|
|
||||||||


* Department of Cancer Biology, Lerner Research Institute, and
Computer Core, Cleveland Clinic Foundation, Ohio;
Department of Genetics, Case Western Reserve University, Cleveland, Ohio; and
Department of Laboratory Medicine and Pathobiology, University of Toronto, Ontario, Canada
Correspondence: Bryan R. G. Williams, Ph.D., Chairman, Department of Cancer Biology NB 40, Lerner Research Institute, Cleveland Clinic Foundation, Cleveland, OH 44195. E-mail: williab{at}ccf.org
| ABSTRACT |
|---|
|
|
|---|
Key Words: cytokine Janus kinase protein kinase RNA-regulated
| INTRODUCTION |
|---|
|
|
|---|
subtypes, IFN-
and IFN-ß, and the type II species IFN-
. IFNs
play an essential role in innate immunity by inhibiting the replication
and spread of viral, bacterial, and parasitic pathogens. They also
modulate immune responses and exert antiproliferative effects in some
cell types. As a result of these functions, IFNs are used in the clinic
to treat certain viral infections, some cancer types, and multiple
sclerosis [reviewed in ref. 1
]. IFNs mediate their effects by binding
to cell surface receptors activating members of the JAK kinase
family of proteins. Activated JAK kinases phosphorylate the signal
transducers and activators of transcription (STAT) family of
transcription factors. The STAT proteins homo- or heterodimerize and
form complexes with other transcription factors to activate
transcription of IFN-stimulated genes (ISGs) [2
]. The
gene products regulated by IFNs are the primary effectors of the IFN
response. Although the functions of most ISGs remain to be elucidated,
some of the best studied ISGs play pivotal roles in host defense. Experiments using mice indicate the IFNs are essential for innate immunity against viral infections. Mice with a targeted disruption of the type I or II IFN receptor genes are extremely susceptible to viral infections. These mice have multiple defects in host defense and show enhanced viral replication in many tissues [3 4 5 ]. IFNs induce the production of several known antiviral proteins including the double-stranded RNA-dependent kinase "protein kinase RNA-regulated" (PRKR), a family of 2',5'-oligoadenylate synthetases that lead to the activation of RNase L and the Mx proteins, all of which have been shown to restrict the growth of certain viruses [1 ]. However, the inhibition of viral replication induced by IFNs is only partially dependent on these particular ISGs, because mice triply deficient for PRKR, RNase L, and Mx1 genes retain partial responsiveness to the antiviral effects of IFNs [6 ]. These results imply that other as-yet-unidentified ISGs are also potent antiviral effectors.
To identify ISGs and perhaps elucidate new functions for IFN, we
undertook extensive microarray analysis of RNA samples collected from
experiments on human and murine cell lines treated with IFN-
,
IFN-ß, or IFN-
. Previous work from our laboratory identified 122
ISGs, using HT1080 as the cell type and oligonucleotide microarrays
[7
]. Here we extend and confirm these results using
other microarray-screening methods. By combining data from all sources
listed in Table 1
, we have screened several thousand individual sequences and have
extended the initial 122 ISGs to over 300. To uncover new IFN
functions, each ISG was assigned to a series of defining functional
categories. Furthermore, the categorized ISGs were assembled into a
database that contains gene names, descriptions, accession numbers, and
links to other databases containing nucleotide and protein sequence
information. The genes were placed into progressively more specific
functional categories that are fully searchable. Our laboratory is
drawing from the microarray data to construct a cDNA microarray of
human and murine ISGs.
|
| CONSTRUCTION OF THE ISG DATABASE |
|---|
|
|
|---|
at least twofold in
two IFN-treated samples or threefold in a single treated sample. We
found that this subset includes nearly all the known ISGs that were
included on the Affymetrix arrays [7
]. Only experiments
that showed at least 1 in 3 genes to be present and a detection limit
of between 1 and 5 pM, determined using spiked RNA controls, were
included in the analysis. The INCYTE data sets were obtained using the
UniGEM V microarrays (INCYTE, St Louis, MO). The sample RNA was
harvested from cells treated as outlined in Table 1
, by Trizol
extraction according to manufacturer instructions (Life Technologies,
Gaithersburg, MD). The purified RNA was then shipped to INCYTE Corp.
for hybridization and analysis. To be included in the database, genes
must have increased in expression at least 2-fold in the
IFN-
-treated sample or 1.5-fold by IFN-ß twice during treatment.
