Essentialsin"_ •

s_'E]ll

Poster Presentation

11th International Symposium on GlycoconjugatesJuly, 1991

Poster # 14.14

Analysis and Purification of SyntheticOligonucleotides, DNA Restriction

Fragments, PCR Products and PlasmidsBy HPLC

W. Warren, G. Vella & K. O'Connor

Millipore Corporation, Waters Chromatography Division34 Maple Street, Milford, MA. 01757 U.S.A.

r

Waters Chromatography Division WatersWaters. The absolute essential Mi,_poreCorpora.onfor bioresearch. 34 Maple St. Divisionof MILLIPORE

Milford, MA 01757(508) 478-2000

Introduction

The purification and analysis of synthetic oligonucleotides, DNArestriction fragments, polymerase chain reaction (PCR*) products andplasmids are of growing importance in both academic and industrialsettings. Traditionally, techniques such as gel electrophoresis andultracentrifugation have been used for the isolation and analysis of thesemolecules. Unfortunately, these techniques are time consuming (i.e. hoursto days) and labor intensive. In addition, product recoveries are often lowand may contain interfering substances (e.g. cesium chloride, ethidiumbromide or polyacrylamide) derived from the separation matrix.

High performance liquid chromatography (HPLC) provides a singlestep alternative to gel electrophoresis and ultracentrigugation for thisapplication and has been proven effective in the analysis and purification ofa variety of biomolecules including proteins, peptides and amino acids.Furthermore, because the separation and detection techniques are non-destructive, direct isolation of biologically active material is easilyaccomplished. This poster will describe the principles involved with thischromatographic technique and will detail its application to the purificationand analysis of synthetic oligonucleotides, DNA restriction fragments, PCRproducts ar'd plasmids.

* See U.S. Pate 't No. 4683202 to Cetus Corporation

b

Synthesis of Oligonucleotides : -

The advent of automated, solid-phase DNA synthesizers of improvedcoupling efficiency and multi-strand synthesis capability has eased thepreviously time-consuming and complicated process of constructingbiologically active synthetic oligonucleotides. In brief, synthesis involves thesystematic stepwise addition of specific nucleotides [protected at their 5'hydroxy end with a dimethoxytrityl group (DMT)] to a solid phase supportcontaining the previously attached nucleotides. Yet despite numeroustechnological advancements, coupling efficiency in each synthesis cycleremains below 100% resulting in contamination of the desired syntheticoligomer product by sequences of shorter length. The amount of productcontaminants is proportional to the length of the synthesis product asshown in Figure 1.

Figure 1: Synthetic Oligonucleotide Length Compared toTheoretical Yield at Various Coupling Efficiencies

9o

8o

7o

20-

I0- _

0s I i I I

0 10 ,80 120 160 200

Length of Oligonuclemide

• !

' Rapid Oligonucleotide Purification:,i Using Disposable Columns

Purification of full length oligonucleotide product from failuresequencescan be performedwithin30 minuteson an Oligo-PakTM

column and syringe. Between 10 and 50 O.D. 260nm units (10.D.260nm unit = 33 pg oligonucleotide) of DMT-protected crude materialcan be purified in a single run. In addition, on-column productdetritylation can also be performed on this acid and alkali stablechemistry. The final product is of suitable purity for use as primers forsequencing or PCR or as hybridization probes for southern blotanalysis.

These single use, disposable columns separate via the principleof reverse phase chromatography. Separations are based upon therelative hydrophobicities of the different species present in themixture. The least hydrophobic molecules elute prior to the morehydrophobic species. For this method, the oligomer must beproduced such that the 5' DMT group is left on the oligomer uponcompletion of the synthesis and removal from the synthesis support.The DMT _,roup renders the product more hydrophobic, allowing it tobe sepam _d from the failure sequences which have lesshydrophol __, 5' hydroxy groups present. Since these failuresequence _:are less retained on the cartridge they can be easilyremoved _om the DMT protected product. Figure 2 compares areversed-,_hase HPLC separation from a 20mer synthesis (DMT on)before and after purification on an Oligo-Pak Column showing theeffectiveness of this technique.

