APOTEX ET AL. - EXHIBIT 1057
`Apotex Inc. et al. v. Novartis AG
`IPR2017-00854
`
`

`

`
`
`.MEDIcALscIENCES
`
`
`
`Table 1. Anthropometric parameters, leptin, and IFN-y measurements in RRMS patients and
`NIND controls
`
`Parameter measured
`
`No. of patients
`Sex (male/female)‘
`Age, yr
`Height, m
`Mass, kg
`BMl, kg/rnZ
`Serum leptin, pg/ml
`Serum leptin/BMI
`CSF leptin, pg/ml
`CSF leptin/BMl
`CSF leptin/serum leptin
`CSF leptin/CSF albumin
`CSF leptin index’
`CSF lFN-y, pg/ml
`
`RRMS patients
`
`126
`58/68
`36.5 : 9.5
`1.65 : 0.07
`65.9 I 9.6
`24.1 : 3.3
`21,5170 : 15,6760
`900.0 : 650.0
`1,143.1 : 1,389.5
`47.7 : 57.3
`0.09 t 0.18
`5,7 t 6.5 X10"S
`20.7 : 53.4
`3.9 I 3.1
`
`NIND controls
`117
`52/65
`38.2 1 15.5
`1.66 I 0.07
`65.4 t 10.7
`23.4 : 3.2
`11,7270 2: 13.0570
`488.6 : 482.0
`205.3 : 222.5
`8.2 1 8.7
`0.03 i 0.06
`1.2 : 2.1 x 10‘6
`7.0 I 12.2
`0.45 : 1.3
`
`P
`
`0.30
`0.52
`0.65
`0.12
`0.0001
`0.0001
`0.0001
`0.0001
`0.001
`0.0001
`0.008
`0.0001
`
`*The CSF leptin index is a measure of in situ synthesis of leptin in the CNS, calculated with the following formula:
`(CSF leptin/CSF albumin)/(serum leptin/serum albumin).
`
`Flow Cytometry. Immunophenotypic analysis of peripheral blood
`from RRMS patients and healthy controls was performed with an
`EPICS XL flow cytometer (Beclunan Coulter) using the Beckman
`Coulter software program XL SYSTEM 11. Triple combinations of
`various anti—human mAbs were used (Coulter Immunotech, Mar-
`seille, France). All samples were analyzed within 3-4 h of sampling,
`and staining was performed according to standard procedures as
`described in ref. 17.
`‘
`
`CD4+CD25+ TchS in various mouse strains were analyzed by
`flow cytometry with a FACSCalibur flow cytometer (Becton
`Dickinson) and the Becton Dickinson software program
`CELLQUEST. mAbs were added to single cell suspensions of lym—
`phocytes obtained from spleens and lymph nodes after lysis with the
`ACK buffer [0.15 M NH4Cl/10 mM KHCO3/0.1 mM NazEDTA
`(pH 7.4)]. The analysis and quantification of the TRCgs population
`was obtained by gating on CD4‘ T cells.
`
`Human Myelin Basic Protein (hMBP) T Cell Lines. hMBP—specific
`short—term T cells lines were generated according to a method
`reported in ref. 18. The T cell lines were derived from peripheral
`blood lymphocytes of three naive-to-treatment RRMS patients.
`
`Proliferation and Suppression Assays. For in vitro blocking experi-
`ments, Abs against human leptin provided by Radek Sokol (Bio-
`Vendor, Brno, Czech Republic) and mAb against the human leptin
`receptor (R & D Systems) were used at a final concentration of
`10—25 ug/ml; the control was irrelevant IgG Ab (BioVendor).
`The in vitro suppressive capacity of TMS isolated from RRMS
`patients and healthy controls was measured after magnetic cell
`sorting by using the Dynal CD4‘CD25' Tch kit (Dynal, Oslo).
`Briefly, CD4‘CD25‘ T cells (5 ><'10" cells per well) were cocul-
`tured with CD41CD25+ (5 X 104 cells per well) in a 1:1 ratio (both
`98% pure) and stimulated for 3 d in the presence of anti-CD3/
`CD28 Dynabeads (0.1 bead per cell) (Dynal). In mice, TRCgs were
`isolated with the Regulatory T Cell Isolation kit (Miltenyi Biotec,
`Gladbach, Germany) and stimulated with anti-CD3 antibody (2C11
`hybridoma) at 200 ng/ml final concentration and irradiated (30 Gy)
`T cell—depleted syngeneic splenocytes (1:1 ratio, 5 X 104 cells per
`well).
`
`Immunocytochemistry. T cells cultured, or not, with hMBP were
`washed twice with PBS on d 5 of culture, spotted onto glass slides,
`and fixed with methanol for 2 min. Leptin and ObR were detected
`with polyclonal Abs (Santa Cruz Biotechnology) (3).
`
`Mice. Female ob/ob (C57BL6/J-0br/0b), WT controls (C57BL6/J-
`WT), female leptin—receptor-deficient (db/db) mice (C57BL—Ks—
`
`db/db), C57BL—Ks—db/+ controls (db/+), and SJL/J mice (all 6»8
`weeks old) were obtained from Harlan Italy (Corezzana, Italy).