In general, the INCYTE method underestimated the fold induction of ISGs
compared with the Affymetrix method, even when identical RNA templates
were analyzed (R. H. Silverman, unpublished results). The array
format, cell type, and treatment applied are detailed in Table 1
. The Affymetrix human 6800 chip set identified 122 ISGs [7 ] that were subsequently included in the functional groupings. This study showed that IFN positively influenced the expression of approximately 1 in every 55 human genes represented on the chip. By extrapolation, this suggests that there may be as many as 6362,180 ISGs, assuming that the number of genes in the human genome is between 35,000 and 120,000. Because all the data sets are redundant and include both human and murine genes, no overall figure on the number of individual genes screened was determined. Combining the results from the three array experiments and applying the above cutoff criteria, 335 unique ISGs with high homology to known genes and 78 expressed sequence tag (EST) sequences were identified.
The genes were grouped into broad functional families and then further categorized into more specific groups. For example, an IFN-induced protease forming part of the proteosomal subunit was placed in the broad category "Host Defense" and then in the more specific group "Antigen Processing," then "Proteosome Subunit," and finally "Protease." Table 2 outlines the functional categories containing over 4 members, defined for the 335 unique ISGs identified by the microarrays. These groups were assigned by using resources from the following databases: GenBank (http://www.ncbi.nlm.nih.gov/Genbank/index.html), PubMed (http://www4.ncbi.nlm.nih.gov/entrez/query.fcgi), and Gene Cards (http://bioinfo.weizmann.ac.il/cards) [8 ].
|
|
| FUNCTIONS OF ISGS |
|---|
|
|
|---|
[10
, 16
]. Phosphorylation of
eIF2
halts cellular translation, thus inhibiting host and viral
protein synthesis in infected cells [12
,
16
]. We recently showed that PRKR also phosphorylates the
B56
subunit of pyrophosphatase 2A (PP2A), which reduces
translation of a luciferase reporter, presumably by inhibition of PP2A
dephosphorylation of the eIF4E translation initiation factor
[17
]. Thus PRKR might inhibit translation and viral
replication through multiple mechanisms.
Inhibition of translation by dsRNA was the assay used to clone PRKR
[16
]; however, the availability of mice with a targeted
deletion in the PRKR gene [18
] has allowed new roles for
the PRKR protein to be elucidated. PRKR-null mice develop normally and
have no overt defects [18
]; however, murine embryonic
fibroblasts (MEFs) derived from PRKR-null mice are less sensitive to
some apoptotic stimuli than those obtained from isogenic normal mice
[19
]. Also, overexpression of PRKR in 3T3 cells induces
apoptosis [20
, 21
]. Recent evidence has
indicated that PRKR is required for the activation of the nuclear
factor-
B (NF-
B) transcription factor by some inducers
[22
, 23
]. NF-
B is a latent transcription
factor that is sequestered in the cytoplasm by the I
B
subunit.
The IKK
and IKKß kinases phosphorylate the I
B proteins, which
targets them for degradation and releases the NF-
B transcription
factor, which then translocates to the nucleus and activates gene
expression [24
25
26
].
The genes induced by NF-
B play an important role in apoptosis,
immune modulation, and induction of inflammatory cytokines
[27
]. Others and we have shown that activation of
NF-
B by dsRNA depends on PRKR activation of the IKKß kinase
[22
]. In PRKR-null cell lines, dsRNA failed to stimulate
IKK activity compared with cells from an isogenic background that are
wild type for PRKR. Coimmunoprecipitation assays showed that PRKR was
physically associated with the IKK complex and transient expression of
a dominant negative mutant of IKKß, or the NF-
B-inducing kinase
inhibited dsRNA-induced gene expression from an NF-
B-dependent
reporter construct [22
]. Taken together, these results
demonstrate that PRKR-dependent dsRNA induction of NF-
B is mediated
by NF-
B-inducing kinase and IKK activation. Whether PRKR kinase
activity is required for this function remains a subject of debate
[28
, 29
].
