American journal of Pathology,
`Vol. 144. No. 5, Maci 1994
`Copyright © American Society for Investigative Pathology
`
`Constitutive Secretion of Soluble Interleukin-2
`Receptor by Human T Cell Lymphoma
`Xenografted into SCID Mice
`
`Correlation of Tumor Volume with Concentration of
`Tumor-Derived Soluble Interleukin-2 Receptor in
`Body Fluids of the Host Mice
`
`Mariusz A. Wasik,** Nicholas Sioutos,*1
`Melissa Tuttle,* Janet R. Butmarc,*
`William D. Kaplan,tt and Marshall E. Kadin*t
`From the Department of Pathology,* Beth Israel Hospital,
`Boston; Oncologic Nuclear Medicine.t Dana Farber Cancer
`Institute, Boston; and Harvard Medical School,t
`Boston, Massachusetts
`
`Increased serum concentration ofsoluble a-chain
`receptorfor interleukin-2 (sIL-2R) has been noted
`in patients with a variety of inflammatory condi-
`tions and lymphoid malignancies including T ceUl
`leukemia and lymphoma. Elevated sIL-2R serum
`levels seen in lymphoid malignancies appear to
`correlate with the clinical stage of disease. How-
`ever, because sIL-2R is produced by normal ac-
`it has been uncertain
`tivated lymphocytes,
`whether serum sIL-2R in such conditions is de-
`rivedfrom tumor ceUs or normal immune cells re-
`sponding to the tumor. To address this question,
`we used a model of human (CD30+) anaplastic,
`large TceU lymphoma transplanted into immuno-
`deficient SCID mice. Reverse transcriptionpolym-
`erase chain reaction of tumor RNA showed that
`the tumor, designated mJB6, contains mRNA for
`a-chain of human IL-2R. Furthermore, 15 to 25%
`of tumor cells stained with anti-human IL-2R
`a-chain mAb. Solidphase ELISA analysis ofserum
`samples from mice bearing mJB6 lymphoma
`showed high concentrations of human sIL-2R.
`None of the control mice without lymphoma or
`with human nonlymphoid tumors (prostatic car-
`cinoma, ovarian carcinoma, and glioblastoma
`multiforme) showed detectable human sIL-2R.
`The sIL-2R serum titers ofmJB6-bearing mice cor-
`
`related strongly with tumor volume (P < 0.0001).
`Tumors as small as 0.4 to 0.8 mm3 could be de-
`tected by this method. The sensitivity of sIL-2R
`ELISA exceeded at least 150 times the sensitivity
`of conventional radioisotopic tumor detection.
`Total resection of mJB6 tumors resulted in com-
`plete clearance of sIL-2Rfrom the murine serum
`within 48 hours with a haf-life of 6 hours. Ac-
`cordingly, partial resection led to a significant de-
`crease in sIL-2R followed by gradual increase
`with tumor regrowth. sIL-2R was also detected in
`the urine ofmJB6-transplanted mice. As in serum,
`urine concentrations ofsIL-2R wereproportional
`to tumor mass (P < 0.02). Based on thesefindings
`we postulate that malignant cells are a major
`source ofserum sIL-2R in patients with lymphoid
`tumors. In addition, our data further support
`monitoring sIL-2R concentration in bodyfluids as
`a sensitive method to detect change in tumor vol-
`ume in such patients.
`(Am J Pathol 1994,
`144:1089-1097)
`
`The a-chain of interleukin-2 receptor (IL-2R) is ex-
`pressed on the surface of T and B lymphocytes in two
`different forms.1 When expressed alone, a-chain
`binds IL-2 protein with low affinity. Combined with 3-
`and, as recently reported,2 y-chain, a-chain forms a
`high affinity IL-2R capable of transducing a cell-
`activating signal.3'4 No IL-2R a-chain is detected on
`
`Supported by Council for Tobacco Research grant 2630 (MEK).
`Accepted for publication January 6, 1994.
`Address reprint requests to Dr. Mariusz Wasik, Department of
`Pathology and Laboratory Medicine, Hospital of the University of
`Pennsylvania, 3400 Spruce Street, Philadelphia, PA 19104.
`
`1089
`
`NOVARTIS EXHIBIT 2112
`West-Ward v. Novartis, IPR 2017-01592
`Page 1 of 9
`
`

`

`Wasik et al
`1090
`AJP May 1994, Vol. 144, No. 5
`
`resting T lymphocytes, whereas cell activation results
`in expression of the a-chain at high concentration.