`Experiments were performed following the guidelines of the Isti—
`tuto Superiore di Sanita, Rome.
`
`EAE Induction and Treatment with the Fusion Protein of ObR and the
`
`Fc Fragment of IgG (Obnch). The peptide used for EAE induction
`in SJL/J female mice was the proteolipid protein peptide
`(PLP)139_151) (HSLGKWLGHPDKF). The peptide was synthe-
`sized by INBIOS (Pozzuoli, Italy), purity was assessed by HPLC
`(>97% pure), and amino acid composition was verified by mass
`spectrometry. For EAE induction, mice were immunized so in the
`flank with 100 [.L1 of complete Freund’s adjuvant (Difco) emulsified
`with 100 ug of PLP139_151 peptide on d 0 and with 200 ng of pertussis
`toxin (Sigma) i.p. on d 0 and d 1. Mice were scored for clinical
`symptoms and weighed daily according to a system described in refs.
`2 and 3. Brains and spinal cords were dissected 1520 d after
`immunization and fixed in 10% formalin. Paraffin—embedded sec
`tions of 5 pm thickness were stained with hematoxylin/eosin, and
`sections from 4—10 segments per mouse were examined blindly for
`the number of inflammatory foci by using a scoring system de—
`scribed in ref. 3.
`
`The chimeric fusion protein OszFc (R & D Systems) in 200 [.L1
`of PBS was injected i.p. at a dose of 100 pg per mouse per day for
`three consecutive days. Thus, treatment with OszFc of SJL/J mice
`was performed on d ~1, d 0, and d +1 both before and after
`PLPthsl immunization. The same amount of control IgG was
`injected i.p. in the control SJL/J mice.
`
`Real-Time Quantitative PCR. mRNA was extracted from purified
`CD4‘CD25‘ cells (98% pure by FACS analysis) by using the
`MicroFastTrack 2.0 kit
`followed by cDNA synthesis with
`the SuperScript System (Invitrogen). Expression levels of the
`transcription factor FoxP3 were analyzed by real-time quan—
`titative PCR (TaqMan gene expression assay) by using an ABI
`PRISM 7700 thermal cycler (Applied Biosystems). TaqMan
`primers and probes for FoxP3 and for the housekeeping gene
`GAPDH were purchased as premade kits (Applied Biosys-
`tems). For relative quantitation of gene expression to the
`endogenous control, the comparative CT method was used in
`accordance with the manufacturer’s guidelines. Results are
`expressed as the percentage of FoxP3 increase compared with
`CD4‘CD25‘ effector T cells.
`
`Statistical Analysis. Nonparametric analyses were performed by
`using the Mann—Whitney U test for unrelated two-group analyses.
`The AN OVA test was used to assess differences between groups.
`
`Matarese et al,
`
`PNAS
`
`| April 5, 2005
`
`| Vol.102
`
`| no.14 l
`
`5151
`
`a 7
`
`1
`fl
`
`
`
`

`

`a
`RRMS Manta-125)
`mm} y-7ousxusom
`A
`3 n2=m-
`.
`E E6ml “‘01
`.
`3mg 19-,”
`
`'
`
`~-
`
`‘
`
`'
`
`mun comm (n I117)
`b
`10080 y-1mw’32usn
`sew viz-0.2
`Rum
`
`
`
`uts l7.5 20 225 25 27.5 30 32.5
`
`0‘5 175 20 R5 25 27.5 36 32.5 H 37.5
`35 37.5
`Body mass Index (mm?)
`
`C
`
`7WD:
`Emmi
`moi
`
`,
`
`_
`
`-
`
`.
`
`.
`
`.
`anus mum (n - ml
`y - ms: O um
`R - u:
`Iii-0.0“"
`.P; 0.1
`
`7m
`3cm
`50°”
`
`nun Willis-111’)
`y = 1515:4551:
`R - u
`RZIIM
`P- 0.0001
`
` 015 £75 20 22.5 25 275 30 32.5 35 375
`
`Body mass index (kglmz)
`
`.
`
`'
`;
`
`.
`
`'
`
`-
`
`'
`
`mm
`
`e
`9
`E
`gm
`5 1,0000
`.
`s
`5 30°00 .
`a 20900 ‘
`
`was WM! (n' m)
`y- m: . mom
`- R:- 0.1
`n - 0.3
`‘ P-nnooi /,
`‘
`.
`//’.
`.
`
`c
`
`f
`mom
`
`m
`5mm
`W
`
`
`
`
`NIND controls (n = in;
`yI31.7ZI 0 5214.4:
`R1 - 0.:
`n . 0.54
`P-um
`
`3am .
`
`2mm 100m
`
`
`20m1050.. ll!
