Interferon-α resistance in renal carcinoma cells is associated with defective induction of signal transducer and activator of transcription 1 which can be restored by a supernatant of phorbol 12-myristate 13-acetate stimulated peripheral blood mononuclear cells

Therapy of selected human malignancies with interferon-α is widely accepted but often complicated by the emergence of interferon-α resistance. Interferon is a pleiotropic cytokine with antiproliferative, antitumour, antiviral and immunmodulatory effect; it signals through the Jak-STAT signal transduction pathway where signal transducer and activator of transcription 1 plays an important role. Here we report both, a lack of signal transducer and activator of transcription induction in interferon-α resistant renal cell carcinoma cells and signal transducer and activator of transcription 1 reinduction of phorbol 12-myristate 13-acetate-stimulated peripheral blood mononuclear cells supernatant. Preliminary experiments on the identification of the molecules that reinducing signal transducers and activators of transcription 1 indicate that interferon-γ may be the responsible candidate cytokine, but several others may be involved as well. This work provides the basis for therapeutic strategies directed at the molecular modulation of interferon-α resistance in human neoplasms. British Journal of Cancer (2002) 86, 449–455. DOI: 10.1038/sj/bjc/6600066 www.bjcancer.com © 2002 The Cancer Research Campaign

Interferon-a (IFN-a) plays an important role in the treatment of various human malignancies, among them renal cell carcinoma (Dorr, 1993); however, response to IFN-a is often impaired by the development of IFN-resistance (Devita et al, 1989), mechanisms of which are poorly understood.
Interferon-a belongs to a group of cytokines with antiviral, antiproliferative, antitumour and immunmodulatory activities (Pestka et al, 1987). Binding of IFN-a to the IFN Type I receptor results in oligomerization of the receptor subunits and subsequent transphosphorylation of receptor-associated Janus-kinases Jak1 and Tyk2; activated Jak1 and Tyk2 subsequently phosphorylate tyrosine residues on the associated receptor chain. Signal transducers and activators of transcription (STAT) 1 and 2 can then bind to the receptor by their SH2 domains which are thereupon tyrosine phosphorylated by the receptor-associated Janus-kinases; thereafter, the STATs are released from the receptor and form STAT1-STAT2heterodimers which translocate to the nucleus where they bind with p48 to form the interferon stimulated gene factor 3 (ISGF3). ISGF3 binds to the interferon stimulated response element (ISRE) in the promoter of IFN-induced genes resulting in transcription of interferon-stimulated genes (ISG) (Schindler and Darnell, 1995;Haque and Williams, 1998).
There is evidence that IFN-a resistance is associated with defective components of the Jak-STAT-Pathway (Pansky et al, 2000) e.g., defective activation of ISGF3 (Xu et al, 1994;Wong et al, 1997), lack of STAT1 expression (Sun et al, 1998) or STAT3 induction (Yang et al, 1998). It has been reported that sequential treatment of interferon resistant cells with retinoic acid or tamoxifen followed by interferon-a up-regulates STAT1 expression and ISGF3 activation, respectively, in cells which do not respond to either single agent (Kolla et al, 1996;Lindner et al, 1997).
Here we sought (a) to characterize STAT1 deficiency associated with IFN-a resistance and (b) to identify modulators of STAT1 induction. These studies provide the basis for a potential modulation of resistance to IFN-a in renal cell carcinoma ex-vivo as well as in renal cell carcinoma patients receiving systemic IFN-a.

Cells and culture conditions
A-498 and Caki-2 cells were obtained from ATCC (Rockville, USA) (Cat. No. HTB 44). The fresh RCC cells were established according to Ebert et al (1990). Cells were grown in RPMI 1640 supplemen-ted with 10% heat-inactivated FCS, 2 mM L-glutamine, 50 mg ml 71 streptomycin and 50 IU ml 71 penicillin at 378C in an atmosphere containing 5% CO 2 and passaged once a week.
A relative resistance to the antiproliferative effects of interferona was best produced by continuous incubation of the A-498 and fresh renal cell carcinoma (RCC) cells with 1000 IU ml 71 interferon-a over 3 -4 months.

Preparation of PMA-stimulated PBMC supernatant
Peripheral blood mononuclear cells (PBMC) were isolated from buffy-coat leucocyte concentrates of healthy donors by Ficoll gradient centrifugation. The cells were cultured in RPMI 1640 supplemented as indicated above and stimulated with 10 ng ml 71 PMA for 4 days. After centrifugation for 10 min at 225 g the media was decanted and again centrifuged for 10 min at 2000 g. The supernatant was diluted 1 : 2 with RPMI 1640 and added to the cells.

