4892
`
`Biochemistry 2000, 39, 4892-4899
`
`. --···-----·------
`
`-----llliiili-~
`
`Decreased Secretion of ApoB Follows Inhibition of ApoB-MTP Binding by a
`Novel Antagonistt
`
`Ahmed Bakillah,* Neeru Nayak,* Uday Saxena,§ Russell M. Medford,§ and M. Mahmood Hussain*·*J1
`
`Departments of Parhology and Biochemistry, School of Medicine, MCP Halinemann University,
`Philadelphia, PennsylvQllia 19129, Atherogenics Inc., Norcross, Georgia, and Departments of Anatomy and
`Cell Biology, and Pediatrics, SUNY Health Science Center. Brooklyn. New York 11203
`
`Received October I 5, 1999; Re11ised Manuscript Received January 20, 2000
`
`AB~ACT: Apolipoprotein B (ar)olJ) and n~crosomal triglyc~ride tra11sfer:protein (MTP) are essential for
`the efficient assem.bly of triglyceride-1ich lipoproteins. Evidence has been present_ed for physical intcraction_s
`between tht;Se_ proteins. To ·study the.importance of apoB.:.,.MTP binding in apoB secretion, we, have
`identified· a compound; AQI-St 7. that inhibited (60.,.... 70% at 40 µM) th~ binding of various apoB peptides
`to.MTP.but·rioi tQ.an anti~apoB·inonoclonal antibody,- IDl-;-whoseepitope overlaps with an M_TP binding_
`.si_te in apoR AGI-$17 ha<;! :no significant effect on the lipid tnmsfer. activity of. the_ purified MTP.- Ill
`c<;>ntra5t,,. an~~et,.ailtagoru~· '13.MS.:20Ql5Q;,dfrl~ot .. affeet: aiJQB-MT~ binding bu~ irihi.bited MTP' s lipid. .
`frajlsfer' activity, Thi.': diff~rential effects of .these ,iilhibjfors .suggest. two functionally. independent,. apoB
`· binding and lipid .transfer.-.doin~n~·,in MfP; .AGI-S 1:7 was then. used to $tudy jts ¢ffect on the lipic;f -transfer. ·
`. ·and .. apoB.biodirig-attiviti~s.'of,¥fP in}IepG2:cells;·AOI~Sl7. had:i10 effect on celhila~ lipi~ ti;ansfer .
`· .activities,. but it -inhibited ·c~iinmunoprecipitation of apoB v-titb. -MTP: These .studi¢s indicate· that .. A(jf- ·
`· .S17 iohibits apoB~MTP· tji~djfig: but J:ms no -~ffect.on -:~~sJipid.-transfei: :activity. ~~rimei'l.~'l. were
`. : ihen _performed to _sµid.y lb~ eff~t. of i_n~ibitiun o.f apoBLlyITP bjnding on apoB. secrctj()n ·in HepG2
`· .c¢1Js. p.rn:.s 17· (40 µM)'qid riot iifferit· c.eJI"proteiri levels:but ·decteas¢.tnhe total mass· of apoB >s¢~reted .·
`b.Y 10~8~70.: Sii;Ili'Jarly, t,\OFS:t:? iri~ibit~. tt_ie secretil)ri of IUlsceri~ ~J><>B. l>y 60.,..,80%, J>ut :did, ~ot at;ect
`·albumin's"e:Cretion.-.The!ie,siu,diesJnUidit~ .. tOOt AGI-S.F decr~ses· apoB secreti.on:r1Jost likely by inhi.biting ..
`apoB-MTP·interactions._ Thtis;·the:.biqding· of·MTP: to apoB may be im'Portant .for the-assembly_ and
`sec;:retion ·of apo'6-coiitaiJ1jng ·Jipoprote_i'ns·and can be ·a.- po~~ntial target for -the deveiopme11t ·of. lipid(cid:173)
`lowering drugs. Ii-is proposed ihat the.apoB binding may represem_MTP's chaperope activity th'at assists
`~n the t~nsfer fr<)in the menibraJie :_to the. luiµen: of the endopJasmic reticulum- and· in the. net)ipidation of .
`·nascent ·apoB, and may -be_ essential for· lipbprotein assembly and secretion.
`: · .... :··.'.· .. ·_ ..
`~ : ..
`. . · ..... :, ~
`
`'The presence of it- neutral lipici.tra:nsfor. activity .In _the
`lumen of the endoplasmic reticulum was: iderttifi'ed 'based
`on in vitro lipid transt~r assays using. hi~eoaH:ontcnts of
`.the endoplasmic retlculuni (/, .i) .. Sub~equently~· th~ protein
`. responsible for' the lipid -lra~sfer adivity,_ lhe -mierosomal .
`trigiycende· tiansfer protein (MTP); 1. was purified· to hOtnO(cid:173)
`gcncity and was shown t1»t:onsisl 'of. lwo polypeptides of 97
`and 55· kDil' subunits." The 55 kDa· polyPt:ptide was shown
`to be a ubiquitous enzyine, protein ·disulfiqe isoinerase, that.
`is present · in · the· endopl&Sinic reticwwii. The . 97 kDa
`polypeptide.is-unique and has ~n shown to be responsible
`
`i Supported by. granls from the National I~stitutes of Health (DK-
`46900. and HL-64272) and the American Heart Association. National
`Center. M.M.H. is on established investigator of the American Heart
`Association.·
`• Address correspondence to this author at !he Departments of
`Anatomy and Cell Biology and of Pediatrics, SUNY Health Science
`Center, 450 Clarkson Ave.; Box 5, Brooklyn, NY 11203. Phone: .718- ·
`270-4790; Fax.: 718-270-3732; E·mail: mnhnioodhussain@netmail.
