`
`Winnewisser, W. (1972), Doctoral Dissertation, University
`of Munich, Munich, Germany.
`
`Ziegenhorn, J. (1970), Doctoral Dissertation, University of
`Munich, Munich, Germany.
`
`Covalent Coupling of Ribonucleic Acid to Agarose”
`
`Donald L. Robbersont and Norman Davidson
`
`ABSTRACT: A procedure for coupling RNA to a modified agar-
`ose resin is described. The amino group of NH2(CH2)6C02-
`CHI is coupled to the agarose by an alkaline CNBr proce-
`dure. The ester group is converted to a hydrazide by reaction
`with hydrazine. RNA is oxidized with IO4- to give a 3’-ter-
`minal aldehyde and then coupled to the resin by hydrazone
`formation. This coupling reaction is very slow unless nega-
`tively charged carboxyl groups on the resin surface are
`blocked by amide-bond formation with glycinamide as di-
`rected by a water soluble carbodiimide. The carbodiimide-
`glycinamide step seems to introduce some unidentified basic
`positively charged groups onto the resin surface, thus causing
`
`I n the development of techniques for the isolation of mole-
`
`cules which interact specifically with a given biological macro-
`molecule, it is often useful to attach the latter to a solid sup-
`port. A number of methods for attaching nucleic acids to
`solid supports have been developed and applied recently
`(Gilham, 1968, 1971; Nelidova and Kiselev, 1968; Alberts et
`ul., 1968; Litman, 1968; Alberts and Herrick, 1971; Jovin
`and Kornberg, 1968; Bonavida et a[., 1970; Poonian et al.,
`1971). We describe here a procedure for covalent coupling of
`RNA molecules via their 3’ terminus to an agarose gel. Our
`final objective, the isolation of high molecular weight single
`strands of DNA containing rDNA genes, has not yet been
`achieved, but the coupling method may be useful for other
`applications and we accordingly describe it here.
`The procedure involves the following steps: (1) coupling
`of NHz(CHJ5CO2CH3 (e-aminocaproic acid methyl ester) to
`agarose which has been activated by treatment with alkaline
`CNBr. The use of an 6-aminocaprolyl derivative to provide a
`functional group sterically separated from the agarose and
`therefore more accessible to reagents in solution was sug-
`gested by Cuatrecasas et al. (1968). The activation of a poly-
`saccharide support by alkaline CNBr for coupling to a nucle-
`ophilic reagent was described by Axen et al. (1967); (2) con-
`version of the caprolyl ester function to a hydrazide agarose,
`NH(CH2)C0NHNH2. This we call “hydrazide agarose.”
`(4) The next step in the procedure as originally planned was
`
`~
`
`~~~
`
`* From the Church Laboratories of Chemical Biology, California
`Institute of Technology, Pasadena, California 91 109. Receioed Sepfem-
`ber 10, 1971. This research has been supported by NIH Grant GM
`10991. D. R. has been supported by Training Grant GM 00086.
`t Present address: Department of Pathology, Stanford University
`Medical School, Stanford, Calif. 94305; to whom to address correspon-
`dence.
`
`~~
`
`nonspecific ionic binding of RNA to the resin. The nonspe-
`cific binding accelerates the rate of hydrazone-bond formation.
`Excess nonspecifically bound RNA is released by raising the
`pH above 9. Poly(U) on the resin surface can hybridize with
`poly(A) in solution, and the poly(A) can be eluted in a dena-
`turing MezSO solvent. With Escherichia coli 16s rRNA
`coupled to the resin, a 20-fold enrichment of rDNA from
`sheared denatured E. coli DNA of single-strand molecular
`weight about 7 x 104 was achieved by hybridization, followed
`by elution with NaOH. The procedure was not successful
`for enriching rDNA in a preparation of high molecular
`weight DNA.
