stained for IFNGRa

(Texas red) and STATla (FITC). At the same time, DAPI, a dye that stains nuclear

DNA, was used to delineate the nuclear volume in the cells. As can be seen in

Figure 12, over time the IFNGRa subunit showed increasing nuclear accumulation,

demonstrating a time-dependent nuclear translocation of this subunit in these cells.

The translocation was discerable within 10 min of stimulation of cells with IFNy,

peaked at 20 min, and started to decline thereafter, reflecting the exit of IFNGRa

from the nucleus (Figure 12A). This translocation of IFNGRa was closely

paralleled by the nuclear accumulation and nuclear exit of STATIa in these same

cells, and the pattern of staining for STATla appeared to overlap the pattern of

staining for IFNGRa at each time point observed. A quantitative analysis of the

nuclear fluorescence staining for IFNGRa and STATla within individual cells

(Figure 12B) shows that rates of entry and exit of IFNGRa and STATla are

coincidental, and that a constant ratio of IFNGRa to STATla is maintained

throughout their traverse of the nucleus at any given time. These data suggest that

the nuclear localization of STATla and IFNGRa may be linked. It is known that

IFNy is internalized via receptor mediated endocytosis involving complexation with

the cell surface receptors. It is this ligand/receptor complex which is then

internalized. Therefore, the receptor subunits which are undergoing changes in

Figure 12. Time-course analysis of the nuclear translocation of IFNGRa and STATla in WISH cells treated with IFNy.
(A). Cells were either left untreated or treated with IFNy for the indicated time periods, as labeled in the Figure. Cells were
then fixed and stained immunofluorescently with antibodies to IFNGRa (rabbit anti-IFNGRa) and STATla (goat anti-
STATla). Nuclei in these doubly stained cells were then stained with DAPI to delineate the nuclear volume. (B).
Quantitation images presented in (A) as described in 'Materials and Methods'. The average fluorescence ratio Fn/Fc for each
field shown is plotted against time of stimulation with IFNy.



10 min



IFN y-treated,
20 min

IFN -treated,
30 min



45 min

Figure 12 continued:





Time, min

Figure 12 continued:

their localization are likely to be those on the surface which are directly exposed to

the IFNy. Again, no evidence was found for the nuclear accumulation of the

IFNGRp (data not shown). Thus, the IFNGRa subunit and STATIa, but not

IFNGRp, were found to translocate to the nucleus with an apparent fixed

stoichiometry and similar kinetics following IFNy treatment of cells.

Dose response experiments examining the relationship of the nuclear

translocation of IFNGRa and STATIa with respect to increasing concentrations

(antiviral units) of IFNy are presented in Figure 13. As can be seen in Figure 13A,

the accumulation of IFNGRa and STATla increased in a dose-response fashion.

Moreover, IFNGRa and STATla appeared to co-localize in the nucleus in a dose-

dependent fashion. In these experiments, we found that the time taken for both

IFNGRa and STAT1 a to enter and exit from the nucleus was itself a direct function

of the concentration of IFNy. Higher concentrations of IFNy resulted in maximal

nuclear accumulation much earlier than the 30 min used in these experiments (data

not shown). Quantitation of the images (Figure 13B) again confirmed that the

relative amounts of IFNGRa and STATIa present in the nucleus paralleled each

other for each concentration tested. This is consistent with the conclusion drawn

from the data in Figure 12 that the nuclear translocation of IFNGRa and STATla

appear to be coupled. Further, since the amounts of IFNy used were titrated as units

of antiviral activity it appears that there is a close relationship between biological

activity and the coupled activation and nuclear translocation of IFNGRa and


Figure 13. Dose-response analysis of the nuclear translocation of IFNGRa and
STAT Ia in WISH cells treated with IFNy. (A). Cells were treated for 30 min at
37C with the indicated concentrations of IFNy and cells were then stained as in
Figure 4, except that DAPI was not added. (B). Quantitation of images in Figure
5A was as for Figure 4, except that Fn/Fc was plotted against antiviral units of IFN



IFRy, 30 U/ml

IFWy, 300 U/ml



IFNy, 3000 U/ml





IFN units/ml

Figure 13 Continued:


Immunoprecipitation of isolated cytoplasmic and nuclear extracts confirm the
selective nuclear accumulation of IFNGRa

To further confirm the differential nuclear localization of the IFNGR subunits,

extracts from nuclei isolated from untreated and IFNy treated cells were examined

along with the corresponding cytoplasmic (including membrane) extracts by

immunoprecipitation and immunoblotting with antibodies to IFNGRa, IFNGRp,

and STATlca.