Figure 2: Reverse Phase HPLC Analysis of Tritylated 20merBefore and After Oligo-Pak Column Purification

1.500 20met:)roducl

$1mpll: 1.41 O.O.260 unite I_om 20met gynthlllll (DMT on)Column: Oe_fl-lPllk TM CllL 5_, 3bOA 13.gramI 1,'10mm)Eluenl A: 0.1M Tflethylemmonium ACelmtm,pH 65Eluenl B: A¢,tonllrllll MllllQ TM Wuler 195:5)Grmdlant: _1* 40%8 In ,tOrain., LInmafFlow: t0 mllmln.Tlmp: 30"C

E

Failure sequences

20 40

Minutes

Semele: 0.17 o.e. 260 un¢ll of ellgo-Fik Column pullllldIlmpll from 20air Synlhllll lout on):

Column: Oelli-PIk TM Clio 5., 309A (3.gram • 1S0mmlElulnl A: 0fM lt|liflyllhnmanlum A¢iIIIi, pH 6.SEIuenl e: A,=llonllflllU MIIIIQTM Water (95:5)Orlldlinl: 5 - 40%8 In 40 rain., Llnlm'FlOw: |.0 mltmln.

E Temn: 30"C¢NNN

0.000

0 20 40

Minutes

.. Reverse Phase HPLC Purification.. of Synthetic Oligonucleotides

Reverse phase HPLC of DMT-protected syntheticoligonucleotides operates on the same principle as that employedusing Oligo-Pak columns. However, the increased resolving powerobtained with HPLC gradient separations can also be effectivelyused to separate smaller DMT containing sequences which can arisevia product depurination or incomplete end-capping from the DMT-protected full length material.

Quantities up to 100 O.D. 260nm units of DMT-protectedreaction mixture can be chromatographed in a single 40 minute runusing the high-resolution Delta-Pak TM C18,300A chemistry(Figure 3). For purification of multimilligram quantities, larger columnconfiguratiions are available and have been successfully used topurity up to_3000 O.D. 260nm units of a phosphorothioated 21 merusing a sir_,gle25mm x 100mm PrepPakTM cartridge column(Figure 4).

Figure 3: Reverse Phase HPLC of Tritylated 50mer

Sample: 1.83 O.D._ unils flora 50met synlhesis (DM[ on)Column: Della-Pak+MCle , 5 tun, 300,_,, 3.9ram x I S0mmEluenl A: O. I M hielhyJammonium acelale, pH 6.5

0.400 Eluenl B: Acelonillile/MIIli Q_ walnr, 95/5 2Gradient: 5 AO%B in dO mir,, li.ern

Flow: I.O 'ml/min I. Non DMI conl+lining lailufe sequencesIemperalure: 30°C 2. DMI ptolected 5Omer

E ;:"C /''

0 _

C

0

O.000 i v m v0 20 40

Mlnules

Figure 4: Reverse Phase HPLC of DMT protectedPhosphorothiated 21 mer

Failure Sequences 21 mer Product

2.000 Sample: 3000 O.D. 260rim units fromphosphorothioated 2 ! mer synthesis

Column: Della-Pak PrepPak, C18,300A, 15gin(25mm x 100ram)

Eluent A: 100raM TEAA, pll 7.0 (95%)/Acetonitrile (5°,'0)

"_ Eluent B: Acetonilrile (95%)/Milll Q1_1water (5%)_ "= Gradient: 10 - 50%B in 30 minutes

E ', i.:ti Flow: 8.0 mls/min.

,__ ,ii),_" Temp: Ambient

r_

om

.D<

0.000

0 15 30Minutes

Figure 7: Autoradiogram of Fractions Collected fromGen-Pak FAX Purification of Detritylated 20mer

(D

1-- O,I 00c c co o o c

LL LL I.L

Analysis and Purification of Restriction Fragmentsby Anion Exchange HPLC

High-resolution, anion exchange HPLC on the Gen-Pak FAXcolumn is well for the analysis and purification of restrictionfragments. Using a simple gradient of increasing salt concentration,DNA fragments elute from this chemistry in order of increasingnegative charge which is proportional to increasing DNA length. Theseparation ofla _ III digest of PhiX 174 DNA is shown in Figure 8.As noted, the separation is monitored at 260nm which provides ahigh level of detection sensitivity without requiring the use of toxicadditives such as ethidium bromide, In addition, useful preparativeapplications involve the isolation of small insert DNA (e.g. <1000 basepairs) from the larger vector fragments (e.g. >3000 base pairs) or inpurifying various sized restriction fragments less than 1000 basepairs in length.

