Vol . 2, No . 4, 1991
`
`ISSN 0923 -179X
`CODEN BISPE4
`
`International Journal of Separation Science
`in Biotechnology
`
`Kluwer Academic Publishers
`
`1
`
`MTX1037
`
`

`

`Copies of articles from this ·
`publication are now available from
`the UMI Article Clearinghouse.
`For more information about ~he Clearinghouse, please fill out
`and mail back the coupon below.
`·
`
`The UMI Article Clearinghouse offers articles from more
`than 11 .000 copyright-cleared periodicals in a w ide range
`of subjects . You can place you r orde rs electronica lly, as
`\'/ell as by phone , mail, and telefacsimile. For more infor(cid:173)
`mation, please complete and mail this coupon to UMI Arti(cid:173)
`cle Clearinghouse. 300 North Zeeb Road. Box II , Ann Ar(cid:173)
`bor, Ml ~8106 USA. Or cart toll-free for an immediate
`response: 800-521 -0600. From Alaska and Michigan call
`collect 3 13-761-4700 . From Canada . call toll-free
`800-343-5299.
`
`U·Ml
`
`A Be!t & Howell Company
`300 Nort h Zeeb Road
`Ann Arbor, Ml 48106 USA
`
`I 'd like to know more about UMI A rt icle
`YES!
`Clearinghouse.
`
`Name ________________________ __
`Title ________________________ _
`
`Company/Institution ________________ _
`Add ress ______________________ _
`City/State/Zip ____________________ __
`Telephone 1---1---------------
`
`Photocopying. In the U.S .A.: This journal is registered at the Copyright Clearance Center, Inc., 27 Congress Street, Salem,
`MA OI970.
`Authorisation to photocopy items for internal or personal use, or the internal or personal use of specific clients, is granted by
`Kluwer Academic Publishers for users registered with the Copyright Clearance Center (CCC) Transactional Reporting
`Service, provided that the base fee of$ I .00 per copy plus $ 0. I 5 per page per copy is paid directly to CCC. For those
`organisations that have been granted a photocopy I icence by CCC, a separate system of payment has been arranged. The fee
`code for users of the Transactional Reporting Service is 0923-I 79X/ I99 I/$ I .00 + 0. I 5.
`Authorisation does not extend to other kinds of copying, such as that for general distribution, for adve rtising or promotional
`purposes, for creating new collective works, or for resale.
`In the rest of tire world: Penn iss ion to photocopy must be obtained from the copyright owner. Please apply to Kluwer
`Academic Publishers, P.O. Box 17, 3300 AA Dordrecht, the Netherlands.
`
`Bioseparation is published bimonthly.
`Subscription prices for 1991, Volume 2 (6 issues each) is :
`For institutions Dfl. 5 12.00/US$ 290.50 including postage and handling.
`For individuals Dfl. 265.00/US$ 150.50 including postage and handling.
`Second-class postage paid at Rahway, N.J. ISSN 0923-179X.
`U.S. Mailing Agent: Expediters of the Printed Word Ltd., 2323 Randolph Ave., Avenel, NJ 07001, U.S.A.
`Published by Kluwer Academic Publishers, Spuiboulevard 50, P.O. Box 17,3300 AA Dordrecht, the Netherlands, and 101
`Philip Drive, Norwell, MA 0206I, U.S.A.
`Postmaster: Please send all address corrections to: Bioseparation, c/o Expediters of the Printed Word Ltd., 2323 Randolph
`Ave., Avenel, NJ 07001 , U.S .A.
`
`Printed on acidjree paper
`
`2
`
`