`
`the oxidation of RNA at its 3’ terminus with IO4- and reac-
`tion of the oxidized RNA with the hydrazide function to give
`a hydrazone. Low yields were observed in this step and were
`attributed to electrostatic repulsion between carboxyl groups
`on the surface of the resin and the negative polyelectrolyte
`RNA. Therefore step (4) was inserted into the sequence of re-
`actions. (The coupling of RNA to polyacrylhydrazine-agar
`has been described by Nelidova and Kiselev (1968).) (3) In
`order to neutralize these putative carboxyl groups, we have
`treated the hydrazide agarose with glycinamide and a water-
`soluble carbodiimide, thus forming amide bonds to the car-
`boxyl groups (Hoare and Koshland, 1967). The product is
`called “blocked hydrazide agarose.” The resulting product
`is coupled to RNA according to step 4 above. In addition
`the carbodiimide-glycinamide
`to amide-bond formation,
`treatment evidently introduces some positive charges onto the
`resin. We find that the product is capable of binding un-
`oxidized RNA at pH’s below 7. The polynucleotide so bound
`is released by treatment at pH 8 or greater, leaving only oxi-
`dized RNA covalently attached by the hydrazone bond to the
`agarose.
`
`Materials and Methods
`Activation of Agarose. Agarose (100 ml; Sepharose 4B-200
`from the Sigma Chemical Co.) was washed by filtration on a
`coarse sintered-glass funnel with several changes of water and
`suspended to a total volume of 200 ml in water. Cyanogen
`bromide (10 g), dissolved in 200 ml of HzO, was added to the
`200-ml suspension of agarose in H20. The reaction flask was
`immersed in an ice bath and the agarose was kept in sus-
`pension by magnetic stirring. The suspension was immediately
`adjusted to pH 11 and maintained at this pH for 9 min by add-
`ing 4 N NaOH. It was then washed onto a coarse sintered-
`B I O C H E M I S T R Y , V O L . 1 1 , N O . 4, 1 9 7 2 533
`
`Exhibit 2103 Page 1
`
`Enzo Exhibit 2103 - Hologic, Inc. v.
`Enzo Life Sciences, Inc. - Case IPR2016-00822
`
`
`
`glass funnel with 2 1. of ice-cold 0.1 M NaHC0, buffer (pH 9).
`The activated agarose was suspended to a volume of 170 ml
`in 0.1 M NaHCO, buffer (pH 9). A solution of 9 g of t-amino-
`caproic acid methyl ester hydrochloride (Cyclo Chemical
`Co.) in 30 ml of 0.1 M NaHC0, buffer was readjusted to pH
`9 by the addition of 4 N NaOH. The ice-cold solution of the
`amino ester was added to the activated agarose suspension;
`gentle stirring at 4" was continued for 24 hr.
`Hydrazinoloysis. The resin from the above treatment was
`washed with 2 1. of ice-cold water and suspended to a total
`volume of 200 ml. A 20-1111 aliquot was removed for control
`assays and the remainder of the suspension in a 500-ml erlen-
`meyer flask was cooled on ice and 140 ml of hydrazine hy-
`drate, 99-100% (Matheson Coleman & Bell) was added
`slowly. The solution was then incubated at 70" for 15 rnin. The
`resin was maintained in suspension by gentle manual swirl-
`ing. The reaction temperature was observed with a thermom-
`eter. The mixture was then cooled to 25-30' and washed
`onto a sintered-glass funnel with 2 1. of water. A resin sus-
`pension should now be at pH -7. The resin was suspended in
`water to a volume approximately equal to twice the resin vol-
`ume for convenience in pipetting. It could be stored at 4" for
`several weeks without significant loss of aldehyde coupling
`capacity.
`Blocking of the Carboxyl Groups. The carbodiimide-cou-
`pling procedure is basically that of Hoare and Koshland(l967).
`Approximately one-third of the hydrazide agarose prepared
`as described above (-30 ml of settled resin) was washed with
`500 ml of water and resuspended in water to a total volume
`of 67 ml. Glycinamide hydrochloride (5.3 g; Cyclo Chemical
`Co.) was added with magnetic stirring and the pH adjusted
`to 4.75 with 1 N NaOH after which 0.67 g of 1-ethyl-3-di-
`methylaminipropylcarbodiimide . HC1 was added. The pH
`was maintained at 4.75 by addition of 1 N HC1 for 2 hr.
