Leukemia Research 23 (1999) 373–384
`
`Differences in phosphorylation of the IL-2R associated
`JAK:STAT proteins between HTLV-I ((cid:27)), IL-2-independent and
`IL-2-dependent cell lines and uncultured leukemic cells from
`patients with adult T-cell lymphoma:leukemia(cid:19)
`Qian Zhang a, Benhur Lee a, Magda Korecka a, Gong Li c, Charles Weyland b,
`Steven Eck a, Antoine Gessain d, Naochimi Arima e, Stuart R. Lessin c,
`Leslie M. Shaw a, Selina Luger b, Malek Kamoun a, Mariusz A. Wasik a,*
`a Department of Pathology and Laboratory Medicine, Uni6ersity of Pennsyl6ania Center, 7.106 Founders Bldg., 3400 Spruce Str. Philadelphia,
`PA 19104, USA
`b Department of Hematology-Oncology, Uni6ersity of Pennsyl6ania Center, 7.106 Founders Bldg., 3400 Spruce Str. Philadelphia, PA 19104, USA
`c Department of Dermatology, Uni6ersity of Pennsyl6ania Center, 7.106 Founders Bldg., 3400 Spruce Str. Philadelphia, PA 19104, USA
`d Institut Pasteur, Paris, France
`e Kagoshima Uni6ersity, Kagoshima, Japan
`
`Received 24 August 1998; accepted 7 October 1998
`
`Abstract
`
`To determine activation status of the IL-2R-associated (Jak:STAT) pathway in the HTLV-I infected cells, we examined tyrosine
`phosphorylation of Jak3, STAT3, and STAT5 in several HTLV-I ((cid:27)) T-cell lines and in uncultured leukemic T cells isolated from
`patients with adult T-cell lymphoma:leukemia (ATLL). Constitutive basal phosphorylation of Jak3 and, usually, STAT3 and
`STAT5 was detected in all four IL-2-independent cell lines tested, but in none of the three IL-2-dependent cell lines. Similarly,
`there was no detectable basal phosphorylation of Jak3 and STAT5 in the leukemic cells from ATLL patients (0:8 and 0:3,
`respectively). However, stimulation with IL-2 resulted in Jak3 and STAT5 phosphorylation in both leukemic ATLL cells and
`IL-2-dependent lines. Furthermore, expression of SHP-l phosphatase which is a negative regulator of cytokine receptor signaling,
`was lost in most IL-2 independent cell lines (3:4) but not in the leukemic ATLL cells (0:3). Finally, the HTLV-I ((cid:27)) T-cell lines
`(313) but not the control, HTLV-I ((cid:28)) T-cell lines were resistant to rapamycin and its novel analog RAD. We conclude that (l)
`HTLV-I infection per se does not result in a constitutive phosphorylation of the Jak3, STAT3, and STAT5 proteins; (2) malignant
`transformation in at least some cases of ATLL does not require the constitutive, but may require IL-2-induced, activation of the
`IL-2R Jak:STAT pathway; and (3) there are major differences in T-cell immortalization mechanism(s) which appear to involve
`SHP-l and target molecules for rapamycin and RAD. © 1999 Published by Elsevier Science Ltd. All rights reserved.
`
`Keywords: IL-2R signaling; Malignant T cells; JAK5 kinase; STAT 5 protein; SHP-1 phosphatase
`
`1. Introduction
`
`IL-2 is a key cytokine involved in proliferation and
`differentiation of T lymphocytes and other cells of the
`immune system. IL-2 signaling involves dimerization of
`the b chain and common g chain (gc) of the IL-2
`
`(cid:19) Part of this work was presented at the XXXVIII Annual Meeting
`of the American Society of Hematology.
`* Corresponding author.
`
`receptor (IL-2R) [1]. In addition to IL-2R, the gc is a
`component of several other cytokine receptors; it can
`co-dimerize with cytokine-specific chains to transduce
`signals mediated by IL-4, IL-7, IL-9, and IL-15 [2–5].
`Signaling by cytokine receptors involves sequential acti-
`vation of the Janus-family tyrosine kineses (Jaks) and
`signal transducer and activator of transcription proteins
`(STATs) (reviewed in [6]). Binding of IL-2 to the IL-2R
`results in activation of Jakl and Jak3 kineses and
`tyrosine phosphorylation of several substrates, includ-
`
`0145-2126:99:$ - see front matter © 1999 Published by Elsevier Science Ltd. All rights reserved.
`PII: S 0 1 4 5 - 2 1 2 6 ( 9 8 ) 0 0 1 7 3 - 8
`
`Breckenridge Exhibit 1063
`Zhang
`Page 001
`
`

`

`374
`
`Q. Zhang et al. :Leukemia Research 23 (1999) 373–384
`
`ing Jakl and Jak3 themselves, as well as the IL-2R b
`and gc chains [7]. The phosphorylated IL-2R chains
`recruit proteins such as STAT5 [8,9] and,
`in phyto-
`hemagglutinin preactivated T-cell blasts, STAT3 [8,10].