`Similarly to other cytokine receptors, the extracellular
`fragment of IL-2R a-chain is secreted in a soluble
`form.5-1 1 Although the function of slL-2R remains un-
`certain,12 this receptor can be detected at low con-
`centrations in blood and urine of normal individu-
`als.12'13 Elevated serum concentrations of slL-2R
`have been detected in patients with a broad range of
`diseases such as leukemias14-16 and lympho-
`mas,17'18 certain autoimmune disorders, viral infec-
`tions including AIDS, and during rejection of trans-
`planted organs.12'13 In lymphoid malignancies slL-2R
`serum concentrations appear to correlate with extent
`or stage of disease.14-18 It remains uncertain, how-
`ever, whether elevated slL-2R levels result from se-
`cretion of the receptor by malignant cells or by nor-
`mal, activated lymphocytes responding to the tumor,
`because in vitro studies show that both the malignant
`cells and normal activated lymphocytes can secrete
`slL-2R.5'19 To address this question, we used a model
`of human anaplastic T cell lymphoma xenografted
`into SCID mice. Results of our study show that the
`lymphoma cells constitutively release slL-2R in this in
`vivo model. Furthermore, there is a direct correlation
`between tumor volume and slL-2R levels in blood and
`urine of the host mice.
`
`Materials and Methods
`Origin of the mJB6 Tumor
`The JB6 cell line was obtained by in vitro culture of
`peripheral blood leukemic cells from a 12-year-old
`boy with advanced anaplastic large cell lymphoma.20
`The T cell origin of the tumor was established based
`on the immunohistochemical (positivity for CD2, CD7,
`and Bf1 antigens) and molecular (clonal rearrange-
`ment of 1-chain of T cell receptor genes) findings.
`After several months in culture, 108 JB6 cells were
`injected intraperitoneally into SCID mice. One mouse
`developed a widespread tumor. Fragments of mouse
`abdominal tumor mass (mJB6) were inoculated sub-
`cutaneously into several other SCID mice and main-
`tained as described below.
`
`mJB6 Tumor-Bearing Mice
`C.B-17 scid/scid mice used in these experiments
`were kindly provided by Dr. M.J. Bosma, Fox Chase
`Cancer Center, Philadelphia, PA, and were kept in the
`animal facility at Joint Center for Radiation Therapy,
`New England Deaconess Hospital, Boston, MA.
`
`mJB6 Tumor Transplantation
`Passage of the mJB6 tumor was performed under
`anesthesia (ketamine/xylazine mixture) by subcuta-
`neous injection in the right chest wall of 6 to 7 small
`(0.1 cm in the largest dimension) tumor tissue frag-
`ments, unless stated otherwise (see Results). Mice
`implanted with the mJB6 tumor were maintained up to
`6 weeks until tumor growth reached approximately 2
`cm in the largest dimension. Even the largest tumors
`did not have a visible adverse effect on the host mice.
`
`Measurement of the Tumor Volume
`Volume of the tumor was determined from the equa-
`tion, volume = 0.4 x ab2, where a and b designate
`long and short tumor diameter, respectively.
`
`Collection of Blood and Urine
`Blood was collected under anesthesia from the orbital
`sinus of the right eye or by terminal cardiac puncture.
`Urine was collected by suprapubic urinary bladder
`puncture after terminal bleeding. Control serum and
`urine from athymic nude (NIH-nu/nu) mice bearing
`human ovarian carcinoma (2008), prostatic carci-
`noma (PC3), and glioblastoma multiforme (T98G)
`were kindly provided by Dr. B. Teicher, Joint Center for
`Radiation Therapy, Dana Farber Cancer Institute,
`Boston, MA. Serum, heparinized plasma, and urine
`samples were stored at -80 C for up to 4 months
`without detectable loss of activity.
`
`Reverse Transcription-Polymerase Chain
`Reaction
`The reaction was performed as described.21'22
`Briefly, cDNA was obtained from total RNA by incu-
`bation with murine Maloney leukemia virus reverse
`(GIBCO-BRL,
`MD),
`Gaithersburg,
`transcriptase
`oligo(dT), deoxynucleoside triphosphates, bovine se-
`rum albumin, DTT, and RNAsin. Aliquots of synthe-
`sized cDNA were amplified for 30 cycles in DNA ther-
`mal cycler (Perkin-Elmer Cetus, Norwalk, CT) in the
`presence of Taq polymerase (Perkin-Elmer Cetus),
`deoxynucleoside triphosphates, and primer pairs
`specific for human IL-2R a-chain (CD25), murine
`IL-2R a-chain, murine 1-actin (Clontech, Palo Alto,
`CA), or human 13-actin (synthesized on a Cyclone
`DNA synthesizer, Bioresearch, San Rafael, CA) and
`purified using NENSORB PREP columns (DuPont Co.,
`Wilmington, DE) as described.22 Polymerase chain
`reaction-amplified DNA was visualized in ultraviolet
`
`NOVARTIS EXHIBIT 2112
`West-Ward v. Novartis, IPR 2017-01592
`Page 2 of 9
`
`

`

`1091
`Secretion of sIL-2R by Human T Cell Lymphoma Xenografted into SCID Mice
`AJP May 1994, Vol. 144, No. 5
`
`light after electrophoresis on 2% agarose gel con-
`taining ethidium bromide. The product was identified
`by a predicted molecular weight of the amplified DNA
`fragment. Its identity was confirmed by restriction
`mapping. Samples in which no reverse transcriptase
`was added served as controls.