Figure 8: Separation of DNA Restriction Fragments on theGen-Pak FAX Anion Exchange HPLC Column

CL

tOoO

0.30

Sample: 3.0 fig ot BstN 1 di_lest of pBR322Column: Gen-Pak ,MFAX (4.6ram x 100mm)Eluent A: 25mM Tris/CI, lmM EDTA, pH 8.0Eluent B: 25raM Tds/CI, lmM EDTA, 1.0M NaCI, pH 8.0Gradient: 25 - 75%B in 30 rnin., Linear Q. 1:31_Flow: 0.75 ml/min. .D .oTemp: 30°C (_ ooil f.,o

cTJO

EC

Cl.

0i '_,, , _ J _ , | i, , | i i i i i '' | _ i _ I'' i | i i i I i ," 35

1 8 Minules

if;Qualitative and Quantitative Analysis ofPCR;Reaction Mixtures by Anion Exchange HPLC

The ana!ysis of PCR amplification mixtures can also be rapidlyobtained on thi_:Gen-Pak FAX column as illustrated in Figure 9.As noted, ample component resolution is afforded by this method. As such,this rapid qualitative analysis can be used in the optimization of reactionconditions for PCR amplifications. In addition, quantitative analysis is alsoeasily obtained. Based upon peak area determinations and calculationsagainst an external standard, it is possible to determine the amount of aspecific PCR product contained in an amplification mixture. In this example,it was determined that 323ngs and 6.6ngs of the desired 465 base pair PCRproduct were contained in amplification mixtures containing3ng (Figure 9:Top) and 1 fg (Figure 9:Bottom) of Hepatitis B Virus (HBV)S-gene template.

%1

Figure 9: Chromatography on Gen-Pak FAX Column of aPCR Amplification Mixture Generated using "3ng and lfg of HBV S-gene Temp

dNTPs and Taq enzyme 465 bp PCR loroduct

Sample: PCR amplificalion mixture using

0.040 3 ng HBV-S gene lemplateColumn: Gen-Pak TM FAX (4 6ram x 100mm)

Eluent A: 25ram Tds/CI. ImM EDTA. pH 8 0

Eluan! B: 25mM rds/CI, lmM EDTA, 1 0M NaCI, pl-! 80Gradmnt: 40 - 75%B In 30 rain. Unear

Flow: 0.75 ml/min.

Temg: 30°C

Hon-specilic DI'.IA

"-_ _sequences0.000

1 I 1 r [ l I I

0 15 35Minules

Sample COlrnesy el Dr Peter Knudsen Cotumbta UnlvefSily

Sample: PCR amphfic.atlon rn=xlure

usrng 1 fg HBV-S gene template Primers andColumn: Gen-Pak n'' FAX (4 6rnm x t0Omm|

rI Eluent A: 25mM Tds/CI. lrnM EDTA. DH B 0I Eluenl B: 25rnM FrL_CI. IrnM EDTA. I OM NaCI Dtl 8 0

0 008 I Gradient: 4(_ - 75%B in 30 rnm. bnear| Flow: 0.75 rnl/rnm

: 3°"C Non-specllic DHA

sequences

7 j

• 465 bD PCrl produ_

0.000

I -----r ........ T _ - r ........ T ......... T-- ' ' l

0 15 35

Figure 10: Purification of a 344 bp PCR Product FromGen-Pak FAX Column Prior to Sequencing

dNTPs and Taq enzyme 500ng of 344 bp PCR productfrom human papilloma virus

0.038 r- Sample: PCR amplification mixturecontaining 344 bp product

Column: Gen-Pak TM FAX (4.6mm x t00mm)Eluent A: 25ram Tris/CI, lmM EDTA, pH 80Eluent B: 25mM Tris/CI, lmM EDTA. 1,0M NaCI, pH 8,0Gradient: Load at 40%B

Then, 40 - 55%B in 0.01 rain.. StepThen, 55 - 65%B in 30 rain., Linear25 - 75%B in 30 min., Linear

E I Flow: 0.75 ml/min.

c /Temp: 30°C

O

Non-specific DNA

0.000 , I _ i * ! , i ' i ' w

0 15 30Minutes

Samplecourtesyof Dr.JayDoniger,GeorgelownUniversilyMedicalCenter

Anion Exchange HPLC Purification of Plasmids

The purification of plasmids from bacterial proteins,chromosomal.'DNA and RNA is generally accomplished using cesiumchloride gradient centrifugation. Typical run times range from 5 to 48hours, deper_ing upon rotor design, and must frequently be repeatedto affect complete plasmid separation from the bacterial cellcontaminanESl