The PRKR protein is required for efficient activation of some
stress-activated protein kinases (SAPKs). The cytokines tumor necrosis
factor-
(TNF-
) and interleukin (IL)-1ß and pathogenic factors
such as lipopolysaccharide (LPS) and dsRNA induced phosphorylation and
activation of mitogen-activated protein (MAP) kinase kinase (MKK6),
p38
MAPK and Jun N-terminal kinase SAPKs in normal MEFs but not in
PRKR-null MEFs [30
]. This did not reflect a global
failure to activate the SAPKs in the PRKR-null background because other
physiochemical stressors activated the SAPKs normally
[30
]. The failure to activate SAPKs also appeared to be
important in vivo, because induction of two SAPK target genes, IL-6 and
IL-12, by LPS was reduced in PRKR-null mice when compared with normal
mice [30
]. NF-
B and p38
can synergize to induce
inflammatory cytokines, implying that PRKR may be an important mediator
of the inflammatory response induced by foreign pathogens.
Recent work has identified that PRKR-null mice are highly sensitive to
vesicular stomatitis virus (VSV) or influenza virus. It is interesting
that the differences were apparent only when these viruses were
administered intranasally and not when they were administered
systemically [11
]. The lungs from infected PRKR-null
mice showed much higher titers of virus compared with wild-type mice.
The reason PRKR-null mice are more sensitive is not entirely clear; it
may be the failure of PRKR to halt protein synthesis in the lungs of
PRKR-null mice, allowing increased viral replication, although other
eIF2
kinases are known to exist. A lack of PRKR could result in a
reduction of apoptosis, allowing the virus to replicate to higher
levels in cells, or stimulation of the immune system may be defective
in PRKR-null mice through reduced NF-
B or SAPK activity.
Alternatively it has been demonstrated that MEFs derived from PRKR-null
mice fail to produce nitric oxide (NO) in response to LPS, IFN
and
dsRNA [31
]. NO is an important inhibitor of VSV
replication [32
], and a defect in NO production in the
lungs of PRKR-null mice may explain the sensitivity of the lungs to
VSV. We are currently using various methods to determine the functions
of PRKR which are important in host defense against viral and bacterial
infection.
Another transcript encoding the antiviral OAS2 protein was induced by IFN in the array experiments. The OAS proteins are activated by dsRNA and produce 2'5'-oligoadenylates that bind and activate the latent ribonuclease RNase L. The RNase L protein cleaves viral and cellular mRNA and ribosomal RNA, halting cellular production of protein [1 ]. The OAS proteins form a gene family that resides on chromosome 12q24.1 and likely evolved from a single ancestral gene through gene duplication (32a). The p46 (OAS1) and p69 (OAS2) synthetase proteins have antiviral activity through the activation of RNase L after viral infection. However, the recently identified and largest member of this family, p100 (OAS100), does not activate RNase L but rather potentiates apoptosis in cells exposed to dsRNA. Overexpression of p100 but not other OAS family members sensitizes cells to apoptosis by dsRNA [32a ].
The array experiments showed induction of a family of large guanosine
triphosphatases (GTPases) known in the mouse as Mx1 and Mx2 and in the
human as MxA and MxB. The Mx genes are mutated in most inbred
laboratory mouse strains that are highly sensitive to the influenza
virus [33
]. The Mx proteins belong to the dynamin family
and inhibit the replication of some RNA viruses by binding to viral
ribonucleoprotein structures and preventing transcription of viral RNA
or movement of viral subparticles within the cell [34
,
35
]. Another recently described IFN-induced
guanylate-binding protein (GBP) 1 was identified using the arrays and
has been shown to have innate antiviral activity [36
].
Overexpression of GBP1 inhibited the replication of both VSV and
encephalomyocarditis virus in 3T3 cells. Expression of antisense GBP1
also reduced the antiviral effect of IFN
but not IFN
[36
]. The GBP1 gene was preferentially induced by
IFN
.
The above ISGs are all previously characterized mediators of innate immunity; other ISGs must be involved in the inhibition of viral replication induced by IFNs because mice triply deficient for PRKR, RNase L, and Mx1 genes retain partial responsiveness to the antiviral effects of IFNs [6 ]. The GBP1 gene product also only partially protects cells from viral killing [36 ]. The complement of the intracellular innate immune response is the humoral immune response, which is mediated by immune effector cells that respond to and clear the infectious agent. The categorization of the ISGs showed a large subgroup of host defense gene-induced humoral immune responses. These immunomodulatory genes included the chemokines, which are small proteins that recruit lymphocytes to sites of inflammation or infection [37 38 39 ]. In all, genes for six chemokinesMIG, EBI1, SCYA2, SCYA5, SCYB10, and IL-8were identified as ISGs. IFN also induced the expression of four genesICAM1, SELL, CD47 (see Fig. 2 ), and ALCAMthat promote lymphocyte adhesion to endothelial cells. Adhesion of lymphocytes to vessel walls is an important first step in the trafficking of lymphocytes to areas of infection. This is the first report we know of that details the induction of the CD47 protein by IFNs. The CD47 protein associates with integrins and plays roles in cell adhesion signaling [40 , 41 ], and CD47-null mice are extremely susceptible to bacterial infection primarily caused by a failure to recruit neutrophils [42 ]. Thus, IFN induces numerous genes that enhance recruitment of immune effector cells to the site of production.
