All oligonucleotides were synthesized by Eurogentec (Seraing, Belgium) (Table 1 ). Oligo A is a 15 base phosphodiester deoxyribonucleotide which is complementary to the 5'-region of the initiation codon AUG of the Env mRNA from Friend retrovirus. The two phosphorothioate sequences selected have the same base sequence as oligo A, but with two internucleotide phosphorothioate linkages at the 3'-end (oligo 2-S-A) or all internucleotide phosphorothioate linkages (oligo all-S-A). The sense target, a 21mer phosphodiester deoxyribonucleotide, is composed of a 15mer complementary to oligo A plus a 6mer (AAA AAA) at the 5'-end, used as a spacer arm between the surface of the plate and oligo A. The tracer was 5'-biotinylated oligo A. 3' and 5' oligo A deletion oligomers (N-2, N-4 and N-6) were synthesized. Compounds K and KS were respectively 15 base phosphodiester and phosphorothioate deoxyribonucleotides and were selected arbitrarily because their sequence is different from that of oligo A and because they do not interfere in the assay. They were necessary to avoid non-specific adsorption of the phosphodiester or phosphorothioate sequence to the microwells.

Table 1 . Sequences of the oligodeoxynucleotides tested and cross-reactivity coefficients measured in mouse plasma

Name	Sequence	Cross-reactivity (%)
Oligo A	5'-TGAACACGCCATGTC	100
3' Deletion of oligo A
3' N-2	5'-TGAACACGCCATG	48
3' N-4	5'-TGAACACGCCA	<0.5
3' N-6	5' TGAACACGC	<0.5
5' Deletion of oligo A
5' N-2	5'-AACACGCCATGTC	48
5' N-4	5'-CACGCCATGTC	3
5' N-6	5'-CGCCATGTC	<0.5
Oligo 2-S-A	5'-TGAACACGCCATG*T*C	154
Oligo all-S-A	5'-T*G*A*A*C*A*C*G*C*C*A*T*G*T*C	6
Sense target	5'-pAAAAAAGACATGGCGTGTTCA	100
Tracer	5'-biotinTGAACACGCCATGTC
Oligo K	5'-CAG GTG TAT TGC ACT	<0.5
Oligo KS	5'-C*A*G*G*T*G*T*A*T*T*G*C*A*C*T*	<0.5

*Phosphorothioate linkage.

Blank mouse plasma was collected from OF 1 male mice (Charles River, Saint Aubin, France). The G4 form of acetylcholinesterase (AChE, EC 3.1.1.7) covalently bound to streptavidin was kindly provided by SPI-BIO (Massy, France) and was used as enzymatic reagent. Enzyme activities were measured as previously described for enzyme immunoassays (13 ) using Ellman's reagent, an AChE substrate consisting of 2.2 g acetylthiocholine and 1 g 5,5'-dithio-bis-(2-nitrobenzoic acid) in 200 ml 0.05 M phosphate buffer, pH 7.4.

Assay buffer consisted of distilled water containing 0.75 M NaCl, 5 mM EDTA, 0.1% Tween 20, 10 [mu]g/ml sheared, denatured herring sperm DNA, 1 [mu]g/ml deoxyribonucleotide K and 5 * 10^-3 M phosphate buffer, pH 7.

Apparatus

Assays were performed in microtiter plates (Covalink NH modules; A/S Nunc, Polylabo, Strasbourg, France). Experiments were performed using an Autowash 96 microtiter plate washer (Labsystem, Helsinki, Finland), an Autodrop solution distributor (Titertek Flow Laboratories, Ayrshire, UK) and a Multiskan RC plate reader (Labsystem).

Coating of the plates

Deoxyribonucleotide A was covalently immobilized on the surface of wells of Covalink NH microplates as described by Rasmussen et al. (9 ) and Chevrier et al. (8 ). The 5'-end terminal phosphate groups were grafted onto the polystyrene surface via amino groups using 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide. Briefly, phosphorylated sense target was denatured by heating for 10 min at 95^oC and was then immediately chilled on ice. Denatured phosphorylated sense target (5.5 nM in 100 [mu]l cold 10 mM 1-methylimidazole buffer, pH 7) and 1-ethyl-3-(3-dimethyl- aminopropyl)-carbodiimide (770 [mu]g in 20 [mu]l cold 100 mM 1-methylimidazole buffer, pH 7) were added to each well. Plates were kept at +4^oC until use.