As can be seen in Figure 14A, in cytoplasmic extracts the IFNGRa subunit,

consistent with its glycosylated nature, migrated as a diffuse band in both untreated

and IFNy treated cells. By contrast, IFNGRa was detected in nuclear extracts of

only IFNy stimulated cells (Figure 14B), consistent with the ligand-dependent

accumulation of this subunit in the nucleus. Likewise, in these cells, STATla is

also translocated to the nucleus only after IFNy treatment (Figure 14E and 14F).

Moreover, the time-course for the entry and exit of STAT1a parallels the time-

course of IFNGRa. Thus, these data support the immunofluorescence experiments.

The relative absence of IFNGRa and STATIa in nuclear extracts of untreated cells,

compared to their strong presence in cytoplasmic extracts, demonstrates that

minimal cross-contamination of nuclear extracts occurred. Cross-contamination

was also ruled out by comparing the specific activity of the cytoplasm-specific

enzyme lactate dehydrogenase. No detectable activity was seen in nuclear extracts

(see Materials and Methods).

Unlike the IFNGRa subunit, no accumulation of IFNGRp chain was seen in the

nuclei of IFNy-treated cells in immunoblots comparing nuclear and cytoplasmic

(A) (B)
(A) Cytoplasmic (B) Nuclear
0 15 30 60 0 15 30 60

1144418- ,mI lmmR

(C) (D)
25000- 20000-

20000 ooo
15000- 10
S10000 -

0 0 O

time (min) time (min)

Figure 14. Immunoprecipitation of IFNGRa, but not of IFNGRp, from nuclear
extracts of IFNy treated WISH cells. Cytoplasmic and nuclear extracts from WISH
cells, either left untreated (0) or treated with IFNy for 15 min (15), 30 min (30), 60
min (60) or 90 min (90), were prepared as described in "Materials and Methods".
Extracts were immunoprecipitated and immunoblotted with antibodies to IFNGRa
or STATIa as follows: Cytoplasmic (A, C) and nuclear extracts (B, D) were
immunoprecipitated and immunoblotted with antibodies to IFNGRa (A, B) and
STATla antibodies. (E, F). Immunoblots for STATla from cytoplasmic extracts
and nuclear extracts presented in (G) and (H) were probed with antibodies specific
for Tyr70' phosphorylated STAT a. (M). Cytoplasmic extracts were
immunoprecipitated with antibodies to IFNGRa and analyzed by immunoblotting
with antibodies to STATla. All antibodies were developed in rabbits. For these
experiments equal amounts of protein were used within a given blot. Quantitation
of the nuclear blots was also performed by densitometric analysis


0 15 30 60

(G) Cytoplasmic

- STAT1 --

0 15 30 60

(H) Nuclear

0 15 30 60 0 15 30 60

I STAT 1- Y701 )H

(I) (J)

Rn -





time (min)

time (min)


time (min)





0 ID 0 0i

time (min)

Figure 14 continued:




E 5000-
A 0-

extracts (data not shown), although ligand-specific STATla accumulation in the

nuclei of the same cells was again demonstrated (data not shown).

Figure 14G and 14H show the tyrosine phosphorylation status of STATIa in the

cytoplasm and the nucleus, respectively. The antibody used specifically recognizes

STATIa molecules tyrosine phosphorylated on Tyr701, which is the only tyrosine

residue that is phosphorylated in response to IFNy (Shuai et al., 1992; Shuai et al.,

1993). Tyrosine phosphorylation of the target Tyr701 residue of STATla could be

detected both in the cytoplasm and nucleus. The level of nuclear tyrosine

phosphorylation was consistent with the limited portion of the total cellular

STAT1a that was present in the nucleus at any given time. In particular, in the

nucleus the tyrosine phosphorylation profile of STATla (Figure 14H) mimicked

that of the entry and exit of total nuclear STATla (Figure 14F), which in turn

overlaps with the nuclear entry and exit of the IFNGRa chain (Figure 14B). Thus,

the nuclear translocation of IFNGRa coincides specifically with the nuclear entry of

STAT a that is activated on the critical Tyr701 by recruitment to the IFNGRa chain.

This raises the possibility that Tyr701-activated STATla that is bound to the

receptor IFNGRa may be translocated to the nucleus as a complex with IFNGRa.

The data in Figure 14F and 14H show that, under these conditions, we do not

see sustained presence of tyrosine phosphorylated STATla in the nucleus even in

the presence of continued activation by ligand (compare Figure 14E and 14G). This

is consistent with the earlier studies (Haspel et al., 1996; Koster and Hauser, 1999).