`

`Bioseparation 2: 207-215, 1991.
`© 1991 Klu wer Academic Publishers . Printed in th e Netherlands.
`
`Oligonucleotide analysis by anion exchange HPLC
`
`L.L. Lloyd 1, F .P. Warner' and J.F. Kenned /
`
`'Polymer Laboratories Ltd. , Essex Road , Church StJ·etton. , Shropshire SY6 6AX, UK; 2Research
`Laboratory for the Chemistry of Bioactive Carbohydrates and Proteins , School of Ch emistry ,
`University of Birmingham , PO Box 363 , Birmingham B15 2 TT, UK
`
`Received 24 April 1991 ; accepted in revised form 13 August 1991
`
`Key words:
`
`anion exchange HPLC , oligonucleotides, wide pore polymer matrix
`
`Abstract
`
`The ability to resolve and purify synthetic oligonucleotides by high performance anion exchange
`chromatography was evaluated using two wide pore polymeric HPLC matrices . The materials used are
`rigid macroporous copolymers which have a fully quaternised polyethyleneimine coating to provide a
`strong anion exchange , quaternary amine , functionality.
`Oligomers of poly(rA) , poly(rC) and RNA produced by alkaline hydroylsis of the polymers were
`chromatographed to evaluate the selectivity of the system prior to the analysis of synthetic oligonu(cid:173)
`cleotides produced using a commercial oligonucleotide synthesizer
`
`Introduction
`
`T he production of high purity synthetic DNA
`strands is important for a variety of applications
`including their use as linkers in recombinant
`DNA procedures (Stiege et a/. , 1988) , as the
`precursers of very large DNA fragments for the
`synthesis of whole genes (Theriault eta/., 1986) ,
`as primers in sequencing techniques (Sanger et
`a!., 1977) , as hybridisation probes in gene clon(cid:173)
`ing experiments (Joudrier et a!. , 1987) and to
`provide site-specific base alterations in selective
`mutagenesis studies (Garcia eta!. , 1987) . With
`the commercial availability of automated solid
`phase DNA synthesizers , it is possible to manu(cid:173)
`facture on a relatively large scale , biologically
`active synthetic oligonucleotides in a reasonably
`short period of time with high base coupling
`efficiencies . However , when the 'product' is re(cid:173)
`moved from the solid phase support , it is never
`100% pure , short chain failure sequences will
`also be present .
`The requirements for purity of the synthetic
`DNA will vary according to the end-use. When
`
`the oligonucleotide is to be used as a probe or
`primer , then a small level of contamination with
`failure sequences may not be critical. However,
`if it is to be used in structural studies, then high
`purity may be essential. When purity I homo(cid:173)
`geneity of the synthetic material is required, then
`one or more of the many purification procedures
`available must be employed.
`Properties / characteristics of the DNA strand
`which can be used to obtain a separation include
`differences in molecular weight , base composi(cid:173)
`tion , charge , specific sequences or end sequences
`and the techniques used can be non-chromato(cid:173)
`graphic or chromatographic. Of the non-chro(cid:173)
`matographic methods , polyacrylamide gel elec(cid:173)
`trophoresis (PAGE) dominates for both purifica(cid:173)
`tion and as an identification technique (Ikakura
`et a/. , .1984). However , there are several dis(cid:173)
`advantages with the technique. It is limited for
`purification procedures as it is both difficult to
`automate and scale-up. Also , it is necessary to
`visualise the oligonucleotides in the gel after
`separating using staining, autoradiography or
`UV shadowing prior to recovery by diffusion or
`
`3
`
`