HPLC separations on the Waters Protein-Pak_MDEAE 8HRchemistry are performed within 60 minutes using a simple NaCIgradient in a Tris/EDTA, pH 8.0 buffer. This cllemistry is capable ofpurifying microgram to milligram quantities of plasmid DNA startingfrom an alkaline lysed, RNase and Proteinase K digested,polyethylene glycol precipitated sample. As indicated in thechromatogram of the purification of 2.4 mgs of a 2939 base pairplasmid (pRNH124), contaminating RNA is well resolved from theplasmid (Figure 12). Agarose gel electrophoresis of the HPLC purifiedplasmid confirmed that it was free of contaminating RNA.Furthermore, lhe isolated plasmid was biologically active as indicatedby its ability to be cut with restriction enzymes as well as to transformcompetent .E. coli cells.

Purification of PCR Products by Anion Exchange HPLC

The Gen-Pak FAX column is also ideally suited for PCR productpurifications as exemplified by the isolation of a 344 base pair DNAfragment produced from an amplification of a human papilloma virusintegrated within a human exocervical epithelial cell line (Figure 10). Asindicated, the target DNA is easily isolated from contaminating dNTPs,primer, primer-dimers and non-specific DNA sequences using a simpleNaCI gradient in a Tris/EDTA buffer system.

The quality of the HPLC purified 344 base pair PCR product wasconfirmed by performing standard dideoxynucleotide chain terminationsequencing. A segment of the autoradiogram from the sequencing gel isshown in Figure 11 indicating excellent readability of the sequence withresults comparable to those obtained using M 13 subcloned DNA.

,e

Figure 11: Sequencing Auloradlogram Irom ttPLC Purllled PCR Producl

A C G T

|

_,-7 _J i

1

[ ,:

r

* #

i

Figure 12: Autoradiogram of PCR Amplification MixturesWhere Various Concentrations of Hepatitis BVirus S Gene Templates Were Used

_- 0 0 _ 0 0 _ C

Autoradiogram courtesy of Dr. Peter KnudsenCollegeof Physiciansand Surgeons,ColumbiaUniversity

Conclusions

The Waters Gen-Pak FAX column is useful in purifying target DNA fromdNTPs, primers, primer-dimers, Taq polymerase enzyme and non-specific sequences enabling direct sequencing without subcloning.

On-Line, high-sensitivity detection and quantitation of PCR products at260nm on. the Gen,Pak FAX column is an attractive alternative tosouthern blot analysis for some applications.

References

1) Thompson, J.A. 1987. Isolation, purification, and analysis ofDNA restriction fragments. Biochrom. 2(1):4-18.

J

2) Goebel, U., R. Maas and A. Clad. 1987. Quantitativeelectroelution of oligonucleotides and large DNAfragments from gels and purification by'electro-elution. J. Biochem. Biophys. Methods. 14(5):245-260.

3) Oliver, R.W.A. HPLC of macromolecules: A practicalapproach. (IRL Press Ltd., Oxford, England, 1989) pp. 195-196.

4) Maniatis, T., E.F. Fritsch and J. Sambrook. MolecularCloning. (Cold Spring Harbor Laboratory, New York,1982).

5) Hines, R., K. O'Connor, G. Vella and W. Warren. InPreparation.

Anion, Exchange HPLC for the Analysis of Oligonucleotides

The rapid analysis of synthetic oligonucleotides can be readilyaccomplished using high resolution, anion exchange HPLC of theGen-Pak TMFAX column. Separations rely primarily upon theinteraction of the negatively charged phosphate groups on the DNAbackbone with the positively charged cations contained on the anionexchanger and are best performed on detritylated samples. Via theuse of a gradient of increasing ionic strength, detritylatedoligonucleotides elute in order of increasing chain length, often withN from N-1 resolution (Figure 5). On-line 260nm absorbancedetection allows real time, high sensitivity analysis of oligonucleotidemixtures and eliminates the use of post-run visualization techniquesas employed using gel electrophoresis. This rapid chromatographictechnique is thus useful for monitoring the day to day efficiencyperformance of the DNA synthesizer used in the laboratory.