Standard solution, preparation of control samples and plasma samples

Oligo A and phosphorothioate analogs were diluted first in water, then in blank mouse plasma heated for 15 min at 60^oC and containing 10 [mu]g/ml denatured herring sperm DNA, 5 mM EDTA and 1 [mu]g/ml oligo K (or oligo KS) for phosphodiester assay (or phosphorothioate assay). Herring sperm DNA, EDTA and oligo K (or KS) were added to plasma samples or plasma control samples before assay to ensure that the media of the standard and samples were identical.

Competitive hybridization assay

The oligonucleotide tracer was diluted in assay buffer at a concentration of 55 pM. Before use the coated plates were washed with 0.01 M phosphate buffer, pH 7.4, containing 0.05% Tween 20 (washing buffer) using the Autowash apparatus (300 [mu]l/well, 10 cycles) and stored on ice. The assay was performed in a total volume of 150 [mu]l (50 [mu]l tracer and 100 [mu]l samples, standard or control). The plates were covered with plastic adhesive sheet and left for 15 min in a 60^oC water bath, which was then switched off and the plates were left overnight in water. The plates were then washed as described above. Streptavidin-acetylcholinesterase conjugate (0.2 U/ml) was dispensed (100 [mu]l) into each well. The plates were covered with a plastic adhesive sheet and left for 4 h at room temperature protected from light. They were washed again and Ellman's reagent (200 [mu]l) was dispensed into each well using the Autodrop apparatus. During the enzymatic reaction the plates were protected from light. Absorbance at 414 nm was measured in each well after 4 h. All measurements for standards, control samples and samples were made in duplicate. Immunofit EIA/RIA (Beckman, CA) software with a four parameter logistic transformation was used for calibration and IC₅₀ measurement. IC₅₀ is the concentration of analyte giving 50% of the signal achieved in the absence of analyte.

Validation procedure

The intra- and inter-assay accuracy and precision of the method were assessed from control samples which were included in each assay following the recommendations for analytical methods validation in pharmacokinetic studies (14 ). Three concentrations of quality control samples were selected from the range covered by the assay according to the precision profile (Fig. 2 B). The quality control samples were aliquoted and stored frozen (-80^oC) with plasma samples until assay. They were used to estimate the accuracy in the upper, middle and lower parts of the calibration curve, which corresponded to three different slopes. The concentrations of the control samples were determined for each assay and these data were accumulated over five assays.

The specificity of the competitive hybridization assay was assessed by comparing standard curves obtained with deoxyribonucleotide A to those obtained with oligo A-related oligomers deleted at the 3'- or 5'-end (N-2 to N-6) in mouse plasma. IC₅₀ values were used to evaluate the relative affinity of the tested sequence for the sense target. Results were expressed in terms of percentage cross-reactivity by dividing the IC₅₀ value for oligo A by the IC₅₀ for the tested oligomer and multiplying by 100. `Cross-reactivity' is a term commonly used in the field of immunoanalysis (15 ) to measure the antibody response to substances other than the analyte. This term contrasts with `assay specificity', which describes the ability of an antibody to produce a measurable response only for the analyte of interest. In a wider sense, considering the analogy between antigen-antibody reactions and sense-antisense hybridization, we use the term cross-reactivity to describe the ability of substances other than the analyte to hybridize to the immobilized sense sequence.

The limit of quantification was defined as the lower concentration of quality control samples of oligo A with a coefficient of variation (CV, calculated from intra- and inter-assay accuracy) <25%.

In vitro plasma stability

An aqueous solution (30 [mu]l) containing oligo A (272 nM), oligo 2-S-A (272 nM) or oligo all-S-A (2.72 [mu]M) was incubated with 270 [mu]l 100% non-heat-inactivated blank mouse plasma at 37^oC in a water bath. Two tests were performed for each deoxyribonucleotide (A, 2-S-A and all-S-A). At the indicated times (0, 5, 15, 30 and 60 min) a 50 [mu]l fraction of each incubation solution was removed and added to 450 [mu]l assay buffer. The samples were analyzed by the competitive hybridization assay as described above.