`
`onlmmmrmmwmmm
`ODIWMM‘OOOSGOOWFW
`CSF lentil) (99ml
`
`Minn control: (n - 117}
`y amnion)
`R2 - (worm
`R - M17
`P - rm
`
`Runs mm in - m;
`y- 554.39"an
`iu - 91
`R . us
`P a «.0001
`
`.
`.v_,/
`
`.._/’/
`
`'
`
`52
`
`14
`
`15
`
`h
`mac
`sum
`SM
`mu
`
`3000
`
`CSF interferonq (pglml)
`
`Fig. 1. Naive‘to-therapy RRMS patients show an increased secretion of leptin in
`serum and CSF that correlates with lFN-y production in CSF. Statistical analyses of
`these data are summarized in Table 1. (a and b) Simple regression analysis
`between serum leptin and BMI in RRMS patients (n = 126) and NlND controls(n :
`117). The correlation was lost in RRMS patients, whereas correlation was main-
`tained in NlND controls. (cand d) The correlation between CSF leptin and BMI was
`lost in RRMS patients whereas correlation was very strong in NlND controls. (e and
`f) Significant correlation between serum and CSF leptin in both RRMS patients
`and NIND controls; the correlation was stronger in patients than in controls. (9
`and h) Simple regression analysis between the CSF leptin and the lFN—y levels in
`CSF in both RRMS patients and NIND controls. Only in RRMS patients was there a
`statistically significant positive correlation between the CSF leptin and the lFN—y
`levels in CSF ln NIND controls, the lFN»y average levels were very low (see Table
`1), and no correlation was observed with CSF leptin.
`
`Simple regression analysis and the Pearson’s correlation coeffi—
`cients were adopted to study the relationship between different
`variables. The program used was STATVIEW (Abacus Concepts,
`Cary, NC). Results are expressed as mean : SD; P < 0.05 was
`considered statistically significant.
`
`Results
`
`increased Serum and CSF Leptin in Naive-to-Treatment RRMS Patients
`Correlates with lFN-y Production in CSF.
`We found that leptin was increased in both serum and CSF of
`naive—to-therapy RRMS patients (Table 1 and Fig. 1 a—d). These
`
`5152
`
`i www.pnas.0rg/cgi/doi/‘l0.1073,i’pnas.0408995102
`
`differences were maintained even when serum and CSF leptin
`were normalized for BMI (Table 1 and Fig. 1 a—d). In addition,
`as expected, serum and CSF-leptin secretion positively corre—
`lated with BMI
`in NIND controls (Fig.
`1 b and d). This
`correlation was lost in RRMS patients (Fig.
`1 a and c). Con—
`versely, the correlation between serum leptin and CSF leptin was
`maintained in both RRMS patients and NIND controls; how?
`ever, this correlation was stronger in NIND controls than in
`RRMS patients (Fig. 1 e and f). We also compared the CSF-
`leptin/serum-leptin ratio and observed a statistically significant
`increase of this value in RRMS (Table 1). This evidence was
`further supported by the lack of increase of albumin in the CSF
`of RRMS patients, a marker of blood-brain-barrier (BBB)
`damage. In addition, we calculated the CSF-leptin/CSF-albumin
`ratio as a further indicator of BBB integrity and the CSF leptin
`index, calculated as the (CSF leptin/CSF albumin)/(serum
`leptin/serum albumin),
`to evaluate the in situ production of
`leptin by CNS. As shown in Table 1.
`the CSF-leptin/CSF-
`albumin ratio and the CSF leptin index were higher in RRMS
`patients (Table 1), suggesting the production of leptin by CNS in
`RRMS.
`
`Finally, we measured the amount of IFN-y and IL—4 in CSF and
`observed a significant increase in IFN-y (Table 1) and a positive
`correlation with CSF leptin secretion in RRMS patients only (Fig.
`1g and h). IL—4 did not show any significant increase in CSF, and
`the concentration of IL-4 was always below the detection limit of the
`assay in both RRMS and NIND controls (data not shown).
`
`anti-Leptin
`RRMS Patient '1 E150
`
`E100
`El
`50
`
`0
`
`g 200
`
`Medium
`
`memo pglml)
`
`Mun cum: Anson mm
`($5!qu (10W) (10149”!
`map (10ugiml)
`
`lines against hMBP derived from nai‘ve-to-treatment RRMS
`T cell
`Fig. 2.
`patients produce immunoreactive leptin, up-regulate the ObR, and are inhibe
`ited in their proliferation by anti-leptin or anti—Ieptin-receptor blocking an-
`tibodies, (a—c) Expression of leptin in T cells from a naive—to~treatment RRMS
`patient in the presence of medium alone (a) or after activation with hMBP (b
`and c). Leptin was detectable only after activation in the cytoplasm of T cells.