`This product, the blocked hydrazide resin, was washed at
`room temperature with several liters of distilled water and
`suspended in a convenient pipetting volume in 5 x 10-4 M
`EDTA (pH 7.2) for storage at 4".
`RNA. tRNA and synthetic polyribonucleotides were pur-
`chased from Schwarz-Mann. Some of the 16s and 23s E.coli
`rRNA used in this study was prepared from ribosomal sub-
`units by centrifugation of the latter in sucrose gradients con-
`taining SDS (Jeanteur et al., 1968). Another portion of 16s
`rRNA was a generous gift of Dr. T. Yamane; it had been iso-
`lated by phenol extraction of subunits of E. coli B ribosomes.
`Alkaline Phosphatase Treatmeqt. A stock solution of alka-
`line phosphatase was prepared by diluting 20 p1 of concentra-
`ted enzyme suspension (Worthington Biochemical Corp.)
`into 5 ml of 0.2 M Tris-acetate-0.01 M magnesium acetate
`(pH 7.7). This solution (1 ml), after heating at 90" for 10 min,
`released ca. 20 pmoles of P,/hr at 37" from p-nitrophenyl
`phosphate. (The enzyme is stable for 30 min at 85" in lo-?
`M Mg?+ (Torriani, 1966); it was our hope that the heat treat-
`ment might inactivate contaminating RNases.) To one vol-
`ume of RNA (0.5-3 mg/ml) in 0.2 M Tris-acetate-0.01 M mag-
`nesium acetate (pH 7.7) was added an equal volume of the di-
`luted alkaline phosphatase solution. The solution was incu-
`bated at 37" for 1 hr, chilled in ice, and treated with two vol-
`umes of ethanol. The RNA precipitate was collected by cen-
`trifugation for 5 min at 8000 rpm in the SS-34 rotor of a
`Sorvall centrifuge. The RNA pellet was redissolved in an
`amount of 0.1 M sodium acetate buffer (pH 5) equal to its
`initial volume and stored at -20". This treatment was used
`for all RNA samples except tRNA.
`Oxidation of RNA. The oxidation procedure is as described
`534 B I O C H E M I S T R Y , V O L . 1 1 , N O . 4, 1 9 7 2
`
`R O B B E R S O N A N D D A V I D S O N
`
`by Hunt (1965) and McIlreavey and Midgley (1967). To 1 ml
`of alkaline phosphatase treated RNA or tRNA (0.5-3 mg/
`ml) was added 0.14 ml of a fresh solution of 0.2 M NaIO,. The
`solution was allowed to stand for 1 hr at room temperature
`in the dark. The reaction was then stopped by addition of
`0.08 ml of ethylene glycol, followed by incubation in the dark
`for an additional 15 min at room temperature. The oxidized
`RNA was then either precipitated with ethanol, as described
`above, or dialyzed against several changes of 0.1 hi sodium
`acetate buffer (pH 5) at 4" over a 24-hr period. Such a step
`for the removal of formaldehyde, produced by the reaction
`between excess NaIO, and ethylene glycol, is essential before
`treatment with the hydrazide resin. The oxidized RNA sam-
`ples were stored at - 20" and used over a 3-day period. Control
`samples of unoxidized RNA were carried through all the
`above steps except for the addition of NaI04, for which dis-
`tilled water was substituted.
`Coupling of R N A to Resins. These reactions were conveni-
`ently carried out in 13 x 100 mm Kimax screw-cap glass
`culture tubes. The resins can then be separated from reactants
`and washed by centrifugation for several minutes in a clinica:
`centrifuge. Small portions of the resin suspension, prepared
`as described above, were washed several times in 0.1 h$ sodium
`acetate (pH 5) prior to the addition of RNA. The RNA solu-
`tion was added. After appropriate incubation time at 25" the
`supernatant solution was separated. The absorbance of the
`supernatant solutions at 260 nm minus its absorbance at 330
`nm was measured to determine RNA uptake. After taking
`absorbance readings, the supernatant was returned to the re-
`action tube and the resin resuspended. Reaction mixtures
`were kept in suspension by gentle rocking on a wrist-action
`shaker. Generally speaking, resin and RNA concentrations
`were adjusted so that the reactions would be complete within
`several hours.