Animal studies

Oligo A, 2-S-A and all-S-A were injected at a dose of 120 nmol/kg via the tail vein into OF 1 male mice weighing ~24 g. Animals were anesthetized by halothane at the indicated times (3, 6, 12 and 24 min). Blood was collected by abdominal vein puncture after halothane anesthesia. It was transferred to (5 mM) EDTA-coated collection tubes, centrifuged at +4^oC to obtain plasma samples and stored frozen (-80^oC) until assay. Three mice were used at each time point. Pharmacokinetic analysis was performed using SIPHAR software (SIMED SA, Créteil, France). Non-compartmental analysis gave a rough estimate of the half-life (T_1/2) and of the area under the plasma concentration-time curve (AUC) extrapolated to infinity. Plasma clearance (Cl) was calculated by dividing the dose by AUC.

RESULTS AND DISCUSSION

Analytical method

The aim of this work was to develop a competitive hybridization assay directly derived from the immunoenzymatic model by replacement of the antigen-antibody interaction by a sense-antisense oligonucleotide reaction. The principle of the method is illustrated in Figure 1 .

Figure 1. Principle of hybridization competitive assay. Assays were performed in 96-well plates. The sense target was covalently bound to the microtiter wells. 5'-Biotinylated oligo A was used as tracer. The assay involved competitive hybridization of tracer and antisense nucleotide to the solid phase-immobilized oligonucleotide. Solid phase-bound tracer oligonucleotide was measured after reaction with a streptavidin-acetylcholinesterase conjugate. After washing, solid phase-bound enzyme was measured using the colorimetric method of Ellman.

Figure 2. (A) Calibration curves of phosphodiester oligo A (crosses and dotted line) and of phosphorothioate analogs oligo 2-S-A (filled squares) and oligo all-S-A (empty circles). (B) Precision profile corresponding to six calibration curves for each analyte. CVs were determined using back-calculated concentrations for each standard point using a four parameter logistic transformation.

This principle was applied to measurement of plasma concentration of a 15mer phosphodiester oligonucleotide (oligo A) and of two modified sequences with phosphorothioate linkages (oligo 2-S-A and oligo all-S-A) using the same sense tracer. The hybridization could take place directly in plasma without any extraction procedure. Typical standard curves recorded in mouse plasma (Fig. 2 A) showed that there was competitive hybridization with target sense oligo A between the tracer and the three tested antisense sequences (oligo A, oligo 2-S-A and oligo all-S-A). As in competitive enzyme immunoassays, the colorimetric signal was inversely related to the amount of analyte present in the sample. Interestingly, using the same conditions the mouse plasma concentration of the sense target could be determined by this technique. Calibration curves (data not shown) were equivalent to those of oligo A. In this case there is competition between solid phase-immobilized sense target and soluble sense target for hybridization with the tracer.

Table 2 . Effect of tracer and sense target on signal intensity at zero oligo A concentration

Biotinylated tracer (pM)	Signal intensity Concentration of sense target (nM)
	1000	100	10	1	0.1	0
2200	0.890	0.907	0.440	0.055	0.007	0.003
220	0.677	0.691	0.329	0.030	0.007	0.001
55	0.373	0.370	0.147	0.009		0.002
11	0.055	0.073	0.019			0.003

Signal intensities were measured after 10 min reaction with Ellman's reagent.

Table 3 . Effect of tracer and sense target on signal intensity at 2.18 nM oligo A

Biotinylated tracer (pM)	Signal intensity (%) Concentration of sense target (nM)
	1000	100	10	1	0.1
2200	103	100	72	48	40
220	95	80	21	14	7
55	80	82	16	5
11	54	42	0

Signal intensities were measured after 10 min reaction with Ellman's reagent and are expressed as percentages of the signal intensity at zero concentration.

