CA2033182A1 - Nonglycosylated human interleukin-3 analog proteins - Google Patents

Abstract

TITLE

Nonglycosylated Human Interleukin-3 Analog Proteins ABSTRACT

The present invention provides a pharmaceutical compositions comprising as active ingredients certain truncated purified IL-3(Pro8 Asp15 Asp70) analog proteins expressed by transformed yeast of the species Saccharomyces cerevisiae, which when administered to a primate do not result in detectable urticana.

Description

2~33~2 TITLE
Nor~lycosylated Human Ir~terleukin-3 Analr~ Protains ~:
E! ~ CKGROUt~D OF 'rHE I~JVENTII~
~p~lion~bte~ era-ytobn~ines, auld~yto pharm2~1 ~dions colr~ris;ng sebct~ ~unca~ed nonlalyoosyla1ed anal~
ir~e~n-3 ~L~) p~otoins wh~ ~xhibit ben0ncia) cSfical and toxicob~ical properUes when cornpared to hulL-3 proteins produced in E. co1i.
The dilferentiation and prs~eration r~ hernatopoietic cells is r~ulatod by sscreted ~co~t~ns con~c~ve~ known as cobny ~Urnui~tir~ lactors (CSFs). In r~rine and human ~ems, ~e p~teins m~ ~anubcyt~rr~ro~ bny ~bmulatin~ factor (GM^CSF), p~es El anub~e and ma~oph~e ~n ~T~m r~nr~1 b~ne marrow, and which a~o appears1~ re~late the activny ot matuTe, ~merenniated ~ranulocytes andmacrophages. Oth~r CSFs include macrophage CSF (M~SF or CSF-1~, which ir~uces the ~elecU~e prolUera~ir n of macropha~es, anci granulocyte CSF (G-CSF) which induces dsvetDpment o1 ~ranubc~e pro~enltors Irom bon0 rnarrow precursors. An addilional CSF, isolated firsl In ~rine systems and n~re recently ~rom human cell sources, has been deslgrlated IL-3 or ~u~SF.
~wine JL-3was orl~inaHy identlned by Ihle 0~ al., J. In~munoL 126:2~8~i (1981~ as a lactor which induc~J e~pression ol a T cell associated enzyme, 20e-hydroxysteroid dehydro~enase. The lador w~s pufi1iad to hom~eneily and shown to re~ulate the groynh and ddferentiation ot rumerous subclasses ~ ear~ h~ opo;e1ic and Iymphold pro~enitor cel~s. cDNA cbnes correspon~ng ~D mufine ~ were ~st i~ola~ed by i ung e~ al., Nature 2~ t984) and Yd~ta et ~., Pr~c. ~tL Acad S~:i. USA ~1:1070 (19~4). Gibbon and h~man genon~c DNA hom~bgues ot me ~ -3 sequence W8rel discbsed by Yang e~
al., Cell47:3 (19B6). The human sequence reported byYan~ et al. included a serine residu0 a~ position B o~ the mature proteln sequence. FollowinQ this 1indin~ lhree ~roups reported Isolatlon ol pro8 hulL-S cDNAs, Including Dorssers e~ al., Gene 55:115 tl987); Otsuka et al., l h7n7~ ~4~8 ~1988~; and G~is et al., Behfing Inst. Mit~. B~.1 (1988).
~ sunre~ ~ ~duals ~o dete~mlne the lrc4uency o~ tha ~roB allele was reoorted by sc~al.,sup~a. b~w~e~rased~ainrflactionwasernployedtoamplifylhe .
DNA sequenoes tlankir~ ~e p~Uon B b~ adiolabeted oli~onucleD~ide probes cQm~lemen~ tollle Ser8 and pro8 10nns were then used ts pr~be arnplified DNA isola~ed 1rom thirteen gen~0cally vMelate~ hdvid~a s. A.T p~tied DNA was Immobilized on nitroce)~bse and analyzed by hybridization under conditions ol increasing stringency. The results indicated tha~ o~ ~he 13 persons examined, each was po~Uive lor DNA er coding the pro8 version ol hulL-3. Three were also positive lor Ser8 hulL-3 at this locus (i.e., were . .
. ,., . ~ ~,. . . .
.. ~ . ~ , . ... -. . -. . :
: . .. .
: . . .
... ... ~. . . . . , -2 ~ 2 hetarozygoles), indicatirlg a moderate Isvel of polymorphism o7 the human IL 3 ~ene in this test population.
Preclinical and clinical studies of human IL-3 proteins produced by various manutac~urers h3s reve31ed su1prisin{1 differer~es in clinical utilny and toxidty. Deveiopmen~
5 of non-toxic, tolerable iorrns of hurnan IL-3 for the7apy is important because recorrbinant human IL~ appears to ba the first cytokine capa~ irmJ ~tir~ gr~rwiopoiesis, arythropoissts and, most Irr~oTtan~, 1hrombopoiesis in vivo.
Severe riL-3 toxlcity was reported i~y Valent et al. at the F-lrst Intemational Syrnposium:
Cytobnss in Hemopoi~sis, Oncology andAlDS, held In Hanover, Federal Republic ot 10 Germany, June 14-17, 1i 98g. A comi~sition r~ hulL-3 (Ser8) produced in E. cob was svalu~ted in v~vo by adntinistra~i~n to rhesus monkeys ~n=10) ~hrice daily at different dosages , 33, and 1~ ~Ighg/day, s.c.) 10r 14 days lollowed by consecutive GM-CSF application (days 14-28, 5 llg/lcg thfice daiO. All monkeys responded to rhulL-3 by a 2-3101d whi~e blood cell (WBC) increase by day 14 due to an increase in eosinophils and basophils. Intracellular 15 histamine tlH) in WBC was lound to risa continuously atter IL-3 treatmcnt until day 8-10. This excess of IH was accompanied by a rise in plasma h~stamine vaiues and an urticaria-iike exanthem revealing innitra~ion ot mast cells and Iymphocytes In~o the demmis observed in si<in ibhpsy specimsns. The cl)nical s~de eHec~s o~ ~h~s rhulL-3 composi~on were attribu~ed ~o basoph~Umas~ cell act~vat~on and histam~ne-produc~ng et~ects ot this cytoklne.
In contras~, the protein compositions ot the present inventlon did not sxhiblt any detectable urticafia or stimulate infiHration ot mast cells or Iymphocytes into the dermis when administered to cynomolgus monkeys or human patients. The resuits ot these toxicological and clinical studies are reported in the Examples.
SUMMARY OF TiHE INVENTION
The present invention provides a pharmaceutical composition comprising as activeinsredien~ e or rnwe more ipuriDed recombinarlt human IL-3 ~pro8 Aspl 5 Asp70) analog proteins expressed by a t~anstormed y~ast host of the species Saccharomyc~s cerevisiae, said proteins salecled from the group consisting ot:
~a) i~et3 rhulL~ Ip~o8 Asp1~ Asp7û);
~ Tt~r4 rhulL 3 (pro8 Asp15 Asp70); and (c) Thr6 rhulL-3 (pro8 Asp15 Asp70).
Such composiUons, when in~used into a p~rnate, do not resuit in obsentable urticaria or intiitration of rnast cells and Iymphocyles into the dermis.
: .
DETAILED DESCRIPTION OF THE INVENTION
The hulL-3 protein compositions of the present invention comprise selected truncated hulL-3 analog proteins which vary slightly at the N-terminus trom the native protein.
Such tn~ncated proteins were observed as componen~s of mixtures of proteins which 2~3~2 .