`(d—f) Expression of ObR on T cells in the presence of medium only (d) or after
`activation with hMBP (e and f). The ObR was expressed at very low levels
`before activation and was significantly up—regulated on the cell membrane
`after activation with hMBP (e and 1‘). All photos show immunoperoxidase
`staining with diaminobenzidine chromogen (brown) and hematoxylin coun-
`terstaining (violet). The open squares in b and e represent the zone of higher
`magnification shown in c and 1‘, respectively. (Magnification: a, b, d, and e,
`x100; c and 2‘, X400.) (9) Anti~hMBP short-term T cell lines secrete immuno—
`reactive leptin. (h) The anti-hMBP proliferative response ofT cells is inhibited
`by the addition to cell cultures of either of the anti‘Ob or anti-ObR antibodies.
`The data shown are from one representative experiment of three.
`
`Matarese et al.
`
`a a '
`
`3
`
`

`

`.9
`
`Et
`
`.
`8
`:5
`2
`s
`3
`o
`as
`
`a
`
`‘
`
`_r:
`§ 40000
`E 30000
`‘3 20000
`10000
`0
`
`b mono
`.
`A 60M I .
`-
`
`50000 ‘
`
`
`
`
`90
`so a
`70 o
`5°
`i:
`so 8
`‘°
`‘
`30 g
`20
`10
`a
`
`RRMS
`
`Healthy controls
`
`0
`
`.
`I
`
`RRISpitlmhtnIM}
`y.4_m.5u,15
`Rum
`Rae,”
`P-am
`
`
`
`0
`
`2.5
`
`I5 I75 20 22.5
`125
`10
`7.5
`5
`at, cmtcozsi oicm+ cells
`
`C mono
`A 60000
`E
`3 some
`E .0...
`5 some
`as
`20000
`10000
`
`0
`
`g
`
`Mycoritmbtn-m
`y-lmihrtifll
`m-m
`R=Ml
`P 0.
`'
`2
`
`°
`
`0°
`°
`
`0
`
`0
`
`e
`8
`0
`0
`
`ea c
`on o
`o
`o
`2.5
`5
`7.5
`in
`its 15
`17.5
`20
`22.5
`at cm’tcozs+ oi 004‘ cells
`
`d raccoo-
`
`4
`
`E
`% iCOO’fl
`9
`E
`%
`5 amour
`
`f
`T
`
`
`
`Q RRMS patieniur
`I RRMS patientitz
`I Healthy control
`I
`r-Wfi
`seam—“a
`r
`
`'52%
`
`
`CDd'CDZS' W’CDfi'
`cwcozs'
`carcms‘
`Beads aniiCDSICDZE slimulaflon
`
`
`
`
`
`
`in
`Inverse correlation between serum leptin and circulating TReg,
`Fig. 3.
`RRMS patients. (a) The immune phenotype of circulating lymphocytes in RRMS
`patients selected on the basis of their increase in serum leptin (RRMS patients
`with a serum leptin increaseto 22.5~fo|d higherthan the mean ofserum leptin
`observed in NIND and healthy controls) revealed a significant reduction in the
`percentage and the absolute number of circulating TReg,_ (t, P : 0.0001 and *,
`P : 0.0001, respectively). (b and c) Astatistically significant inverse correlation
`was observed between serum leptin and circulating TRegs in RRMS patients (b),
`whereas no correlation was observed in healthy controls (c). (d) Functional
`analysis of CD4 ‘CD25’ TR,395 of two RRMS patients selected on the basis of an
`increase in serum leptin. The proliferative response was inhibited upon addi~
`tion of CD4'CD25' cellsto the CD4tCD25‘ responder population at a 1:1 ratio
`in normal controls (black bars). CD4‘CD25' cells from two naive-to—therapy
`patients with RRMS exhibited significantly less suppressor activity (white and
`gray bars). *, P = 0.03. CD4’ CD25' cells alone were unresponsive upon
`stimulation as reported in ref. 9. The numbers above the bars represent the
`percent of inhibition of proliferation in the experiment. The data shown are
`from one representative experiment of five.
`
`OszFc Soluble Chimera Increases the Number of Thy, and Amelio-
`rates Clinical Course and Progression of Relapsing EAE (R-EAE). Treat—
`ment of normal R—EAE-susceptible SJL/J mice with anti-leptin
`
`PNAS
`
`[ April 5. 2005
`
`|
`
`vol. 102
`
`i
`
`no. 14
`
`|
`
`5153
`
`
`
` MEDICALsciences
`
`RRMS Patient-Derived T Cell Lines Activated with hMBP Produce
`
`Immunoreactive Leptin and Up-Regulate the ObR. To investigate
`whether leptin could be secreted by hMBP-activated autoreactive T
`cells present in the CNS, we generated short-term T cell lines from
`RRMS patients and stained them with leptin- and ObR-specific
`antibodies. As shown in Fig. 2 a—f, hMBP—activated T cells from
`three naive-to—therapy RRMS patients produced consistent
`amounts of leptin and up—regulated the ObR. The production of
`leptin was also confirmed with the measurement of immunoreac—
`tive leptin in the culture medium by a human-leptin-specific ELISA
`(Fig. 2g).