`The STATs, upon phosphorylation, presumably by the
`Jaks, translocate into the nucleus and bind to DNA to
`initiate transcription of the IL-2 responsive genes. In-
`volvement of Jak3 is crucial for transduction of signals
`mediated by rc because mutations of Jak3 result in
`severe immunodeficiency in patients [11,12] and mice
`[13,14] which mimics the immunodeficiency seen in
`mutations of the gc itself [15–18]. The immunodefi-
`ciency in the Jak3-deficient mice can be reversed by
`transfection of the hematopoietic [19] or embryonic [20]
`cells with functional, wild-type Jak3. Stat 5 appears
`also critical for activation of normal, postthymic T
`cells, because mature T cells derived from mice deficient
`in both a and b isoforms of STAT5 failed to yield
`proliferative response upon stimulation [21].
`Adult T-cell lymphoma leukemia (ATLL) is a malig-
`nancy affecting mature T lymphocytes. In most cases,
`the malignant ATLL cells display the CD3(cid:27), CD4(cid:27),
`CD8(cid:28), CD7(cid:28), and T-cell receptor (TCR) a:b((cid:27))
`phenotype. Characteristically, they express activation
`antigens such as HLA-DR and CD25 (IL2Ra chain)
`[22]. Numerous epidemiological and clinical studies
`have established the association of HTLV-I with ATLL
`as well as other diseases including tropical spastic para-
`paresis:HTLV-I associated myelopathy (TSP:HAM)
`[23]. Depending on the clinical course, the extent of the
`disease and the serum calcium level, ATLL can be
`divided into four clinical subtypes: acute, chronic,
`lymphomatous, and smoldering [24]. In all these vari-
`ants, patients have serum antibodies to HTLV-I and
`clonal integration of one or few copies of the virus in
`the DNA of the malignant cells [25,26]. Patients with
`detectable monoclonal or oligoclonal populations and
`elevated PBMC counts are at increased risk of develop-
`ing an overt ATLL disease [27]. However, the low
`frequency of ATLL (4–5%) among HTLV-I infected
`individuals [28] and the long average time interval
`between the occurrence of infection and the develop-
`ment of malignancy (20–30 years) indicate that addi-
`tional events are required for malignant transformation
`of T-cells.
`Experiments with HTLV-I transformed T-cell lines
`suggested that the virus may induce basal constitutive
`activation of the IL-2R associated Jak:STAT pathway,
`and that this pathway may be involved in HTLV-I-me-
`diated T-cell transformation [29,30]. However, we pre-
`viously
`found
`that
`constitutive
`activation
`of
`IL-2R-associated Jak:STAT signaling pathway also oc-
`curs in HTLV-I ((cid:28)) malignant cells from patients with
`cutaneous anaplastic large T-cell
`lymphoma (ALCL)
`[31]. This finding indicated that the constitutive activa-
`tion of IL-2R Jak:STAT pathway in transformed T
`
`cells may not be due to to the HTLV-I infection. To
`explore further the putative role of HTLV-I infection in
`the activation of this pathway, we examined several
`HTLV-I ((cid:27)) T-cell lines that differ in their IL-2 depen-
`dency for the basal and IL-2-induced phosphorylation
`of the Jak3, STAT3, and STAT5 proteins. Further-
`more, we also examined uncultured leukemic cells iso-
`lated directly from patients with ATLL. Our data
`demonstrate that the constitutive phosphorylation of
`the IL-2R associated Jak and STAT proteins is de-
`tectable only in the IL-2 independent HTLV-I ((cid:27)) cell
`lines. It is not seen in the IL-2 dependent HTLV-I ((cid:27))
`lines and the leukemic ATLL cells. Both these cell
`types, however, phosphorylate Jak3 and STAT5 in
`response to IL-2. Furthermore, most of the IL-2 inde-
`pendent cell lines, but none of the leukemic ATLL cells,
`lacked expression of SHP-1 which down-regulates
`phosphorylation of Jak3. Finally, the HTLV-I ((cid:27))
`lines, in contrast to the control, HTLV-I ((cid:28)) T-cell
`lines showed resistance to the IL-2R-signaling in-
`hibitors rapamycin and its novel analog RAD. Implica-
`tions of these findings for pathogenesis of the HTLV-I
`infection and ATLL are discussed.
`
`2. Materials and methods
`
`2.1. Patients
`
`A total of eight ATLL patients, two from the US and
`six from Japan, were tested. Peripheral blood mononu-
`clear cells (PBMC) were obtained by Ficoll centrifuga-
`tion [31,32]
`from two patients diagnosed at
`the
`University of Pennsylvania with ATLL based on clini-
`cal, histopathological, and immunophenotypic criteria.