Two parameters were identified as governing the sensitivity of the method: the concentration of sense target covalently linked to the microwells and the tracer concentration. To optimize these parameters experiments were conducted with or without oligo A and using various tracer concentrations (from 11 to 2200 pM) and various concentrations of sense target (from 0 to 1000 nM) (Tables 2 and 3 ). In the absence of competitor oligo A the signal is proportional to the sense or tracer concentration (Table 2 ). The same signal was observed for a 100 and a 1000 nM sense target concentration (Table 2 ). This could be related to saturation of the coupling site on the well surface. In the presence of oligo A a decrease in signal corresponds to an increase in assay sensitivity. The lower the concentration of sense target or tracer the greater the sensitivity of the method (Table 3 ), but the longer the reaction period with Ellman's reagent to differentiate signal from background. In the light of these results, sense target and tracer concentrations were arbitrarily set at 5.5 and 55 pM respectively.

Validation parameters were assessed with three quality control samples in mouse plasma for each tested sequence (oligo A, oligo 2-S-A and oligo all-S-A) (Table 4 ). Acceptable intra- and inter-assay CV (<20%, 25% for lowest quality control concentration) were obtained in the range of calibration: 0.9-3.6, 0.2-1.8 and 4-36 nM for oligo A, oligo 2-S-A and oligo all-S-A respectively (Table 4 ). Compared with the nominal value for each antisense quality control the results showed biases below 20% (25% for lowest quality control concentration).

Table 4 . Accuracy and precision^a

	Quality control value (nM)	Intra-assay CV%	Inter-assay CV%	Bias (%)b
Oligo A	3.6	6.7	12	4.7
	1.8	11	3.8	13
	0.9	11	7.8	21
Oligo 2-S-A	1.8	11	6.1	2.1
	0.6	10	9.8	16
	0.2	22	25	18
Oligo all-S-A	36	9.5	17.8	11.9
	12	17	11	19
	4	1.7	10.6	16

^aEach value is the mean of five assays.
^bBias was calculated using inter-assay experiments.

The specificity of the method was assessed in plasma by comparing the ability of oligo A and oligo A-related oligomers deleted at the 3'- or 5'-end to compete with tracer (see Materials and Methods and Table 1 ). Cross-reactivity of 3' or 5' N-2 deleted oligomers was 48%, whereas it was below 3% with N-4 and N-6 3' or 5' oligomers. These results are in total agreement with those described by de Serres et al. (12 ) with scintillation proximity competitive hybridization assay of a phosphorothioate antisense oligonucleotide in plasma. Concentrations determined by this assay are reported as intact oligo A (or 2-S-A or all-S-A). We assumed that these concentrations could be overestimated by the presence of N-1 or N-2 or perhaps N-3 metabolites. Work is in progress to quantify this overestimation. Nevertheless, our experiments demonstrated that these metabolites were able to hybridize with the sense target in biological fluid (plasma) and hence, like oligo A, they should be endowed with pharmacological activity. In a sense our test could be considered as measuring the total `hybrido-reactivity' of the sample, which is very likely closely related to the biological activity of the oligonucleotide.

Application

In vitro screening of oligonucleotide hybridization potency. The ability of various unrelated sequences to hybridize with the sense target was evaluated using the hybridization competitive assay in a biological fluid (mouse plasma) by determining cross-reactivity coefficient (see Materials and Methods and Table 1 ). As expected, oligo K (or KS), whose sequence is not complementary with the sense target or antisense oligo A, was very poorly recognized by the assay (cross-reactivity <0.2). The phosphorothioate deoxyribonucleotide with all internucleotide phosphorothioate linkages (oligo all-S-A) was a weak competitor, with a cross-reactivity of 6%. It is very likely that this is at least partly due to the lower affinity of the phosphorothioate chimera (T_m < T_m of phosphodiester oligo A) for the sense target. Under these conditions competition is not favored and it would be interesting to use a phosphorothioate tracer which may increase (but not obligatorily) assay sensitivity. Surprisingly, the best competitor was phosphorothioate deoxyribonucleotide 2-S-A, with two internucleotide phosphorothioate linkages at the 3'-end, which showed 154% cross-reactivity. It is thus detected more sensitively than oligo A, which is entirely composed of phosphodiester bonds. This result is rather puzzling because oligo A is supposed to have a greater affinity for the immobilized sense sequence. However, it is worth noting that the standard curves established in mouse plasma not only reflect the hybridization potency of antisense oligonucleotide but may also be influenced by protein binding, oligo degradation and other uncharacterized interactions. Some of these interactions may be responsible for the difference observed between oligo A and oligo 2-S-A.