`hscbklyn.edu.
`'MCP Hahnemann University.
`l Alherogenics Inc ..
`11 SUNY Health Science Center.
`1 Abbreviations: apoB, apolipoprotein B; LDL, low-density lipo(cid:173)
`pmwin(s); MTP, microsomal lriglyceric.le transfer protein; PBS, phosphate(cid:173)
`buffered saline; PBS-Tween, PBS containing 0.05% Tween-20.
`
`.
`.
`.
`. .
`for the lipid transfer acliv_ity rtor reviews.- see (3, 4)J ... 'rite
`· importance of the 97 kDa subm).it in the ns$embl)' and
`secretion of apoB-containini£-liP<>proteins was established
`usi_ng three independent approaclies.-Fi~st. -mu~ations· in· the
`97 kPa pi>lypeptid~ were·showri to· be ·~ponsible-.for the
`lack: of. apoB::coritaiillng · 'liphproiefos · in'· th~ ~las.ma of
`individuals :With abet.alipoproteinemi':i. [for. reviews, see. (3,
`4)]. Second, coexpressfon of the 97 kDa polypeptide with
`;ipoB i~ cells that do not secrete lipoproteins has been shown
`to be sufficient for:. lipoprotein assembly and secretion (5:_
`7). _Third, compound~- that inhibit MT.P's' lipid transfer
`activity in vitro and decrease. ap0B secretion in cell ~ulrures
`have been identified (8-13). Recently, second-generation
`inhibitors have been synthesized that decrease triglyceride
`secretion in raL'>, and lower plasma lipid levels in hamsters
`(14). More interestingly, they lower lipid levels (14) in
`·Watanabe heritable hyperlipidernic rabbits (15. 16) that
`express low levels ofLDL receptors due to a genetic defect.
`
`Jn addition to its lipid transfer activity, MTP has been
`shown to interafa physically with apoB (17-23). ApoB and
`MTP bind with high affinity, and these interactions are
`affected by the length and degree of lipidation of apoB ( 19).
`Lysine and arginine residu~s in the N-tenni11al 18% of apoB
`
`J0.1021/bi9924009 CCC: $19.00 © 2000 American Chemical Society
`Published on Web 0313 tnooo
`
`1 of 8
`
`PENN EX. 2151
`CFAD V. UPENN
`IPR2015-01836
`
`

`
`....,,.__,_.---------"--'-'------'---'----------·---·····-·······-····--··---------------~·
`
`Physiological Significance of ApoB-MTP Binding
`
`are critical for its binding to MTP (20). Within this region.
`apoB and MTP can interact at multiple sites \22. 23). ~ne
`·of the MTP binding sites has been recently localized to ammo
`acids 270-570 (21)_ In the present investigations, we have
`attempted to understand the importance of apoB-M!P
`bindino. For this purpose. we identified a novel antagcm1st,
`A"GI~S j 7, ·th~t iiiliibits ai>oB-MTP binding without affecting
`MTP's lipid transfer activity in vitro and in cells. AGl-S 17
`decreased apoB secretion in hepatoma cells, indicating that
`these interactions play an important role in the assembly and
`secretion of apoB-containing lipoproteins.
`
`,.
`I
`
`MATERj.ALS·AND METHODS
`
`Material,\·. AGI-S 17 (602 Da) and ·AGl~3 (616 l)a) are
`synthetic organic ·conipounds of proprietary nature -froin
`Atherogenics Inc. Other .compounds, BMS-200150 and CP-
`10447, were kindly provided by Bristol-Myers Squibb ·and
`Pfi7.er, respectively: Stocks (10 ni1\lf)·o(these co~11pounds
`wcre·prep3red·iii dimelhytsulfoxide. Antibodies·against MTP
`were. kindly· provided by. [),r. ·Ha_ris .JaIJJiJ. <:Qristpl~fyl.ycrs
`Squibb), Egg phosphatidylchoJme, <,":ardiolipin,· and triglyc(cid:173)
`erides: and [14C]triolein and [3H]phqsph~tiqylcholif.1~. u_sed ..
`·-ror .v¢~i~te. prbparai~on~ .wcm;:.pureti~~d.Jro~:A:vanti. f>olar- ·
`·Lipids (Alabii~~ter~ A:l;-) _an~ .f\'ew En~fand. ~uc!~r,J}3ofton,
`. ~1A)1 , .re,~pec#vefr ... :qt.her ·chemicals \v~.R< f r()m: '$1gr:na
`Chemical Co .. (St. :li>uis;· M.O).
`.
`· · .
`·
`. . .
`.
`· : &11s: .H:@ieeii~ ~¢re._obJliil)Cd 1~9in· the A!rieri~ari.Tyi)e·
`C~ltu~ Collectio~ (Roekvillc, lv1D), and cultured in mi.nimal
`essential medium. conttrlning 10% fet~l bOvine. Sefllm '(Bio(cid:173)
`fluids, Maryland; MD) and I% antibiotic-antimycotic (Life
`Technologies. Grand. Islands. NY). McA-RH7777. cells stab.IY ·
`transfected with human.apoB l 8 or apoB28 cDNAs (24; 25)
`were grown in Dulbecco's modified 01edium. (DMEM)
`containing· 10% fetal bovine serum; !0% horse serum, and
`··
`· · · .
`.·
`.
`.