`With the unblocked resins, the resins, after reaction, were
`washed exhaustively with 0.1 M sodium acetate buffer (pH
`5.0). Bound RNA was measured by digestion with pancreatic
`RNase at 50 pg/ml in standard saline citrate (pH 7.0) for one
`hour at 25". The supernatant absorbance at 260 nm was mea-
`sured to determine the amount of coupled RNA released. In
`all cases, there was good agreement between the amount of
`RNA coupled as estimated from the supernatant absorbance
`as a function of time during reaction and from the amount of
`coupled RNA released by RNase digestion.
`For experiments with the blocked hydrazide agarose, the
`uncoupled, nonspecifically bound RNA was released from
`the surface of the resin by treatment with NaHCOa buffer
`(pH 9). Nonspecifically bound RNA was estimated from the
`absorbance of the supernatant solutions from this treatment.
`Covalently bound RNA was assayed by washing exhaustively
`in 0.1 M NaHC0, buffer (pH 9) and subsequent digestion
`with RNase as described above. In all cases, the data from
`RNase digestion were corrected for hyperchromicity due to
`hydrolysis. These hyperchromicities at 260 nm in standard
`saline citrate at 25" were found to be 21.6z (tRNA), 6 . 4 z
`(poly(U)), 46.0% (poly (C)), and 11 % (16s rRNA).
`Coupling of Small Molecular Weight Aldehydes. The hydra-
`zide content of resins was checked by coupling to benzalde-
`hyde dissolved in 0.1 M sodium acetate (pH 5). The superna-
`tant absorbance at 248 nm was determined as a function of
`time. The molar extinction coefficient of benzaldehyde was
`measured as 12 X lo3.
`Solutions of the mononucleotide (5 '-UMP) were oxidized
`with equivalent amounts of NaI04 as previously described
`(Hunt, 1965). The coupling to the several resins was measured
`
`Exhibit 2103 Page 2
`
`
`
`C O V A L E N T C O U P L I N G O F R N A T O A G A R O S E
`
`TABLE I : Properties of the Coupling Reaction of Various Compounds to Hydrazide-Derived Sephar0ses.a
`
`~~
`
`~
`
`mmoles
`Coupled per rnl
`Vol of Reaction
`Settled
`Vol
`of Settled Resin
`at Satn
`Resin
`Compound
`Resin (ml)
`(ml)
`Initial Concn (M)
`1 x 10-4
`Benzaldehyde
`0
`0.13
`2.00
`Agarose (untreated)
`1 . 1 x 10-3
`1 x 10-4
`0.13
`2.00
`Benzaldehyde
`Hydrazide agarose
`2 .oo
`1 x 10-4
`0.26
`Unoxidized UMP
`Hydrazide agarose
`0
`1.04 x 10-4
`0 . 6 x 10-3
`0.26
`2.00
`Oxidized
`Hydrazide agarose
`2.64 x
`0
`0.26
`1.52
`Unoxidized tRNA
`Hydrazide agarose
`2.52 x 10-6
`8.0 x 10-6
`0.26
`1.52
`Oxidized tRNA
`Hydrazide agarose
`1.09 x 10-7
`Unoxidized 16s rRNA
`0.26
`1.52
`Hydrazide agarose
`0
`1 .oo x 10-7
`1 .o x 10-7
`Oxidized 16s rRNA
`0.26
`1.52
`Hydrazide agarose
`1 x 10-4
`Benzaldehyde
`0
`0.13
`2.00
`Agarose (untreated)
`0 . 8 x 10-3
`1 x 10-4
`0.13
`2.00
`“Blocked” hydrazide agarose
`Benzaldehyde
`1 x 10-4
`0.26
`2.00
`“Blocked” hydrazide agarose
`Unoxidized UMP
`0
`1 . 1 x 10-3
`1 x 10-4
`0.26
`2.00
`“Blocked” hydrazide agarose
`Oxidized UMP
`6 . 8 x
`0.026
`0.52
`Unoxidized tRNA
`“Blocked” hydrazide agarose
`0
`6 . 8 x
`87 x
`0.026
`0.52
`18
`Oxidized tRNA
`“Blocked” hydrazide agarose
`2 . 8 x 10-7
`Unoxidized 16s rRNA
`0.021
`0.52
`“Blocked” hydrazide agarose
`0
`c
`3 . 5 x 10-6
`2 . 8 x 10-7
`Oxidized 16s rRNA
`9
`0.021
`0.52
`“Blocked” hydrazide agarose
`a All reactions were carried out at ambient temperature (25 & 1 ”). * Reaction time required to attain one-half the maximum
`amount of RNA coupled for the reaction conditions indicated. 0 A low level of unoxidized rRNA (<lo% of the input rRNA)
`bound to the surface of this resin was not released by washing at pH 9 (see Text).