Figure 3. In vitro degradation of phosphodiester oligo A without (crosses and dotted line) or with 5 mM EDTA (crosses and full line) and of phosphorothioate analogs oligo 2-S-A (filled squares) and oligo all-S-A (empty circles) in mouse plasma. Each point represents the mean for two experiments.

In vitro stability. The hybridization competitive assay was used to examine the stability of the three tested compounds in mouse plasma (Fig. 3 ). Oligo A was degraded in mouse plasma [>85% in 1 h, apparent half-life (T_1/2) ~30 min]. Addition of 5 mM EDTA prevented degradation. Using HPLC analysis and radioactive nucleic acids, Sands et al. (16 ) observed the same half-life for a 20 base phosphodiester sequence.

Phosphorothioate sequences (oligo S-A or oligo all-S-A) were more stable in mouse plasma, with an apparent T_1/2 of ~1 h. In contrast, Sands et al. (16 ) noted no degradation of a phosphorothioate 20mer sequence in mouse plasma. This discrepancy concerning in vitro stability of phosphorothioates has been observed before and explained by Crooke et al. (17 ) as a consequence of nuclease inhibition by phosphorothioate. Our in vitro stability studies were performed with phosphorothioate plasma concentrations 4 (oligo all-S-A) to 40 (oligo 2-S-A) times lower than those used by Sands et al. (16 ).Plasma pharmacokinetics. The profiles of the first phase of plasma disapearance for the three tested antisense sequences are reported in Figure 4 . Oligo A cleared rapidly from blood, with a T_1/2 of ~4.8 min and approximate AUC and clearance (Cl) of 114 min ng/ml and 4.4 l/kg/min respectively. To our knowledge these results constitute the first pharmacokinetic data relative to intact phosphodiesters in mouse.

Figure 4. Plasma concentration of phosphodiester oligo A (crosses and dotted line) and phosphorothioate analogs oligo 2-S-A (filled squares) and oligo all-S-A (empty circles), i.v. administered to mice at a single dose (0.5 mg/kg). The bars represent the standard deviation of the mean values from three mice.

The introduction of two phosphorothioate linkages (oligo 2-S-A) did not affect the pharmacokinetic profile or parameters: T_1/2, AUC and Cl were 7.4 min, 188 min ng/ml and 2.7 l/kg/min respectively. The replacement of all phosphodiester linkages with phosphorothioate (oligo all-S-A) greatly increased AUC (3257 min ng/ml, by a factor of 29), decreased clearance (0.2 l/kg/min, by a factor of 22) and had no effect on T_1/2 (5.2 min). As a consequence of the introduction of phosphorothioate linkages, stronger plasma protein binding of oligo all-S-A may explain the discrepancy in T_1/2 between oligo 2-S-A and oligo A.

Half-lives observed in vitro are greater than those observed in vivo. As previously suggested (18 ), in vivo plasma metabolism plays a minor role in overall in vivo elimination. This observation confirms the good specificity of the competitive hybridization assay and strengthens our assumption that N-1 and N-2 and, maybe, N-3 metabolites are minor constituents in the intact nucleotide pool in plasma.

CONCLUSION

This work demonstrated that the competitive hybridization assay can be easily applied to in vitro screening and pharmacokinetic studies of phosphodiester oligonucleotides and analogs in biological fluids like plasma, without the need for radiolabeling or extraction. The technique is convenient and can be optimized quickly. Only a few days are necessary to synthesize the sense target and tracer oligonucleotide sequences. We are able to perform >100 assays in a single day. The method could easily be developed further using an automated workstation. This assay will undoubtedly constitute a useful tool for pharmacokinetic studies of modified antisense sequences in man.

REFERENCES

1 Uhlmann,E. and Peyman,A. (1990) Chem. Rev., 90, 543-584.

2 Crooke,S.T. (1995) Pharm. News, 2, 22-26.

3 Putnam,D.A. (1996) Am. J. Hlth Syst. Pharm., 53, 151-160.

4 Crooke,R.M. (1991) Anti-Cancer Drug Des., 6, 609-646.

5 Srinivasan,S.K. and Iversen,P. (1995) J. Clin. Lab. Anal., 9, 129-37. MEDLINE Abstract

6 Agrawal,S., Temsamani,J., Galbraith,W. and Tang,J.(1995) Drug Dispos. Clin Pharmacokinet., 28, 7-16.

7 Agrawal,S. and Temsamani,J. (1996) In Agrawal,S. (ed.), Molecular Medicine: Antisense Therapeutics. Human Press, Totowa, NJ, pp. 247-270.