`Both patients developed anti-HTLV-I antibodies as de-
`termined by Western Blot of serum proteins, peripheral
`white blood cell count greater than 50(cid:29)103:ul with a
`predominant lymphocytosis (greater than 50%), abnor-
`mal pathognomonic cells with multi-lobulated, flower-
`like nuclei on peripheral blood smear, serum LDH
`greater than 1.5 times upper limit of normal, and
`corrected Ca2(cid:27) of greater than 14 mg:dl. The above
`findings fulfilled criteria for the acute form of ATLL
`c
`[24]. Serum concentrations of soluble IL-2R were
`c
`59 520 U:ml (patient
`1) and 45 740 U:ml (patient
`2) (normalB1000 U:ml) indicating high tumor bur-
`dens ([33–35]). Flow cytometry analysis revealed that
`greater than 95% of PBMC from these patients had
`phenotypes
`consistent with ATLL cells
`(CD3(cid:27),
`CD4(cid:27), CD7(cid:28), CD25(cid:27), HLA DR(cid:27)). Fig. 1 shows
`c
`the cells with characteristic flower-like nuclei and
`salient flow cytometry data from patient
`1. Six
`Japanese patients were diagnosed with ATLL (five
`acute and one chronic form) at the Kagoshima Univer-
`sity using the same criteria as described above. Ficoll-
`
`Breckenridge Exhibit 1063
`Zhang
`Page 002
`
`

`

`Q. Zhang et al. :Leukemia Research 23 (1999) 373–384
`
`375
`
`isolated PBMC were cryopreserved in DMSO:FBS
`containing medium and thawed shortly before being
`used for experiments.
`
`Table 1
`HTLV-I status of the cell populations examined
`
`Cell types
`
`HTLV status
`
`Method of detec-
`tion
`
`PCR tax
`PCR tax
`PCR tax
`PCR tax
`PCR tax
`
`PCR tax
`Immunofluores-
`cence
`Immunofluores-
`cence
`
`PCR tax
`PCR tax
`PCR tax
`PCR tax
`Positive serology*
`
`(cid:28)(cid:28)
`
`(cid:28)
`(cid:28)
`(cid:28)
`
`(cid:27)
`(cid:27)
`
`(cid:27)
`
`(cid:27)(cid:27)
`
`(cid:27)
`(cid:27)
`(cid:27)
`
`Control cell lines
`YT (NK-like)
`Sez-4
`PB-1
`2A
`2B
`IL-2 dependent cell lines
`Boul
`Laf
`
`Cor
`
`IL-2 independent cell lines
`ATL-2
`C91PL
`C10MJ2
`HUT102
`ATLL Patients (8)
`
`* Cells were also histopathologically and immunophenotypically
`consistent with HTLV-I infected cells.
`
`Syndrome) and bears close morphological, pheno-
`typic, and genotypic resemblance to the fresh tumor
`cells [42]. The Sez4 line requires IL-2 (50–100 U:ml)
`for continuous proliferation. The YT line [43], a hu-
`man NK cell line, was kindly provided by J. Yodoi,
`Kyoto University, Kyoto, Japan. PB-1, 2A, and 2B
`T-cell
`lines which were established from a patient
`with a progressive cutaneous T-cell
`lymphoprolifera-
`tive disorder have been described in detail previously
`[31,44,45]. The PB-1 cell line was obtained at a rela-
`tively early stage of
`the patient’s cutaneous T-cell
`lymphoma from neoplastic T-cells circulating in pe-
`ripheral blood. The 2A and 2B lines were established
`at a later, more aggressive stage from two separate
`skin nodules, which represented a high-grade, T-cell
`anaplastic large-cell
`lymphoma. All five control cell
`lines were determined to be HTLV-I ((cid:28)) by PCR
`detection of the HTLV tax gene (Table 1). All the
`cell lines were propagated in a complete RPMI 1640
`medium containing 10% FBS (Hyclone, Logan, UT),
`1% L-glutamine
`(M.A. Bioproducts, Walkersville,
`MD), and 1% penicillin:streptomycin:fungizone mix-
`ture (M.A. Bioproducts).
`
`2.2. Cell lines
`
`lines were used:
`Two types of HTLV-I ((cid:27)) cell
`IL-2 independent
`cell
`lines derived mostly from
`ATLL patients and IL-2 dependent cell lines derived
`from nonleukemic, TSP:HAM patients. The HTLV-I
`((cid:27)), IL-2 independent cell lines were ATL-2: CD4(cid:27)
`CD8(cid:28) T cells originally cultured in IL-2 from
`PBMC of a patient with acute ATLL [36]; C91PL:
`cord blood T cell line established by co-culturing cord
`blood cells with known ATLL cells in the presence of
`IL-2 [37]; C1OMJ2: established from HTLV-I infected
`lymphocytes
`in a patient with ATLL [38]; and
`HUT102B: constitutive producer of HTLV-I derived
`from the lymph node of a patient with HTLV-I, also
`initially dependent on IL-2 [39,40]. HTLV-I ((cid:27)), IL-
`2 dependent cell
`lines, Boul, Laf, and Cor, were
`derived from HTLV-I
`((cid:27)) non-leukemic patients
`with TSP:HAM [41]. These cell lines required 50–100
`U:ml IL-2 for optimal growth and did not become
`IL-2 independent even after multiple passages. For
`controls, five HTLV-I ((cid:28)) cell lines were used. The
`Sez4 line, kindly provided by T. Abrams, Hahnemann
`University was derived from a patient with a
`leukemic phase of cutaneous T-cell lymphoma (Sezary
`
`c
`Fig. 1. Left panel: a representative peripheral blood smear from
`ATLL patient
`1 showing pathognomonic large cells with flower-
`like nuclei (Wright–Giemsa, magnification 1000(cid:29)). Right panel:
`flow cytometry analysis of the patients PBMC which shows that
`\95% cells exhibit a CD3(cid:27), CD4(cid:27), CD8(cid:28), CD7(cid:28), HLA-DR(cid:27)
`and CD25(cid:27) phenotype consistent with ATLL. Representative dot
`plots are shown where \95% of cells are CD4(cid:27):CD25(cid:27)and
`CD7(cid:28).