`
`Immunohistochemical Analysis of the
`mJB6 Tumor
`
`Immunohistochemical staining of the mJB6 tumor
`was performed as previously described.2324 In brief,
`4-p sections of frozen, periodate-lysine-paraformal-
`dehyde-prefixed frozen tissue were cut and fixed ad-
`ditionally for 5 minutes in cold acetone. Immuno-
`localization was performed with anti-human IL-2R
`a-chain (CD25, Tac, Becton Dickinson, San Jose,
`CA) and CD30 (BerH2, Dako Corporation, Carpen-
`teria, CA) monoclonal antibodies using a three-step
`immunoperoxidase method. Slides were developed
`tetrahydrochloride.
`using
`3,3'-diaminobenzidine
`Staining was enhanced with osmium tetroxide and
`slides were counterstained with methyl green.
`
`ELISA for Human slL-2R
`The slL-2R ELISA plates were obtained from T cell
`Diagnostics, Cambridge, MA, and the assay was per-
`formed as recommended by the manufacturer as pre-
`viously described.25 In brief, 50 pl of standards, se-
`rum, or urine diluted 1:5 to 1:40 were applied in
`duplicate into wells precoated with anti-human IL-2R
`monoclonal antibody. Next, 50 p1 of a monoclonal an-
`tibody against a different IL-2R epitope conjugated
`with horseradish peroxidase were added and the
`plates were incubated for 3 hours at room tempera-
`ture on a rotator set at 150 rpm. After washing of the
`wells, 100 p1 of the O-phenylenediamine solution was
`applied into each well. Plates were incubated at room
`temperature for 30 minutes and then 50 pl of 2 N
`H2SO4 were added to inhibit the enzymatic reaction.
`The optical density value was determined using an
`automated ELISA reader (Molecular Devices Corp.,
`Menlo Park, CA) and data were processed and plot-
`ted using Softmax (Molecular Devices Corp.) and Mi-
`crosoft Excel (Microsoft Corp., Redmont, WA) soft-
`ware. The data are shown as a mean of the duplicate
`samples (SD was <10%).
`
`Surgical Removal of the Tumor
`The implanted mJB6 tumors were either partially or
`totally resected in a laminar flow hood under sterile
`
`conditions. The dermal wounds were closed with ster-
`ile surgical staples that, in case of partial resection,
`were removed after 7 days.
`
`Radioisotopic Detection of the Tumor
`mJB6 tumors were implanted under the skin over the
`right scapula and allowed to grow to approximately 1
`to 2 cm in diameter. Tumor-bearing mice were then
`injected into the tail vein with 250 pCi of Ga-67 as the
`citrate. Nine-minute posterior whole body images
`(approximately 8 x 104 counts each) were obtained
`at 72 hours after injection using a pinhole collimator
`(4-mm aperture) fitted to a small field of view gamma
`camera (Siemens LEM, Hoffman Estates, IL). Scinti-
`graphic data were collected on a PCS 512 computer
`(Picker International Inc., Highland Heights, OH) in a
`128 x 128 matrix for further evaluation.
`
`Statistical Analysis of the Results
`The data were analyzed using the Prophet system, a
`national computer resource sponsored by the Divi-
`sion of Research Resources, NIH. The correlation be-
`tween tumor volume and slL-2R concentration in se-
`rum and urine was evaluated using the Pearson's
`product moment (simple) correlation coefficient. Dif-
`ferences between serum and urine sIL-2R concen-
`tration were analyzed using slope and intercept co-
`variance analysis.
`
`Results
`Expression of mRNA for IL-2R by mJB6
`Tumor
`To examine whether the mJB6 tumor cells produced
`a message for IL-2R a-chain, we performed a reverse
`transcription-polymerase chain reaction test with
`primers specific for the human a-chain. Three addi-
`tional sets of primers directed against human ,B-actin
`and murine IL-2R a-chain and ,B-actin served as con-
`trols. As shown in Figure 1, total RNA isolated from the
`mJB6 cells contained transcripts for human IL-2R
`a-chain and 13-actin but no detectable message for
`their murine counterparts. In contrast, murine spleen
`contained no transcripts recognizable by both human
`oligoprobes and contained mRNA for murine IL-2R
`a-chain and f-actin.