8 Chevrier,D., Rasmussen,S.R. and Guesdon,J.-L. (1993) Mol. Cell. Probe, 7, 187-197.

9 Rasmussen,S.R., Larsen,M.R and Rasmussen,S.E. (1991) Anal. Biochem., 198, 138-142. MEDLINE Abstract

10 Nikiforov,T.N. and Rogers,Y.-H. (1995) Anal. Biochem., 227, 201-209.

11 Temsamani,J., Kubert,M. and Agrawal,S. (1993) Anal. Biochem., 215, 54-58. MEDLINE Abstract

12 de Serres,M., McNulty,M.J., Christensen,L., Zon,G. and Findlay,J.W.A. (1996) Anal. Biochem., 233, 228-33. MEDLINE Abstract

13 Pradelles,P., Grassi,J., Chabardes,D. and Guiso,N. (1989) Anal. Chem., 61, 447-453. MEDLINE Abstract

14 Shah,V.P., Midha,K.K., Dighe,S., McGilveray,I.J., Skelly,J.P.,Yacobi,A., Layloff,T., Viswanatan,C.T., Cook,C.E., McDowall,R.D., Pittman,K.A. and Spector,S. (1992) J. Pharm. Sci., 81, 309-312.

15 Davies,C. (1994) In Wild,D. (ed.), The Immunoassay Handbook. Stockton Press, New York, NY, pp. 89-91.

16 Sands,H., Gorey-Feret,L.J., Cocuzza,A.J., Hobbs,F.W., Chidester,D. and Trainor,G.L. (1994) Mol. Pharmacol., 45, 932-943. MEDLINE Abstract

17 Crooke,S.T., Graham,M.J., Zukerman,J.E., Brooks,D., Conklin,B.S., Cummins,L.L., Greig,M.J., Guinosso,C.J., Kornbrust,D., Manoharan,M., Sasmor,H.M., Schleich,T, Tivel,K.L. and Griffey,R.H. (1996) J. Pharmacol. Exp. Ther., 277, 923-937. MEDLINE Abstract

18 Shaw,J.P., Fishback,J.A., Cundy,K.C. and Lee,W.A. (1995) Pharmacol. Res., 12, 1937-1942.

---

*To whom correspondence should be addressed. Tel: +33 1 69 08 77 07; Fax: +33 1 69 08 59 07; Email: deverre@dsvidf.cea.fr
CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				G. Dai, K. K. Chan, S. Liu, D. Hoyt, S. Whitman, M. Klisovic, T. Shen, M. A. Caligiuri, J. Byrd, M. Grever, et al. Cellular Uptake and Intracellular Levels of the Bcl-2 Antisense G3139 in Cultured Cells and Treated Patients with Acute Myeloid Leukemia Clin. Cancer Res., April 15, 2005; 11(8): 2998 - 3008. [Abstract] [Full Text] [PDF]

				B Tavitian In vivo imaging with oligonucleotides for diagnosis and drug development Gut, June 1, 2003; 52(90004): iv40 - 47. [Abstract] [Full Text] [PDF]

				A. Bochot, E. Fattal, V. Boutet, J. R. Deverre, J. C. Jeanny, H. Chacun, and P. Couvreur Intravitreal Delivery of Oligonucleotides by Sterically Stabilized Liposomes Invest. Ophthalmol. Vis. Sci., January 1, 2002; 43(1): 253 - 259. [Abstract] [Full Text] [PDF]

This Article Abstract Print PDF (94K) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Search for citing articles in: ISI Web of Science (26) Request Permissions Commercial Re-use Guidelines for Open Access NAR Content Google Scholar Articles by Deverre, J. R. Articles by Grognet, J. M. Search for Related Content PubMed PubMed Citation Articles by Deverre, J. R. Articles by Grognet, J. M. Social Bookmarking          What's this?