`
`Neutralization of Leptin or Its Receptor Inhibits T Cell Activation of
`hMBP-Specific T Cell Lines Derived from RRMS Patients. We measured
`the proliferative response against-hMBP on T cells from three
`naive-to-treatment RRMS patients and added either an anti-leptin
`or an anti-leptin-receptor blocking antibody to the culture medium
`(Fig. 211). We observed a significant reduction in the proliferative
`response of all three patients tested, ranging from 45% to 60%
`inhibition of proliferation (Fig. 211).
`
`in
`Inverse Correlation Between Serum~ Leptin and Circulating Tneg,
`Naive-to-Treatment RRMS Patients. The analysis of the immune
`phenotype was also performed on the peripheral blood of 31
`individuals from the naive-to—therapy RRMS patient population,
`selected on the basis of increase in serum leptin concentration (a
`serum leptin increase to 22.5—fold higher than the mean serum
`leptin observed in NIND and healthy controls). We compared these
`phenotypes with the immune phenotype of 27 healthy controls
`matched for age, sex, and BMI. The relative percentage and the
`absolute cell count per mm3 of the CD3‘, CD4‘, CD8‘, CD19*,
`CD3'CD16+CD56‘, and CD4+CD25+ TRegs subpopulations were
`performed (see Table 3, which is published as supporting informa-
`tion on the PNAS web site). Interestingly, naive-to—therapy RRMS
`patients, selected on the basis of their serum leptin increase, showed
`a significant reduction in the percentage and absolute number of
`TM, in the peripheral blood (Fig. 3a and Table 3), whereas no
`difference was observed in the frequency of the other cell subv
`populations (Table 3). TM, measurement in healthy controls was
`in agreement with that found in other studies (14). Regression
`analysis between serum leptin and the percentage of Tch, showed
`an inverse correlation in RRMS patients (Fig. 3b) but not in the
`controls (Fig. 3C). In vitro analysis Of ”FMS-mediated suppression in
`RRMS patients indicated a reduced ability to suppress T cell
`proliferation as compared with healthy controls (Fig. 3d), as
`reported in ref. 15. Moreover, the addition of leptin (100 ng/ml) to
`human '1"chS alone, or during coculture with CD4‘ CD25‘ effec~
`tors, did not alter significantly either proliferation or the suppressive
`capacity of Tchs (see Fig. 5 a and b, which is published as supporting
`information on the PNAS web site).
`
`ob/ob and db/db Mice Have Increased Tnegs. To analyze in more
`detail
`the effect of leptin on the generation of TM,
`in the
`periphery, we measured the effect of chronic leptin deficiency on
`the number of TRCE, in 0b/0b mice. These mice showed an increased
`frequency of TRCgS in lymphoid organs when compared with normal
`WT mice (10.4 : 3.7% vs. 4.7 i 1.7%, respectively; P < 0.02). In
`addition, we counted TRCgS in the lymphoid organs of db/db mice
`and, again, observed an increased percentage of TREE, when com-
`pared with db/+ heterozygote controls (13.9 : 1.9 vs. 7.9 i 0.9,
`respectively; P < 0.01). Finally, the suppressive capacity and phe—
`notype of TRegs from db/db mice were evaluated. No significant
`differences were observed in terms of either suppressive capacity or
`hyporesponsiveness of TRCg, (see Fig. 6 a—c, which is published as
`supporting information on the PNAS web site). In addition, ex~
`pression levels of FoxP3 in Tm, of 017/01) and rib/db mice were
`comparable to those in normal control mice (Fig. 6d).
`
`Matarese et at.
`
`a V
`
`A
`'3
`
`
`
`
`
`

`

`
`
`
`
`CTR-Ab ObR:Fc
`in viva treatment from day -1 to #1
`
`
`ill
`
`ill,
`
`
`
`
`
`Meanclinicalscore
`
`
`
`Percentageinitialbodyweight
`
`Fig. 4. Neutralization of leptin with ObR:Fc increases the number of TRegs
`and ameliorates the clinical course of R-EAE.
`(a) Treatment of R-EAE-
`susceptible SJL/J female mice with ObR:Fc induced a significant increase in the
`circulating TRegs- *, P = 0.01. (b) Mean clinical score (bars) and body weight
`(curves) of SJL/J female mice pretreated with the ObR:Fc (white bars and
`squares) or the CTR-Ab (black bars and squares) on d —1, d 0, and d +1 and
`immunized with the PLPiggzm on d 0. Statistical analyses of these data are
`summarized in Table 2. The data shown are from one representative experi-
`ment of two (r1 = 6 mice per group). t, P 2 0.01; H, P = 0.002.
`
`blocking ObR:Fc soluble chimera induced an increase of the
`percentage of TRC,gs in the periphery (Fig. 4a). To test whether this
`treatment could also modify the induction/progression of R-EAE,
`we pretreated SJL/J female mice with ObR:Fc chimera before
`immunization with the encephalitogenic peptide PLP139_151. The
`treatment was performed from d —1 to d +1 with i.p. injection of
`100 pg per day per mouse of ObR:Fc chimera dissolved in PBS (Fig.