`.·
`·
`1 % antibiotic-antir:nyi;:otic:
`.. MTP 's (.ipid. Tran..efet Activity and ApoB'":--_MTP iJinding. •
`Heter~dimeric.MTP was purified and llSSayed using synthetic
`uniiameuar· donor and ·acceptor vesides as descriped (1. · 2,
`9. 19; i6). Vaiiotis cbhceutrations of different ~II1pounds
`were foduded .during :assays/arid· inhibitioi:i of the._transf~.
`of radiolabeled. triglycerid~s arid phospholipids was deter(cid:173)
`mined by scintiliatfon· counting. For -apoB-MTP binding,
`ELISA pliites were coated with purified MTP ( 1 µg/wc~l)
`and incubated in triplicate. with human LDL ( 1 µg/weU) m
`the presence of the indicated concentrations of various
`qJmpounds as described l:lefore (19-21). Microtiter wells
`were ~ashed (3x) with PB~·Twecn, and apoB bound was
`quantified by a sandwich ELISA (25; 27).
`.
`Cellular Lipid Transfer Activities and ApoB-MTP Bind(cid:173)
`ing. HepG2 cells (25 t..m2 . flasks) . were incubated with
`different concentrations of AGI-Sl7 for 6 h. Cells were
`collected in PBS, pelleted, homogenized in 1 mL of 50 mM
`KC!. 5 mM EDTA, leupeptin·(IOO /tg/mL), l mM phenyl(cid:173)
`methylsulfonyl fluoride. 50 mM Tris-HCI buffer, pH 7A.
`and incubated on ice for 30 min_ Cell lysates were adjusted
`to J mg of protein/mL, 0.054% sodium ·deoxycholate, pH
`7.4, incubated on ice for I h, and centrifuged to remove
`cellular debris. The supernatant was dialyzed extensively
`against 15 mM Tris-HCI buffer, pH 7.4. containing 40 mM
`NaCl, I mM EDT A, and 0.02% sodium azide. Dialyzed cell
`lysates ( 100 µg of protein) were used to measure the transfer
`
`·:·
`. ·.· . . : .
`·-~;~:.
`~~­
`. ·~f:
`
`;f.:-.
`
`. ~~­!.·
`~f
`it
`
`i.f.
`
`l'· . it. I ~\
`
`. :]~(
`
`~:;::·.
`
`f~}}~
`.1~·
`. ~j?;,
`. ~~-'
`
`Biochemi.wy, Vol_ 3.9, No. 16, 2000 4893
`
`of [3H]triolein from the donor to acceptor small unilamellar
`vesicles (9. 26. 28).
`The effect of AGI-Sl7 on intracellular apoB-MTP
`binding was evaluated by coimmunoprecipitation ( / 7J. First.
`we delem1ined the effect of differenl conccntrntions of AGl(cid:173)
`S 17 (0-4.0 µM) o.n the binding .of polyclonal .anti.bodies to
`MTP using Western analysis and the chemiluminescence
`detection system. AGI.-S17 did not inhibit the binding of
`polyclnnal antibodies to the 97 kDa subunit (data not shown).
`We also studied the effect of AGI-Sl7 on the binding
`between ap0B and. its polyclonaJ antibody using ELISA
`Again. AGl-Sl7 did not i~1hibit the recognition of ap_oB by
`its polyclonal antibodies (data not s.~own). These expenmentc;
`indicate that AGl-Sl7 _does nor inJ:iibit antigen-antibod}'.
`interactions: Next, cells W<:re 'i1icubated with different
`concentrations of AGI-Sl7 for 6 h. collec~ed in a nondena(cid:173)
`ru~'ing buffer (62.5. nlM sucrose. 0.5% sodium d.eoxycholate,
`0.5% Triton X-102, 150 mM NaCl, 5 mM EDTA, 50 mM
`. Tris"HCI, pH 7.4, _and prot~a..;c! in~ihitors); and immunopre(cid:173)
`cipitatcd with.an_ti-97 kDa MTf·subw1it antibody: lmm\.1-
`. noptecipita~e$.\\fere.separat¢d on SDS,-ps)ly~cryla~ide gel~!'
`transferred to nitrocellri!9s.e J,Ilembranes.; and ·~loelced wjth
`f>ss:Tween: ~ntaini.ng· 5%. noijfaf ;dfi¢ _milk. flifemh~anes ·
`. w~c prt>&:d.:~iih.·~h~~p .. anti-hµnian ;"lj>oB ·antib()dics ( J :moo.
`dilutio11}11nd. Jiero~idase•conjµgated'"l~«>tein.A· U.i5~ dilu(cid:173)
`tion). Th~ pre~en:c~ .. of apo~ ~.C!.O~ b~?s w~s vis\ialized. using
`enhanced :c.t:ie1nilwni11escence.re_8gents (Amers_ham, Piscat- ·
`··: ·
`. .
`.
`.
`away; Ni) and, fiuerograpby:, · . .
`·Effect of AGl~SJ 7 011 the Se_cretiqn:of Apo8_ by Cultured
`Cells. To detemiine tile effect ofAGl-Sl7 on-the.secretion
`of nasc~nt apoB, HepG2 ccJls (25 cni2 tlasks) were prciry(cid:173)
`cubaled in serum<. cysteine-, and methioni"ne-free mediu~
`for 2 h and radiolabeled. wit)l a mixture of radiolabeled amino
`aciCls (EXpre35S3-~S. :New. Engl!ind Nucfoit): in· t~e presence
`of various concentrations of. AGJ-S 17 for 6 ,h. _ApoB or
`albumin was immun_tiprCC_ipitatCdfrpm ceJI~ and ril(:dia using_
`sheep :imti~human apo!J or imti-hum:in albJ.iffiin anlibodies
`.(Boehringer Marillhei1ri); sepaiai:ed Ofl polyacrylaini~ .. geJs,
`exposed to Pbpsphoriniager sereens; an£! visuali~d using
`Phosphorimager 445SJ. ·(Moi~cuJ~ ·,Dy~afllics, S~r)Jiyvnle?