`
`2.3. Flow cytometry
`
`Flow cytometry immunophenotyping of PBMC was
`performed using a standard panel of T- and B-cell
`reactive mAbs including the ones which recognize ac-
`tivation antigens HLA-DR and CD25
`(Becton-
`Dickinson).
`
`Breckenridge Exhibit 1063
`Zhang
`Page 003
`
`

`

`376
`
`Q. Zhang et al. :Leukemia Research 23 (1999) 373–384
`
`2.4. IL-2, antibodies (Ab), and immunosuppressi6e
`agents
`
`Recombinant human IL-2 was kindly provided by C.
`Reynolds, NCI, Frederick, MD. Rabbit polyclonal Abs
`against JAK3, STAT3, STAT5, and SHP-1 were pur-
`chased from Santa Cruz Biotechnology (Santa Cruz,
`CA). Anti-phosphotyrosine 4G10 murine monoclonal
`Ab was purchased from UBI (Lake Placid, NY). Perox-
`idase-conjugated donkey anti-mouse and goat anti-rab-
`bit Abs were obtained from Jackson Jackson Immuno
`Research (West Grove, PA). Rapamycin and SZS RAD
`were kindly provided by, respectively, Wyeth-Ayerst
`(Princeton, NJ)
`and Novartis
`Pharma
`(Basel,
`Switzerland).
`
`2.5. PCR:Southern blot analysis
`
`A total of 1.0 ug of genomic DNA was added to 50
`ml of standard buffer containing 1.5 mM MgC12, 1.25
`mM dNTP mix, 15 pmol of 3% and 5% primers, and 2.5
`U Taq polymerase (Perkin-Elmer, Norwalk, CT). Reac-
`tion mixtures were amplified for 30 cycles of denatura-
`tion at 94°C for 30 s, annealing at 55°C for 30 s, and
`extension at 72°C for 30 s. Oligonucleotide primers for
`conserved sequences of the HTLV-I:II tax gene, SK43
`and SK44 [46], synthesized by Research Genetics,
`Huntsville, AL, were used for amplification of HTLV-I:
`II tax gene sequences. Amplification products were
`separated on 2% agarose gels, blotted and probed with
`32P-labeled SK45 probe [46] from Research Genetics as
`described before [47]. DNA from HTLV-II infected cell
`line (MoT) served as positive control.
`
`2.6. Protein expression and phosphorylation
`
`These assays were performed as described [31,48]. In
`brief, the cells (l0(cid:29)106) were washed, exposed for 5
`min to medium or 500 U IL-2, lysed for 20 min in l ml
`ice-cold lysis buffer (0.5% NP-40, 10 mM Tris–HC1
`(pH 7.4), 150 mM NaCl, 0.4 mM EDTA, 1 mM
`sodium orthovanadate, 0.5 mM PMSF, 10 mM NaF,
`and 3 ug:ml each of pepstatin, leupeptin, chymostatin,
`and aprotinin: Sigma). The lysates were centrifuged at
`15 000 rpm for 10 min. Next, the supernatants were
`precleared overnight at 4°C with protein A-sepharose
`(Sigma, St Louis, MO), incubated with the anti-Jak3,
`-STAT3, -STAT5, or SHP-1 Ab, and protein A-sep-
`harose, washed, boiled, suspended in reducing SDS
`loading buffer, separated on a 10% polyacrylamide-
`SDS gel, and transferred electrophoretically to hy-
`bridization transfer membranes. The membranes were
`blocked with 2% bovine serum albumin in TBST buffer
`(l0 mM Tris–HCl (pH 7.4), 75 mM NaCl,
`l mM
`EDTA, 0.1% Tween 20) for at least 2 h at room
`temperature or overnight in a cold room. To detect
`
`protein phosphorylation, the membranes were incu-
`bated with 4G10 Ab, washed, incubated with donkey
`anti-mouse, peroxidase-conjugated Ab and washed
`again. To detect protein expression the membranes
`were incubated with the same Jak3, -STAT3, -STAT5,
`or -SHP-1 Abs which were used for precipitation. Blots
`were developed using the ECL chemiluminescence
`reagents (Amersham Life Science, Arlington Heights,
`IL).