`
`Detection of IL-2R a-chain Protein in
`mJB6 Tumor Tissue
`To determine whether mJB6 tumor cells synthesize
`IL-2R a-chain protein, we performed an immunocy-
`
`NOVARTIS EXHIBIT 2112
`West-Ward v. Novartis, IPR 2017-01592
`Page 3 of 9
`
`

`

`Wasik et al
`1092
`AJP Mav 1994, Vol. 144, No. 5
`
`......................
`
`Figure 2. Immunohistochemical detection of IL-2R a-chain protein in mJB6 tuimor tissue. Sections of periodate-lysine-paraformaldehvde-prefixed
`frozen mJB6 tuimor tissue were fixed in cold acetone. Antigent immunolocalizationi was performed by thrcee-step immunoperoxidase method. The
`staining was enhanced with osmium tetroxide and methyl green was used as a counterstain. A: Morphologic appearance of mJB6 tumor (hema-
`toxylin and eosin stain). B: Immunohistochemical staining uwith anti-hulman IL-2R a-chain (CD25) monoclonal antibody. C: staining with anti-
`human CD.30 (BevH2) monoclonal antibody.
`
`,\s\
`yV*. \v
`
`Iq .. \v
`10,
`\.\ O.
`'J'q.
`IVW. \V y
`I.r,.. \VI-,
`.-o
`'6.\
`qp
`WI
`
`-700
`-540
`
`II
`
`mJB6 Tumor
`
`Murine Spleen
`
`tually all tumor cells were positive for the CD30 (Ki-1)
`antigen, which is characteristic of anaplastic large
`cell lymphoma (panel C).
`
`Production of sIL-2R by mJB6 Tumor
`To address the question of whether mJB6 tumor pro-
`duces a soluble form of IL-2R a-chain, we tested se-
`rum samples from mice with two different mJB6 tumor
`
`Figure 1. Expression of mRNA for a-chain of
`huomatn I-2R by mJB6 tutmor. Total RNA de-
`rived from mJB6 tumor or spleetn of the tumor 661-
`host SCID inouse was reversibly transcribed
`and the cDNA uwas PCR amplified in the pres-
`ence ofprimer pairs specificforfragment of hu-
`mlan f-actinl (predicted mnolecular weight 661 398-
`bp), huomani IL-2R a-chain (398 bp), mutrine
`f3-actin (540 bp), or murine IL-2R a-chain
`( 700 bp). The amplified DNA was vistualized in
`uiltraviolet light after electrophore.sis on 2% aga-
`rose gel containing ethidium bromide. Identity
`wa.'aS confirmfed bv restriction
`t)rorlaIyrt
`,,f thz?
`mapping of the amplified fragment (data not
`shown). Tuimor-derived cDNA is in lanes I to
`4, spleen-derived DVA in lanes 5 to 8.
`
`tochemical analysis of tumor tissue sections using
`murine monoclonal antibody specific for human IL-2R
`a-chain (CD25). Figure 2, in addition to showing the
`histological appearance of the mJB6 tumor (panel A),
`revealed strong staining of the tumor cells in a char-
`acteristic membranous pattern (panel B). It is note-
`worthy that only a subpopulation of cells, approxi-
`mately 15 to 25% depending on the microscopic field
`examined, expressed IL-2R a-chain. In contrast, vir-
`
`NOVARTIS EXHIBIT 2112
`West-Ward v. Novartis, IPR 2017-01592
`Page 4 of 9
`
`

`

`0_.I
`
`volumes in a solid phase ELISA specific for human
`slL-2R. As controls, we used serum samples from
`mice with no tumor or with three different types of
`human nonlymphoid tumors. As shown in Figure 3,
`only sera from mJB6-bearing mice contained human
`slL-2R at high concentrations. In contrast, sera from
`mice with no tumor or human prostatic carcinoma,
`ovarian carcinoma, or glioblastoma, all of which had
`a volume comparable to mJB6 tumors, contained no
`detectable human slL-2R. Importantly, the slL-2R
`concentration in mJB6-bearing mice appeared to cor-
`relate directly with tumor volume because the larger
`tumor (3.36 cm3) yielded higher slL-2R serum titer
`(58,683 U/ml) than the smaller one (0.38 cm3 and
`27,548 U/ml, respectively).
`
`Correlation Between mJB6 Tumor Volume
`and slL-2R Serum Concentration
`To confirm the apparent relationship between the tu-
`mor volume and serum level of slL-2R, we tested a
`number of blood samples from five mice with mJB6
`tumors at various stages of tumor growth. The results
`(Figure 4) show a strong direct correlation between
`the tumor volume and slL-2R concentration (P <
`0.0001).