`CA). fudividuat baricis\vere· quariiitat~·and thc·Joeal average
`background was subtra~iCd uslr1g hnagequant Program.
`T~ deteniline the effect ofAGI-S l7 on the total mass ·of
`apoR secreted, HepG2 cells (24-well plates) were incubated
`with various concentrations of AGI-S 17 for 6 h, and the
`amount of apoB secreted was quantitated by ELISA (25, 27).
`Jn addition, the conditioned medium ~as subjected to density
`gmdient. ultracentrifugation (Z9), and apoB in different
`fractions was quanlitated.
`Other Amllyses. Protein. was· determined using ~he Coo(cid:173)
`massie Plus reagent (Pierce Chemical Co., Rockford, Il..} with
`BSA as a standard (30). Optical density in ELISA plates
`was measured at 405 nm using a Dynatech. MRX microplale
`reader (Dynatcch Labs, Chantilly, VA). The data were plotted
`as the mean ± the standard deviation.
`
`RESULTS
`
`Identification of <m Antagonist That Inhibits ApoB-MTP
`Binding and Does Not Affect the Lipid Tra11sfer Actfoity of
`MTP. To evaluate whether apoB-MTP binding might be
`important for the secretion of apoB-containing lipoproteins,
`
`2 of 8
`
`PENN EX. 2151
`CFAD V. UPENN
`IPR2015-01836
`
`

`
`4894 Biochemistry, Vol. 39. No. 16, 2000
`
`Bakillah et al.
`
`A. LDL-MTP binding
`
`A.ApoB100
`
`---AGl-517
`--AGl-3
`- - BMS-200, 150
`
`100
`
`80
`
`c
`0
`E 60
`.Cl
`'.E 40
`.E
`~ 0
`
`---MTP
`__._ 101
`
`70
`
`50
`
`~-:a :c 30
`
`c
`0
`
`.E
`'if.
`
`30
`Compounds IJIMJ
`·e. MTP activity
`
`40
`
`AGl..S17 (JIM)
`
`.-_;:
`
`. .J
`
`100
`c 80
`o·.: .
`. ~·80.
`··;c..
`.
`·:c .. 40
`. .E
`·~
`
`..... '
`
`40 ..
`
`.
`
`.
`
`-
`
`: . .
`
`1:0.
`20
`ccu11~ounds';JJ.&NiJ ·
`AolnU; I: . Eff~ct of ·different compowids on µ)i..-MTP interac(cid:173)
`tions and MJP's lipid transfer acthity. (A) Effect on LDl-MTP
`. binding:-.. He~rodimeric MTP. (1 µg/well) was immobilized am.I
`incubated in aiplicl1~ with LQL ( 1 l'WW.Cll) in the p~e~e aild
`abs~nce .of..variou~ ·concentratiom;"Of different compounds for 2 h
`~ d~bet,hm®r Ml1teri~~ llJld _Meth9ds. ~ Ll;>L bound to~
`w:tS q1iantilatecl by ELISA. Controls did.not contain any drug but
`did contain dinii:tltyl. sulfoxiµe'. The Qata an; rePf~ntative·or six.
`.independent. cxperiiil~Tits-.. (B) Effect- on MTP's lipid: irahsfer
`activity. MTP:s. triglyceride transfe.r· activity:.was determined in
`triplicate using purified MTP (I ;1yassay) and synth~ic lipid
`vesicles in the presence and absence of the indicated concentrations
`of different compoundS as desdibed .under Materials and Methods.
`The data':ai-C representative of three independent experiments. Mean
`±. standard deviati6ns are plotted.
`.
`. .
`. ·.
`·..
`..
`.
`we sc~cened ·.several· compounds. that might inhibit these
`interactions wi~hout affecting the llpid transfer activity of
`MTP. Representative data for three compounds are presented
`(Figure 1 ): In the absence of any compound,. 15.5 ± 0.8 fmol
`ofLDL was bounq to immobilized MTi>, which was similar
`to that reported previously (19). AGl-3 and AGl•Sl.7
`inhibited 72 and 71 %, respeetively, of the binding at 40 µM
`(Figure IA). In contr.1.st.. BMS-200150, an inhibitor ofMTP's
`lipid transfer activity, ·did not affect LDL-:-MTP binding.
`Similarly, other MTP inhibitors, CP-10447, BMS-192951,
`and BMS-197636, did not affect LDL-MTP binding (data
`not shown). Next, we evaluated the effect of these cori1-
`pounds on MTP's lipid transfer activity. As expected, BMS-
`200150 inhibited the triglyceride transfer activity (Figure IB).
`AGI-3 (40 µM) inhibited 32% of the activity at this
`concentration, In contrast, AGI-S17 inhibited neither the
`triglyceride (Figure· lB) nor the phospholipid (data not
`shown) transfer activity of MTP. These studies suggested
`that AGI-3 inhibit~ both LDL-MTP interactions and MTP
`
`7..5
`:e: :.
`'O _.., .. so
`··:s··.·
`.. ;J::: .
`.E.