`4b). ObR:Fc-treated mice showed a reduced peak clinical score, an
`improvement in disease relapses and progression, and a reduction
`in the percentage of body weight
`loss (Fig. 4b and Table 2).
`Moreover, a significant increase in body weight before/during the
`early phases of disease in ObR:Fc-treated mice (on d 9,
`the
`percentage of initial body weight
`in ObR:Fc-treated mice was
`112.4 ': 0.9% vs. 100.6 : 0.7% in control (CTR)—Ab-treated mice,
`P : 0.01) was observed, compared with a classical reduction in body
`
`weight preceding the onset of clinical symptoms in CTR-Abtreated
`mice In addition, ObR: Fc- treated mice showed very rapid reduc-
`tion of body weight after d 10 and a rapid recovery after d 13 of
`disease to a weight
`that was significantly higher than that of
`CTR-Ab-treated mice (Fig. 4b and Table 2). On the contrary,
`control mice showed a more stable body weight
`loss that was
`maintained over the disease course (Fig. 4b and Table 2). Finally,
`CNS inflammatory lesions were also significantly reduced in Ob—
`R:Fc-treated mice (Table 2). A significant increase in Tch5 was
`observed on d 15 of the disease course in mice pretreated with
`ObR:Fc (Table 2).
`
`Discussion
`
`In this report, we analyze the secretion of leptin in serum and CSF
`of naive-to—treatment RRMS patients in correlation with the se-
`cretion of IFN-y in CSF and the percentage of circulating Tchg. The
`data presented here provide evidence that a significant increase of
`leptin secretion occurs in the acute phase of MS and that this event
`positively correlates with the CSF production of IFN- y. Increased
`secretion is present in both the serum and CSF of RRMS patients
`and determines the loss of correlation between leptin and BMI (Fig.
`1 a and c). Moreover, the increase of leptin in the CSF is higher than
`that in the serum (a 5.6-fold increase in CSF leptin vs. a 18-fold
`increase in serum leptin, P = 00001, Table 1), possibly secondary
`to in situ synthesis of leptin in the CNS and/or an increased
`transport across the blood— brain barrier, upon enhanced systemic
`production. Indeed, the CSF-leptin/serum-leptin ratio, the CSF-
`leptin/serum--albumin ratio, and the CSF leptin index all signifi—
`cantly increase in RRMS patients when compared with NIND
`controls (Table 1).
`Recently, gene-microarray analysis of Th-1 lymphocytes and
`active MS lesions in humans revealed elevated transcripts of many
`genes of the neuroimmunoendocrine axis, including leptin (19, 20).
`Leptin’s transcript was also abundant in the gene—expression profile
`of human Th-I clones, demonstrating that the leptin gene is induced
`in and associated with polarization toward Th-l responses, com-
`monly involved in T cell-mediated autoimmune diseases such as MS
`(19, 20). We previously reported in situ leptin secretion by inflam»
`matory T cells and macrophages in active EAE lesions (3). Here, we
`show that autoreactive hMBP—specific T cells from RRMS patients
`can produce immunoreactive leptin and up—regulate the leptin
`receptor after activation (Fig. 2 a—f). possibly explaining, in part, the
`increased in situ CSF leptin levels in RRMS patients. Interestingly,
`both anti-leptin and anti—leptin-receptor blocking antibodies re-
`duced the proliferative responses of hMBP-specific T cell lines (Fig.
`2 h, j, and l), underscoring the possibilities of leptin-based inter-
`vention on this autocrine loop:
`Many questions need to be answered about whether and how
`TRCgs can regulate autoimmunity in humans. In animal models of
`autoimmune diseases, the role of TREE, has been demonstrated (21).
`More recently, a reduced function and/or generation of TRCg, in
`human autoimmune diseases such as systemic lupus erythematosus,
`type 1 diabetes, autoimmune polyglandular syndromc type II,
`
`Table 2. Effect of pretreatment with soluble ObR:Fc chimera on neurological impairment and percentage of CD4+CD25+ during active
`R-EAE induction with the PLP139.151 encephalitogenic peptide in SJL/J female mice
`
`Percentage of
`CD4*CD25+
`No. of
`Percentage of initial
`after
`inflammatory
`body weight at
`Peak clinical Average
`Day of onset
`lncidence,
`______«—___—_________....—__._.—.____—___—_———~—
`Antigen no./tota| (%)
`(range)
`score
`CDI"
`disease peak
`foci
`treatment
`Group of mice
`
`4.5 : 0.7
`308 : 1.8
`89.4 : 0:5
`42.7 t 7.9
`2.8 f 0.7
`8.1 I 0.4 (8—9)
`6/6 (100.0)
`PLPizem
`SlL/J CTR»Ab (d ‘1 to d+1)
`
`
`
`
`10.6 t 2.0 (8—13)6/6 (100.0)PLP1397151SJL/J ObR:Fc (d -1tO del) 11.3 : 4.3‘M1.9 t 0.7' 21.8 : 5.3‘ 108.2 f. 0.7‘ 15.0 :1.5‘
`
`
`
`
`
`The data shown are from one representative of two independent experiments shown in Fig. 4b. CTR»Ab, control Ab.