resuHed Irom expression o~ cDNAs encoding mature hulL-3 proteins In S. c~revisjae, presumably as a result ol proteolysis by the yeast expression host. These mixtures can co T prise one or rnor~ human IL-3 (pro8 Asp15 Asp70) ana!og proteins having N terrrlini at Me~ Thr4 or Thfi of the following amino acid sequence (the anernative N-termini are 5 underlined)~
~}a Pr~ ThL ~ln 5h~ ~hr Pro Ieu Ly~ Thr Ser Trp Val A~p r~ Ser A~n ~e~ Ile ~p Glu Ile Ile Thr ~i~ ~eu Lys Gln Pro Pro Leu Pro Leu L~u A~p Phe A3n A~n ~eu Agn Gly Glu A~p Glu ;~
~p Ile Leu Met Glu Asn Asn Leu Arg Arg Pro A~n Leu Glu Ala Phe Asn Arg Ala Val Lys Ser ~eu Gln Asp Ala Ser Ala Ile Glu Ser Ile Leu Lys Asn L~u Leu Pro Cys Leu P~o Leu Ala Thr Ala Ala Pro Thr Arg His Pro Ile His Ile Lys Asp Gly Asp Trp Asn Glu Phe Arg Arg Lys Leu Thr Phe Tyr Leu Lys Thr Leu Glu Asn Ala Gln Ala Gln G~n Thr Thr Leu Ser L~u Ala Ile Phe.
The lerm rhulL-3 (pro8 Aspt5 Asp70) rneans a recombinant human interleukin-3 polypeptide having a prollne residu0 at position 8 and asparagine residues at posUions 15 and 70 wherein residue 1 refers lo the ~irs~ amlno acld o~ ~he mature naUve or wllct-type human Irnerleuk~n~3. The amlno acld sequence o~ such a polypeptlde is set ~orth abov0. The various N-terrninal sp~cies are Indlcated by designa~ng lhe N-t0rminal resldue b0lore lhe term rhulL-3 0.9. Met3 rhulL-3 (pro8 Asp15 Asp70). The analog protelns of the presentinvention have an aspartic acid (Asp) residue substituted lor each asparagine in the N-~lycosylation sites presenl in lhe native human IL-3 sequence to preclude N-~lycosylation by the yeast host at these sites.
A DNA segmsnt encoding human IL-3 was isolated 1rom a cDNA library as described in lh~ xperimental section below. Site-speciflc oligonucleotide mutagenesis was then p~rlorrned lo provWe a cDNA encoding human ~L-3 (pro8 Asp15 Asp70). Th0 resuHingaHered cDNA was 0mployed to constnuct a yeast expression vector wh~ch was used to 1rans~orrn an appropriate yeast expression strain which was grown in cuHure under conditions prorno~lng derepression of ?he yeast promoler. Th0 resuHlng protein was then purifled by a colTbination o1 teversed-phase hi~h-i~erforrrlance liquid chromatography (RP-HPLC) and ion-0xchan~e chrornalography to provide a purmed product. Assay o~ th~s product using a human -bone rnarrow proR~eration assay and hu~nan IL-3 t~ceptor binding assay confirmed express~on - 35 of an ana~ prod~ct sud1able lor human pl~arrnaceutical use. N-terminal sequencing of the pr~ct rr~ea~ed a rrmtu~e ai ana~og prot ins which varied sfightly from iot to bt. To obtain a composRion consistin~ essentially ! a single species Met3 hulL-3 (pro8 Asp15 Asp70) a second expression vector was constructed wherein lhc sequence encoding the yeast ~ ~ ~

:~ ' ~ ....
~, . ,- -. . . . . . . . . . . .

secreUQn leadsr was diractly tused in Irame to a construct lacking the codons specifying the first two amno acids ot the ma1ure protein.
Preferably, lhe compositions of the invention consist esserltially of M~t3 hulL-3 (pro8 Asp1~ Asp7~). However, ~ an exp ession v~ctor encodin~ the fulHength rnature S protein is employed, a mix~ufe wlU ~ obtained ~omprisin~ Ala1 h~ (pro8 Aspl5 Asp70) in adrrlixture with one or more proteins selected trom ~t-B foregoino ~up, typically corr~rising:
(a) from about 30 to about 40 percer~t A~a1 hulL-3 (proB Asp15 Asp70);
~b) trom about 10 to about 30 psrcent Mat3 hulL-3 (Pro8 Asp15 Asp70); and (c) a rernainder iraction ~ubstantially consistin~ o~ a rnixtur0 ot ThP and Thr6 hulL-3(Pro8A~p15 Asp70) Pr~er~bly, the cDmposilions ot lhs invention exhibit a spscific biological activity in the human bDne maTrow assay of trom about 1.8 to about 7.0 x 107 U/mg~ and preferably 4.0 to about ;
7.0 x 107 U/mg, and a binding afflnity ~or human monocyte IL-3 receptors, expressed as an inhibition constant, of from about 2.0 to abou~ 8.0 x ~0'~ M-1, pro1erably from about 4.0 to .
about8.0 x1010M-'. Itisunders~oodthatminor~ractions(lessthan1 mole%)ofrhulL-3 ~:
protelns havin~ other N-terminl may also be present in the mixtures o~ the pres~nt Invention.
1. Assays ~or hulL-3 BiolQnl~al Activitv Assays used to measure hulL-3 biological activity are described below.
A. HumanBoneMarr~ml~
Freshly isolated human bone marrow cells are preincubated lor 2 hours at 37C, 5%
CO2, In Ussue cuHure flasks containing 2 x 1 o6 calls per ml prewarmed, pregassed senum-free RPMI1640 medium (Gibco, Chagfin Falls, OH, USA) rontaining 50 units/ml penicillin, 5û ~ml streptomydn, and 300 llg/ml fresh L-glutarrline (hereinafler assay lT edium7. After preincubation, nonadherent cells are removed by pipening the media gently over the surface of the flask. Nonadharent cells are collected by centrifugation at 1000 rpm for 10 minutes at 4C, resuspended in a small volume ol assay medium con~aining 10% ~e~al bovlne senum (FBS), and counted usin3 Trypan blue ~or viability and Turks stain for rscovery of whit~ cells.
Cells are kept at about 4C in assay medium contalning 10% FBS until added to assay plates.
50 ~1 assay medium are added to each well of a 96 well ~lat f~tom tissue cuitureplate. 50 ~ of sample diluted in assay medium are added to the firs~ well of each row, and serial dlutions are rnade across each r~w in the usual rnanner. 1.25 x 104 bone rnarrow cells, in a volume of 100 11l, are then added to each well. Plates are incuba~ed for 4 days at 37C, ~% C02, in a plaslic box corlWmng sterile distilled H20 to preverlt desiccation. On day 4, 25 Jllol! as~ dium containing 5% FBS and 80 ~Ci/ml 13HI-~hymidine ~80 CUmrnol) ar~ added 1~ each weil and the plate incubated 5 hr at 37C, 5% C02, in a plastic box. Afler incubation, cells are harvested onto glass fiber mters, washed, and tested for incorporated radioac~ivity using a scintillation counter. Units of hulL-3 activity are calculated by reference to the quantity of hulL-3 which induces 50% ol maximal thymidine incorporation. For ~xarnple, if a 100 ~
'~