`

`208
`
`electroelution. These two processes of visualisa(cid:173)
`tion a nd extraction are laborious and can res ult
`in poor recoveri es or the conta minati on / degra(cid:173)
`dation of the product (Vornham a nd Ke rschner ,
`1986). An altern ative procedure for the purifica(cid:173)
`tion of synthe tic oligonucleotides involves the
`use of high performance liquid chroma tography
`(HPLC ) whe re separations can be achieved
`based on molecul ar size in solution (gel filtra(cid:173)
`tio n) , charge (ion exchange) and hydrophobicit y
`(reve rsed phase) (McLaughlin and Piel, 1982;
`F ritz et a!. , 1978 ; Gabriel and Michalewsky,
`1973).
`T he HPLC se paration mode used will normal(cid:173)
`ly be de te rmined by th e size of the oligonu(cid:173)
`cleotide , the origin and the e nd use of th e prod(cid:173)
`uct. R eversed phase chroma tograph y is used to
`isola te small to medium le ngth oligo nucleotides
`produced by solid phase synthesis which are
`dim e thoxytrityl blocked (trityl-on). By the use of
`' trityl specific' reversed phase HPLC fa ilure se(cid:173)
`que nces a nd byproducts of de protection art!
`sepa rated from the re tained trityl pro tected pro(cid:173)
`duct. H owever , it is not always possible to iso(cid:173)
`late th e product from othe r trityla ted species
`which may be prese nt (Z ieske , 1988; Hill and
`M ayhew, 1990) . Anion exchange chromatog(cid:173)
`raph y, where the separation is obtained based on
`diffe re nces in the ne tt negative cha rge of the
`o ligonucleo tides, can also be used . The synthe tic
`prod uct is separated from
`the earlier eluting
`sm alle r failure seque nces which have a lower
`va lue of nett negati ve charge. Howeve r, co nven(cid:173)
`tio nal small pore wea k anio n exchange rs have
`not bee n abl e to provide th e degree of resolution
`o r loading necessary for the purification of syn(cid:173)
`th eti c o ligo nucleotides (Lawso n et a! ., 1983 ) .
`Qua te rniza tion of the pol ye th yle neimine weak
`to improve
`a nion exchanger has bee n shown
`bo th resolution a nd loading (Drage r a nd R eg(cid:173)
`nier, 1985) . A tte mpts have also bee n made to
`develo p HPLC station ary phases whi ch combine
`bo th io n exchan ge and reve rsed phase separation
`modes for the a nal ys is of oligo nucleotides and
`specific
`tRNAs
`(Bischoff and McLaughlin ,
`1985) .
`th e
`The work prese nted here evaluates
`sui tab ili ty of two wide po re st rong ani on exchan(cid:173)
`ge rs, PL-SAX 1000 A a nd 4000 A, which are
`polyethyle neimine
`coated
`qu ate rnized
`full y
`
`mecha ni call y sta ble pol yme rs fo r the analysis of
`sy nthetic oligonucleotides . The effec t of pore
`size o n the resolution of oligo mers over a range
`of chain le ngths was evaluated using a mi xture of
`o ligo me rs of diffe re nt chain lengths produced by
`contro lled alkaline hydrol ysis of th e homopol y(cid:173)
`me rs po ly(rA) and pol y(rC) . T he use of th ese
`chromatog raphic packings for th e separation of
`sy nthe ti c o ligo nucleotides produced by solid
`phase sy nthesis has also bee n eva lu ated.
`
`Experimental
`
`Chromatograph ic system
`
`The chromatographic app aratus used was a mod(cid:173)
`ul ar sys te m consisting of two reciprocating high
`pressure pumps, mode l 64, a gradi ent fo rmer,
`mode l SOB , static mi xing chamber with a 20 11-I
`inte rn al vo lume , vari able wavele ngt h UV de tec(cid:173)
`tor , model 84 , fitted with a sta nd ard a nalytical
`cell of 10 mm pathl e ngth / 10 11-l volume , twin pen
`cha rt reco rd e r (Kna ue r G mbH , Be rlin , Ger(cid:173)
`man y) and a Rheod yne 7125
`inj ection valve
`fitt ed with a 200 11-I sample loop (suppli ed by
`HPLC T echnology Ltd ., Macclesfi eld , UK) .
`A ll mobile phases were prepa red using HPLC
`grade wa te r and a nalyti cal grade buffe r sa lts
`(FSA Laboratory Supplies , Lo ughborough,
`UK). The samples of poly(rA) , poly(rC) and
`RNA were of high purity (Sigma Che mical Com(cid:173)
`pa ny Ltd ., Poole , U K ) .
`H igh pe rformance ani on exchange chromatog(cid:173)
`raph y was pe rform ed using a pol yme ric strong
`ani on exchange r, PL-SAX , with eithe r a 1000 A
`or 4000 A pore size packed in a 150 x 4.6 mm
`I. D . stainless steel column (Po lyme r Laborator(cid:173)
`ies Ltd ., Church Stretton , UK).
`
`Preparation of the oligomers
`
`A lkaline hydrolysis of the homopolymers unde r
`co ntrolled conditions was used to pre pare the
`spectrum of oligomers. T he method used was as
`fo llows: 10 mg of pol yme r was dissolved in 2 ml
`of distilled wa te r and to this was added with
`mi xing 0.12 ml 5 M KOH and
`the hydrolysis
`mi xture hea ted at 60 oc for 7 min . T he hydrolysis
`was halted by th e additi on of 0 .5 ml 1.0 M T ris
`
`4
`
`