`*Curnulative disease index, sum of daily scores determined for each mouse of that group and averaged.
`'P = 0.01.
`‘P : 0.002.
`
`5154
`
`I www.pnas.org/‘cgi,I'doi/10.1073,r"pnas.0408995102
`
`Matarese et al.
`
`

`

`juvenile idiopathic arthritis, and MS has been described in refs.
`10—15. Recently, this reduction has'been shown, in RRMS, to be a
`functional defect of TRegg rather than a reduced number of Tchs in
`the periphery (15 ). To address whether leptin secretion could have
`an effect on TchS in RRMS patients, we measured the TchS
`frequency in the peripheral blood of nai‘ve-to—treatment RRMS
`patients selected on the basis of an increase in serum leptin to
`22.5—fold higher than levels measured in NIND and healthy
`controls. Here, we show that
`the average percentage and the
`absolute number of TchS in these RRMS patients were significantly
`lower than those of healthy controls (Fig. 3a and Table 3). No
`significant differences in CD3‘, CD4+, CD8+, CD19+, and
`CD3”CD16*CD56‘ cells were observed in either study group
`(Table 3). In addition, our functional data confirmed that, in our
`experimental conditions, RRMS patients showed a functional Tchs
`defect, confirming findings previously reported in ref. 15 (Fig. 3d).
`Administration of exogenous leptin to human Tch, or to suppres—
`sion assays did not alter hyporesponsiveness and suppressive ca—
`pacity (Fig. 5 a and b), suggesting that in vitro leptin is not
`responsible for impaired TchS function. Simple regression analysis
`showed an inverse correlation between systemic leptin concentra—
`tions and TRegs in the naive-to-treatment RRMS population (Fig. 3
`b and c). These data demonstrate an inverse relationship between
`leptin and TRegS in MS and may account for a reduced generation
`of TREES, at least early in the disease, in naive-to—treatment patients.
`Indeed, we hypothesize that, after therapy, these phenomena may
`be masked and overcome by therapy»induced effects. In fact, in the
`case of chronic leptin deficiency, such as in ob/ob mice, we found
`an increased number of circulating TM, and similar results were
`observed in db/db mice. This finding was also confirmed by
`experiments showing a higher recovery and percentage of TRCgs
`from R—EAE—susceptible SJL/J female mice treated with leptin—
`blocking ObR:Fc (Fig. 4a). Also, this pretreatment subsequently
`ameliorated R-EAE onset and progression (Fig. 4b and Table 2).
`The fact that TchS from db/db mice had a similar suppressive
`capacity and phenotype compared with Tch, from normal controls
`(Fig. 6 a—d) suggests that leptin does not affect in vitro suppressive
`funetion but, rather, in vivo expansion/proliferation of TRCES. Fur-
`ther studies need to address this point. Recent reports have shown
`increased secretion of serum leptin before relapses in RRMS
`patients during treatment with IFN-B and the capacity of leptin to
`enhance in vitro secretion of TNF—a, IL-6, and IL—10 by peripheral
`blood mononuclear cells of RRMS patients in the acute phase of
`the disease but not in patients in the stable phase (22, 23). In view
`of the above considerations, we suggest that, in MS, leptin may be
`part of a wider scenario in which several proinflammatory soluble
`factors may act in concert in driving the pathogenic (autoreactive)
`Th-l responses targeting neuroantigens (24). Recently, Hafler er al.
`
`(15) reported a decrease in the effector function and cloning
`frequency of TRegs from the peripheral blood of patients with MS.
`We show here that, in na'ive-to-therapy MS patients, not only the
`function but also the number of TRCgs
`is affected, and, more
`importantly, the finding inversely correlates with the concentration
`of serum leptin. It appears therefore that, early in the disease, the
`effects on Tchs in MS may be different from the effect observed
`after therapy has been initiated. Regarding the correlation with
`leptin,
`it is worth mentioning that strains of mice prone to the
`spontaneous development of autoimmune diseases, such as nono—
`bese-diabetic (NOD) and IL—2-deficient (IL-2””) mice, show re-
`duced frequency of Tchs in the periphery (9) associated with
`abnormal leptin responses due to increased serum leptin concen—
`trations (disproportionate to fat mass) (25, 26). NOD mice have
`higher basal serum leptin levels than normal
`age—,
`sex—, and
`fat-matched controls (25). IL—2‘/‘ mice are prone to spontaneous
`development of inflammatory bowel disease and other autoim—
`mune disorders (26). Whereas in normal mice, serum leptin de-
`creases with fat—mass loss, in IL—2‘/‘ mice there is a paradoxical rise
`in serum leptin compared with control mice, even after starvation,
`which reduces serum leptin (26). These data support the hypothesis
`that a disproportionate response in leptin secretion can correlate
`with a reduction in the periphery of the Tm,gs compartment in these
`two models.