`
`2.7. Proliferation assays
`
`These tests were performed as described previously
`[31,32]. In brief,
`the cell
`lines or PHA-stimulated
`PBMC were cultured for either 10 or 18 h in triplicate
`at 2(cid:29)104 cells:well in the presence of various concen-
`trations of the immunosupressive drugs; rapamycin or
`RAD. After 14 h pulse with 0.5 mCi of [3H]thymidine,
`radioactivity of the cells was measured.
`
`3. Results
`
`3.1. Determination of the HTLV-I infection status
`
`The HTLV-I ((cid:27)) status of all cell populations used
`was determined by detection of viral tax gene in ge-
`nomic DNA, detection of viral gene products via im-
`munofluorescence, or in patients, by clinical, histopath-
`ologic and immunophenotypic criteria for ATLL in
`combination with serological evidence for HTLV-I in-
`fection (see Section 2.1 and Fig. 1). Table 1 summarizes
`the HTLV-I status of the cell populations examined.
`Controls used included cutaneous T-cell lymphoma cell
`lines: Sez4 (IL-2 dependent) and PB-1, 2A, and 2B (all
`IL-2 independent) and an NK cell line, YT. All control
`cell lines were shown to be HTLV-I ((cid:28)) by the PCR
`assay for the HTLV-I tax gene.
`
`3.2. Phosphorylation of IL-2R associated Jak:STAT
`proteins in IL-2 independent T-cell lines
`
`The few HTLV-I ((cid:27)), IL-2 independent T-cell lines
`tested to date have all been reported to display consti-
`tutive activation of the IL-2R associated Jak:STAT
`pathway [26,27]. We explored the extent of these find-
`ings by analyzing four HTLV-I ((cid:27)), IL-2 independent
`T-cell lines (ATL-2, C1OMJ2, C91PL, HUT 102), two
`of which (ATL-2, C1OMJ2) have not been examined to
`date. Fig. 2A and Table 2 show that all the cell lines
`demonstrate a strong basal, constitutive phosphoryla-
`tion of Jak3 with only the HUT102 cell line showing a
`slight augmentation in response to IL-2. STAT3 and
`STAT5 are also strongly, constitutively phosphorylated
`in the ATL-2 and C10MJ2 cell lines (Fig. 2B, C, and
`Table 2). Interestingly, constitutive phosphorylation of
`
`Breckenridge Exhibit 1063
`Zhang
`Page 004
`
`

`

`Q. Zhang et al. :Leukemia Research 23 (1999) 373–384
`
`377
`
`the entire Jak3:STAT3:STAT5 pathway does not seem
`to be a universal feature even in these cell lines because
`two of the lines (C91PL and HUT102B) did not exhibit
`any detectable, basal phosphorylation of STAT5 in
`repeated experiments although they expressed the
`protein and strongly phosphorylated STAT5 in re-
`sponse to IL-2 (Fig. 2C, Table 2). This finding implies
`a dissociation between Jak3 and STAT5 phosphoryla-
`
`tion and suggests that the signals transduced by Jak3
`may not necessarily always depend on phosphorylation
`of STAT5. In addition, one cell line, C91PL, showed a
`lack of STAT3 phosphorylation in the presence or
`absence of IL-2. This pattern of response is similar to
`that seen in resting PBMC rather than mitogen pre-ac-
`tivated T-cells [8,26,28]. The control Sez4 malignant
`T-cell line which is HTLV-I ((cid:28)) and, noteworthy, IL-2
`
`Fig. 2. Phosphorylation of proteins associated with IL-2R signal transduction pathway in HTLV-I positive, IL-2 independent cell lines (ATL-2,
`C9lPL, C1OMJ2, HUT102B) derived from leukemic patients without ((cid:28)) and with ((cid:27)) stimulation by IL-2: (A) Jak3, (B) STAT3, and (C)
`STAT5. The cell lysates were immunoprecipitated with the anti-Jak3, -STAT3, and -STAT5 Ab, electrophoretically separated, transferred to a
`membrane and probed with an Ab (4G10) which recognizes phosphorylated but not non-phosphorylated tyrosines. Loading of equal sample
`volumes was confirmed by the subsesquent immunoblotting with the Ab used for immunoprecipitation after removal of the bound 4G10 Ab.