.. . . . . ., , . , .. .~ . .
.. , . ~ . .

: .
~ . .. , , ~ .

2 ~ ~ 3 ~
.
sarr~1e g0nerates one-half maximal thymidine incorporaUon at a dilution of 1:20, one unit is defined as the activity contained In 1/20 of 100 il~. or 5 il~- The sample would therefore aontaln 1000 divided by 5, or 200 uni~s per milliliter (Ulml) ot hulL-3 activity.
~uma[LlL-3 Recer~t~r Bindirla AssaY
In a ~0U rou~i bonon~ mluo~i~er plate, ten 2-foW dilutions of oach sampl0 are pT~par~d in i~PiUi 1~40 medium beginriin~ ab O~D75 j~gtn~ (50 ~i each). For ~ach assay plate, 1riplicate 50 ~i arlquots o~ rhulL-3 standards ~1.5 ~ ~a~ive controls ~bindir~
IT~dium) are set up. For purposes cl this assay, ~binding nædiumU rneans RPMI 1640 with 2.5% (wh) bovine serum alburnin (BSA), 0.2% (w/v) sGdium æide, ar~ 20 mM HEPES, pH
7.2. 50 il-~ ot 1251 rhu IL-3 stock (1 x 10~9 M in binding medium, specihc activity 4~ x 1015 cp~mole), pr~pared as described in PCT appli~atl~n US ~9102599, are then adcied to eachwen, foibwed by ~3 ~ ~ human monocytes (harvested from culture by centrifugation.
washed twice with PPMI 1640, and resuspended to 4 x 107 cellslml in binding medium).
Each plate is then incubated for 1 hour at 37C on a rotary shaker wi~h vigorous mixing. After incubation, cells and bound 125l rhulL-3 are separated ~rom unbound 125l rhulL-3 by a ph2halate oil centrifugation technique, conducted as follows. First, duplicate 65 il' aliquots are removed from each Incubation mixture, and each aliquot Is layeted onto 200 iul ol a cold phthalate oll mixture i1.5 parts dibutyl phthalate, 1 part b~s(-2-ethylh~xyl)-ph~halate (Eastman Kodak Co., Rochester, NY)I in 400 ul polyethylene centrlfuge tubes and centrifuged for 1 minu~e In a microcentrifuge at 8C. The cells plus bound 1251 rhulL-3 sedlment through the phthalate oil mixture while the less dense binding medium containing unbound 1251 rhulL-3 Is iefl on its surface. 1251 rhulL-3 bound to cells is measured by cutting the tubes in haH and counting top andlip portions1Or t~51, The resuits of the IL-3 receptor bindin~ assay are expressed as inhibition (/O) ca~ula1ed according to Equation 1 below:
EquaUon 1 1 (%) = 10Qx ~Ma~Te~
(M~ - Min) when Max is the level of binding ot 125l IL-3 in ~he presence of medium alone, Min is the level of binding in the presence of 1.5 ilgtml unlabeled and Test is the binciing level in the presence of test sample dilutions. Resuits are analyzed by non-linear least squares ntting ;~
Equaticn 2 bebw:
Equation ~ M (/Oj = M Kl I
1 +K-t:;+K
to the data expressed in the torm given in 1 where M (O/D) is the maximal level of inhibition, K
(M-1 ) is the binciing constant of 1251 rhu IL~, C ~M) is the c~ncentration of 1251 rhu IL-3, 1 (M) is the concentraUon ot the test sample and Kl (M 1 ) is the inhibition (binding) constant of the test 4 0 sample. The values of Kl lor test samples are scaled to that o~ an rhulL-3 standard by dividing [K /Ks] when Ks is the inhibition constant ot the standard, to yield a relative binding activity.

( ` ::

2. Protei~ExDress~ nBec~rTbinarltyeastsysterns R~oombinant Sacc~aromyces cer~visiae host cells are used ~or expression ot ~he racombinant proteins ~l this invention. Prelerrad expression vectors can be derived 1rom pBC102-Kæ (ATCC 67255) Which contains ~NA se~uences from pBR322 ~or setection ar~i repacatlon ~n E. co~ (Apr ~ene ar~ ori~in o~ replication) and yeast ~NA sHquerlces including a ~lucose-repressible alcohol dehydrogenase 2 ~ADH2) prornoter. 1~he ADH2 prorno~er has ibeen r~escribed by Russell et ai., J. Biol. Chem. 2~674 (1982) and Beier ~t al., Natur~ ~ :
300:724 (1982). Plasmid pBC102-K22 also includes a Trp1 gene as a selectable rnarker and ~ :
1 0 the yeast 2 jl ~rigin of replication. Actjacerlt to ~he prornoter is the yeast c -lactor bader r~uerlce enabling secretion ot hetero1O~ous proteins ~rom a yeast host. The cl-tactor bader sequence ~s modifiedlo contain, near Us 3 end, an Asp718 (Kpnl and Asp718 are isoschkomers) restriction site to facilHate fusion of this sequence to 1Oreign genes. A
sequence ooding for the Glu-Ala-Glu-Ala amino acids was ornitted lO allow efficient processing 1 ~ of secreted protein, as described by Brake et al., Proc. Natl. Acad. Sci. USA 81 :4642 ~1984) .
ARernative expression vectors are yeast vectors which cornprise an c~-factor promoter, for example pYc~HuGM (ATCC 53157), which bears the wild-type human GM-CSF
~ene. Others are known to those skilled in the art. The construction of pYolHuGM Is described in published European Patent Appacation No. 18S,350 (8530682.7), the disclosure o~ which ls ~ncorporated by re~erence hereln.
The choice of appropriate yeast s~rains ~or trans~ormation will be determined by the nature of the selectable markers and other Seatures ot the vector. Appropriate S. cerevisiae ~rains for transformation by expression vectors derived from pBl:;102-K22 or pYo~HuGM
irlclude strains X2181-1B, available lrom the Yeast Genetic Stock Center, Berkeley, CA, USA
(see below), having the genotype G~ trpl gal1 ade1 his~, Jt7 (ATCC 52683; c~ his2 ade1 trp1 met14 vra3~; and IL166-58 (ATCC 46183; c~ his1 trp1~. A particulariy preferred expression ~:
strain lor use with pBC102-K22, XV2181, is a diploid forrned by mating two haploid strains, X2181 -1 B, available trom the Yeast Genetic Stock Center, Department ol Biophysics and Medtcal Phys~cs, Univers~ty o~ Calitornia, Berkeley, CA g4702, USA; and XV617-1-3B, available 1rom 1he Department ot Genetics, Unlversity ol Washington, Seattle, WA 98~ 05, USA, or Immunex Corporation, 51 University Street, Seattle, WA 98101, USA. A suitable transforrnation protocol i5 thal dascfibed by Hinnen, et al., Prv~. NE?L Acad. Sci. USA 75:1929 (197~, selec~ or Trp+ translormants ~n a selective medium conslsting o~ 0.67% yeast nilroaen base, 0.5% casamir~ acids, 2% glucose,10 ~/ml adenirle anci 20 llg/ml uradl.
Host strains transtormed by vectors compfising the ADH~ or c~-~actor prornoters are ~rown for expression in a rich medium consisting ot 1 % yeast ex~ract, 2% pep1One, and 1%
glucose supplemented w-ith 8011g/ml adenine and B0 ;lglml uracil. Derepression ol the ADlt2 prornoter occurs upon exhaustion ol medium glucose. Cnude yeast supernatants areharvested by fiitra~ion and lrozen or heW at 4C prior to lurther puri~ication.