`

`HCl, pH 6.5 , 0.1 ml 1.0 M HCl a nd removing
`from the heating bath. The oligome rs were then
`precipitated by
`the ad dition of 13.6 ml of
`etha no l : acetone (1:1 v/v) . After standing for 5 h
`at 4 °C, the residue was separated by centrifuga(cid:173)
`tion at 5000 rpm fo r 1 h at 4 oc and dried in
`vacuo for 5 min . T he oligomers were redissolved
`in 2 ml of distilled wa ter and lyophilized.
`
`Preparation of oligonucleotides of known
`sequence
`
`O ligo nucleo tides of known seque nce were pro(cid:173)
`duced in the University of Birmingham Mac(cid:173)
`romolecular Analysis Service Laboratories using
`a BT8500 DNA synthesiser (Biotech Instruments
`Ltd. , Luton , UK) using the cyanoethyl phos(cid:173)
`phoramidites reaction route (Gait , 1984) . In this
`met hod ,
`the oligo nucleo tides are synthesised
`wit hin disposable micro cartridges . The solve nt
`and reagents are drive n throu gh the cartridges by
`argo n, which both acts as the drive gas and as a
`blanket to maintain absolute anh ydrous condi(cid:173)
`tions. The first residue of the intend ed DNA
`strand is attached to an insoluble support. The
`. chain is then extended base by base to give the
`required sequ e nce. The solid support used is a
`controlled pore glass to which the 3' deoxy(cid:173)
`ribo nucleo tide is bonded via a base labile lin k(cid:173)
`age. Acid is used to remove th e 5' hyd roxyl
`protectin g group and after washing, the nex t
`protected base is added , using tetrazole as an
`activator. The sy nthesis proceeds with washing ,
`capping to block a ny unreacti ve sites and oxida(cid:173)
`tion
`to convert the phosphite
`to phosph ate.
`Afte r further washing , the mate ri al is read y fo r
`
`209
`
`the nex t addition cycle . T he progress of the
`synthesis can be monitored during the deprotec(cid:173)
`tion stage when an intense ora nge colour is
`formed whe n the 5' blocking gro up is removed.
`The completed DNA strand is simultaneously
`removed from th e solid support and deprotected
`by trea tme nt with an aqu eous solution of am(cid:173)
`monia at 55 oc. After drying, the oligonucleo tide
`is ready for purification.
`
`Results and discussion
`
`Analysis of alkalin e degraded poly(rA)
`
`It is known that the partial hydrol ys is of pol y(rA)
`produces a se ries of oligomers, 3'-oligo nucleo(cid:173)
`tides , as a result of the hydrol ys is of the bond
`between the oxygen at position 5 in ribose and
`the phosphorous atom. It could , therefore, be
`ex pected that th e use of a strong anion exchange
`HPLC adsorbent would resolve a series of
`oligomer peaks with the spectrum being depen(cid:173)
`de nt upon the degree of hydrol ysis . Figure 1
`shows the spectrum which is obtained using the
`wide pore PL-SAX 4000 A 8 p.m materi al
`packed in a 150 x 4.6 mm J.D. stainless steel
`column operated with a linear gradient of in (cid:173)
`creasing sa lt concentration , 0-1.0 M KCl , in
`6o min . T here is so me residual intact poly(rA)
`which elutes at the e nd of th e gradient , de(cid:173)
`termined by runnin g a sample of th e homo(cid:173)
`polymer used to produce th e seri es of oligomers .
`A lso resolved is a second se ries of smaller peaks
`in ad dition to th e main oligomer spectrum . If
`after hydrolysis , the 3' -phosphate residue is left
`
`0
`Fig. 1. Separation of th e o ligo mers produced by the alkaline hydro lysis of poly(rA) on a PL-SAX 4000 A 8 p.m 150 x 4.6 mm
`1.0. column . E lue nt A: 0.02 M KH 2P0 4 , pH 5.5, 5 M urea; elue nt 8: A + 1.0 M KCI; grad ie nt : linea r 0-100% Bin 60 min ; fi ow
`ra te: 1.0 mlmin _, ; de tector: UV, 260 nm. Peak I is resid ual poly(rA).
`
`5
`
`