`
`Breckenridge Exhibit 1063
`Zhang
`Page 005
`
`

`

`378
`
`Q. Zhang et al. :Leukemia Research 23 (1999) 373–384
`
`Table 2
`Tyrosine phosphorylation of Jak3, STAT3 and STAT5 in HTLV-I positive cell populations: IL-2-independent and dependent T-cell lines, and
`leukemic cells denved from ATLL patients
`
`IL-2 Independent
`(HTLV I(cid:27), leukemic)
`ATL-2
`C10MJ2
`C91PL
`HUT102B
`IL-2 Dependent
`(HTLV I(cid:27), non-leukemic)
`Boul
`Laf
`Cor
`c
`c
`ATLL patients
`c
`1–
`2
`c
`c
`3
`8
`4–
`Controls
`(HTLV I-)
`Sez-4
`YT (NK-like)
`PBMC†
`PHA blasts**
`
`(IL-2)*
`
`JAK3
`
`STAT3
`
`STAT5
`
`(cid:28)
`
`(cid:27)
`(cid:27)
`(cid:27)
`(cid:27)
`
`(cid:28)
`(cid:28)
`(cid:28):(cid:27)
`
`(cid:28)
`(cid:28)
`(cid:28)
`
`(cid:28)
`
`(cid:28)
`
`(cid:27)
`
`(cid:27)
`(cid:27)
`(cid:27)
`(cid:27)(cid:27)
`
`(cid:27)
`(cid:27)
`(cid:27)(cid:27)
`
`(cid:27)
`(cid:27)
`(cid:27)
`
`(cid:27)
`
`(cid:27)
`
`(cid:28)
`
`(cid:27)
`(cid:27)
`(cid:28)
`(cid:27)
`
`(cid:28):(cid:27)
`(cid:28):(cid:27)
`NDa
`
`(cid:28)
`ND
`ND
`
`(cid:27)
`
`(cid:28)
`
`(cid:27)
`
`(cid:27)
`(cid:27)(cid:27)
`(cid:28)
`(cid:27)
`
`(cid:27)
`(cid:27)
`(cid:27)
`
`(cid:28)
`ND
`ND
`
`(cid:27)
`
`(cid:27)
`
`(cid:28)
`
`(cid:27)
`(cid:27)
`(cid:28)
`(cid:28)
`
`(cid:28)
`(cid:28)
`(cid:28)
`
`(cid:28)
`(cid:28)
`ND
`
`(cid:28)
`
`(cid:28)
`
`(cid:27)(cid:27)
`
`(cid:27)(cid:27)
`(cid:27)
`(cid:27)
`(cid:27)
`
`(cid:27)
`(cid:27)
`(cid:27)
`
`(cid:27)
`(cid:27)
`ND
`
`(cid:27)
`
`(cid:27)
`
`a ND, not done.
`* Cells were exposed in vitro to 500 units of IL-2.
`** Normal PBMC tested after 5 days of stimulation with PHA.
`† PBMC from a healthy donor.
`
`dependent, also showed a lack of STAT3 phosphoryla-
`tion even after stimulation with IL-2. These observa-
`tions indicate that STAT3 activation may not be crucial
`for the transduction of IL-2R:Jak 3 mediated mitogenic
`signals.
`
`3.3. Phosphorylation of IL-2R associated Jak:STAT
`proteins in IL-2 dependent T-cell lines
`
`To determine if HTLV-I infection per se induces
`constitutive phosphorylation of the IL-2R associated
`Jak:STAT pathway, we analyzed several HTLV-I ((cid:27))
`cell lines which are IL-2 dependent. To ensure that any
`basal phosphorylation of Jak:STAT proteins was not
`due to the prior exposure to IL-2, these cells were
`incubated for 4–6 h without IL-2 before the assay. In
`contrast to the results obtained with the HTLV-I ((cid:27)),
`IL-2 independent cell lines, Jak3 was not constitutively
`phosphorylated in any of these lines (Boul, Laf, Cor)
`(Fig. 3A, Table 2). Only one of the lines, Cor, appears
`to have a minimal level of basal phosphorylation in the
`absence of IL-2. However, this phosphorylation de-
`creased to non-detectable levels when time of IL-2
`withdrawl was increased to 16 h (such prolonged IL-2
`removal did not affect the ability of cells to respond to
`IL-2; data not shown). These findings indicate that the
`
`three
`Jak3:STAT3:STAT5 phosphorylation in all
`HTLV-I ((cid:27)), IL-2 dependent T-cell lines was triggered
`by IL-2 and not HTLV-I infection. All cell lines phos-
`phorylated Jak3 in response to IL-2 (Fig. 3A, Table 2)
`which indicates that this pathway is functional and
`presumably required for the IL2 mediated proliferation
`of these cells. STAT3 appears to have, under the testing
`condition, a minimal level of basal constitutive phos-
`phorylation in two lines (Boul and Laf), which was
`markedly augmented by exogenous IL-2 (Fig. 3B, Table
`2). The control, HTLV-I negative, sez4 T-cell line failed
`to phosphorylate STAT3 upon IL-2 stimulation.
`No basal activation of STAT5 was noted in the three
`HTLV-I ((cid:27)) and the control HTLV-I ((cid:28)) Sez4 line,
`but all four lines phosphorylated strongly STAT5 after
`exposure to IL-2 (Fig. 3C, Table 2).