.
-- :;. : .

i: : . . . . ,. .: . :
:- . ,. :
::
-. . ~, , , ?

~, eurincaliQno~lL-3 Recombinarlt hulL-3 resuiting ~rom fermentation of yeast strains can be purHi0d by sir~k or sequential rev~rsed-phase HPLC steps on a preparaUve HPLC column, by methods anabgQus to ~ose des~ed by U~al et al., 1 Chromatog. 296:171 ~1984) ar~i t`,rabstein et al., J. Exp. MBd ~ 14~5 t1986). A pr~ierred purification protwol is described in ~sample 2, below.

Example A. Isolation o~ C~le. encodina hulL-3 Peripheral blood iymphocytes were isolated ~rom buffy coats prepared lrom whole bbDd ~Porlland Red Cross, i~ and, Ore~on, USA) by F~coll hypaque density centrifugation.
T celis were Isola1ed by rosetting w~th 2-aminoethyUhiouronium bromide-treated sheep red blood cells. Cens were cuJlured in 175 cm2 nasks at 5 X 1 o6 cells/ml ~or 18 hour in 100 ml RPMI, 10% ~etal oalf senum, 50 ~ B-mercaptoethanol; 1% phytohemagglutinin tPHA) and 10 r~/ml phorbol 12-myristate 13-acetate (i A). RNA was extracted by the guanldinium CsCI
method and poly A+ Ri~lA prepared by oligo-dT cellulose chromatography (Maniatis et al., Molecular Clonjng: A Laboratory Manual, Cold Spr~ng Harbor, 1982). cDNA was prepared 1rorn poly A' RNA essentially as described by Gubler and HoHman, Gene 25:263-269 (1983).
The cDNA was rendered double-stranded using DNA polymerase 1, blunt-ended with T4 DNA
polymerase, methylated with EcoRI methylase lo protect EcoRI cleavage sites within the cDNA, and ligated to EcoRI linkers. These construc~s were digested with EcoRI to remove all but one copy ot the linkers at each end o~ the cDNA, ligated to EcoRI-cut and dephosphorylated arms o~ bacteriophage ~gt10 (Huynh et al., DNA Cloning:A Practical Approach, Glover, ed., IRL Press, pp. 49-78), and packaged into ~phage extracts (Strata~ene, San Diego, CA, USA) according to the manutacturer s instructions. 500,000 recombinants were plated on E. colistrain C600M1- and screened by standard plaque hybridization techniques using the ~ollowing probes.
Two oligonucleotides were synthesized, with sequences complementary to selected 5 and 3 sequences o~ the hulL-3 gene. The 5 probe, complementary to a sequence encoding part ot the hulL-3 leader, had the sequence 5 -GAGTTGGAGCAGGAGCA-GGAC-3 .
The 3 probe, corresponding to a region encoding amino acids 123-130 of the mature protein, had the sequence 5 -GATCGCGAt3GC TCAAAGTCGT-3 . The method o~ synthesisw~s a standa~ ;a~mated ~nester meShod substantially similar to that disclosed by Sood et al., JU~. Acids Res. 4:2557 (1977) and Hirose et al., Te~. Lelt. 28:2449 ~1978). Folhwing synthesis, oligonucleotides were deblocked and purified by preparative gel eleclrophoresis.
For use as screening probes, the oligonucleotides were terrninally radiolabeled wHh 32P-ATP
and T4 polynucleotide kinase using techniques similar to those disclosed by Maniatis et al.
The E. co0 strain used ~or library screening was G600htl- (Huynh et al., 1985, supra).

,.. ... .. .
~: , . - :
; :

'. ' . ~ .:
-2 ~ c~3~ 2 Thineen positive plaques were purified and re-probed separately with the two hybridi~ation probes. Eleven clones hybridized to bo~h oligonucleotides. The ci~NA ins0rts 1r~m several positive recornbinant phage were sutxloned into an EcoRI-cut deriva~iv0 (pGEMi3L~8) of the standard cloning vec~or pBR322 containing a polylinker having a unique 5 EooRI site, a i9amH1 site ar~ct numerous other unique restriction sites. An exemplary vector ot lhis type, pEM~L, is describad by Dente et al., Nucl. Acids Res. 1~:1645 (1983), in which the promot~rs tor SP6 and T7 polymerases flank the multiple ci~n~ns sltes. The nudeotide se~tuences of seler,1ed clones were determined by the chain tennination method.
Spedficatty, panial EcoR1 digestion o~ IGT1 O:IL-3 clones 2, 3, 4 ~md 5 yielded fragrnents ranginQ ~rom 850 bp to 1,000 bp in size which were separately sut~c~ned into the EcoR1 site of iæEMBL18. The inserts of the pGEMBL:rhulL-3 subclones were sequenoed usin~ a uni\fersal primer that binds adJacent to the rnuttiple cloning sne o~ pGEMBL18, ar~ syrthetic oligonucleotide primers derived from the hulL-3 sequence.
Example 1: ModHication ot t~i-Glycosylation Sites ~ncoded bv hulL-3 CC~NA and Assemblv ol Ex~r~ion Vec~or for rhulL-3 ~8 AsDl5 ASD71 The two asparagine-tinked giycosylation sites present in the natural prolein (Asn1 5 and Asn70) were altered by changing the codons at these posi~ions to ones that encode aspartio acid. This prevents N-linked glycosylation (often hyper~lycosylation) of the secreted 20 protein by ~he yeast cells, and a more homogeneous product ~s obtalned. These changes were made as described below upon subclonlng the hulL-3 cDNA ~nto ~he yeas~ expresslon vector plXY1 20.
The yeast expression vector plXY120 is substantially identical to pBC102-K22, described in EPA 243,153, except that the tollowing synthetic oligonucleotide containing 25 multiple cloning sites was inserted from the Asp718 site tamino acid 79) near the 3 end ot the o~-factor signal peptide to the Spel site contained in the 2jl sequences:
A~p718 GTACC~TTGGATAAAAGAGACTACAAGGACGACGATG-GAAACCTATTTTCTCTGATGTTCCTGCTGCTAC-NcoI
-ACAAGAGGCCTCCATGGATCCCCCGGGACA
-TGTTCTCCGGAGGTACCTAGGGGGCCCTGTGATC
~BamHI SpeI