`
`tl/j
`(min)
`
`65
`
`60
`
`65
`
`70
`
`120
`
`120
`
`by observing the absorbance at 262 nm of the resin superna-
`tant as a function of time. In all cases, spectra were taken
`throughout the course of the reaction to verify that the uv
`absorbance was associated with the compound of interest.
`The molar extinction coefficient of UMP was taken as 1.0 X
`lo4 at 262 mw.
`
`4.0 -
`3.2 -
`-
`2.4 -
`
`1.6-
`
`‘260
`Vol.
`
`Results
`The results are presented in Table I. The hydrazide agarose
`is capable of binding approximately
`mmoles of C6&
`CHO or oxidized 5’-UMP per ml of resin. Only about 0.01 as
`many of the hydrazide groups react with oxidized tRNA, and
`the coupling of 16s RNA is lower by an additional factor of
`100. The blocking treatment with carbodiimide and glycin-
`amide does not significantly affect the binding capability for
`small molecules, but the coupling of tRNA is increased by a
`factor of 10, and that of 16s RNA by a factor of 35. Further-
`more, the time to reach the plateau level of binding was de-
`creased by the blocking reaction. The control studies show
`that in all cases oxidation of the 3‘ terminus of the RNA is
`necessary for reaction.
`Amino acid analysis, determined in a Beckman amino acid
`analyzer after hydrolysis of the blocked resin in 1 N HC1 at
`100” for 24 hr, indicated 2.43 X 10-4 mmoles of glycine/ml
`of settled resin resulting from the blocking reaction. (We are
`
`FIGURE 1: Release of tRNA nonspecifically bound to blocked hy-
`drazide agarose as a function of pH. A sample of unoxidized tRNA
`(250 pg) was incubated with 0.53 ml of (volume of settled resin)
`blocked hydrazide agarose in 0.1 M sodium acetate buffer (pH5.0)ina
`total volume of 2.0 ml at 25’. The amount of tRNA bound reached
`saturation after 20 min. The resin was then washed with carbonate
`free water and the pH adjusted to 4.0 in a total volume of 2.0 ml. The
`pH was increased by adding microliter quantities of 0.015 M NaOH
`solution. After each addition of base the pH was recorded and the
`solution centrifuged to permit the Atsa of the supernatant to be
`determined. Each point then represents the observed pH and ob-
`served
`of the supernatant multiplied by the volume of the
`solution at that point. The midpoint for release of tRNA from the
`resin corresponds to the titration of a monobasic acid with a pK. of
`1.54.
`
`FIGURE 2: Effect of RNA size on the amount nonspecifically bound
`to blocked hydrazide agarose. Samples of unoxidized RNA were
`nonspecifically bound to blocked hydrazide agarose in 0.1 M sodium
`acetate buffer at pH 5 and 25”. For each RNA. the amount that is
`bound is plotted against the reciprocal of sedimentation coefficient
`for that RNA, which is a measure of the hydrodynamic size of the
`RNA. The data presented satisfy the relationship n -16 X 10-s/s~~.w.