|
Signaling
IFN induced 53 genes involved in modulating other signaling
pathways, the largest single category of genes induced by IFN. The
signaling category contained several genes involved in inflammatory
cytokine signaling such as MYD88 [47
48
49
],
JUN, RELA, MAP2K1, MAP3K8,
and TRADD. The MYD88 protein is an important link in
toll-like receptor signaling pathways [49
,
50
], as well as in the proinflammatory cytokine IL-1
signaling pathway. The TRADD protein is an adapter in the TNF-
and
IL-1 signaling pathways [51
]. Both the MYD88
and TRADD genes are important in activation of NF-
B by
foreign pathogens and proinflammatory cytokines. Induction of these
signaling proteins by IFN could potentially lead to an increased
response to ligand binding, which may have implications for the
clinical side effects of IFN, including high fever
[52
]. IFNs may alter the tissue specificity of some
ligand responses through the induction of a crucial signaling protein
in cell types where it is not normally expressed.
IFN also induced expression of an anti-inflammatory protein, LGALS3B
[cyclophilin-c-associated protein (CYCAP) in the mouse; see Fig. 2
].
Mice with a targeted deletion in the CyCAP gene are very
susceptible to LPS killing [53
] and express elevated
levels of IFN-
, IL-12, and TNF-
. Thus CYCAP may function to curb
the proinflammatory effects of IFN. The interplay between the various
ISGs that mediate inflammation may be important in reducing the toxic
side effects of high-dose IFN therapy, but this requires further study.
Another class of signaling proteins was the G proteins. These included RAN, RANBP, NET1A, and GEM. G proteins mediate many intracellular functions, including cytoskeletal remodeling, vesicle transport, and growth [these functions are reviewed in ref. 54 55 56 ]. Potential modifiers of the Ras G protein also were induced by IFN; RAN, ras homologue ARHC, ras-related rab-8 MEL, and ras GTPase-activating protein IQGAP1 were all identified as ISGs. The ras signaling pathway leads to activation of cell growth, and overstimulation of the ras pathway is a common defect in many cancers [57 , 58 ]. Activation of signaling pathways stimulated by mitogenic growth factors is also a common property of many cancers. IFN induced 15 genes that are involved with growth factors and growth factor signaling. These included VEGF, FGF, VRP, PDGFRL, ECGF1, EREG, and CTGF. Most of these growth factors are mitogenic and have been implicated in the control of angiogenesis and cancer [59 60 61 62 ], raising the interesting possibility that IFNs may influence angiogenesis. IFN is a potent antitumor agent in a limited number of cancer types [63 ]; however, if the induction of growth factors is confirmed in vivo, IFN treatment of cancers that rely on growth factors for survival might be detrimental.
The induction of signaling proteins by IFN was mirrored by the induction of a large number of transcription factors, 37 in total. Of the 37 transcription factors 30 activated transcription whereas only 8 were confirmed to repress transcription. Some of the genes categorized under transcription factors both activated and repressed transcription, and the direct role of others was unknown. The array experiments identified six members of the IFN response factor (IRF) family of proteins: IRF1, IRF2, IRF3, IRF4, IRF5, and IRF7. The IRF proteins are a family of secondary effectors that mediate immune modulation, IFN production after viral infection, and IFN signaling [64 65 66 67 ]. This family of proteins, which is crucial to the biological response to IFNs, was the most represented family of transcription factors identified by the array experiments.
Other interesting transcription factors induced by IFN were the
hypoxia-inducible factor (HIF1A) gene (see Fig. 2
) and the Myc
promoter-binding protein (MPB1). The HIF1
protein stimulates
transcription of genes that mediate the intracellular response to
anoxia and induces angiogenesis [68
, 69
].