In aWition, a 514-bp DNA tragment derived flom the single-s~randed bacteriophage35 71 r,ontaining the origin ot replication and intergenic region was inserted at the Nru1 site in the pBR322 DNA sequences. The ,~esen~ Di th~ f1 o~igin o~ replication enables generation of sln~stlanded copies of the vector when transformed into appropriate (male) strains of E
coli and superintected with bacteriophage t1. This capability facilitates DNA sequendng ot the vector and allows the possibility of in vitro mutagenesis.
The yeast expression vec~or plXY~20 was digested with the restriction enzymes Asp718, which cleaves near the 3 end ot the a-factor leader peptide (nucleotide 237), and . . . . . . . .. . . . .. . .
:. . :
,, :
:::.,. : : ,. -.. .. :
.
. . . . .

,:~' . : ' . ' " " ' ., - .
. .'~ . ~, 2~33~ ~2 BamH1 which cleaves in ths polylinker. The large vector Iragment was purified and li~ated to 1he tolbwir~ DNA tragmerlts: (1 ) a hulL-3 cDNA tragmerlt derived ~rorn plasmid GE~aL18:huJL-3 1rom Ihe Cla t site (r~ucleoUde 58 oS mature hulL-3) to the BamH1 sUe (3 to the hulL-3 cl~NAin a polyfiT~ker); and ~2) ~he 1OIIowing synthetic oligonucleotide linker A:
GTA CCT TTG G~T A~A AGA GAC ~AC AAG G~C GAC GAT GAC AAG GCT CCC ATG
GA AAC CTA TTT TCT GTG ATG TTC CTG CTG CTA CTG TTC CGA GGG TAC
ACC CAG AC~ ACG CCC TTG AAG AC~ AGC TGG GTT ~AT 'rGC TCT AAC ATG AT
TGG GTC TGC TGC GGG AAC TTC TGG TCG ACC CAA CTA ACG AGA TTG TAC TAG C

Oligonucleotide A regenerates the sequence encoding the C-~erminus o~ the o~-1actor leader i~iJtide ar~ fu5ing it ir~frame to the octapeptide DYKDDDDK which is in tum hs~f to the N-~emunus ot mature r~ is fusion ~o the rilulL-3 protein allows detection wlth antiioody specific tor the octapeptide and was used initially for moni~oring the expression and purificaUon of rhulL-3. This oiigonucleotide also encodes an amino acid change at position 15 (Asn'5 to Asp15) to aHer this N-linked glycosyiation site. The underlined nucleotides in oligonucleotide A represent changes from the wild type cDNA sequence. Only the A to G
and C ~o T changes at nucleotides 43 and 45 respectively (coun~ing ~rom th0 codon corresponding to the N-terminal alanlne ot the mature hulL-3 molecule) resuH in an amino acld change (Asp15). The other base changes ~n~roduce conven~ent restrict~on sites (Ahall and Pvull) without aHering ~he amino acid sequence. The resuHing plasmid was deslgnated p~XY139 anci contains a rhulL-3 cDNA with one rema~ning N-l~nked glycosylation consensus sequence (Asn70).
Plasmid plXY139 was used to perform oligonucleotide-directed mutagenesis to r0rnove the second N-linked glycosylation consensus sequence by changing Asn70 to Asp70. The in vitro mutagenesis was conducted by a method similar to that described by Walder and Walder Gene42:133 (1986). The yeast vector plXY139 containsthe origin of replication ~or the single-stranded bacteriophage f 1 and is capable of generating single-3 O stranded DNA when present in a suitable (male) strain of E. coli and superinfecled with helper phage.
Single-strarlded DNA was generated by transforrning E coli strain JM107 and superlntcctlng with helper phage IR1. Slngle-stranded DNA was isolated and annealed to the 1Olbwin3 mutagenlc oligonucteotide B GTC AAG AGT TTA CAG ~AC GCA TCA GCA AAT
G, which provides a codon switch subsU~vtir~ Asp for Asn at position 70 ot mature hulL-3.
Annealing and yeast translorrr~tion cDrdil~sns were don~ as described by Walder and Walder supra. Yeast transtonnarns were selec~ed ~ ~rowth on rnedium lacklng tryptophan pooled and DNA extrac~ed as described by Holm et al. Gene 42:169 (1986). This DNA
contair!irlg a mixture ot wild type and mutant plasmid DNA was used to trans~orrn E. cOn RR1 to ampicillin resistance. The resulting colonies were screened by hybridization to radiolabeled .... . .. . , - . , . -,. , . ~ . , :: . . - .
: . , .
:~. . . , . . .: : -:: . , . : . :
:: ~ - . . - -. . . ,. . .. , -. .... . . . . . :