Whole cell extracts
Cells were treated with IFN-a, IFN-g and supernatant as single agents, as well as with combinations of IFN-a and supernatant for various times as indicated in Figures 1 -5, 7. IFN-a or IFN-g was added during the last 30 min of incubation with supernatant. Cells were washed with ice-cold phosphate-buffered saline (PBS) and resuspended in RSB buffer (10 mM Tris (pH 7.4); 10 mM NaCl; 1.5 mM MgCl 2 ; 10 mM NaF; supplemented with 0.15 mM PMSF; 1 mM DTT; 0.2 mM Na 3 VO 4 ) and incubated 10 min on ice for swelling. After centrifugation at 15 800 g for 10 s at 48C, cells were resuspended in buffer C (20 mM HEPES; 420 mM NaCl; 1.5 mM MgCl 2 ; 0.2 mM EDTA; 25% Glycerin; 10 mM NaF; supplemented with 0.15 mM PMSF; 1 mM DTT; 0.2 mM Na 3 VO 4 ), sucked up and down 5 -10 times in a 26 G syringe and incubated on ice for 45 min. Whole cell extract was obtained by centrifugation at 15 800 g for 10 min at 48C. The supernatant (=whole cell extract) was shock frozen in liquid nitrogen at 71758C and stored at 7808C until use.

STAT1-reinduction experiments of g-interferon
Based on experiments of Venkataraman et al (1999) IFN-g-1b was used in a concentration of 10 ng ml 71 . Experiments were performed in duplicate.

Statistical analysis
Assays were regularly performed in triplicates and statistical means were established.

IFN-a associated antiproliferative effect and STAT1 induction in RCC cells
The RCC cell lines A-498 and Caki-2 and fresh RCC cells were assessed for the antiproliferative effect of IFN-a by a cell proliferation ELISA based on BrdU incorporation. In A-498 and fresh RCC cells, IFN-a inhibited cell proliferation in a dose-dependent manner while in Caki-2 cells, IFN-a showed no antiproliferative effect (data not shown).
As IFN-a is known to induce STAT1 in various cell types, cells were investigated for STAT1 induction. Cells were incubated for 30 min with 1000 IU ml 71 interferon-a and assessed for STAT1 induction by EMSA. We found that in A-498 and fresh RCC cells IFN-a induced STAT1 but not in Caki-2 cells, suggesting that Experimental Therapeutics primary resistance towards the antiproliferative effect of IFN-a is associated with defective STAT1 induction ( Figure 1). Identity of STAT1 was confirmed by adding STAT1-mAb which selectively blocked the STAT1-band.

IFN-a resistant A-498 and RCC cells are defective in the induction of STAT1
IFN-a resistance of A-498 and fresh RCC cells, respectively, was induced by culturing the cells over 3 -4 months in 1000 IU ml 71 IFN-a. A relative resistance to IFN-a in A-498 and fresh RCC cells was subsequently confirmed by cell proliferation ELISA.
An EMSA for determination of STAT1 activity was performed as indicated above. IFN-a failed to induce STAT1 induction in resis-

PMA-stimulated PBMC supernatant can reinduce STAT1 in IFN-a resistant A-498 and fresh RCC cells
The supernatant was prepared as described below. A-498 cells were treated for 1, 4, 6 and 24 h either with the supernatant alone or with supernatant followed by 30 min 1000 IU ml 71 interferon-a as indicated in Figure 2A,B.
In the sensitive cells, the supernatant alone induced a visible band after 1 h of incubation, only. In the presence of IFN-a, a STAT1-band was detectable at all time points (Figure 2A). The supernatant alone and its combination with IFN-a were able to restore STAT1 induction in resistant cells. The most intense STAT1-band was detected after 1 h of stimulation, it was significantly less intense 4 and 6 h and disappeared at 24 h ( Figure  2B). When repeated with the PBMC supernatant of another donor, this assay led to the identical results (data not shown).
As we sought to exclude a cell line specific effect of the supernatant, we treated fresh RCC cells in the same manner. These cells were first derived from one RCC patient, and an IFN-a resistant variant which also lacks STAT1 induction was subsequently generated.
In sensitive cells, the supernatant alone induced a weak STAT1band after only 1 h. As expected in the presence of interferon-a, STAT1 was induced at all time points ( Figure 3A). The present supernatant reinduced STAT1 in the IFN-a resistant cells with a clearly detectable band after 1 h ( Figure 3B); Interferon-a treated A-498 S cells served as positive control for STAT1.

In Caki-2 cells, PMA-stimulated PBMC supernatant cannot reinduce STAT1
As the supernatant induced optimal STAT1 in A-498 and fresh RCC cells after 1 h, Caki-2 cells were treated for 1 h with supernatant alone or in combination with IFN-a. Neither the supernatant alone nor its combination with interferon reinduced STAT1 in Caki-2 cells (Figure 4).

Confirmation of supernatant-induced STAT1 via competition with STAT1-mAb and unlabelled oligo
To confirm that the band induced by the present supernatant correlated to STAT1, a competitional assay was performed with STAT1-mAb and with 50-, 100-and 300-fold excess of unlabelled oligo, which all blocked the STAT1-band ( Figure 5).