`
`Effect of Surgical Resection of mJB6
`Tumor on slL-2R Serum Concentration
`To further investigate the relationship between the tu-
`mor growth and slL-2R in serum, we performed partial
`and total resection of the several mJB6 tumors (Figure
`5). Total resection of three small tumors (volume 0.18
`to 0.23 cm3) resulted in a rapid decrease in circulating
`slL-2R. A 50% decrease occurred within 6 hours after
`
`Secretion of slL-2R by Human T Cell Lymphoma Xenografted into SCID Mice
`1093
`AJP May 1994, Vol. 144, No. 5
`
`90.000
`
`80,000
`
`70,000
`
`60,000
`
`50,000
`
`40,000
`
`30,000
`
`20,000
`
`10,000
`
`0-
`
`EB
`
`m N
`
`2
`
`2.5
`
`0
`
`0.5
`
`1.5
`1
`TUMOR VOLUME (cm3)
`Figure 4. Correlation between mlB6 tminor volume antd silL-2R con-
`centrationI in serum. Sernum .samples from five nmice bearing mjB6 tol
`mor of' dfferenit volumes u'erce collectced from each mouse on three
`separate occasions at 7- to 10-day intervals. Thrcee tnice (labeled +,
`*. and A) had small tumors at the time of'fi'rst bleedinig, the remnain-
`inig tuwo (O and *) had larger tumors. Concentratiotn of tumor-
`derived sIL-2R in serum of the mice uwas determined by ELISA specific
`for hmnan sIL-2R.
`
`resection with complete serum clearance by 48
`hours. Similarly, a marked decrease in circulating
`sIL-2R was seen 24 hours after partial resection of two
`relatively large tumors (0.63 and 1.04 cm3). Regrowth
`of these partially resected tumors resulted in a con-
`tinuous increase in slL-2R serum concentration.
`
`Sensitivity of slL-2R ELISA in Detection of
`mJB6 Tumor
`Next, we tested the limit of sensitivity of tumor detec-
`tion by measurement of serum slL-2R. For this pur-
`pose mice were injected subcutaneously with a single
`fragment of tumor rather than with 6 to 7 fragments as
`in the previous experiments (see Materials and Meth-
`ods). We reasoned that such an approach would per-
`mit better vascularization of the small tumors. Inad-
`equate angiogenesis is not only a factor limiting the
`tumor growth26,27 but in this case it might also de-
`crease serum slL-2R concentration in the smallest
`palpable tumors. As shown in Figure 6, tumors as
`small as 0.80 mm3 were easily detectable by slL-2R
`ELISA method (386 U/ml). Even a smaller tumor (0.40
`mm3) could be detected yielding 193 U/mI of slL-2R
`compared with 0 U/ml for the serum from a mouse with
`no palpable tumor. This, however, appears to be the
`limit of sensitivity for this system because in some
`experiments a nonspecific background of up to 100
`U/ml for serum from mice with no injected tumor was
`noted (data not shown). The sensitivity of slL-2R
`ELISA for detection of mJB6 compared very favorably
`with a radioisotopic method that is routinely used for
`
`10.016
`
`- 40,000
`
`ia
`
`20,000
`
`10,0066
`
`HeTiu
`
`Wm-Tsn
`H 46.39)
`
`S-T?7
`(3.86
`
`---I
`onstmle
`(a@)(6.
`TYPE OF TUMOR (volum: mS)
`
`0,,W
`
`SI
`
`NISU
`
`Figure 3. Production of sIL-2R a-chain by rnJB6 tumor. Concentra-
`tion of tumor-derived sIL-2R in serum of the host mice u'as deter-
`mined by solid phase ELISA using plates prcecoated with anti-hutman
`sIL-2R monoclonal antibody and horseradish peroxidase conjugated
`to the second monoclonal antibody against a different epitope of hul-
`man slL-2R. Mice with no tumor or xenografted with human pros-
`tatic carcinoma, ovarian carcinoma, and glioblastoma multiforme
`senred as controls.
`
`NOVARTIS EXHIBIT 2112
`West-Ward v. Novartis, IPR 2017-01592
`Page 5 of 9
`
`

`

`Wasik et al
`1094
`AJP May 1994, Vol. 144, No. 5
`
`20,000 -_
`
`18,000 -_
`
`16,000
`
`14,000
`
`E 12,000
`;C 10,000
`
`0
`
`4
`
`8
`
`1 2
`
`1 6
`
`20
`
`28
`24
`TIME (hours)
`
`32
`
`36
`
`40
`
`44
`
`48
`
`8,000
`
`6,000
`
`4,000
`
`2,000
`
`0
`
`30,000
`
`25,000
`
`20,000
`
`1 5.000
`
`E-c
`
`c
`N~
`
`A
`
`.0 Ann
`
`1,6000
`
`14,000
`
`12,000
`
`a. 10.000
`
`E o
`
`0,000
`
`6,000
`
`4,000
`
`2,000
`
`a
`u
`
`106.00
`
`176.40
`
`51.20
`
`4.80
`32.00
`TUMOR VOLUI (mmp)
`Figure 6. Sensitivity of sIL-2R ELISA in detection of mJB6 tumor. Se-
`rum samples from eight mice were collected 5 to 10 days after trans-
`plantation ofthe tumor and testedfor concentration of human sIL-2R
`by ELISA.