2~3:~2 ,~
oligonucleotide B usir~ standard techniques. Plasmids comprising DNA encoding hulL-3 Asp70 were identified by ~he hybridization to radiolab~led oligonucleotide B under strin~ent c~nditions and verified by nucleotide sequencing.
The resulting yeas2 e~ression plasrnid was designated plXY138, and r,ornained the hulL-3 ~ene encoding the Asp15 Asp70 amiro acid changss andl the octapeptlde DYKDDDDK a1~7e N-l~rrnirws. The f~l yeas~t expression piasmd i~ idenlical to plXY138 sxcept that it lacks the nucle~ide sequences codin~ fw the octapeptide, thus generating mature rhulL-3 as the product.
-~
Tho hnal yeast expression plasmid was constructed as descri~d b~e~w. The yeast expression vector pl)CY120 was cleaved with the restriction enzymes Asp718 anci BarnH1 as -descnbed above. Th~ larS~e vector 1ragment was ligated to~ether with (1 ) a hulL-3 cDNA
1~nentderived1romplasrnidplXY138tha~extended1romtheAhallsite(whichcleavesa ra~ciec~lide 19 d mature hUIL-3) 1r~ the BamH1 slte 3~ to the cDNA, and (2) the followinçJ
syrnhetic oligonucle~tide C:
-GTA CCT TTG GAT AAA AGA GCT CCC ATG ACC CAG ACG A
GA AAC CTA ~TT TCT CGT GGG TAC TGG GTC TGC TGC
Pro ~eu A~p Lys Arg Ala Pro Met Thr Gl~ Thr Thr O~lgor~c~eotide C re~enerates ~he 3 end o1 ~he ~-1actor bader peptide from th0 Asp718 site ~Yhe amino aclds P~-Leu-Asp-Lys-Arg) and the N-terminal seven am~no aclds of hulL~ to the Ahall site. The resuiting plasmid was designated plXY151. This vec~or, when present In yeast, allows glucose-regulated expression and secretion of rhulL-3 (pro8 Asp15 Asp70). In :
order 20 provide relatively homoyeneous compositions consisting largely of Met3 rhulL-3 ~pro8 Aspl5 Asp70), a similar expression vector, designated plXY290, was constructed using a substitute oligonucleotide identical to Oligonucleotide C but lacking the codons specifying ~ ~:
the Ala and Pro residues present in the native human IL-3 protein.
When~ranstected m~o a suitable host which is then grown under appropriate conditions, this vector is capable of directing expression of rhulL-3 preparations which are substantially completely cornposed of the Met3 rhulL-3 (pro8 ASP15 Asp70) product (see Example 3, below).

ExamPle 2- Exoression ar~ Purification o~ rhl~lL-3 ~PrQ8 Asp15~7 The host strain, XV2181, a diploid S. cerevisias strain, was formed by rnating XV617- :
1-3B ~o~, his6, leu2- 1, trpl- 1, ura 3, StB~, obtaincd from th0 Universlty o1 Washington, Department ot Genetics Yeast Strain Bank, Seattle, WA, USA, and X21 81 -1 B ~ rpl- l, gal1, ade1, hisZJ, obtained 1rom the Yeas~ Gen~Uc Stock Cen~er, University of Califomia, ~erkeley, CA, USA. Tlte hosl strain 7s transformed with the expression plasmid by the method of Sherman el al., La~oratory Course Manual for Methods in Yeast Genetics, Co~d Spring Harbor Laboratory, 1986.

,. :. ~ .. , ; ., . . .,, , . ~ .;, .. . . . . . .

2~3~
, i Il ;
Yeast containing ~he expression plasmid plXY151 (see Example 1 above) are n~intained on YNB-trp agar plates stored a~ 4C. A preculture is started by inoculat~n~ several isolated r~combinant yeast colonies into one liter of YNB-trp medium t6.7~L Yeast Nitrogen Base 5 ~L casamino ~cids 40 m~L adenine 160 mg/L urac~l and 200 mg/L tyrosine) and is ~r~wn overrb0ht in h40 2-~er tlasks at 30C with v~g~rous shaking. By rnomin~ the cuiture is s~urated in station~y phase at an ODôoo ot 2 t~ 7. The ferrnentets ~hrse machines of t O
i-iter workir~ volume) previously cleaned and sterilized are nlled to 80% O~ their working -~
capadty with Si~2 rnedium (4.0 glL ammonium suifate 3.2 yL monobasic potassium phosphate 3.0 ~L yeas~ ex~ract 1.0 g/L citric acid 0.1 ~L sodium chloride 5 mVL 2%
caicium chbride 2.5 mVL vitamin 101 solution 0.5 mVL trace elements solution 0.5 mUL 20%
magnesium suHa~e 2.0 mVL ~lucose) and main~ained at 30C with 500-600 rpm agitation and t~16 Ipm aeration. The inoculum is added. After two hours o~ ~rowth a nutrient feed of 50%
glucose is be~un at a rate such that 50 g/L is added over a period o~ 10-12 hours. The nutrient feed is then shilted to 50% e~hanol added at 30-40 mlhr ur~il harvest.
Total elapsed Ume o~ fermentation is approximately 20 hours after which optical density (600nm) ranges from 30 ~o 45. The ~emmenters are ~hen cooled ~o 20C pH of ~he yeast beer is adjusled to 8.0 by ~he addition o~ 5 M NaOH and ~he resuiting material filtered throu~h a Mlllipore Pellicon filter system equipped with a 0.45 jlm nlter cassene and collected In a sterile 10 L carboy.
The rhulL-3 contalned In the yeast broth Is applied to a 5 cm X 30 cm column packed wUh 15-20 jlm ~4 r~ersed-phase s~lica (Vydacl Separations Group Hesperial CAI USA) by ;wrr~r~ the flite~ed yeast broth directly onto ~he column. The column is equilibrated in 0.1%
tfifluoroacetic acid tn water (Solvent A) prior to application of the yeast broth and is flushed with this solverd following ~he compleUon of ~he application ot the bro~h ~o the column until tha op~ical absorbance of ~he eflluent approaches base line values. At ~his Ume a gradient of 0.1% tfifluoroacetic acid in aceton-nrile (Solvent B) is established from 0% B to 100% B al a rate of change of 2% B per rninute and at a flow rate ol 100 mUrninute. A pause of five minutes when the gradient reaches 30% B is programmed into the gradient controller. Twenty mlnutes after the InUiation o~ the gradient~ one minute fractions are collected. Aliquots ot the 3 0 ~ractions ar~ analyzed for pr~tein con~ent by a tluorescamine assay using bovlne senJm ailburriin leSA) as a stand~rd. RhulL-3 was found ~o elute in fractions 8 9 and 10. Fractions contai~ ~hulL-3 trorn ~he first step are stored at 4C ~n polyé~hylene bottlss unUI
a~p~oxSma~ely 1 ~ram ~o~al pro~ein is obtained. To ~he pool is added 2 volumes o~ 0.1%
tritluoroacetic acid in wa~er. Th~s solution ~s ~hen purnped onto a secor~ 5 cm X 30 cm column packed unth 15-20 il C-18 silica (Vydac Separations Broup Hesperia CA USA) that is ~quiiibrated in Soivent A (0.1% TFA in water). Following application of the material the column is tlushed wUh Solvent A and ~hen a gradient identical to that described above is estabOshed at a ra~e o~ change o~ 1% Solvent B per minute and a~ a fbw ra~e o~ 100 ml/rnin.