We also found the same to be true for the IFNGRa chain (compare Figure 14B and

14F). In the case of STATla this has been shown to be due to the dynamic

equilibrium, at the nuclear level, between the activation and nuclear translocation of

STAT1a by JAK kinases and the deactivation of STAT a by a nuclear phosphatase

(Haspel et al., 1996). This is manifested as an appearance in the nucleus of total

serinee plus tyrosine) phosphorylated STATIa following activation, and the

subsequent rapid exit of nuclear STATla following dephosphorylation (Haspel et

al., 1996). The exit phase from the nucleus is accompanied by accumulation of

total serinee plus tyrosine) phosphorylated STATIa in the cytoplasm (Haspel et al.,

1996), the significance of which is not known, but does suggest that in the

continued presence of ligand, JAK activation of STATs occurs over a time course

that probably exceeds the requirements for immediate-early gene activation.

The dynamics of IFNGRa and STATla association in the cytoplasm was also

examined in the above context. As shown in Figure 14 M, cytoplasmic extracts of

untreated and IFNy treated cells were immunoprecipitated with antibodies to

IFNGRa and following Western blotting probed with antibodies to STATla. Low

levels of STATIa were completed with IFNGRa in the cytoplasm (Figure 14 M) at

times when the levels of nuclear accumulated STATla (Figure 14 F) and nuclear

IFNGRa (Figure 14 B) were relatively high, consistent with the rapid transport of

IFNGRa and STAT1a to the nucleus, possibly via a complex of IFNGRa and

STATla. Also, increased STATlo/IFNGRa complexes were seen during the slow

cytoplasmic accumulation phase of tyrosine phosphorylated STATla that follows

the exit of both IFNGRa and STATla from the nucleus. Thus, the

immunofluorescence and immunoprecipitation data suggest that the IFNGRa

subunit may play a role in the transport of STATIa to the nucleus. However, the

VI) Cytoplasmic
0 15 30 60 90

W. 4M0 m STAT 1





* 2000



time (min)

Figure 14 continued:

0P 0

two molecules appear to dissociate in the nucleus since STATla did not co-

immunoprecipitate with the IFNGRa chain in the nucleus (data not shown).

The findings reported here showing association of IFNy activated IFNGRa and

STATla up to nuclear translocation may appear at odds with models of IFNy

signaling, where STATla is postulated or shown to dimerize, dissociate from the

receptor, and then undergo nuclear translocation (Bach et al., 1997; Pestka et al.,

1997). Our data suggest that some, but not all, STATlc remains associated with

the IFNGRa subunit upon IFNy stimulation and probably undergoes nuclear

translocation associated with this subunit. These findings do not preclude the

presence of free STATI ca homodimers as reported by others. Our data do, however,

provide one approach for determination of the mechanism of nuclear translocation

of STAT transcription factors.

The C-terminal region of IFNy has biological significance

A number of earlier studies have shown that truncations in the C-terminus of

IFNy that destroy the polybasic region, which we have identified as a NLS, lead to

drastic loss in the biological properties of IFNy (Arakawa et al., 1986; Dobeli et al.,

1988; Arakawa et al., 1989; Wetzel et al., 1990; Lundell et al., 1991; Slodowski et

al., 1991). These studies suggest that the NLS is critical for the biological activity

of IFNy. As a starting point to analyzing the role in signal transduction of the NLS

in the C-terminus of IFNy, we re-examined one such mutant, IFNy (1-123), that is

deleted from residues 124 (see Table 3 for sequences) onwards including the IFNy

NLS. To determine the structural and functional properties of IFNy (1-123), we

Table III. Sequences referred to in this study.


IFNy (122-132) (human) AAKGIKRKRS

a The sequences are shown only for the relevant C-terminal portions of the intact IFNs. Sequences are derived from the mature
form of IFNy. NLS sequences are in bold.

compared it with wild-type IFNy in in vitro binding assays and antiviral assays.

Fig. 15 shows the Scatchard analysis of a standard "cold" saturation experiment on

binding of IFNy(1-123) to WISH cells and its comparison with wild-type IFNy

(Green et al., 1998). Scatchard analysis of the binding showed that IFNy (1-123)

bound to receptors on WISH cells with a Kd that was very similar to that of wild-

type IFNy. Thus, IFNy (1-123) was just as competent as wild-type IFNy in binding

to receptors on intact cells. Deletion of the C-terminal amino acids, including the

NLS, from IFNy did not significantly affect the affinity of IFNy (1-123) for the

receptor, showing that the C-terminal amino acids in general, and the NLS in

particular, do not contribute significantly to high-affinity binding to the receptor

complex. This conclusion is further supported by recent studies using surface

plasmon resonance to address the binding of IFNy (1-124) to the extracellular

domain of IFNGRa (Sadir, et al., 1998), and by data from the X-ray crystal

structure of the complex between IFNy and the extracellular domain of IFNGRa

(Walter et al., 1995).