This adds further support for a role of IFN in stimulating
angiogenesis. HIF1
is overexpressed in some cancer types, where it
is thought to be important for protecting the tumor cells from anoxia
and stimulation of angiogenesis into the tumor [70
]. The
MPB1 protein binds the Myc promoter and represses transcription
[71
], which might be one of the mechanisms IFN uses to
down-regulate myc expression and reduce proliferation in
certain cell types [1
]. This protein may play a role in
the antitumor activities of the IFNs. Thus it appears that IFN induces
proteins involved in both growth activation and attenuation. It will be
interesting to see whether the induction of these genes is cell type
specific and correlates with the effects of IFN on the cell type.
IFN also induced many proteins involved in the posttranscriptional
regulation of gene expression. These include genes involved in
translation, such as those encoding elongation and initiation factors:
EIF2A, EIF2B, EIF2S2,
EIF3S10, and EIF3S6 and the genes encoding the
translational inhibitors IFN-inducible 56K (IFI56)
[72
, 73
] and PRKR [10
]. It
is unclear whether induction of the elongation and initiation factors
increases translation of proteins, because IFN has never been shown to
enhance translation. The role of induction of translation elongation
and initiation factors in the IFN response remains to be studied. The
dsRNA- and IFN-induced IFI56 protein appears to inhibit translation
after IFN stimulation of cells, through sequestration of the
translation initiation factor eIF-3 [73
]. The PRKR
protein kinase is activated by dsRNA and inhibits translation by
phosphorylation of eIF2
.
IFN also regulated many genes involved in protein degradation with seven proteases and five ubiquitin related genes. Three of the proteases are catalytic components of the proteosomal subunits which, coupled with the ubiquitination pathway, are responsible for the targeted degradation of many proteins. As our understanding of the posttranslational modification of proteins expands, the roles it plays are becoming increasingly important [25 , 44 , 74 ]. IFN also regulated several RNA-interacting proteins, consisting of two helicases (DDX3 and DDX21) and four genes shown to play a role in RNA splicing (SFPQ, SFRF2, SF3A3, and SF3A1). Thus, IFN may modulate the production of functional proteins at several levels, transcriptional induction of mRNA, the splicing and processing of this RNA, translation, and finally degradation of the protein.
The apoptosis category contained 19 ISGs. Apoptosis is the result of a proteolytic cascade of cysteine proteases (caspases) which leads to cleavage of important substrates and subsequent cell death [75 ]. IFN induced mostly proapoptotic proteins including CASP4 (see Fig. 2 ), CASP8, trail (TNFSF10), BAK1, and Fas or CD95 (TNFRSF6). All these proteins are involved in the activation of apoptosis by numerous inducers [76 77 78 79 ]. CASP4 and CASP8 are caspases that actively cleave other caspases and transduce the proteolytic cascade. The TNFRSF6 protein is a death receptor protein involved in signaling T-cell killing [76 ], while the TNFSF10 protein is a soluble TNF-like molecule involved in the induction of apoptosis when it binds its receptor [78 ]. Also, the phospholipid scramblase (PLSCR1) protein, a new ISG, was induced by IFN and is implicated in moving phosphatidyl serine to the outside of the plasma membrane in apoptotic cells [79 ]. This is a critical step in facilitating the recognition and destruction of the apoptotic cell by immune effector cells and is also involved in blood clotting [79 ]. The IFNs have been reported to be proapoptotic cytokines; they induce apoptosis in cells infected with certain viruses [80 ] and can also cause apoptosis in some transformed cell lines [81 , 82 ].
To aid in analysis of the IFN-response, we are arraying as many ISGs as
we can obtain onto a single ISG-chip. A prototype of this array has
been screened using cDNA from a renal cancer cell line, RCC1, that was
either treated with 100 IU of IFN
2b for 16 h or left untreated.
Total RNA was isolated using the Trizol method (Life Technologies). A
total of 1 µg of each RNA was amplified using two rounds of ds-cDNA
synthesis followed by T7-driven in vitro transcription
[83
]. The untreated RNA was labeled with the Cy5 (red)
fluorophore, and the RNA from IFN-treated cells was labeled with the
Cy3 (green) fluorophore (Amersham Pharmacia Biotech, Little Chalfont,
Buckinghamshire, England) [83
]. The hybridized
and scanned array and a selection of the results that were obtained are
shown in Figure 3
. The genes induced by IFN show increased green
fluorescence while those suppressed by IFN show increased red
fluorescence. Those that remained unchanged during this experiment are
yellow.