`Because of the influence of leptin on food intake and metabo-
`lism, the findings reported here underscore the role of molecules at
`the interface between metabolism and immunity in the control of
`not only inflammation but also autoimmune reactivity (24, 27).
`Recently, molecules with orexigenic activity, such as ghrelin and
`neuropeptide Y (NPY), have been shown to mediate not only
`effects opposite to those of leptin on the hypothalamic control of
`food intake but also on peripheral immune responses (28, 29).
`Indeed, ghrelin blocks the leptin—induced secretion of proinflam-
`matory cytokines by human T cells (28), and NPY ameliorates the
`clinical course and progression of EAE (29). Given these consid-
`erations, we may envisage a situation in which the influenccs
`exerted by several metabolic regulators,
`including leptin, might
`broadly influence vital functions not limited to caloric tuning but,
`rather, affecting immune responses and the interaction of the
`individual with the environment. Although additional studies are
`needed, our data provide direct evidence of a negative association
`between leptin secretion and Tchs in the early stages of an
`autoimmune disease characterized by Th—1 autoreactivity, such
`as MS.
`
`This work is dedicated to the memory of Eugenia Papa. This work was
`supported by Fondazione Italiana Sclerosi Multipla Grants 2001/R/68
`and 2002/R/55 (to S.Z.) and by a National Institutes of Health grant (to
`A,L.C.).
`
`(St Matarcse, G. (2004) Nat. Rev. Immunul. 4, 371—379.
`1. La Cava, A.
`2. Matarcsc, G.. Di Giacomo, A., Sanna, V., Lord, G. M., Howard, ,1. K, Di Tuom, A., Bloom,
`S. R., lecblcr, R. 1., Zappacosta, S. & Fontana, S. (2001)]. Immunol. 166, 5909—5916.
`3. Sanna, V., Di Giacomo, A., La Cava, A., Lecblcr, R. 1., Fontana. 5., Zappacosta, S. &
`Malarcsc, G. (2003) J. Clin. Invest. 111, 241—250.
`4. Busso, N., 50, A., Chobaz-Pcclat, V., Morard, C., Martinez-Soriu, E., Talabot-Ayer, D. &
`Gabay, C. (2002)]. Immunol. 168, 875—882.
`_
`5. Sicgmund, 13., Lchr, H. A. & Fantuzzi, G. (2002) Gastroenlemlogy 122, 201172025.
`6 Siegmund, B.. Sennello. .1. A., Jones-Carson. 1.. Gamhnni-Robertson, F., Lchr, H A., Batra,
`A., Fedke, 1.. ZeitL. M.
`(i: Fantuzzi, G (2004) (£141 53, 9654972.
`7 Zamvll, S. S. St Slemmun, 1.. (199(0/1/11111 Rm‘
`lmmurml.
`11, 579-621
`8. Brown, A. M. & MeFarlin, D. (1981) Lab. Invest. 45, 278-281.
`‘) Sakaguchl. S. (21104)Arim/ Rm IH‘UYHHH), 22, 5.11562
`10. Kukreja, A.. Cost, 0., Marker, 1., Zliang, C. Sun, / LIn-Su, K., Ten, 5., Sanz, M., Exlcy,
`M., Wilson, B . ('l 11/. (2002) J. (Int. Invest. 109, 1317140.
`11. Baecher-Ailan, C. & Hafler, D. A. (2004)]. Exp. Med. 3, 273—276.
`12. Crispin, J C. Martinez, A. & Alcoccr-Varcla,1. (2003) J. Autoimmun. 21. 27L276.
`13. dc cher,1.M.,Weddcrhurn,L.R.,Taams,L.S.,Pate1,A.,Varsani,1—1.,chin. M., dc Jagcr,
`W., Pugayung, G., Giannoni, F., Rijkers, (1,211)]. (2004)]. Immunol. 172, 6435—6443.
`14. Kricgel, M. A., Lohmann, T.. Gabler, (7., Blank. N., Kalden,1. R. & Lorenz, H. M. (2004)
`J. Exp. Med. 9, 12854291.
`15. Viglictta, V., Baecher~Allan, C, Weincr, H. L. & Haflcr, D A. (2004) I Exp. Med. 5, 9717979
`16. McDonald, W. 1., Compston, A., Edan, G., Goodkin, D., Hartung, H. P., Luhlin, F. D.,
`McFarland. H. F., Paty, D. W., Polman, C. H., Reingold, S. C... 61 al. (2001)/im1. Neural.
`1, 1217127.
`
`l7. Pontieiellu, A., Pcrna, F., Sturkenboom, M. C.. Marcbctiello, 1., Bocchino, M. & Sanduzzi,
`A. (2001) Int. J. Tuberr. Lung Dir. 5, 114871155.
`18. Montanaro, D.. Sanna, V., Matarese, G., Lurby, B. 8., Racioppi, L., Carrieri, P. 13., Bruno.
`R., Davey, N. J., Zappacosta, S. & Fontana, S. (2001) Clin. Exp. Immunol 123, 2887203.
`19. Rogge, L.,Biancl1i.E., Biffi,