When compared in antiviral assays, however, IFNy (1-123) was found to be

drastically reduced in its biological activity. IFNy (1-123) had less than 0.5% of the

specific antiviral activity of wild-type IFNy when compared on the same WISH

cells (data not shown). This is consistent with the earlier functional studies on C-

terminal deletion mutants (Arakawa et al., 1986; Dobeli et al., 1988; Arakawa et al.,

1989; Wetzel et al., 1990; Lundell et al., 1991; Slodowski et al., 1991). Thus, while

the NLS-containing C-terminal region does not seem to contribute significantly to

high-affinity interactions for binding to the receptor complex on WISH cells, it

o IFNy

0.05- U IFNy(1-123)


0.03- a


0.01 i I -
0 0.25 0.5 0.75 1

B (nM)

Figure 15. Comparison of binding constants for IFNy and IFNy(1-123) on WISH
cells. Binding was evaluated in a standard "cold" saturation experiment using 1251-
IFNy (5 nM). Data were analyzed using the LIGAND computer program, and data
from Scatchard analysis of binding have been replotted here. Samples were run in
triplicate. Specific binding of '5I-IFNy, determined in the presence of a 100-fold
excess of unlabeled IFNy, was found to be > 80%.

appears to play a crucial role in the ability of IFNy to induce a biological response

in these cells. The possibility, however, that this region is involved in low-affinity

interactions with the extracellular domains of the receptor cannot be ruled out.

Since nuclear translocation of STATla is dependent on phosphorylation of

STATIa by JAK1 and JAK2, we used nuclear translocation of STAT Il as a marker

of receptor activation. Cells were treated with IFNy (1-123) and nuclear

localization of STATIa was compared with that of wild-type IFNy. As can be seen

in Fig. 16, IFNy (1-123) was impaired in its ability to induce STATla nuclear

translocation, compared to IFNy. Thus, the deletion of the NLS in IFNy is

coincident with the loss of ability to induce the activation and nuclear translocation

of STATla. Since STATla's presence in the nucleus is required for biological

activity this is consistent with the poor biological activity of IFNy (1-123).

Coprecipitation of STAT1 and IFNy with importin a.

Nuclear transport of STATla occurs through the interaction of activated

STAT1 a with the importin-a analog Npi-1. Thus, since IFNy was found completed

with STATla we determined whether IFNy was completed with Npi-1. To further

gain insight into the role of the C-terminal NLS of IFNy in this complexation and

regulation, we also used the human deletion mutant IFNy (1-123) that lacks the NLS

in similar studies and compared the results with wild-type IFNy. '25I-labeled human

IFNy and human IFNy (1-123) were used to treat human WISH cells and lysates

were immunoprecipitated with Npi-1 antibodies. As seen in Figure 17 (lower

panel), wild-type IFNy was recovered as a complex with Npi-1, however, the NLS-

mutant IFNy (1-123) was not found to be completed with Npi-1. These data



Figure 16. Deletion of the NLS in IFNy inhibits nuclear translocation of STAT Ia.
WISH cells were treated for 30 min at 37C with equal amounts (2.5 ng/ml) of
IFNy or IFNy(1-123), or left untreated, as indicated, before being fixed and
subjected to immunofluorescence staining for STAT a localization using rabbit
anti-STATIa antibodies.


show that the NLS is required for complexation with Npi-1. When these complexes

were examined for the presence of STATla (Figure 17, upper panel), STATIa was

found to co-immunoprecipitate in significant amounts only in cells treated with

wild-type IFNy. Thus, the interaction of the NLS of IFNy with the importin Npi-1 is

required for the formation of a stable complex between STATIa and its nuclear

transporter Npi-1. Npi-1 immunoprecipitated from cells treated with IFNy (1-123)

showed a low and transient signal for STATla, suggesting some complexation of

STATla with Npi-1 could occur in a NLS-independent manner but this complex is

weak and short-lived. These data strongly suggest that the NLS of IFNy directly

regulates the Npi-l-mediated entry of STAT1 a into the nucleus.

In the latter experiments, a small amount of phosphorylated STATla was

detectable in the nucleus of cells treated with IFNy(1-123) (data not shown) that

coincided with the transient Npi-1:STATla complex seen in Figure 17 in similarly

treated cells. The kinetics of this transport differed from that of wild-type IFNy.

This suggests that small amounts of STATla can translocate to the nucleus through

other mechanisms. This may be related to the fact that a second weaker NLS in

human IFNy exists upstream of the one studied here (see also Bader and Witzerbin,

1994). It remains to be determined if this NLS can function similarly to that

described here.