Figure 5: High Resolution Anion Exchange HPLCAnalysis of Oligonucleotide Standards

Sample: 1 pg pdlAI 5, 10, 12-18metColumn: Gen-PakTM FAX, 4.6mm x 100ramEluent A: 25 mM Tds/CI, I mM EDTA, pH 8.0/Ace/onihile, 90/10Eluent B: 25 mM Tris/CI, 1 mM EDTA, 1.0 M NaCI, pH 8.0/Acelonihile, 90/I 0Gradlenl: 10-60%B in 30 rain, linear, hold for I0 rainFlow: 0.75 ml/minTern )eralure: 30°C

2O. lO0-

h

- 1. 5met- '_ 2. ] Omer- 1 3. I 2metE= - 4. 18mer

O,o

¢:

£< -

D

O.0OO_/

l I I '1 I I " U r" i "'i t i = I I' ,_O0 2O

Minutes

Ani_,n Exchange HPLC Purification of Oligonucleotides_._ _.

Depu_ination as well as incomplete end-capping during theoligonucleotide synthesis reaction can result in hydrophobic DMT groupson partial length sequences. Although these DMT-containing sequencescan be purilied from full length tritylated product via reverse phase HPLC,anion exchange chromatography provides an alternative chromatographictechnique. As indicated by the separation of a detritylated 20mer sample(Figure 6) with subsequent gel electrophoresis analysis of collectedfractions (Figure 7), the Gen-Pak FAX column provides N from N-1resolution from such synthesis mixtures. The maximum load capacity onthis column is 20 O.D. 260nm units. Although larger masses ofoligonucleotides can be purified in a single run on the Protein-Pak IM DEAEor Q HR anion exchange chemistries, less component resolution isobtained compared to chromatography on the Gen-Pak FAX column (Datanot shown).

Figure 6: High Resolution Anion Exchange HPLCPurification of Detritylated 20mer

Sample: 1.30.D. 26Ohm units from 20mer SynthesisColumn: Gen-Pak TM FAX (46ram x f00mm)Eluent A: 25raM Tris/Cl, 1raM EDTA, pH 8 0(90%)/

Acelonilrlle(10%)Eluent B: 25raM Trls/CI, traM EDTA, 10M NaCt, pH 80(90%1/

Acetonitrile(10%)Gradient: 15 - 35%B in 2 minutes

Then, 35%B - 55%B in 30 minutes

0.200 Flow: 0.75 ml/min.

Temp: 80°C _ _

O o o

EJ _ r_IL LL

Eo

%

0.000

0 15 30Minutes

362 P. P. Rop et at. / J. Chromatogr. 615 (1993) 357-364

' ,, _ ,_ ' _ ' _ ' : , ! 1:__ ot. ": i .: : | "• + ,,,+ i ! i ! i L

_................. :.......................,..............?..............:...............+..............!.......:

.. i!kfL;++ i + L:................_ I t .........._..............t..............i..............÷.............._......t.:.

(A) z..............._"- _.......":............"."............":............._.............._.............!.......I

i i i\1 D

i-J.............-_..... ,-....+---...._,-_.._..._...... +...........................,.............._......

1 . '. . _ . :. . _ . '. . _",..,,-"-,_"

i...... _1 .......

o°

.... -+1....... it• • _+

• • _li _ i i i it............... a (B) _ "'"""'""__--'_-'_ ........_"......['_.

• i iI......:' .......": i '_ i i _ i i !g ....................................................................... :4'

]l , _

: i- _ • o: _• : i _ .:

f_ ' i b.+\ : _ "

Fig. 2. Chromatograms and corresponding UV spectra of (A) a necropsic blood sample from the first case and (B) a necropsic bloodsample from the second case. Peaks: 1 = dextromoramide; 2 = norpropoxyphene; 3 = propoxyphene; 4 = methadone (I.S.).

Several drugs were also tested with the pro- fere with norpropoxyphene, and desipramine andposed method for possible interference (Table I). cyamemazine with propoxyphene• ThereforeAmong these compounds pethidine might inter- propoxyphene and/or its metabolite cannot be

P. P. Rop et al. ! J. Chromatogr. 615 (1993) 357-364 363

TABLE III

CONCENTRATIONS OF DEXTROMORAMIDE, PROPOXYPHENE AND NORPROPOXYPHENE IN COMPARISON

WITH THOSE OF PREVIOUS REPORTS (ng ml-z)

P = Plasma; B = blood; Ther = therapeutic; Tox = toxic.