`
`3.4. Analysis of Jak3, STAT3, and STAT5
`phosphorylation in leukemic T cells deri6ed from ATLL
`patients
`
`Because HTLV-I ((cid:27)) cell lines examined thus far
`showed either constitutive or IL-2 inducible phosphory-
`lation of Jak3, STAT3, and:or STAT5, we were inter-
`ested in determining which of these two patterns is
`present in the uncultured, malignant T cells derived
`
`Breckenridge Exhibit 1063
`Zhang
`Page 006
`
`

`

`Q. Zhang et al. :Leukemia Research 23 (1999) 373–384
`
`379
`
`directly from ATLL patients. Results from eight ATLL
`c
`patients are summarized in Table 2 and the representa-
`tive data from patient
`1 (see Section 2.1 and Fig. 1)
`are shown in Fig. 4. There was no evidence for basal
`phosphorylation of either Jak3 (0:8 pts), STAT3 (0:2
`pts) or STAT5 (0:3 pts) in the absence of IL-2. How-
`ever, phosphorylation of Jak3 and STAT5 occured in
`response to IL-2 in all patients tested. These findings
`indicate that there is no basal constitutive activation of
`Jak3 and STAT5 in the leukemic ATLL cells tested but
`these cells are able to activate the IL-2R Jak:STAT
`pathway when exposed to IL-2.
`Interestingly, no
`STAT3 phosphorylation could be observed in response
`to IL-2 in the two patients tested. This pattern of
`response resembles the one seen in the normal resting
`PBMC,
`rather
`than in mitogen-preactivated T-cell
`blasts, which do activate STAT3 upon stimulation with
`IL-2 [8,29,31].
`
`Fig. 3. Analysis of Jak:STAT phosphorylation in HTLV-I positive
`IL-2 dependent cell lines (Boul, Cor, Laf) derived from non-leukemic
`patients before ((cid:28)) or after ((cid:27)) stimulation with IL-2: (A) Jak3, (B)
`STAT3, (C) STAT5. The cell immunoprecipitates obtained with the
`anti-Jak3, -STAT3, and -STAT5 Ab were electrophoretically sepa-
`rated, transferred to a membrane and probed with the 4G10 Ab.
`Loading of equal sample volumes was confirmed by reblotting with
`the Ab used for immunoprecipitation (data not presented).
`
`Fig. 4. Phosphorylation of the IL-2R-associated Jak:STAT pathway
`c
`in freshly isolated leukemic cells from a representative ATLL patient
`(
`1) before ((cid:27)) or after ((cid:28)) stimulation with IL-2. The membrane-
`immobilized cell immunoprecipitates were separated, transferred, and
`probed with the 4G10 Ab. Equal sample loading was confirmed by
`reblotting with the Ab used for immunoprecipitation (data not pre-
`sented).
`
`3.5. Analysis of SHP-1 expression in HTLV-I ((cid:27))
`T-cells
`
`It has been shown that a protein tyrosine phos-
`phatase, SHP-1 (also known as PTP1C, SHP, and
`SHPTP1), binds to several different cytokine receptors
`[6]
`including IL-2R [49]. SHP-1 appears to act by
`dephosphorylation of Jak kineses [50]. Noteworthy,
`dysfunction of SHP-1 as seen in a natural (‘motheaten’)
`SHP-1 knock-out mice, resuts in a hyperplasia of the
`erythroid and lymphoid lineages [51,52]. Furthermore,
`it has been found that some HTLV-I ((cid:27)) cell lines
`failed to express SHP-1 protein [49]. Because these
`HTLV-I ((cid:27)) T-cell lines displayed constitutive activa-
`tion of
`the
`IL-2R-associated Jak:STAT pathway
`[29,49], this finding suggested a casual relationship be-
`tween the lack of SHP-1 expression and the constitutive
`Jak:STAT activation. To explore further this apparent
`relationship, we examined SHP-1 expression in our
`panel of HTLV-I ((cid:27)), IL-2 independent T-cell lines
`which displayed constitutive phosphorylation of Jak3,
`STAT5, and:or STAT3. Furthermore, we examined
`SHP-1 expression in the previously uncultured leukemic
`cells from ATLL patients which, as described above,
`showed phosphorylation of Jak3 and STAT5 only after
`stimulation with IL-2. As shown in Fig. 5, three out of
`four HTLV-I ((cid:27)) T-cell lines showed lack of SHP-1
`protein expression. Only one line; HUT102B expressed
`SHP-1; this observation was made also by others [49].
`In contrast to the cell lines, uncultured leukemic cells
`from three ATLL patients expressed SHP-1. These
`findings provide additional evidence that there is an
`inversed correlation between the SHP-1 expression and
`constitutive IL-2R Jak:STAT activation and support
`the hypothesis that there indeed may be the cause-and-
`effect relationship between these two events.