`

`210
`
`two additional
`the nucleoside ,
`to
`attached
`species would be expected to be present , the 2',
`3' cyclic phosphate derivative in which the phos(cid:173)
`phate residue is attached to both the C2 and C3
`of the ribose unit and the 2' phosphate residue
`which is formed from it , Fig . 2. This may ac(cid:173)
`count for what appears
`to be two series of
`oligomer peaks .
`In an attempt to identify the individual peaks
`the oligomer spectrum , a small pore size
`in
`strong anion exchanger which has a higher ionic
`capacity, PL-SAX 1000 A 8 ,urn material was
`used with a shallow gradient to improve the
`resolution of the early eluting peaks. Figure 3a
`shows the separation of a mixture of four syn(cid:173)
`thetic monophosphate standards , 2' adenosine
`monophosphate, 3' adenosine monophosphate ,
`5' adenosine monophosphate and the 2'3' cyclic
`adenosine monophosphate and Fig. 3b the first
`part of the e lution profile of the oligo(r A) sam(cid:173)
`ple. The use of the shallow gradien t with the
`
`higher capacity adsorbent enables resolution of
`the 4 monophosphate nucleotide standards to be
`achieved, the elution order being: 2'3' cyclic
`phosphate, 5' monophosphate, 2' monophos(cid:173)
`phate and 3' monophosp hate . Comparing the
`elution times of the standards with the oligo(rA)
`profile indicates the presence of low levels of 2',
`3' cyclic phosphate , 2' monophosphate and 3'
`monophosphate but not , as would be expected ,
`any 5' monophosphate. In order to further iden(cid:173)
`tify the series of oligomers , th e shallow gradient
`was used with the wider pore material , PL-SAX
`4000 A 8 ,urn , which has improved mass transfer
`and resolution for the higher oligomers. The
`elution profile is shown in Fig. 4. It is known that
`the interconversion of the 3' monophosphate
`into the 2' monophosphate occurs via the cyclic
`phosphate and the relative proportions of th e
`cyclic phosp hate (peaks 2c, 3c, 4c, ... etc.) fol (cid:173)
`lows the same order as the major peaks, 1, 2, 3,
`4, etc. assuming similar detector responses . The
`
`-
`
`Fig. 2. Interconversion of 3' adenosine monophosphate to 2' adenosine monophosphate via the 2', 3' cyclic ade nosine
`monophosphate.
`
`6
`
`

`

`4
`
`211
`
`[A]
`
`[B]
`
`16
`
`0
`
`40 minutes
`
`Fig. 3. (A) Separation of a mi xture of 4 adenosine monophosphates (B) oligo(rA) using the PL-SAX 1000 A 8 J.Lill 150 x 4.6 mm
`f. D. column . Eluent A: 0.02 M KH 2 PO, , pH 5.5, 5 M urea ; eluent B: A + 1.0 M KCI ; gradient: linear 0-100% Bin 200 min ; flow
`rate: J .0 ml min -\ ; detector: UV, .260 nm . Peak identification : peak 1, 2', 3'c AMP ; peak 2, 5' AMP; peak 3, 2' AMP ; Peak 4, 3'
`A MP
`
`contribution of the cyclic phosphate component ,
`compared to the 2' and 3' monophosphates in
`the equilibrium , is always minor . It is possible to
`partially resolve the 2' adenosine monophos(cid:173)
`phate , peak la , and the 3' adenosine mono(cid:173)
`phosphate , peak lb . However , for the dinu(cid:173)
`cleotide, where the adenosine bases would have
`more effect on the retention than the single
`
`terminal phosphate residue , the 2' phosphate
`peak is seen only as a shoulder. With the higher
`oligomers , where the effect of the phosphate
`group is minor compared to the bases , no dif-
`... ferentiation of the 2 ' and 3' phosphate is ob(cid:173)
`tained. The cyclic phosphates appear to be dif(cid:173)
`ferentiated up to n = 12.
`Measurement of the retention volume for the
`
`4
`
`5
`
`3
`
`2b
`
`2a
`
`10
`
`15
`
`200
`minutes
`0
`Fig. 4. Se paration of oligo(rA) sample using the wide pore PL-SAX 4000 A 8 J.Lill 150 X 4.6 mm I. D. column. Conditions as Fig.
`3. Peak ide ntification: peak 1c, 2', 3' cAMP; peak 1a , 2'AMP; peak 1b, 3'AM P; peak 2c, 2', 3' (cAMP) 2; peak 2a, 2' (AMP) 2;
`peak 2b , 3' (AMP) 2; peak 3c, 2', 3' (cAMP) 3: peak 3, 2' and 3' (AMP) 3; peak 4c, 2', 3' (cA MP) 4; peak 4, 2' and 3' (AMP) 4;
`peak 5c, 2', 3' (cAMP) 5; pea k 5, 2' and 3' (AMP) 5; peak lOc , 2', 3' (cAMP) 10; pea k 10, 2' and 3' (AMP) 10; peak 15 , 2' and 3'
`(AM P) 15 ; pea k 20, 2' and 3' (AMP) 20; peak 25, 2' and 3' (AMP) 25.
`
`7
`
`