`
`0.80
`
`0.40
`
`0.00
`
`100.000
`
`140,000
`
`A
`
`Iblme
`
`(')
`* :,
`
`.IU_1W
`11.311
`
`_J&T4I3
`PamXms
`P.1)1
`imm
`(3S.
`-TUWORVMUtmk.: erns)
`n
`
`13.023
`
`gbMmIsi
`
`cas-e)
`
`B
`
`120,000
`
`E 100,000
`3
`
`v o,no
`
`40,@00
`a.
`
`20,000
`
`0
`
`169.000v
`tom adba _
`
`Tf
`
`140,000
`
`128,001
`
`a,
`
`00,6000
`
`6@0.000
`
`sale"0@
`
`40,000
`
`30.000
`
`_
`
`0.05
`
`0.6
`
`60.63
`TUMR VOLUME (M3)
`
`1.22
`
`4.60
`
`Figure 7. Detection1 of sIL-2R in urine of mJB6-bearing mice in re-
`gard to tumor type (A) anid volume (B). Concentration of human
`sIL-2R in urine of noJB6 tumor-bearing mice and control mice u'as
`determined by ELISA.
`
`lymphoid human tumors, contained high levels of sIL-
`2R. Furthermore, urine sIL-2R concentration corre-
`lated directly with mJB6 tumor volume (P < 0.02;
`Figure 7B). Interestingly, concentrations of sIL-2R in
`urine appeared markedly higher than slL-2R in serum
`for comparable tumor volumes (Figures 3, 4, and 7).
`
`10,000
`
`5,000
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`0
`
`1
`
`2
`
`3
`
`4
`
`5
`
`8
`6
`7
`TIME (days)
`Figure 5. Effect of total (A) and partial (B ) resection of m,JB6 tumor
`on serum concentration of sLL-2R. 7Three mice underwent a total and
`two mice a partial mJB6 tumor resection. Serum concentration of tu-
`mor sIL-2R was established by ELISA. A: Five serum samples were col-
`lectedfrom each mouse (0, A, and *) at 4- to 24-hour intervals. B:
`Three serum samples from one mouse (0) andfour samples from the
`other mouse (A) were obtained at 1-, 6-, and 7-day intervals.
`
`detection of sites of involvement by lymphoma in pa-
`tients. Indeed, only large tumors (2.16 and 2.25 cm3)
`could be visualized after intravenous injection of ra-
`dioactive gallium, whereas smaller tumors (115 and
`144 mm3) were not visualized (data not shown). Thus,
`it appears that slL-2R ELISA was at least 150 to 200
`times more sensitive (see Figure 6) than conventional
`isotopic method in this experimental system.
`
`Detection of sIL-2R in Urine of
`mJB6-Bearing Mice
`
`It has been shown that slL-2R and also soluble re-
`ceptors for some other cytokines can be detected in
`human urine in both health and disease.12,13 It was
`therefore of interest to determine whether slL-2R pro-
`duced by mJB6 tumor could be detected in the urine
`of host mice. Figure 7A shows that this indeed was the
`case. In similarity to serum data, only urine samples
`from mice bearing mJB6 but not samples from mice
`without tumor, or from mice with three different non-
`
`NOVARTIS EXHIBIT 2112
`West-Ward v. Novartis, IPR 2017-01592
`Page 6 of 9
`
`

`

`Secretion of slL-2R by Human T Cell Lymphoma Xenografted into SCID Mice
`1095
`AJP May 1994, Vol. 144, No. 5
`
`Discussion
`
`This study shows that human T cell lymphoma xe-
`nografted into SCID mice constitutively secretes sIL-
`2R. The tumor-derived slL-2R was detected in serum
`and urine of host mice. Concentration of the slL-2R in
`these body fluids strongly correlated with the tumor
`volume. Tumors as small as 0.4 to 0.8 mm3 were de-
`tected by this method. The half-life of slL-2R in the
`serum was approximately 6 hours, as determined by
`total resection of xenografted tumors. It may be ar-
`gued that normal, reactive tumor-infiltrating lympho-
`cytes derived either from the host mice or carried over
`from the patient with tumor cells could have contrib-
`uted to the slL-2R concentration in serum of the host
`mice. However, several lines of evidence indicate that
`the tumor cells are the sole source of slL-2R detected
`in our experimental model. The arguments against the
`contribution to slL-2R levels by murine reactive lym-
`phocytes are as follows. First, the ELISA system used
`by us employs two different murine monoclonal anti-
`human slL-2R antibodies that do not cross-react with
`murine slL-2R. Accordingly, no slL-2R was detected
`in control SCID mice without the tumor and athymic
`mice bearing human nonlymphoid tumors (ovarian
`and prostatic carcinomas and malignant astrocy-
`toma). Second, no mRNA for murine slL-2R was found
`within the tumor (Figure 1). Third, morphological
`evaluation showed that the tumor tissue was com-
`posed almost entirely of enlarged, atypical malignant
`cells with no component consistent with reactive cells
`(Figure 2A). This impression was further strengthened
`by immunohistochemical study showing that all the
`cells with exception of blood vessel lining cells
`strongly stained with murine anti-human CD30 mono-
`clonal antibody (Figure 2C). Fourth, no murine T or B
`cell tumor-infiltrating lymphocytes should by present
`in SCID mice, which limits significantly potential en-
`dogenous sources of slL-2R in these mice. However,
`SCID mice are able to produce some slL-2R (Figure
`2) due to the presence of nonlymphoid sources, eg,
`macrophages, and sometimes also due to the "leak-
`age" of the SCID phenotype.