Compound Present study Levels from previous reports

Case I Case 2

Dextromoramide 194 Ther(P): 10-80[7]

Tox (B): > 40 [5]

Post-mortem (B): 280-984 [2,5,7]

Propoxyphene 614 4330 Ther (P): 50-570[5]

Tox (B): > 2000 [5]

Post-mortem (B): 1000-6000 [5];

476-4284 or 1.4-12.6 _Lmot/1 [9]

Norpropoxyphene 1100 3800 Ther (P): 600-3000 [5]Post-mortem (B): 1400-5900 [5]

quantitated in samples that also contain pethi- concentration in whole blood was found to bedine. desipramine or imipramine (a precursor of 194 ng m1-1. It was about 2.5 times as great asdesipramine), or cyamemazine, the highest plasma therapeutic level, 10-80 ng

ml -t [7], but lower than blood post-mortemApplication to necropsic whole-blood samples range (280-984 ng ml-1 [2,5,7]). The concentra-

Two cases of possible overdose were condi- tions of propoxyphene (614 ng m1-1) and nor-dered. Only dextromoramide, propoxyphene and propoxyphene (1100 ng ml- i) were both in the

norpropoxyphene were determined by the de- therapeutic range. So this death could be attri-scribed procedure. Nevertheless the possibly as- buted to the combined effect of dextromoramidesociated drugs encountered in these two cases and propoxyphene with possible intervention ofand some of their metabolites were tested for pos- other substances.sible interference (Table I). No interferences were Case 2. An ex-abuser, 30 years old, was treatedobserved, for addictionwithpropoxyphene(Antalvic),dia-

Case 1. A 25-year-old man who had abused zepam (Valium) and zolpidem (Stilnox). He wasdextromoramide by intravenous injection was found dead in a hotel. At autopsy, regurgitationfound dead in his home with syringes and empty was observed. The results are presented in Tablephials of dextromoramide bitartrate (Palfium). III and Fig. 2B. The concentrations of propox-He was previously treated with dextromoramide yphene (4330 ng ml-1) and norpropoxyphenein association with propoxyphene plus acetami- (3800 ng ml- 1) in whole blood were greater thannophen (Diantalvic), codeine plus ethylmorphine the highest therapeutic level and within the blood(Neocodion), flunitrazepam (Rohypnol)and pra- post-mortem range (Table III). This death waszepam (Lysanxia). The results are presented in certainly caused by an overdose resulting fromTable III and Fig. 2Aa. The dextromoramide the abuse of propoxyphene.

ACKNOWLEDGEMENTS"Figs. 1B and 2A seem to have different slopes. In fact, Fig. 2A

was from a necropsic blood sample in which endogenous sub-

stancescouldaffecttheabsorbance,whereasFig. IBwasfrom The authors express their gratitude to Dela-a blank sample of whole blood without the same deterioration, lande and Houd6 Labs. for providing them with

some of the standard compounds.

364 P. P. Rop et al. / J. Chromatogr. 615 (1993) 357-364

REFERENCES 6 L.P. Hackett, L. J. Dusei and K. F. Ilett, J. Anal Toxicol., 11

(1987) 269.

1 K. Verebely and C. E. Inturrisi, J. Chromatogr., 75 (1973) 7 P. Kintz, P. Mangin, A. A. Lugnier and A. J. Chaumont, J.

195. Chromatogr., 432 (1988) 329.

2 A. J. Chanmont, P. Mangin and A. A. Lugnier, J. M_d. LEg. 8 P. Kintz, A. Tracqui, P. Mangin, A. A. Lugnier and A. J.

Droit Mbd., 24 (1981) 687. Chaumont, J. Anal ToxicoL, 13 (1989) 238.

3 P. A. Margot, D. J. Crouch, B. S. Finkle, J. R. Johnson and 9 P. Theilade, Forensic Sei. Int., 40 (1989) 143.

M. E. Deyman, J. Chromatogr. Sei., 21 (1983) 201. 10 Pok Phak Rop, F. Grimaldi, M. Bresson, J. Spinazzola, J.

4 R. L. Kunka, C. L. Young, G. F. Ladik and T. R. Bates, Z Quicke and A. Viala, J. Chromatogr., 573 (1992) 87.

Pharm. Sci., 74 (1985) 103. 11 H. M. Lee, E. G. Scott and A. Pohland, J. PharmacoL Exp.

5 Clarke's Isolation and Identification of Drugs in Pharmaeeu- Ther., 125 (1959) 14.

ticals, Body Fluids and Post-Mortem Material The London 12 R. E. McMahon, A. S. Ridolfo, H. W. Culp, R. L. Wolen

Pharmaceutical Press, London, 2nd ext., 1986, pp. 521-523. and F. J. Marshall, Toxicol. AppL PharmaeoL, 19 (1971) 426.