The carboxy terminus of HulFNy contains a functional NLS

The competence of the two putative nuclear localization sequences within

the COOH-terminal domain of human IFNy (HulFNy) was evaluated by testing its

ability to mediate the nuclear import of a heterologous protein. This was performed

Figure 17. The formation of an Npi-1/STAT1a complex requires the IFNy NLS.
Human WISH cells were treated with 0.33 jg/ml each 25I-IFNy or '25I-IFN(1-123), as
indicated, for 7 min or 15 min at 370C. Control cells were incubated with the appropriate
ligands at 40C Cells were lysed and immunoprecipitated with antibodies to Npi-1. IFNs
were followed by autoradiography, while STAT 1a was detected by immunoblotting with
anti-STATla antibodies. Quantitation of STAT blots was performed by densitometric

IFNy (1-123)

time (min):

0 7 15 0 7 15

STAT1a-.J -g a.d 4 i I


o i' U> 0 Z. I>
w4 .4
IFNy IFNy (1-123)
time (min)


using the standard in vitro nuclear transport assay in digitonin permeabilized human

HELA cells (Adam et al., 1992).

In this study we used as a substrate a peptide corresponding to amino acids

122-132 of HulFNy coupled to the heterologous autofluorescent APC for use in

import assays. Figure 18 shows that essentially all of the Hela cells show nuclear

accumulation of HuIFNy (122-132). It also shows that this nuclear accumulation is

inhibited at 40 and is dependent on reticulocyte derived cytosolic factors

establishing the fact that this HulFNy (122-132) is capable of mediating the nuclear

import of a heterologous protein coupled to it.

The transport of large molecules across the nuclear pore is a strictly energy

dependent process, dependent on both ATP and GTP. It has been shown that in the

absence of these energy providing compounds a "nuclear rimming" pattern can be

observed (Newmeyer and Forbes, 1988). Figure 19 shows that the import of

HulFNy (122-132) into the nucleus is dependent on the addition of both ATP and

GTP to the cytosolic extracts. The absence of ATP and the absence of GTP with

the addition of the nonhydrolyzable analog GTPyS both result in the inability of

HulFNy (122-132) to mediate the translocation of APC to the nucleus. In Figure 19

the documented "nuclear rimming" can be observed (in particular no GTP). Again,

these data demonstrate that HulFNy (122-132) contains an NLS that functions in an

energy dependant function similar to other nuclear import signals.

These data serve to demonstrate that the domain of HulFNy represented by

HulFNy (122-132) contains an NLS capable of functioning in an energy dependent

capacity similar to other nuclear import signals. HulFNy (122-132) is capable of

IMry (122-132)APC

At 4 degrees C No Lysate

Figure 18. The peptide HulFNy (122-132) mediates the nuclear import of the
heterologous protein APC. Digitonin permeabilized human HELA cells were
incubated for the duration of the assay (30 min) with the complete import reaction
mix containing HuIFN-y (122-132)-APC as substrate at 300C (top), or at 4C
(bottom left). The picture labeled "no lysate" depicts cells incubated with an
import mixture devoid of reticulocyte lysate.

iFNy (122-132) APC


Figure 19. Nuclear import directed by HulFNy (122-132) is strictly energy dependent.

Cells were incubated, as in Figure 18, either with complete import mixture or in the

absence of ATP or GTP as described under material and methods.

mediating the nuclear import of a heterologous protein in an energy dependent


The SV40 large T-antigen NLS is one of the best understood and established

simple polybasic NLS sequences. It has been shown that carboxy- terminus of

murine interferon gamma contains a polybasic NLS sequence capable of

competition with the SV40 NLS for the cellular machinery necessary for nuclear

import. We therefore wanted to see if the same could be said for the human

counterpart. Figure 20 clearly shows that excess SV40 NLS peptide is capable of

completely inhibiting the import of HulFNy (122-132) conjugated APC into the

nucleus. These data demonstrate that HulFNy (122-132) utilizes the same nuclear

import pathway used by SV40 large T-antigen.

IFWy (122-132) no competition

Excess unlabeled IFRy (122-132) SV40 T-NLS

Figure 20. Nuclear import directed by HulFNy (122-132) is sequence-specific and
inhibited by the SV40 T-NLS. Import reaction mixtures containing the substrate
IFNy (122-132)-APC were incubated either in the absence of competitor peptides
or in the presence of the cognate peptide or the SV40 T-NLS peptide. Competitor
peptides were incubated at a 600-fold molar excess, with respect to the substrate
HulFNy (122-132)-APC, in the reaction mixture 5 min before addition of the



In collaboration with others I have shown that after the receptor-mediated

endocytosis of IFNy, the IFNy molecule and the IFNy receptor play an active role in

cellular signaling. Using the subcloned and expressed cytoplasmic domain of the

HulFNGRa without its high affinity extracellular domain and the soluble

MulFNGRa, we correlated the previously observed biological activity of

internalized IFNy and IFNy peptides and relative affinity for the cytoplasmic


Because glycosylation can affect protein solubility, structure, antigenicity,

and the observed Mr, we evaluated our protein for oligosaccharides. Treatment to

remove N-linked carbohydrates did not affect the observed Mr. In addition, our

observed Mr of 43 KDa is consistent with the Mr observed for the intact,

deglycosylated HulFNyR isolated from placental tissue (Calderon et al., 1988).