`Some of the above findings (morphological and im-
`munohistochemical evaluation and the lack of detect-
`able slL-2R in mice with three different human non-
`lymphoid tumors) argue against the presence of
`human, patient-derived tumor-reactive lymphocytes
`in our experimental model. Furthermore, additional
`immunohistochemical studies with several mono-
`clonal antibodies against human antigens CD3, 4, 8,
`and 20 were negative demonstrating lack of normal T
`and B lymphocytes. Finally, we find it highly unlikely
`that human tumor-reactive lymphocytes would sur-
`
`vive multiple in vitro and in vivo passages without ei-
`ther overcoming the tumor or being eliminated by the
`host mice.
`Synthesis of soluble receptors for cytokines is a
`well-documented phenomenon. 12 The soluble forms
`are produced either by proteolytic cleavage of the
`membrane-bound receptor, as it is thought to be with
`slL-2R,28 or by alternate splicing of receptor gene
`transcripts.6'9'11 Concentration of soluble cytokine re-
`ceptors is markedly increased in body fluids of pa-
`tients with a wide spectrum of diseases12 including
`lymphoproliferative disorders, certain autoimmune
`diseases, viral infections, septic shock,29 and graft
`rejection episodes. The exact biological
`role of
`soluble cytokine receptors remains unclear. Several
`possible, apparently contradictory functions have
`been suggested. These include 1) inhibiting cytokine
`activity,12 13'29 2) acting as cytokine carriers,30 and 3)
`mediating cellular signaling after forming a complex
`with the cytokine.31 In this context it is interesting that
`the xenotransplanted mJB6 T cell lymphoma secretes
`slL-2R constitutively. At least two explanations for this
`may be considered. First, that the lymphoma cells
`reflect the phenotype of their normal counterparts
`"frozen" at the activation stage associated with se-
`cretion of slL-2R. Accordingly, normal lymphocytes
`activated by mitogens have been shown to synthe-
`size sIL-2R.5 Second, IL-2 is a potent growth factor for
`immune lymphocytes responding to tumor (tumor-
`infiltrating lymphocytes).32 Because slL-2R has been
`shown in vitro to block biological activity of IL-2,33'34
`it is possible that constitutive secretion of slL-2R in
`vivo confers a growth advantage on lymphoma cells
`by inhibiting IL-2-dependent activation of tumor-
`infiltrating lymphocytes. The fact that the mJB6 tumor
`secretes only slL-2R but not IL-2, as shown by lack of
`cytoplasmic staining of the tumor by anti-IL-2 anti-
`body and absence of human IL-2 in serum of tumor-
`bearing mice (data not presented), argues in favor of
`this possibility.