`
`Breckenridge Exhibit 1063
`Zhang
`Page 007
`
`

`

`380
`
`Q. Zhang et al. :Leukemia Research 23 (1999) 373–384
`
`3.6. Effect of immunosuppressi6e drugs on proliferation
`of HTLV-I ((cid:27)) and HTLV-I ((cid:28)) T-cells
`
`Our previous study [31] has demonstrated that cell
`lines derived from HTLV-I ((cid:28)) cutaneous large T-cell
`lymphoma which display IL-2 independent growth and
`a constitutive activation of the IL-2R Jak:STAT signal-
`ing pathway, are sensitive to an immunosuppressive
`agent rapamycin. Rapamycin has been shown to act as
`an inhibitor of the IL-2R signaling by interfering with
`activation of one of
`the down-stream signaling
`molecules: p70 S6 kinase [53,54]. It was interesting,
`therefore, to test the effect of rapamycin on the HTLV-
`I ((cid:27)) T-cell lines which also show an IL-2 indepen-
`dence and constitutive activation of
`the
`IL-2R
`Jak:STAT pathway. In addition to rapamycin we used
`its novel analog RAD. As can be seen, both rapamycin
`(Fig. 6A) and RAD (Fig. 6B) profoundly inhibited
`growth of the HTLV-I ((cid:28)) T-cells. To exert their effect
`both drugs required a relatively long incubation period
`with the target cells. Whereas 24 h exposure resulted in
`the maximal
`inhibition of 40–80% for the various
`HTLV-I(cid:28)T-cell lines tested, 32 h exposure resulted in
`80–95% inhibition. In striking contrast, none of the
`three HTLV-I ((cid:27)) cell lines was markedly inhibited by
`either rapamycin or RAD (Fig. 6C and D, respectively)
`at even the highest doses after the 32 h exposure
`(0–30% inhibition). Interestingly, addition of RAD,
`particularly at the shorter exposure time, resulted in a
`mild (0–30%) augmentation rather than suppression of
`the proliferative rate. Taken together these data suggest
`that there are important differences in mitogenic signal-
`ing events downstream of IL-2R between the examined
`here HTLV-I ((cid:27)) and HTLV-I ((cid:28)) T-cells despite
`their similarity in constitutive activation of the IL-2R
`Jak:STAT pathway.
`
`Fig. 5. Expression of SHP-1 phosphatase in HTLV-I ((cid:27)) cell lines
`c
`(ATL-2, C91PL, C1OMJ2, HUT102), uncultured leukemic ATLL
`cells (pts
`1, 3, and 5), and control PBMC and PHA-prestimulated
`T-cell blasts (PHA bl). The cell lysates were electophoretically sepa-
`rated, transfered, and probed with an anti-SHP-1 antibody. Equal
`sample loading was confirmed by reblotting with an anti-Jak3 Ab
`(data not presented).
`
`4. Discussion
`
`Two previous reports have shown that HTLV-I ((cid:27))
`T-cell lines have constitutively activated IL-2R-associ-
`ated Jak:STAT signal transduction pathway [29,30].
`This finding suggested that HTLV-I infection may acti-
`vate this signaling pathway. In addition, they implied
`that activation of the IL-2R-associated Jak:STAT path-
`way may be critical in the pathogenesis of ATLL. This
`conclusion was further supported by a recent report
`[55] which described a basal, apparently constitutive
`activation of this pathway in uncultured leukemic cells
`from eight out of twelve ATLL patients. However, our
`data demonstrate clearly that HTLV-I infection per se
`does not result in a constitutive activation of the IL-2R-
`associated Jak:STAT pathway. Although all HTLV-I
`((cid:27)), IL-2 independent cell lines tested so far indeed
`displayed a strong, constitutive phosphorylation of
`Jak3, STAT3, and:or STAT5 (29, 31, Fig. 2, and Table
`2), the HTLV-I ((cid:27)) T-cell lines which require IL-2 for
`their growth, showed phosphorylation of these proteins
`only after stimulation with IL-2 (Fig. 3 and Table 2).
`Withdrawal of IL-2 invariably resulted in the profound
`decrease in the Jak3:STAT3:STAT5 phosphorylation,
`usually to the undetectable levels (Fig. 3, Table 1, data
`not presented). Furthermore,
`in our experiments the
`IL-2R associated Jak:STAT pathway was also quies-
`cent in uncultured leukemic cells isolated directly from
`all eight ATLL patients studied. This pathway was,
`however, fully functional
`in such leukemic cells as
`demonstrated by strong phosphorylation of Jak3 and
`STAT5 upon stimulation of the cells with IL-2. In
`addition to providing additional evidence that HTLV-I
`infection per se does not result in a basal constitutive
`activation of the IL-2R associated Jak:STAT signaling
`pathway, these data indicate that constitutive activation
`of this pathway may not be required for malignant
`T-cell transformation in at least some cases of ATLL.
`Because the leukemic ATLL cells responded to IL-2 in
`virtually all cases, it is likely that this cytokine plays a
`role in the pathogenesis of ATLL. Previous studies
`(reviewed in [56]) have shown that malignant ATLL
`cells derived from lymph nodes rather than peripheral
`blood, tend not only to respond but also produce IL-2.
`This IL-2 production which was not seen in the circu-
`lating
`leukemic
`cells,
`requires
`the presence of
`macrophages. These findings, combined with our data,
`suggest