`B I O C H E M I S T R Y , V O L . 1 1 , N O . 4, 1 9 7 2 535
`
`Exhibit 2103 Page 3
`
`
`
`R O B B E R S O N A N D D A V I D S O N
`
`(a)
`
`(C)
`
`A
`
`"
`w Unax p . r U .. BHS
`0.45 A268 UnlfS
`W p . r C = BHS
`0.06 A,,,
`units
`0-0 p . r U = BHS 0.78
`units
`0.1 M NoCl 0.01 M Cacodylate pH 6.98
`1 1 T = 25°C Tot vol. = I.Oml 0.13 rnl Settled Resin
`
`A256
`.5
`
`0
`
`0
`
`~
`
`"
`
`"
`
`IO0
`'
`
`~
`
`"
`
`200
`~
`'
`t (min)
`
`"
`
`300
`"
`~
`
`A-J
`
`'/IO00 1200
`
`
`
`(b)
`
`DMSO Denaturation of p ' r A : r U
`T = 25OC 0.1 M NaCl 0.01 M Cacodylate pH 6.98
`1.81 Tm + 65.6% DMSO
`
`I .6
`
`t
`
`h260 1
`4l
`2 :__dp,
`
`0
`0
`
`20
`20
`
`40
`40
`'10 DMSO
`
`1
`60
`60
`
`80
`1
`80
`(d 1
`
`DMSO Denaturotion of prA. rU = BHS
`T = 25°C 0.1 M NaCl 0.01 M Cocodylate p H 6.98
`TmZS 66.5% DMSO
`
`-
`
`p . r A : r U = BHS
`w p r U = BHS
`W Unox D,rU .. BHS
`
`. _ .
`
`1.6 -
`
`1.2-
`
`.8 -
`-
`.4
`
`)/--
`
`n
`
`-
`
`0.45
`
`u n i 5
`
`w p ' r C = BHS
`Unax p . rU .. BHS 0.06 A2so units
`M P ' r U BHS 0.78 A 2 6 0 UnltS
`0.1 M NoCl 0.01 M Cacodylate
`pH 6.98
`50% DMSO
`..
`
`"
`
`!-
`
`T 2 5 ° C Tot vol. = 1.0rnl 0.13ml Settled Resin
`
`0
`
`20
`
`60
`
`1
`80
`
`01 1
`0
`
`'
`
` I
`
`'
`
` " '
`IO0
`
`I
`
`'
`
`'
`
`G k S k
`
`' f
`
` '
`'
`200
`40
`% DMSO
`t ( m i d
`FIGURE 3 : Hybridization of poly(rA) to resin-coupled poly(rU). (a) Hybridization of poly(rA) to poly(rU) which had been coupled
`to the blocked hydrazide agarose (BHS). The concentration of poly(rA), indicated by the absorbance at 256 nm in the supernatant, was de-
`termined as a function of time (0). The buffer used was 0.1 M NaCl4.01 M sodium cacodylate (pH 6.98). The hybridization was carried out
`at 25" in a total volume of 1.0 ml of which 0.13 ml represented the settled resin to which polynucleotide has been coupled. At the indicated
`times after addition of p .rA, the suspension was centrifuged and the supernatant absorbance at the indicated wavelength was determined.
`The supernatant was then returned to the reaction tube and the resin resuspended. The resin was kept in suspension by a wrist-action shaker.