`

`212
`
`various peaks obtained for the degraded poly(cid:173)
`(rA) sample confirm th at the major peaks form
`part of a series in which th e parti ally resolved
`do uble t l a and lb is the first member. As this
`doublet has the same elution position as adeno(cid:173)
`sine 2 ' and 3' monophosphate, th e major peaks
`wo uld appea r to represent th e oligonucleotide
`series with terminal 2 ' and 3' phosphates (resolu(cid:173)
`tion is onl y possible up to the dinucleotide) . The
`minor peaks, labelled lc , 2c, 3c, etc. in Fig. 4,
`where pea k lc is the first member and corre(cid:173)
`sponds to ade nosine 2', 3' cyclic monophosphate
`th e o ligomers with
`are
`te rmin al cyclic phos(cid:173)
`phates. Figure 5 is the plot of number of base
`residues vs. the elution volume for the two series
`of oligo mers.
`
`A sample of poly(rC) was degraded by al(cid:173)
`kaline hydrolysis under the same conditions as
`the poly(rA) to compare the profi les obtained for
`pyrimidine bases , oligo(rC) and purine bases ,
`oligo (rA) . Figure 6 shows the elution profile of
`the oligo(rC) sample . From comparison with the
`oligo(rA) profile
`it was established
`th at
`the
`pyrimidine based oligonucleotides elute earlier
`than the purine based ones. Again , two series of
`pea ks are evident. The larger peaks pres um ably
`being du e to the 2' and 3' monophosphates and
`the secondary series the 2 ', 3' cyclic mono(cid:173)
`phosphates.
`T he alkaline degradation of the sample of
`RN A under the same conditions produces, as
`wo uld be expected , a very complex elution pro-
`
`140
`
`120
`
`100
`'-='
`...£,
`QJ 80
`E
`.2
`0 60
`>
`c:
`.2
`.... 40
`:I
`d)
`
`20
`
`,
`){
`;t
`.:
`' '
`
`~
`'
`'
`
`I
`
`20
`15
`peak number
`Fig. 5. Plot o f e lution volume vs. o ligo me r num ber fo r th e seri es of peaks pro duced by alka line hydrolysis o f oligo(rA) using th e
`PL-SAX 4000 A 8 Jl- 111 150 x 4.6 mm I. D. column . Conditions as Fig. 3.
`
`25
`
`5
`
`10
`
`200
`minutes
`Fig. 6. Sepa rati o n of o ligo me rs prod uced by the alka line hydro lys is of poly(rC). Column a nd cond itio ns as Fig. 3.
`
`8
`
`