`'if. 25
`
`·40
`
`20
`AGl-517 [µM)
`F1GL'RE 2:·' Effect of AGl~S17 on 1he ~inding of different apoB
`polypeptides to MTP. (A) Effect on apoBiOO,binding. Serum-free
`conditioned medium (24 h) obtained from ·Hepq~ cells that secrete
`only apoB J 00 .was incubated in· tripli.ca1c ~ith. iminobilized ·ID I
`. or MT.P for 2 h in the presence of the :indicated .. concen1ralions of
`· AGl-Sl7. ApoB bourid was quantitatcdby ELiSA. and.1he.percent
`inhibition was plotted. (B) Eftect on·apo818 .. binding. MTP was
`iinmobilized and incubated with:coriditioned media obtained from
`McA-RH7777 cells stably transfect~cfwith apoB18 in.the preilence
`of various <..'Oncentrations•of AGl-SJ7 for 2·h as described under
`MatCri'ais and Methods. The np0B.h0und ~va." quantitaled.by 'ELISA::
`E.ach point [!:prcscn.1s a meurj ± standard deviation.
`
`activity~ whereas AGl-S.17 inhibits LDL-MTP binding
`without affectfog MTP's lipid transfer activity.
`To detennine the specifi<;ity of the effect of AGI-Sl7 on
`
`the· binding of apojJ to MTP: we. studied the effect of AGI(cid:173)
`S 17 on the. binding of. apoB polypeptides to· MTP and a
`monoclonal antibody, ID l °(Fig.ure · iA). lDJ . recognizes
`nmino acids 474-539 in apolf (31, 32), and· this epitope
`overlaps wiih an MTP binding site, amino acids 270-570,
`in apoB (21). AGJ-S17 (40 µM) inhibited the binding of
`apoBIOO to MTP by abOut 60%but had no inhibitory effect
`on its binding to IOI (Figure 2A). A likely explanation is
`that AGI-S 17 interacts wirh MTP, but not wirh ID l. and
`inhibits its binding lo apoB.
`To examine whether the continuous presenee of AGl-Sl 7
`is required for the inhibition of LDL-:--MTP binding, the
`immobilized MTP was preincubated with AGI-Sl 7, washed,
`and then incubated with LDL. No inhibition for the binding
`of LDL to MTP was observed (data not shown), indicating
`
`3 of 8
`
`PENN EX. 2151
`CFAD V. UPENN
`IPR2015-01836
`
`

`
`Physiological Significance of ApoB-MTP Binding
`
`Biochemistly, Vol. 39, No. 16, 2000 4895
`
`................................. ___ __; _______ '--......_ _______ ~
`
`that the presence of AGl-S 17 is necessary for the inhibition
`of LDL-MTP binding.
`Next, we studied the effect of AGl-S 17 on the binding of
`apoB18 to MTP (Figure 2B). Among the several Lruncaled
`fonns of apoB tested, maximal binding to MTP was observed
`for apoB 18 (19, 20). A progressive increase in the concentra(cid:173)
`tion of" AGI~S 17 resuited. In increa~e<l inhibition of ap(i.B 18-
`MTP binding. Al 40 JlM, it inhibited 60% of the binding
`between apoB18 and immobilized MTP. We have previously
`shown that AGI.,S 17 inhibits the biiiding of a FLAG/apoB
`chimem expres~ing amfoo acids.8:270-570 to MTP but not
`to M2, a monoc1on~I antibody that recogni1..es .the FLAG
`epitope (21). These sn1dies showed that AGl-SHspecifically
`inhiQits the binding of different apoB polypeptides to MTP
`but does _noi affect their bfuding to lDl..Since {Qe binding
`epitopes or' IDI and MTP are overlapping (21, 32), it is likely
`·that AGl-Sl7 interacts with MTP and prevenls the binding
`of different apoB p01ypeptides to MTP.
`·
`·
`Iii Vivo ¥!feet of A_Gl"SJ7 on the Upid Transferand ApoB
`· Bindi1ig Activities o/MTP:To study tht! effect of AGf-Sl7
`on MtP's li1M·iransf'eractivity. FiePdi c~Us ·were.incilbated
`with diffe'rerif concentratioils of AGl-S i7for:6 h (Figure.3A):
`.· C.ell.Jys~~ -vi.er~.µ~ l() ass~y ~ot ihe_Hpl4 .lrii;isf~i;;a~tivity
`/:.
`(9, 26)} Jncubatioii of. cells withdiff~ren(c9neerifrauonS"of
`. ~f;·
`·. ~,:·· ... : AGi:-Si.7 bad 'lio.e(fect on ¢e. uigly.ceride li:Msfer activity
`· of 1hich>soinar-M1i> '(Fig~re'.jAj.'"sltDifai-i:Y: .. ptfos~lloHplCi
`· ~,..
`~·~_.'·
`transfer ·a~vity· in ce!1Jy$ates wa~ not :affe'Ct~ .. (data n()I
`.siio·wn): As ex~cte~ incu.6ation of. ceJl lysates ~1th .QMS~
`~·
`200150 resulted in signilicanl inhibition of the Jipid-.transfer
`l(
`activities (data not sJiown): These studies show that AGI-
`2t
`SI 7 has no effect ()n cellular lipid transfer activity.
`Next, we studied the effect of different concentration~ of·
`AGl~S17 on intracellular apoB-MTP binding (Figure 3B).
`.Ceils· were incubated w.iui different concentra~ons of AGI-
`S17 ·for 6 h, and· MTP was _irri'11unopredpitated under.