Thus, we concluded that the difference in the predicted Mr of 30 KDa and the Mr of

43 KDa determined for the receptor cytoplasmic domain expressed in Pichia

reflected a difference in the predicted and observed Mr seen with the intact receptor

a chain (Aguet et al, 1988; Calderon et al, 1988).

The species non-specific nature observed for HulFNy binding to the

HulFNyR and MulFNyR cytoplasmic domains may be explained by the 88%

amino acid sequence homology the receptors share in their membrane proximal

cytoplasmic region. The site-specific binding of HulFNy, C-terminus

HulFNy (95-134) and C-terminus MulFNy (95-133) to the membrane proximal

region of each receptor cytoplasmic domain was confirmed by dose-dependent

binding inhibition and subsequent saturation of the available binding sites with

these ligands. The affinity constants for HulFNy binding to HulFNGR and

MulFNGR cytoplasmic domains obtained from these assays, 3.7 x 10-8 M and 7.2 x

108 M, respectively, resembled those for SV40 large tumor antigen NLS binding to

the nuclear import protein importin (Hubner et al., 1997). The similarity in Kd

values suggest that the affinity with which HulFNy binds the cytoplasmic domain is

sufficient to effect cytosolic and nuclear transport.

Past and present work in our laboratory has helped delineate a functional

role for cytoplasmic domain binding in the biological response to IFNy. Two of the

three components known to be required for IFNy signal transduction, tyrosine

kinases JAKI and JAK2 (Wilks et al., 1991; Harpur et al., 1992; Valazquez, 1992;

Muller et al., 1993; Watling et al., 1993), bind to specific sites on the cytoplasmic

domain of the IFNyR (Szente and Johnson, 1994; Szente el al., 1994). JAK1 and

JAK2 then effect tyrosine phosphorylation of the third component, transcription

factor STATI, which translocates to the nucleus to initiate transcription of IFNy

inducible genes (Fu, 1992; Schindler et al., 1992; Schindler et al., 1992; Shuai et al.,

1992; Decker et al., 1991; Igarashi et al., 1993; Greenlund et al., 1994, Farrar et al.,

1992). Thus, it appears that IFNy, JAK1, JAK2, STAT and other components

required for inducing the biological response attributed to IFNy must associate

directly or indirectly with the receptor cytoplasmic domain. We propose that this

complex could then be translocated to the nucleus via a nuclear localization motif

we have identified in the amino acid sequence of IFNy (Subramaniam et al., 1998).

Considered with the biological response elicited to IFNy and C-terminal IFNy

peptides delivered to the cytoplasm by pinocytosis, liposomes or microinjection and

the site specific binding of JAKI and JAK2 in the cytoplasmic domain, our results

establish a physiological connection between specific receptor-like binding of IFNy

to the cytoplasmic domain and IFNy induced biological activity.

Having obtained this relationship between the cytoplasmic domain of the

IFNyRa and IFNy induced biological activity, we subsequently examined the

activities of IFNyRa following IFNy stimulation. Using the IFNy receptor complex

on intact cells, we have shown for the first time the selective internalization and

nuclear transport of one subunit of a multimeric cytokine receptor complex,

following ligand-dependent receptor activation. Among the two subunits of the

IFNy receptor complex, IFNGRa and IFNGRp, only the IFNGRa subunit is

selectively translocated to the nucleus in a ligand-dependent fashion. The nuclear

translocation of IFNGRa is rapid, and occurs within the same time frame as the

activation and nuclear translocation of the latent cytoplasmic transcription factor

STATla associated with the receptor. This possibly suggests a role for the

IFNGRa subunit in signal transduction events leading to immediate-early gene


These studies strengthen the assumption that following ligand binding the

IFNGRa chain is a central player in the transduction of the signal intracellularly.