`

`213
`
`0
`minutes
`200
`Fig. 7. Separa tion of o ligome rs produced by th e alkaline hydro lys is of RNA. Column and condi tio ns as Fig. 3 .
`
`fi le , see Fig. 7. T he degradation has proceeded
`to a greater extent , smalle r chain le ngth oligo(cid:173)
`mers produced, than that of th e homopolynu(cid:173)
`cleotides. This would be expected du e to the
`differing bond stre ngths of th e linkages between
`the various nucleo tide units. The four nu cleo(cid:173)
`tides present
`in RNA , adenine , cytosine ,
`gua nine and uracil can be joined in a variety of
`sequ e nces such that 64 trime rs, the basis of the
`ge ne ti c code, can be formed , each one with the
`va ri ous terminal phosphate st ructures , 2', 3' and
`2 ', 3' cyclic. In Fig. 7 , four gro ups of peaks
`whi ch rep rese nt the mono , di , tri and tetramers
`are evide nt. Within the gro up of monomer peaks
`it is possible to observe peaks which have equiv(cid:173)
`ale nt re te ntion volumes to 2' and 3' ade nosi ne
`monophosphate and 2' and 3' cytidine mono(cid:173)
`phosphate wit h the cyclic adenosine monophos(cid:173)
`phate being observed amongst th e ea rli er run(cid:173)
`ning peaks. It is also possible to ide ntify amo ngst
`the dimer , trim e r and tetrame r groupings, peaks
`whose elution vo lumes correlate with those of
`the (rA) and (rC) homo oligomers.
`
`A nalysis of synth etic oligonucleotides
`
`T he hi gh resolution separations achieved using
`the PL-SAX 1000 A and 4000 A mate ri als fo r the
`se paratio n of th e oligomers of poly(rA) , poly(rC)
`
`and RNA produced by alkaline hydrolysis indi(cid:173)
`cates th a t th ese materials would also be applic(cid:173)
`ab le for the analysis of sy nthetic o ligonucleo(cid:173)
`tides. A numbe r of synthetic oligonucleotides of
`simila r chain le ngth were produced to evalu ate
`th e selectivity fo r crude sa mples. The chroma(cid:173)
`tography was pe rformed using the high capacity
`PL-S AX 1000 A 8 f.Lm 150 x 4.6 mm I.D. stain(cid:173)
`less steel colum n.
`...... The first o ligo mer produced was a short chain
`6mer with
`the
`fo llowing base composition
`d(GGATCC) . Figure 8 shows the resolution of
`the 6mer from
`the shorter chain fa ilure se(cid:173)
`que nces a nd the late r eluting sy nthesis contamin(cid:173)
`ants in an analyt ical run of 30 min . To further
`evaluate selectivity for larger sy nthe tic o ligo nu(cid:173)
`cleotides,
`three samples were produced . A
`24me r with
`the sequence d(T AA T ACGAC(cid:173)
`TCACTATAGGGATCC) , a 29me r with these(cid:173)
`que nce d(GATCCATTTGACGTACGTCAA(cid:173)
`A TTT ACCT) and a 30mer with the seque nce
`d(G CGTCCCACGGTTTCGACAGAACAGC(cid:173)
`CGAC). As can be seen from Fig . 9, a 29me r
`oligon ucl eotide , with the wide pore high capacity
`PL-SAX 1000 A material which facilitates good
`so lute permeation I mass
`transfe r with
`these
`medium le ngth oligonucleotides hi gh resolution
`separations with good peak symmetry are
`achieved.
`In amo n excha nge high performance liquid
`
`9
`
`

`

`214
`
`0
`
`minutes
`
`30
`
`Fig. 8. Separation of a 6mer, d(GGATCC) on a PL-SAX
`1000 A 8 ~-tm 150 x 4.6 mm I. D. column. Eluent A: 0.02 M
`KH, P0 4 , pH 5.5 , 5 M urea ; eluent B: A + 1.0 M KCl ; gradi(cid:173)
`ent: linear 0- 100% B in 30 min ; flow rate: 1.0 ml min - 1 ~ ...
`detector: UV, 260 nm.
`
`is achieved,
`the separation
`chromatography ,
`based on the charge to mass ratio of the solute
`and not on absolute size or charge: Therefore, it
`would be expected
`that
`the oligonucleotides
`would not necessarily elute in the order of in-
`
`Fig. 9. Separation of a 29mer , d(GATCCATITGACGTAC(cid:173)
`GTCAAATITACCT). Conditions as Fig. 8.
`
`4
`
`3
`
`2
`
`0
`
`minutes
`
`30
`
`Fig. 10. Separation of a mixture of synthetic oligo-nucleo(cid:173)
`tides. Conditions as Fig. 8. Peak identification: peak 1, 6mer
`d(GGATCC); peak 2, 24mer d(TAATACGACTCAC(cid:173)
`TATAGG-GATCC) ; peak 3, 30mer d(GCGTCCCACGGT(cid:173)
`TTCGACAGAACAGCCGAC) ; peak 4, 29mer d(GAT(cid:173)
`CCATITGACGTACGTCAAATITACCT).
`
`creasing oligomer chain length. Figure 10 shows
`the separation of a mixture of four synthetic
`oligonucleotides, a 6mer , a 24mer , a 29mer and
`a 30mer. It is observed by running the samples
`individually that the 30mer elutes prior to the
`29mer as the 29mer has the higher charge to
`mass ratio at the pH used for the separation.
`
`Conclusions
`
`It has been demonstrated that high performance
`anion exchange chromatography is an alternative
`to non-chromatographic methods and reversed
`phase HPLC
`for
`the analysis of synthetic
`oligonucleotides. The resolution achieved for the
`oligomers produced by alkaline hydrolysis of the
`homopolymers poly(rA) and poly(rC) enables
`the identification , not only of the 2' and 3'
`monophosphate series , but also the 2', 3' cyclic
`monophosphates which are the intermediates in
`the interconversion of the 3' to the 2' mono(cid:173)
`phosphate. The chemical stability of the poly(cid:173)
`meric matrix and strong anion exchange func(cid:173)
`tionality enables aggressive eluents and denatur-
`
`10
`
`