`Regardless of the poorly understood biological
`function of soluble cytokine receptors, their increased
`concentration in body fluids may become an impor-
`tant marker for the detection and monitoring of human
`lymphoid malignancies. Our finding that slL-2R se-
`rum and urine levels correlate directly with tumor vol-
`ume (Figures 4 and 7) support this hypothesis. How-
`ever, some caution is necessary when extrapolating
`results of this study to patients. First, slL-2R is a rela-
`tively nonspecific marker because any activation of
`the immune system appears to increase its concen-
`tration in serum.12,13 Thus, a concomitant inflamma-
`tory disorder such as infection or autoimmune dis-
`ease may affect its level in patient serum. Second,
`
`NOVARTIS EXHIBIT 2112
`West-Ward v. Novartis, IPR 2017-01592
`Page 7 of 9
`
`

`

`Wasik et al
`1096
`AJP May 1994, Vol. 144, No. 5
`
`although our study provides a strong argument for
`tumor cells being a major source of slL-2R in serum
`of patients with lymphoid malignancies, it does not
`exclude in such patients tumor-infiltrating lympho-
`cytes as a possible additional source. The relative
`importance of this putative contribution of tumor-
`responding cells remains to be determined. Finally,
`the exquisite sensitivity of the assay in our system
`(Figure 6) is diminished in humans due to existence
`of slL-2R baseline levels in normal individuals and,
`more importantly, to a dilution of tumor-derived slL-2R
`in a much larger volume of serum in humans com-
`pared with mice. However, the sensitivity of the assay
`may still be superior to the conventional radioisotopic
`scans. In addition, it may be further enhanced by
`screening for a soluble tumor-associated serum
`marker that is made by a large proportion of tumor
`cells rather than only a minority of the cells as it was
`the case in this study (Figure 2). Despite the above
`limitations, measurement of serum concentration of
`slL-2R and other soluble cytokine receptors has a po-
`tential for broad clinical applications. Elevated slL-2R
`serum levels have already been found to significantly
`correlate with adverse prognosis, clinical stage, or
`tumor burden in patients with lymphomas17,18 (MA
`Wasik et al., manuscript in preparation). Measure-
`ment of slL-2R in serum appears to be particularly
`well suited to monitor change in total volume of the
`lymphoma resulting from tumor progression or re-
`gression once the slL-2R serum baseline levels has
`been established in a particular patient (MA Wasik et
`al., manuscript in preparation). Rapid elimination of
`soluble cytokine receptor from circulation, as shown
`for murine slL-4R35 and the human slL-2R (Figure 5),
`support the feasibility of soluble cytokine receptors as
`markers of tumor volume.
`In summary, our data document the constitutive se-
`cretion of slL-2R by xenografted human T cell lym-
`phoma and thus provide supporting evidence that
`malignant cells are a major source of serum slL-2R in
`patients with lymphoid tumors. Furthermore, our re-
`sults provide additional rationale for using measure-
`ments of slL-2R and other soluble molecules to detect
`and monitor changes in tumor volume in such pa-
`tients.
`
`Acknowledgments
`We thank Dr. Beverly Teicher for serum and urine
`samples of mice transplanted with human prostatic
`and ovarian carcinomas and glioblastoma, David
`Northey for technical help with blood collection and
`tumor resection, and Dr. Bernard J. Ransil for assis-
`tance in statistical analysis of the data.
`
`References
`
`1. Tsudo M, Kozak RW, Goldman CK, Waldman TA: Dem-
`onstration of a new (non-Tac) peptide that binds inter-
`in a multichain
`potential
`participant
`leukin-2:
`a
`interleukin-2-receptor complex. Proc NatI Acad Sci
`USA 1986, 83:9694-9698
`2. Takeshita T, Asao H, Ohtani K, Ishii N, Kumaki S,
`Tanaka N, Munakata H, Nakamura M, Sugamura K:
`Cloning of the gamma chain of the human IL-2 recep-
`tor. Science 1992, 257:379-382
`3. Fuji M, Sugamura K, Sano K, Nakai M, Sugita K, Hi-
`numa Y: High-affinity receptor-mediated internalization
`and degradation of interleukin 2 in human T cells. J
`Exp Med 1986, 163:550-562
`4. Arima N, Kamio M, Imada K, Hori T, Hattori T, Tsudo
`M, Okuma M, Uchiyama T: Pseudo-high affinity inter-
`leukin 2 (IL-2) receptor lacks the third component that
`is essential for functional IL-2 binding and signaling. J
`Exp Med 1992, 176:1265-1272
`5. Rubin LA, Kurman CC, Fritz ME, Biddison WE, Boutin
`B, Yarchoan R, Nelson DL: Soluble interleukin-2 recep-
`tors are released from activated human lymphoid cells
`in vitro. J Immunol 1985, 135:3172-3177
`6. Mosley B, Beckman MP, March CJ, Izerda RL, Gimpel
`SD, Vanden Bos T, Friend D, Alpert A, Anderson D,
`Jackson J, Wignall JM, Smith JM, Gallis B, Urdal JE,
`Widmer MB, Cosman D,
`LS: The murine
`Park
`interleukin-4 receptor: molecular cloning and charac-
`terization of secreted and membrane bound forms.
`Cell 1989, 59:335-348
`7. Novick D, Engelmann H, Wallach D, Rubinstein M:
`Soluble cytokine receptors are present in normal hu-
`man urine. J Exp Med 1989, 170:1409-1414
`8. Schall T, Lewis M, Koller K, Lee A, Rice G, Wong G,
`Gatanaga T, Granger G, Lentz R, Raab H, Kohr W,
`Goeddel D: Molecular cloning and expr