`nondenaturing conditjQris O 7, I 8). Iinmunoprecipitates were
`. ~:ro:.1e~1~nte'1:::!~;~.;~~~~1::1:~~1~~~~~:~
`
`polyclona.I antlbOdies:(Figure 3B). Tre~trii~nt ofHepG2.cells
`with ~O and -40 i1M. AGI-Sl7-_re~mlted i1111marke.ci decrease
`(so:..:so% innihition) 1n the· cOiinmunoprecipitation of ap0B . ·
`with 'MTP .. These studies indicate that AGI-Sl7 inhibits
`
`f ~:· mi;;::.!u!/ :~~7~:,~:~:~etion of ApoB by HepG2
`· 1·..
`·.::,,[_ff·~··,,:~-.. ·_:_-..
`~~i~2~~~~~~"~~:J!~~£:E:~:p:;r~
`cells were labeled with pss]methio~ine for' 6 h in the.
`presence and absence· of AGl-Sl7, and the secretion of
`. . . • ::\t~;~_,
`~~~ ~~da:~ ~~~~:)~~: :::~~~:: ~:~~~~n=d~::
`;1£~~E~i:~~:r,~;~::~E~!::r:~:
`',I :J~~~s~i~fi~j~;;~~:=~~ ~{;::EE!;
`
`Ce.tis: We used two indei>endent approaches to study the
`
`under similar conditions (//, 33-36). Different concentra-
`tions of AGI-Sl 7 (0-40 µM) did nolaffect the incorporation
`of radiolabeled amino acids into newly synthesized or
`
`experiment, the amount of apoB or albumin secreted by
`
`'._:: __ ,,_: .. _·:_:.·:_.··.:_·
`
`;:.
`
`.•. : ' ; ;
`
`. . ,f .. ,;_:::·._,;·· .. ::·'._: __ ,:·
`:;\
`
`· frt
`
`:_,.rf,,!1 .. ·-•·-•·
`
`..
`
`·:~;.
`
`• . f.?,·.·.·.·.·.:.··
`
`:.:,'._,:!:_.::_:_:··.
`
`. . .
`
`:··
`, .... ·
`·.~.!.!..'.·.·_,
`.;:
`~: ..
`t .. \!!.·~--
`~-
`,_,..
`. I'.'·:
`.,,,,
`
`A. Microsomal Triglyceride Transfer Activity
`
`0.05
`
`en
`.. ~
`~ 0.04
`.!:-..
`I/) c 0.03
`;$
`0 a. 0.02
`!:; 0
`.....
`~ 0.01
`
`0.0
`
`30.0
`20.0
`10.0
`. AGl-817 [µM]
`
`40.0
`
`B. Coimmunoprecipitation of ApoBIOO with MTP
`
`- - : ..
`
`...
`
`100
`
`::=:.::: .:.
`.... •.,'.' . . . "··.
`~:~:: ;: .. ~ :
`::;4;~~\~!· . ' ::fr~~:~~::
`:. :. :.'.[;Q::•; • :;;:.;·::::' . . .:.~:.~ .. '.:;·~-~~.: ___ :~_:
`~
`. ,:E~;;
`'• . ···;; . .;·- . . ·~:·,..
`'{:·~~-·~.-~
`Q..L-.i:.:..;;....;..J"'-,-L·~~-:~_=~~~~·:~·l-~;:.~·.....i.:..;...;.-L--"'-.;..;.__'--
`20.0 ·, 3o.o · . ·4o.o
`.0.0.
`10.0
`·. AGl-S17 [µMI
`Flffi.!.RE 3: Effect of .'\GI-~ 17 ~ the lipid t{ansfer and apoB binding
`activities of MTP in HepG2 cells, HepG2 cells were incubated with
`· the indicated amounts. of AGI-S17_ in serum~free media containing
`I% bovine serum albumin for 6 ·h a8 ·destriDed under Materials
`and MethOds. (A) Cells were lysed as described. under Ma1erials
`and Metfi_ods,. ili1d HX> µg of protein was:,used tqr i.n vitro lipid
`· transfot.a~tivity using PHJtriolein~ Mean·±. standard.deviaiioni;, n
`. =· 3, are plotted as b¥ graphs and error bars, respectively". (B) Cells
`· \vere Jyi;ed tinder nop~nat\Jring c'ondiii.ons; a:n~ · MTP was. inµm.i(cid:173)
`nopredpitiited with 10.µL of anti~MTP anti~ies. The immuno(cid:173)
`pret.ipitates were wa8hed. eluted in san1ple buffers, scparat~d on .a
`· 5% polyacrylamide gel; tr.msferreq·,to nitrocellulose, and-.reacted
`with anli·h.uman npoB 11ntibodies;. ~d bands were visualized using
`chemiluminescence detection' reagents and fluorognipby. The bands
`were 'scanned; qmintified u'sing a den!!itpnieter;'arid plotted=asbar
`graphs. The data are a representative of .two independent experi ·
`mems: ·
`·
`·
`·
`·
`·
`
`apoB 100 withou_t affecting the secrelion ·of ajbumiri (Figure
`4A). The experiment was repeated using triplicate wells.
`A(JI-S17 (40 µM) inhi.bited apoB· secretion by 61 % ± 9%,
`but had no effect on albumin secretion. These studies indicate
`that' AGI-S 17 inhibitS nascent apoB secretion.
`1n· the second approach, we quantified the effect of AGI(cid:173)
`S 17 on the ~retion of total ma8s of apoB by ELISA (Figure
`4B). Total protein ·content.in HepGi cells was not affected
`by incuba~on. wiQi AQJ-S17,_ but apoB ~ecretion was
`inhibited by 73% at 40 µM {Figure 4B). Taken together,
`Lhese studies show that AGI-S 17 specifically inhibits apoB
`secretion.