`

`ing sa lts such as urea to be used for the a nal ysis
`of oligonucleotides which are self- hybridising
`and may form secondary structures . These sec(cid:173)
`o nd ary structures would otherwise have a detri(cid:173)
`me ntal effect on the chromatography being re(cid:173)
`sponsible for peak tailing or splitting.
`The analysis of the synthetic oligomers used in
`th is study was acco mplished with the high capaci(cid:173)
`ty 1000 A ion exchange material. However , for
`the improved resolution of the longer oligomers
`in
`the hom opolymer series,
`the wider. pore
`4000 A packing was used .
`
`References
`
`Bisc hoff R a nd McLaughlin LW (1985) Ana l. Biochenr. 151:
`526-533.
`Drager RR and Regnier FR (1985) Anal. Biochem. 145:
`47-56.
`Fritz H , Belagaje R , Brown EL, Frit z RH , Jones RA , Lees
`RG and Khorana HG ( 1978) B iochem. I 7: 1257-1267 .
`Gabrie l TF a nd Michalewsk y JE ( 1973) J . Chrom atogr. 80:
`263-265.
`
`215
`
`Gai t MJ (1984) In: Gait MJ (eel) Oligo nucleotide Synthesis:
`A Practical Approac h. IRL Press , Was hington DC.
`Garcia JL , Garcia E and Lopez R ( 1987) Arch. Microbia l. 1:
`52-56.
`Hill TL a nd Mayhew JW (1990) J . Chrom a togr . 512: 415-
`43 1.
`l kakura K , Rossi JJ and Wa llace RB (1984) Ann. Rev.
`Biochem. 53 : 323-356.
`Joudrie r PE , Foard DE , F loener LA and Larkins BA (1987)
`Plant Mol. Bioi. 10: 35-42.
`Lawson TG , Regni e r FE and Weith HL (1983) Anal. Bio(cid:173)
`che m. 133: 85- 93.
`McLaughlin LW a nd Pie! N (1984) In: Gait MJ (eel) Oligonu(cid:173)
`cleotide Synthesis: A Practical Approach. IRL Press ,
`Washington DC.
`Sanger F, Nick le n S and Coulso n AF ( 1977) Proc. Nat !.
`Acad. Sci. 74: 5463-5467.
`Stiege W, Stade K , Schuler D and Brimacombe R ( 1988)
`Nucleic Acid Res. 6: 2369-2388 .
`Theriault NY, Tomich CC and Wierenga W ( 1986) N u(cid:173)
`cleosides a nd Nucleotides 1(5) : 15-32.
`Yornham AV a nd Kerschne r J (1986) A nal. Bioche m. 152:
`221-225.
`Zieske LR ( 1988) Biochromatography 3(3): I 12- 117.
`
`Address fo r correspondence: Prof. JF Ke nn edy, Research
`Labora tory for the Chemistry of Bioactive Carbohydrates
`a nd Proteins, Departme nt of Chemistry , U ni versit y of Bir(cid:173)
`mingham , Birmingham BIS 2TT , UK
`
`11
`
`