The IFNGRa chain has been recognized as the predominant high-affinity binding

subunit for ligand. The intracellular domain of the IFNGRa subunit contains

several elements involved in signal transduction. These include (i) a membrane

proximal cytoplasmic binding site for the C-terminus of the ligand IFNy -- this C-

terminus region of IFNy contains a functional NLS; (ii) a dileucine-containing

membrane proximal domain required for receptor-mediated endocytosis (Farrar et

al., 1991) that overlaps with the binding site for the NLS domain within the ligand

(Szente et al., 1994; Szente and Johnson, 1994; Szente et al., 1995); (iii) a binding

site for the tyrosine kinase JAK1 (Subramaniam et al., 1998; Farrar et al., 1991);

(iv) a tyrosine phosphorylation motif for the docking of the transcription factor

STATIa (Bach et al., 1997; Pestka et al., 1997) to mediate its subsequent activation

by the kinases JAKI and JAK2, and (v) a ligand-induced co-operative binding site

for JAK2 (Szente et al., 1995; Kotenko et al., 1995) which is in turn immediately

proximal to the binding site for the C-terminal NLS domain of IFNy (Szente et al.,

1995). In contrast, the cytoplasmic domain of IFNGRp only contains the site for

the binding of JAK2 (Bach et al., 1997; Pestka et al., 1997), and JAK2 is

constitutively attached to it (Bach et al., 1997; Pestka et al., 1997). Unlike the

IFNGRa subunit, it is not phosphorylated in response to ligand binding. The

attachment of JAK2 to the cytoplasmic side of IFNGRp is, however, essential for

bringing JAK2 into the receptor complex in close proximity to JAKI and STATla

to initiate the signal transduction events. The IFNGRp subunit, thus, serves a

limited but indispensable structural role, that of providing a structural framework

for the recruitment of JAK2 (Bach et al., 1997; Pestka et al., 1997). The specific,

low-affinity but essential interaction of IFNy with the extracellular domain of

IFNGRp is probably responsible for the transfer of JAK2 from IFNGRp to

IFNGRa. This is supported by the finding that while JAK2 can be found

constitutively attached to IFNGRp, JAK2 has been shown to co-immunoprecipitate

with IFNGRa only after ligand binding (Kotenko et al., 1995), and is consistent

with our findings that a co-operative binding site for JAK2 on the cytoplasmic

domain of IFNGRa is induced by the interaction of the C-terminus of IFNy at an

immediately proximal site (Szente et al., 1995). Our studies thus suggest that while

the IFNGRp chain is required only for a limited but essential function, the IFNGRa

continues to play a role in the further steps of signal transduction that may include

the intracellular trafficking and nuclear delivery of signal components like STATla

after their initial activation at the plasma membrane.

In this regard STATlc has been demonstrated to undergo a cycle of

activation, serine and tyrosine phosphorylation, nuclear translocation, and the

subsequent dephosphorylation and exit from the nucleus (Haspel et al., 1996). Our

studies from the immunoprecipitation of cytoplasmic and nuclear extracts confirm

this behavior of STAT a. More interestingly, the cycling of the IFNGRa subunit

appears to closely parallel the cycling of tyrosine phosphorylated STATIa, as

determined by immunofluorescence and immunoprecipitation experiments. The

IFNGRa receptor subunit is seen to accumulate in the nucleus at the same time as

tyrosine phosphorylated STATla, and subsequently exit also at the same time as

STATla. This time-dependent co-localization of IFNGRa and STATla was found

to correlate with the antiviral activity of IFNy, as determined by the

immunofluorescence dose-response experiments, suggesting that both the duration

and magnitude of receptor activation modulate the nuclear accumulation of

IFNGRa and STAT 1 and ultimately the biological response.

The nuclear accumulation of IFNGRa is also coincident with the rapid loss

of tyrosine (Tyr701) phosphorylated STAT1a from the cytoplasm. The appearance

of IFNGRa and STATIa in the nucleus correlates with low levels of

IFNGRa/STAT1a complexes in the cytoplasm. STATla is required to bind the

IFNGRa subunit following ligand binding for it to be tyrosine phosphorylated on

Tyr701, which then activates its nuclear translocation. Thus, this suggests that

IFNGRa and activated STATla probably remain completed up to the nuclear

import of the two molecules. These observations provide a framework for further

studies examining the functional role of the nuclear translocation of IFNGRa,

especially its contribution to the nuclear delivery of activated STATIa molecules.

In addition to establishing the importance of the IFNyR, we have also

provided evidence as to the importance of intracellular IFNy, namely the C-terminal

region. We have shown that internalized IFNy interacts at an intracellular site to

regulate the nuclear translocation of STAT la. This intracellular function of IFNy is

specifically mediated by a C-terminal domain of IFNy encompassed by residues 95-

133, which also contains a NLS that is required for its ability to function

intracellularly. Intracellular IFNy can be recovered as part of a complex with

STATla in IFNy activated cells. This complex also contains the nuclear importin-a

analog Npi-1, which has previously been shown to mediate the nuclear import of

STATla. Further studies showed that the NLS of IFNy is required for the ability of

STATla and IFNy to form this trimeric complex with Npi-1. The formation of the

complex IFNy/Npi-1/STATl a complex and the subsequent nuclear translocation of

STATla were all found to be dependent on the presence of the IFNy NLS. Previous

mutational studies on STATla failed to identify an NLS motif responsible for the

nuclear import of STATla (Sekimoto et al., 1997) via Npi-1. Our data strongly

support the conclusion that the required NLS for the nuclear localization of

STAT1 a is provided by IFNy, and IFNy acts as a chaperone for the nuclear delivery

of STAT a in the form of a IFNy:STATla:Npi-l complex.