.: , . : : ~ : .
:.. . .-.. . . , :

::. .: . ~ :: : . : . -2 ~

Fractions are collected every 0.4 minutes 24 minutes into the gradient. Aliquots of the tractions are analyzed tor protein content by the fluorescamine assay using BSA as a standard.
Peak fractions from ~he C-18 column are pooled and 1/10 volume o~ 0.5 M ;B-alanTne 5 pH 3.6 is add~. A ar~b is taken and then the pool is applled to a 10 ml S-Sepharose Golumn (1 cm X 1 D crn Pham~a~ ~ 5 mUminute. Afler sample application the colulmn is washed with 50 rr~ ~ 1 OrnM ~is pH 7.4 and ~he IL-3 is ~luted with a linaar gradient from 1 OmM
Tris p~i 7.4 ~o 200 mM Tfis pH 7.4 ~200 ml total ~radient volume). Peak fractions of rhulL-3 are then pooied and dialyzed against 100 mM Tris pH 7.4 overnight a~ 4C then sterile 1 0 finered.
The biological activities (units/mg) and binding affinities of three production bts ot ~hulL~ prepared substantially as described above were determined. Binding affinity was assessed by determination of inhibition constants (Kl) using purified proteins. The results are set forth in Table 1 below. In Table 1 all values are expressed as the averages o~ sarnples run 15 in triplicate in two separate assays.
The rhulL-3 ol ~hese ~hree lots (AZZ-0001-0003I was also subjected to protein sequencing by Edman degradation and the integrated value (pmole) of each PTH amino acid residue was compared to the known protein sequence. A line for each concurrent sequence was obtained by linear regress~on analysis and used to estimate ~he prevalence ~%) ot each 20 sequence ~n Ihe preparation The results are set ~orth below.
Table l:
N-Terminal Com;~osition. SpeciIic Activitv and Bindin~ Affinitv Qf rhulL-~pr~L~lgn L~
M~)le% _Specific Activity Binding Affinity LQL Species: ~L1 Met3 [h~4 [hl6 ~2~Q7 ~mg) (x1 01 0~
AZZ-0001 34 23 31 12 4.5 5.58 AZZ-0002 35 15 11 39 5.6 7.14 AZZ-0003 37 18 14 31 5.0 5.95 ~le 3: Ferment~i~ and Purif~a1ion of Met3 rhull-3 (Pro~ ~15 ~ 70) The expression vector plXY290 (see Exampie 1 above) was transfected into Ihe S.
cerevisia~ production strain XV2181 substantially as described ~or plXY151 In i-:xample 2. !
An inocutum was prepared by inocuiating 0.5 mL ot a frozen glycerol stock of theprQcuction strain Into a 6 L flask containing (a) 500 ml of a yeast growth medium containing per liter sterile H20: 20.0 9 (NH4)2S04; 5.0 g KH2P04; 1.0 9 MgS04-7H20; ) 0.1 ~ CaC12-2H20;
40 10.0 g casein digest; 20.0 9 glucose and 20.0 g galaclose; l~) 0.5 ml of a trace saits mix containing per liter sterile H20: 5.0 9 boric acid ~.0 g copper suHate 10.0 9 ferric chloride;
10.0 ~ manganese sul~ate; 0.5 9 sodium molybdate; 1.0 9 zinc suifate; 0.5 9 cobait chloride;
and 100 ml 6N HCI; and (c) 0.5 ml of a vitamin solution containing per liler sterile H20: 1.0 ml : . - . ~ . - . -: , ~ ~. . , : .
;,: - . ~ . , ~, ........ .

:: . ~ . .. . ,::

'2 ~ c~

of a 0.2% solution ot biotin; t,.O g calcium pan~o~henate; 25.0 g myo-inositol; 5.0 ~ niacin; 0.4 g pyfidoxine-HCl; 0.1 ~ folic acid; and 0.5 9 choline chlorida. This inoculum Gulturc was iacubaJ~d at 30C in a rotary flask shaker at 250 rpm tor approxirnately 24 hours.
A ~0 ~ter tennenter was prepared for termentation by filling with 7.5 L deionized H2Q
,, containir~ 50 CJ m~ pc~lassium phosphate,200 ~ ammonium suHata,10 ~
M~S04-7H20, 1 0 ~ CaC~-2HzO, and ~ mi of a antifoamin~ surlactant. Ths ierrr~nter was thsn sterilized by hea~ir~ to 121C ~or ~0 minu~es ~th sUrring. ~fter sterilization, the lerrnerder was supplemented wi~h 35 ml 1 % (wh) thiarnine-HCl, 500 ml l O% caseirl hyrolysate,100 rni 50/0 glucose, and 25 ml each o~ th0 vitamin and trace sai~s solutions 1 0 described above. Fermenter pH was conttolled to pH 5.5 by adclition of 30% NH40i-l, and ,1.
dssolved 2 was spanned at 100 percen~ at 29C. The o ontents ot the inowium were then added to the ~ennenter ancf a 1eed started at about 2.0 ml per ninute of a teed solution consistlng o~ (a) 3600 ml 50% glucose; (b) 625 ml of a yeas~ ~eed sal~s soluUon consisting of (per liter H20): 250 9 ammonium Suna~e, 125 9 monobasic potassium phosphate, and 25 9 1 5 MgS04-7H20; (c) 900 ml 50 E~OH; (d) 60 ml 1 % thiamine-HCl; (e) 450 ml 20/~ yeast extract;
(f) 450 ml 20% peptone; (g) 900 ml 20% casein hydrolysate; and (h) 25 ml each of the vitamin and trace saRs solutlons described abova. The feeci rate was malntained at th~s level for about 20 hours, then adjusted to 3.25 ml per minute for about 20 hours, wheteupon the r~l~ure supematar~ was harves~ed substant~ally as described in i~xample 2, above.
The rhulL-3 contained in the ~reasl broth was lirst reduced by adjusting to 1% beta mercaptoethanol at pH 7.4, then incubating 10r 1 hour at ambient tempera~ure. The resuiting mixture was appOed ~o a 5 cm X 30 cm column packed with 15-20 ~rn C-4 reversed-phase silica (Vydac, Separations Group, Hesperia, CA, USA) by pumping the flltered yeasl broth directly onto the column. The column was equilibrated in 0.1% trHluoroacetic acid in water (Solvent A) pfior to application of the yeast broth and was flushed with this solvent following the completion of the application of the broth to the column untit the optical absorbance of the effluent approaches base line values. At this time a gradient o~ O.1 % tritluoroacetic acid in acetonitfile (Solvent B) was established f rom 0% B to 100% B at a rate of change of 2% B per minu~e and a~ a flow rate of 100 mUminute.
Peak Iractions from the C-4 column were pooled and adjusted to 50 mM sodium acetate, pH 4.7. This solution was ~hen applied directly to a column c;ontainlng Mono S lon exchan~e resin equilibra~ed in 50 mM sodium acetate pH 4.7, and rhulL-3 was eluted with a linear gradient of 1 M NaCl over 20 column volumes. The resuiting rhulL-3 was substantially 100% Met3 rhulL~ (pro8 Asp15 Asp70), and comparable in biological activity 10 the rhulL-3 production hts described in Example 2, above.

. ~
~ :,, . . ~
''~' , :., . :: ,.. ... :.-.. . , . : : : : , , :
:: - ~ : . : : :; . : . :
- :
.: . ~ . : .
.
: ~ ., . :: .-:.: ~ -. ~ . :
::~ , :: . .
: . ... .