`To investigate further the effecL of inhibition of apoB(cid:173)
`MTP biiiding on the flotation properties of secreted apoB,
`conditioned medium· obtained from ce1ls incubated either
`with or without AGl-S 17 was subjecte<J to a density gradient
`ultracentrifugation (Figure 5). Jn the absence of AGl-S 17.
`HepG2 cells secreted apoB mainly as LDL-size particles.
`
`4 of 8
`
`PENN EX. 2151
`CFAD V. UPENN
`IPR2015-01836
`
`

`
`------------------------.-----·--···-······· ······························--------------'·----..,..,;
`..
`
`. ;J
`
`4896 Biochemistry, Vol. 39, No. 16, 2000
`
`Bakillah el al.
`
`A
`
`ApoB
`Albumin
`
`0 20 30 40
`AGI-Sl 7 [µM]
`
`R
`
`100.----------------~
`CJ Cell protein
`-ApoB100
`
`75
`
`. .,
`c:
`0
`:0
`:c ·50
`·c:
`
`~-
`
`25
`
`.
`
`. 30;0·
`. . .
`. .·
`AGl-517 [µM]
`.
`....... .
`f'.19uRE 4: Effect of AGI-S 17 ori .the ~eereiion of apoB pol}tpeptides
`hy HepG2 cells. (A) Effect OJ) tli'e secretion of naseent upon. Cells
`were incubated with different concentrations of AGI-Sl7"in scrum-.
`methionine-; ·and ~ysteine-fiee inCdium conwiling i % bovine serum
`albumin and I 00 µCi/mL Expre35S1~S for: 6 h. ApoB and albumin
`were irnmui•oprecipitnted sequentially. using j>olyclonal antibodies;
`separatt:d on polyacryla_m_ide gels! and _exposed to Pbospl\orhnager
`sereetis. (B) Effect on the secretfon of total apo{J mass. HepG2
`ceHs·. were incubated "inJriplicate ·with different ·concentrations o'
`AOI~Sl 7 in . serum-free mediuin containing I% bovine serum
`albumin for 6 h. ·con_trol i:eUs rei:eiyed .4imethyl sulfoxide on)y.
`Ap0B in the conditioned media was meitsiired by ELISA. Thednta
`are· representative ·or. five indcpendenl experilnents.
`
`LDL
`
`HDt.:·
`
`o Control
`· ;i, BMS,200150_
`• AGl-S17
`
`4
`
`o~~~~'-r~....:....~99i*H~
`0 2. 4 6 8 10 12 14 16 18 20 22
`Fractions
`FIGURE 5: Effect of inhibition of MTP activities on lhe secretion
`of apoB-containing lipoproteins. HepG2 cells were incubated with
`either dimethyl sulfoxide (control), AGI-Sl7 (40 µM). or BMS-
`200150 (10 µM) in DMEM containing 1% bovine serum albumin
`for 6 h. KBr (I g) was ooded to the conditioned media (2 mL).
`sequentially overlayed with 2, 2, 2, arid 3 mLofl.21. 1.063, l.019,
`and 1.006 glmL density solutions, and centrifuged (SW41 rotor.
`40 000 rpm, J 2 °C, 24 h), and different fractions (0.5 mL) were
`collected from the top of the tuhe. ApoB was measured in each
`frnction by ELISA. The distribution of apoB in different fractions
`is presented.
`
`This is in agreement with seveml published studies (29, 37-
`39). In the presence of AGI-Sl7, total apoB secretion was
`
`inhibited by 84%. Under these condilions, no LDL-size
`particles were observed. However, reduced amounts of HDL(cid:173)
`size particles were secreted in the presence of AGl-Sl7. We
`also studied the effect or BMS-200150 on the flotation
`properties of sccrctt.~d lipoproteins. As expected, BMS-
`200150 significantly decreased (73% inhibition) apoB secre(cid:173)
`tion. More important, tile major effect"-was on ihe secretlcm
`of LDL-siie particles. These studies indicated that inhibition
`of either apoB-MTP binding or MTP's lipid transfer activity
`results in decreased secretion of apoB-contaiuing 1ipopro(cid:173)
`teins. Thus, the phenot)ipic effect of inhibiting two indepen(cid:173)
`dent functions of MTP is very similar.
`
`DISCUSSION
`
`·We. have identified a novel antagonist, AGI-Sl7, that
`iJlhi.bi~ed apoB--MTP binding but did not aff~ct MTP's lipid
`transfer· ~ct1vily-.(Figure l). Treatmen~ of H~pG2 cells with
`. AGl:·Sl7 had·oo effect Qn intqicelluliir lipid transf~ activities
`-(Figure :3A)~ HQwe\Ter,. it .. dec~eased: the ait:iourits ·or a~B.
`·cqi.qin)uriop~ipita:te<t, with.MTP (F,igure 3B). Fm1hermore.
`:ii d~~rea~ci aj,oif seeretion iJ1 ·hepat6~1a. cells. ,(Figure 4).
`The. cxt~nf. of)n~1ibition .by A(JI~s 17 (40. /lM) on· the .
`. cqi.ffiiDupopreCipitatio11 .. of ap0B ..yhti Mtf (Figure 38.) .·and
`apoB:_secretioi1 .(Figure 4B) by Hep02 cells. was ·similar .