Earlier studies on the binding of STATla to Npi-1 have suggested that

STATla binds to the C-terminus of Npi-1 (residues 456-538) at a site that cannot

be competed for by the basic SV-40 T-NLS (Sekimoto et al., 1997), suggesting that

STATla binds outside of the "conventional" NLS binding site. However, an

alternate interaction site for basic NLSs has been identified with the C-terminal

residues 501-510 (Moroianu et al., 1996 ), which falls within the STATla binding

region described on Npi-1 (Sekimoto et al., 1997). Interaction of the IFNy NLS at

this site on Npi-1 would provide a mode for binding of an IFNy/STAT a complex

to Npi-1 in an IFNy NLS-dependent fashion. Further, amino acid differences within

NLSs are known to alter both the specificity and affinity for importin binding sites.

Thus, while the SV-40 T-NLSs would qualitatively bind both sites on Npi-1, its

affinity for the C-terminal STATla site on Npi-1 may be too low to compete with

the high specific binding of the IFNy NLS within the IFNy/STATl a complex. This

would explain why in the presence of STATIa already bound to Npi-1, the SV-40

T-NLS cannot displace STATIa, as determined in the earlier studies (Sekimoto et

al., 1997). Different binding sites for IFNy NLS and SV-40 T-NLS are also

consistent with the ability of SV-40 T-NLS to "compete" in functional assays

(Subramaniam et al., 1998), since in functional import assays utilization of Npi-1 by

excess SV-40 T-NLS would prevent import of IFNy. Further studies into the

binding of IFNy to Npi-1 should help provide insight into these phenomenon.

While our data demonstrate that the interaction of IFNy with Npi-1 regulates

the trafficking of STATla to the nucleus, it remains to be determined what the

exact contribution of intracellular IFNy is to signaling events that lead to the

formation of this complex with STATIa. Studies from our laboratory and that of

others provide strong support for one mechanism by which intracellular IFNy may

regulate early receptor events leading to the activation of the JAK/STAT pathway

(Johnson et al., 1998a,b), and the subsequent nuclear translocation of STATla. As

has been mentioned before, the mouse IFNy (95-133) and human IFNy (95-134)

peptides, which contain the NLS, are agonists of IFNy when delivered

intracellularly (Johnson et al., 1998a,b). Since it is well known that the ability of

IFNy to induce MHC Class II and an antiviral state is dependent on a functional

JAK/STAT pathway (Darnell, 1998), the agonist peptides must be able to activate

the JAK/STAT pathway intracellularly to manifest their biological effects. The NLS

motif is also required for the agonist properties of the peptide (Szente et al., 1994).

Thus, the intracellular interaction of the NLS motif of IFNy is required for STATla

nuclear uptake.

As mentioned above STATla appears to be deficient in a NLS that can

mediate its nuclear import (Sekimoto et al., 1997). Hence we propose that the C-

terminal domain of IFNy, subsequent to its role in recruitment of STATla to

IFNGRa and priming of STATla at IFNGRa for nuclear translocation, via its NLS

motif may directly interact with the Ran/importin pathway to mediate the nuclear

delivery of STATla via a IFNy/IFNGRa/STATla complex. This proposal is

consistent with our recent findings that following extracellular binding of IFNy to

the receptor complex IFNGRa is selectively translocated to the nucleus. This

nuclear translocation of IFNGRa occurs with the same kinetics as that of

STATla, and IFNGRa appears to co-localize with STATIa during this time.

IFNGRa and STATla can also be recovered as a complex during these processes

following IFNy treatment. We have shown in this study that IFNy similarly can also

be recovered as a complex with STATla from the cytoplasm of IFNy treated cells.

These data, thus, argue for the presence of a IFNy/IFNGRa/STATla complex in

STAT a nuclear import.

One major advantage of such ligand-assisted nuclear chaperoning of STAT

transcription factors is the high-degree of specificity that is inherent in sequestering

a pool of STAT within the ligand-receptor complex that activated it. This would

imply that ligand and/or receptor are/is involved in the specificity of STATs at the

level of transcription. Further, this would explain why different ligands with

different biological functions on a given cell, activate the same STATs in these


Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E1AH2ZZTR_Q1OGD7 INGEST_TIME 2013-10-24T20:57:27Z PACKAGE AA00017657_00001