|
Figure 4 summarizes the numbers and various functions of ISGs identified by microarray analysis. The analysis of microarray data has offered new insights into IFN biology through the identification of genes such as CyCAP, MYD88, and CD47 and the gene encoding phospholipid scramblase. These genes and others have opened up potential new areas of IFN biology and have provided insights into some of the proteins that are important in the IFN response. The further analysis of genes induced by IFN will lead to a greater understanding of the multitude of effects these cytokines exert on cells. It may aid in the identification of novel therapeutic uses for IFN and should identify new candidates that may prove useful to monitor clinical responsiveness to IFN therapy. It is important that the array experiments, although reliable, are still subject to error; some of the ISGs mentioned may not be confirmed in further studies or may be induced only by IFN in a small number of cell lines. The compilation of the ISGs we have presented in the ISG database provides a new resource that will facilitate further research on IFN-mediated cell responses. It is our aim to maintain and update the database and incorporate newly recognized ISGs as these become known. We will be pleased to add novel ISGs identified by other laboratories and invite investigators to contact us with new information and/or comments.
|
| ACKNOWLEDGEMENTS |
|---|
We thank Mathias Frevel for constructive comments on the manuscript.
Received January 8, 2001; revised April 16, 2001; accepted April 19, 2001.
| REFERENCES |
|---|
|
|
|---|
Related Articles
This article has been cited by other articles:
![]() |
C. M. Cameron, M. J. Cameron, J. F. Bermejo-Martin, L. Ran, L. Xu, P. V. Turner, R. Ran, A. Danesh, Y. Fang, P.-K. M. Chan, et al. Gene Expression Analysis of Host Innate Immune Responses during Lethal H5N1 Infection in Ferrets J. Virol., November 15, 2008; 82(22): 11308 - 11317. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. A. Samarajiwa, S. Forster, K. Auchettl, and P. J. Hertzog INTERFEROME: the database of interferon regulated genes Nucleic Acids Res., November 7, 2008; (2008) gkn732v1. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. J. Wright, J. B. Matthews, I. L. C. Chapple, N. Ling-Mountford, and P. R. Cooper Periodontitis Associates with a Type 1 IFN Signature in Peripheral Blood Neutrophils J. Immunol., October 15, 2008; 181(8): 5775 - 5784. [Abstract] [Full Text] [PDF] |
||||
![]() |
U. M. Nagarajan, D. Prantner, J. D. Sikes, C. W. Andrews Jr., A. M. Goodwin, S. Nagarajan, and T. Darville Type I Interferon Signaling Exacerbates Chlamydia muridarum Genital Infection in a Murine Model Infect. Immun., October 1, 2008; 76(10): 4642 - 4648. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Kisand, M. Link, A. S. B. Wolff, A. Meager, L. Tserel, T. Org, A. Murumagi, R. Uibo, N. Willcox, K. Trebusak Podkrajsek, et al. Interferon autoantibodies associated with AIRE deficiency decrease the expression of IFN-stimulated genes Blood, October 1, 2008; 112(7): 2657 - 2666. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Dolken, Z. Ruzsics, B. Radle, C. C. Friedel, R. Zimmer, J. Mages, R. Hoffmann, P. Dickinson, T. Forster, P. Ghazal, et al. High-resolution gene expression profiling for simultaneous kinetic parameter analysis of RNA synthesis and decay RNA, September 1, 2008; 14(9): 1959 - 1972. [Abstract] [Full Text] [PDF] |
||||
![]() |
K.-C. Lai, K.-W. Chang, C.-J. Liu, S.-Y. Kao, and T.-C. Lee IFN-Induced Protein with Tetratricopeptide Repeats 2 Inhibits Migration Activity and Increases Survival of Oral Squamous Cell Carcinoma Mol. Cancer Res., September 1, 2008; 6(9): 1431 - 1439. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Moserle, S. Indraccolo, M. Ghisi, C. Frasson, E. Fortunato, S. Canevari, S. Miotti, V. Tosello, R. Zamarchi, A. Corradin, et al. The Side Population of Ovarian Cancer Cells Is a Primary Target of IFN-{alpha} Antitumor Effects Cancer Res., July 15, 2008; 68(14): 5658 - 5668. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Saito and M. Gale Jr. Differential recognition of double-stranded RNA by RIG-I-like receptors in antiviral immunity J. Exp. Med., July 7, 2008; 205(7): 1523 - 1527. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. J. Pasieka, B. Lu, S. D. Crosby, K. M. Wylie, L. A. Morrison, D. E. Alexander, V. D. Men |