PMA alone or in combination with interferon-a does not induce STAT1
In order to demonstrate that the effect of STAT1 reinduction was due to the cytokines secreted by PMA-stimulated PBMC and not by PMA itself, A-498 cells were incubated with 10 ng ml 71 of PMA either alone and in combination with IFN-a. PMA alone did not induce STAT1 neither in the sensitive nor in the resistant cells, while the combination of PMA and IFN-a induced STAT1 in the sensitive cells, but not in the resistant cells ( Figure 6A,B).

IFN-g alone does induce STAT1
To investigate the cytokine potentially capable of STAT1 reinduction, IFN-a resistant A-498 cells were incubated with IFN-g 1b (10 ng ml 71 ), which clearly reinduced STAT1. A-498 sensitive and resistant cells treated with IFN-a were used as positive and negative control for STAT1 (Figure 7).

DISCUSSION
Response to IFN-a is often impaired by the development of IFNresistance (Devita et al, 1989). Many mechanisms have been proposed to mediate IFN-a resistance in various different cell types: lack of ISGF3 (Xu et al, 1994;Wong et al, 1997) or one of its components (Dron and Tovey, 1993;Sun et al, 1998), decrease in ISG expression (Affabris et al, 1983;Salzberg et al, 1983), lack of a high affinity receptor site (Hannigan et al, 1984) or the development of anti-interferon antibodies in patients (Quesada et al, 1985;Steis et al, 1988). However, there seems to be no unique mechanism for all cells given reports on interferon-resistant cells which do not inherit one of the above defects (Grups and Bange, 1990;Talpaz et al, 1992). While an interferon-resistant cell line with entire STAT1 induction has been reported (Yang et al, 1998), STAT1 seems to be an essential component of interferon signalling as STAT1 deficient mice do not respond to interferons and are highly sensitive toward viral infections (Durbin et al, 1996;Meraz et al, 1996); in addition complementation of the IFN-unresponsive mutant cell line U3 interferon with STAT1, cDNA constructs reportedly restores ISGF3 formation and transcriptional response to interferon (Müller et al, 1993).
Here, we detected a deficient STAT1 induction in two different RCC sublines, with secondary interferon-a resistance and in a primary resistant RCC cell line. There were several possibilities to explain the defect, e.g.: (a) decreased synthesis, increased degradation, decreased tyrosine phosphorylation of STAT1; (b) malfunction in the preceding steps of the signal transduction pathway; or (c) lack of STAT1 transcript due to a mutation in the STAT1 gene.
In the present experiments, the PMA-stimulated PBMC supernatant reinduced STAT1 in A-498 and fresh RCC cells with secondary IFN-a resistance but not in primary resistant Caki-2 cells. This indicated that in resistant A-498 and fresh RCC cells, the STAT1 Experimental Therapeutics  1 and 4).
gene is still intact while in primary resistant Caki-2 cells, the defect in the signal transduction pathway appears to be more extensive. Previously, Sun et al (1998) reported a lack of STAT1 transcript in IFN resistant lymphoma cells. Here a mutation in or a loss of the STAT1 gene could explain why Caki-2 cells were refractory to PMA-PBMC-mediated STAT-reinduction. It is known that PMA-stimulated PBMC produce a number of cytokines; as observed there are several cytokines including IL-2, IL-6, IL-9, IL-10, IL-11, IFN-g, hormones as growth hormone and prolactin and growth factors as GCSF, EGF, PDGF, CSF-1 and angiotensin which may induce STAT1 in various cell types (Schindler and Darnell, 1995;Briscoe et al, 1996;Ihle, 1996;Leaman et al, 1996).
Here, we demonstrated that IFN-g reinduces STAT1 in IFN-a resistant A-498 cells; this suggested that IFN-g may be the PMAstimulated PBMC derived cytokine capable of reinducing STAT1 but several others may be involved as well. Since biological effects of IFN-a and IFN-g are both mediated by Jak1, it appeared that Tyk2 tyrosine kinase or the receptor may be impaired in IFN-a resistant carcinoma cells upstream of STAT1.
The significance of the additional STAT1 inducers remained unclear as their biological effects are not necessarily mediated by a Jak-STAT signal transduction pathway or require STAT1 induction. Previously, STAT1 deficient mice responded normally to growth hormone, EGF and IL-10 which induce STAT1 in vitro (Meraz et al, 1996). STAT1 induction by cytokines other than interferons were cell-specific (Schindler and Darnell, 1995;Han et al, 1996). Therefore, in RCC some of these known cytokines may actually not be able to induce STAT1, while there might be additional inducers.
Future experiments will focus on further steps toward the identification of those molecular regulators associated with IFN-a resistance in order to provide new modalities of treatment for renal cell carcinoma patients.