Example 4: ~ntraYenOu~miniS~ra~i~ o~rhulL-~ ~PrQ8 ~15~Q70 Compositions tQ ~nomol~us klonk~Ls Three groups of ~hree male and three ~emale cynornolgus monksys (Macat~a ~ascic~bns) ea~h rttceived material trol7~ one of ~he rhulL-3 hts prepared as described in 5 i~am~e 2 v~a intravenous a~rTiinlstration once daily lor 30 cor~cutiv~ days at dose levels of 1 jl~g ~Group 2, bw-dose), 10 ;u9n~9 ~Group 3, mid-dose), and 100 jlg/kg (Group 4, high- ~ -dose). A tourlh ~roup ot three males and thret~ ternales received 0.9% sodium chhride tsaline) and served as the control group (Group 1, control). One animal per sex per group was retaln~d 10r a 17~ay recovery pefiod. Criteria evaluated 10r compou i eltect includ0d `
survival, clifical sJ~ns, physical, electrocadiographic and ophthalmoscopic examinations, .
changes in body tt~npera1ure, body weight, clinical pathology, or~an weigh~ data, ar~i ~ross and n~croscopic pathology.
All animals survived until thelr scheduled sacrifice. Treatrnent-related clinical signs included anorexia. No urticaria was noted. No trea~ment-rela~ed eftects were noted in body 15 weights, physical exarnination, electrocadiographic evaluation or oph~halmoscopic examinations. Clinical pathology parameters evaluated included hematology and ~ i rnyeloid/erythroid (M/E) ratios evaluated trom femoral smears. Senum chemistry and urinalysis was also perfomned. Notable serum chemistry parameters for the mld- and high-dose animals were unremarkable when compared to concurren~ con~rol values or respectlve pretreatment 20 values. Urlnalysls values obtalned followlng treatment were unremarkable.
Intravenous inJection o~ analog rhulL-3 dramat~cally Increased the leukocyt~ counts In mid- and high dose males and ~emdles~ Mean Iymphocyte count values were also increased in cell-treated males and high-dose temales. Increases in segmented neutrophils,eosinophils, and basophils in mid- and high-dose animals in general para~leled ~he increased 25 laukocyte and Iymphocyte counts. An increase was also observed tor the mean ptatelet counts tor the high-dose rnales and compound-treated females during the treatment interval.
The most notable dose-related increases occurred progressively during ~irst and second week with a general decline toward pretreatmen~ levels during post-recovery phase.
Gross pathology noted certain abnormalUies of spleen, thymus, and adrenal glands in 30 Ind~vldual animals recelving ~he highest dosages. However, no ~wo anlmals showed tha same e"fect anci in one case, 1he abnormality was diagnosed as a preexlsting condition. An ir~eased rnean absolute relative thymus weight was noted ~or low- and rnid-cose rnales at terminal sacrifice and one low-dose and one mid-dose temale at post-recovery sacrifice.
Microsoopic Iindings noted in the animals sacriticed at temmination were ~ew in nurrber and 35 were primarily minimal to slight in severity. All histopathoiogy findings were considered incidental in nature and unrel~ed 10 compound administration.

, . ' '. ~ . ' ' . .

, . ,- . . .

- 2~33~ 3~ `
~s ExamDle 5: Cli~ valuation ol IL-3 çomDQ~
The rhulL-3 of one or more ot the production iots prepared as described in Example 2 was administered lor 15 days as a daily suixutaneous inJection to patien~s with advanced progressive neoplasms or bone marrow 1aiiure. Regular routine hematologlcal and 5 biocherr~cal i7~estigations were done and i~one marrow aspirations and t~iopsies were obtainedpriort~an~attertreatmsr~cycles. Atotalof 14paUentsweretreatedat60, 1~5.
250 and 5D0 ~tm2. No patient stopped therapy due to toxicity. In all nine evai~able i~'en'S. substantial increases in WBC, ab olute segrnented ~ranulocyte, eoslr)ophil and mor~te oounts, as well as in reUculocyte and platelet counts, were seen. Hernatoi~ical 10 chan~es, h partir;ular the increase ~n platelet counts, occurred primarily dufir~ ihe second week d ~u~l -3 trea~ment and continued ~or 1-2 weeks afler the ~nd o1 theraw. The mean increase of pla~elet counts was 179,500/~1 (63,000~48,000). In two patients wUh prolongsd thrombocytopenia of 3000/ill and 22,500/ill, respectively, the platelet counts rose to 104,000/111 and 88,000til1 lor an ex~0nded period ol time af~er treatment with 60 iug/m2.
15 Toxicities included fever in three patients (grade 2), headache in one patien~ (grade 2) and iocal ery~hema at the injection site (grade 1) in one patient. Urticaria or InfiitraUon of mast cells o~ iyrnphocytes into the dermis was not observed.

~ , ., ~ . . . .. . .
,''`:~ , , .' " ' ` ' ' ~ `

. : . .. . . .
, ,, . , ~ . .

Claims (9)

What is claimed is:

1. A pharmaceutical composition comprising as active ingtedient one or more morepurifled recombinant human IL-3 (Pro8 Asp15 Asp70) analog proteins expressed by a transforrned yeast host ot the species Saccharomyces cerevisiae, said proteins selected from the group conslsting of:
(a) Met3 rhuIL-3 (pro8 Asp15 Asp70);
(b) Thr4 rhuIL-3 (pro8 Asp15 Asp70); and (C)Thr6 rhuIL-3(pro8 Asp15 Asp70).

2. A pharmaceutical composition according to claim, further comprising Ala1 huIL-3 (pro8 Asp15 Asp70) in admixture with a protein selected from said group.

3. A pharmaceutical composition according to claim 1, wherein the active ingredient consists essentially of Met3 huIL-3 (pro8 Asp15 Asp70).

4. A phammaceutical composition according to claim 1, wherein the active ingredient consists essentiatly ot Thr4 rhuIL-3 (pro8 Asp15 Asp70).

5. A pharmaceutical composition according to claim 1, wherein the active ingredient consists essentially of Thr6 rhuIL-3 (pro8 Asp15 Asp70).

6. A phannaceutical composition according to claim 2, having a specific biological activity in the human bone marrow assay of from about 1.8 to about 7.0 x 107 U/mg.

7. A pharmaceutical composition according to claim 7, having a binding aflinity for human monocyte IL-3 receptors, expressed as an inhibition constant, of from about 2.0 to aboud 8.0 x 1 o10 M-1 I

8. A method for treating neutropenia, anemia or thrombocytopenia in a patient innoed theroof, comprlsing administering a therapeutically effective amount of a composition according to any of claims 1 to 7.

9. The use of a composition according to any of claims 1-7 in the preparation of an agent for the treatment of neutropenia, anemia or thrombocytopenia.