`Thus,: A,GI~~i'7 most .likely· _inhibit~ apoB · secretion· by
`inhibjtiT1g inLrncellu_lar apoB-MTP ·bin~ing m1d may repre(cid:173)
`sent ·a . riovel class ·Of compounds that can . be .useful i_n
`deereasing LDL levels in certain hyperlipidemia~ Independent
`of their etiology.
`·
`. .. Different experimentai . cond~tions arid inhibitors have
`different.effects on lipid transfer and apoB binding activities
`. of MTP. Previously, we had deJllOOStrated that immobil.iza~
`tion of MTP _r~ult~ _in :::::80% loss. of its J~pid.transfor. acti_vity
`(19); whereas the binding of immobilized and soli.1ble MTP ·
`t~ apoB. peptides was _siJllilar (21 ). These ~t!ldies suggest _that
`immobilization of MTP results in p~ial loss (if the lipid
`tnmsfer acti,•ity but has little effect on apoB_l?inding. lT_l the
`present study, we have demonstriited that BMS-2001.50, an
`inhibitor· of MTP's lipid transfer activi_ty, has· no ~(feet on
`a1>9B-:-.MTP _interactfrms, <Fi~ jA).· Simi_larly. :.A.d_1~s11
`inhibits apQB-MTP binding. bu"i.J1as ri<i eff~ oi1.the lipid .
`ln1itsfer activity· ·of 1YfrP (Figm:es l and J). Thus. the
`heterodinieric MT}> conlains a·t least two independent
`fUnctiom!l dom1,1ins that_ are tnvolved in apoB binding and
`lipid transfer activity, re~pectively.
`The present study: shows that AGI-S 17. decreases apoB
`secretion, indicating that apoB-MTP binding is physiologi(cid:173)
`cally imponan·t for its secretion. In ~nother study, mutation
`of an arginine residue that disrupts a salt bridge within apoB
`hai; been shown to decrease the binding of apoB to MTP
`and its secretion (22, 23). In both these studies, the lipid
`transfer activity of MfP was not affected. Thus, the assembly
`and secretion of apoB-containing .lipopruleins may require
`both apoB binding and lipid tr-ansfor activities of MTP. Both
`of these activities may function independent of each other.
`It is well-known that MTP's lipid transfor activity is
`necessary for the lipidation of nascent apoB and lipoprotein
`assembly. What would then be the role of protein-protein
`interactions between apoB and MTP? It has been suggested
`that the nascent apoB interacts with the· inner leaflet of the
`endoplasmic reticulum membrane (40}. ·It is proposed that
`
`5 of 8
`
`PENN EX. 2151
`CFAD V. UPENN
`IPR2015-01836
`
`

`
`·······'·'·-------·· ... ,..................
`
`.. ........................ --'-------~ -·-··---····-·------------------------co
`
`Physiological Significance of ApoB-MTP Binding
`
`Biochemistry, Vol. 39. No. 16~ 2000 4891
`
`ID:+ 6-~-
`MTPwith ~
`
`A.
`
`Rough ER
`with nascent
`ap0B
`
`apoB and
`lipid
`binding
`domains
`
`11t + x•~-
`
`Lipoprotein
`~ Assembly
`
`itin
`
`0
`
`.
`
`~Degradation
`
`),
`
`-
`E
`
`. ApoB binding domain··
`Lipid binding domain
`
`B. tit:+ ey~&--S~····~3. ··.~.·
`IW
`Rougn eR MTPw.~h "·. :~ · · · ·
`. ,with nascent rP,~Band· .· .
`~p_o.B
`. bf~ding: ·
`. sites ·
`
`: _.
`
`;_ MTP.
`
`.
`
`.
`
`·":' ··1· " .. '.
`.. · .. ~4-.· .
`.
`.
`.
`
`~ Core e~pansion @
`"f2}
`f!f)?
`
`5
`
`·..
`
`.~. • .. •~
`
`~
`
`Nascent
`Primor~ial
`lipoprotein
`lipoprot~in
`FIGURE 6: ProposCd roles'. for apoB-MfP binding in lipopr~tein assembly. (A) Role of apoB-MTP binding. First line: The nascent llpoB
`polypcpridc is shown associated Y1ith,the inner leaflet of the ER membrane;. and MlP ._bas been .depicted to contain two i1.~ependenr; lipid
`transfer and apoB bindiµg. domains. When_ the nascenr polypeptide con':3lning _the M1J> bind,ing .,site emerges towanj ~-e lumeila) .side of.
`the ER. it interacts with MTP.;A successful binding between these proteins will re.~ult in complete translation, translocadon into lumen, and
`assembly of apoB·into lipoprotein particlc:s. Second line: .. If.the binding between these prot,e~ns is inhibited (represented bY. "X''.) by the use
`of ·inhibitors, such as AGI-$17,. oi-lmitatioils..in MTP bfodiiig site.~, then the ttansfer..ofapoB from the endoj;las#lic retjculat~membrime to
`the lumen may be inhibited. However, the translation of.apoB ·mRNA may .conti!'lue. The !)&seen! apoB ~Jyj:leptide will then be exposed
`to the cytopl~ic.side of the ER and will bind. to cytoplasmic proteins, such OS ubiquitin.and .l:ieat shock. protein 70 (hsp70), !ind· be
`targeted for degradation by p'roteosome.o;. (B) Hyjiotheticafsteps in lipoproteiri assembly. /. MTP biriding. MT,P may first interact with(cid:173)
`nascent apoB via its apoB binding domain. The lipid. lnmsfer, domain may or may nqt contain lipids. These i'nteractioos may he