`The symbols p.rC = BHS and p.rU = BHS mean oxidized polynucleotides covalently coupled to the resin. The symbol unox p.rU. . BHS
`means unoxidized polynucleotide noncovalently bound to the resin. One control sample consisted of p.rC coupled to BHS (a total of 0.45
`,4268 absorbance unit) (0). The other control sample consisted of BHS resin incubated with unoxidized p'rU which was subsequently washed
`as described in Materials and Methods (A). This resin contained <0.06 ,4260 absorbance unit of p.rU. The BHS sample to which p.rU had
`been coupled contained 0.78 A260 absorbance unit of p.rU. (b) Same as for (a) except that the hybridization buffer is 0.1 M NaCl-O.01 M
`sodium cacodylate (pH 6.98), 50 vol % Me2S0. (c) Solution denaturation as evidenced by the hyperchromicity ( h 6 0 ) of a 1 : 1 complex of
`p.rA and p.rU by increasing concentrations of Me2SO (volume per cent) at constant ionic strength (0.1 M NaCl-0.01 M sodium cacodylate,
`pH 6.98) at 25". The decrease in hyperchromicity at concentrations of MezSO greater than 80% probably indicate precipitation of the poly-
`nucleotide(s). The midpoint of the denaturation corresponds to 65.6% Me2S0. (d) MezSO denaturation at constant ionic strength (0.1 M
`NaCl-O.01 M sodium cacodylate, pH 6.98) of the complex formed between p.rA and BHS coupled p.rU, resulting in release of p.rA to the
`supernatant. The samples used here were derived from part h after exhaustive washing with the hybridization buffer. The amount released
`was measured after centrifugation. The midpoint for release of p .rA to the supernatant corresponds to 66.5 2 Me2SO.
`
`grateful to Mr. Jon Racs for this analysis.) The blocked hy-
`drazide agarose exhibits binding of unoxidized polynucleo-
`tides at p H 5. The "nonspecifically" bound polynucleotide
`can be released from the surface of the resin by raising the p H
`of the washing medium. Figure 1 displays a plot of the
`amount released as a function of pH. The curve has the same
`shape as a titration curve of a monobasic acid with a pK, of
`7.5. More than 95% of the nonspecifically bound tRNA is re-
`leased at p H 9. (The blocked hydrazide resin loses binding ca-
`pacityatarateofabout IO%/month at4Oin5 X 1 0 - 4 ~ E D T A
`(PH 7.2))
`The larger the RNA, the smaller the number of millimoles
`of RNA nonspecifically bound per milliliter of settled resin
`(Figure 2). The data fit the relation (millimoles bound) =
`6 x 1 0 - 5 / ~ ~ ~ . ~ ,
`is the sedimentation coefficient of
`where s
`~
`~
`.
`~
`
`the RNA.
`536
`
`B I O C H E M I S T R Y , V O L . 1 1 , N O . 4, 1 9 7 2
`
`The observations are qualitatively consistent with a picture
`of binding to some positively charged basic sites that titrate
`to uncharged sites with a pK, below 7.5. However, we have
`not been able to identify the positively charged basic groups
`introduced by the carbodiimide-glycinamide treatment.
`RNA and polynucleotides coupled to the blocked resin are
`available, to some extent, for hybridization with comple-
`mentary polynucleotides. Figure 3a,b shows that 0.5-0.7
`A260 unit of poly(A) will bind t o a resin containing 0.78 ,4260
`unit of covalently coupled poly(U), but there is no significant
`binding to a poly(C) resin. The hybridization can be carried
`out in aqueous solution or in 50 % Me2S0.
`A 1 :1 poly(A).poly(U) complex in 0.1 M NaCl at 25" de-
`natures at 65-70 % Me2S0 ; the same Me2S0 concentration
`will cause the hybridized poly(A) to be dissociated from the
`poly(U)-containing resin (Figure 3c,d).
`
`Exhibit 2103 Page 4
`
`
`
`C O V A L E N T C O U P L I N G O F R N A T O A G A R O S E
`
`A sample of 82P-labeled E. coli DNA was sheared to a
`double-strand molecular weight of 1.4 X lo6. Four milliliters
`of solution, containing 50 pg of DNA in 50% Me2S0, 0.1 M
`NaC1, 0.01 M Tris, and 0.001 M EDTA (pH 8), was incubated
`for 12 hr with 0.26 ml of resin containing 48 pg of coupled
`16s RNA. The resin was washed and incubated a second
`and a third time with 4 ml of fresh DNA solution. After fur-
`ther washing, bound DNA was extracted from the resin with
`0.1 N NaOH. A quantity of 0.09 pg of DNA was recovered.
`By membrane filter hybridization, 7 . 2 z of this product, as
`compared to 0.34% of the starting DNA, was complementary
`to 14C-labeled 16s rRNA. Thus, an enrichment of ribosomal
`genes by a factor of 21 with a 15 % yield was achieved. (In a
`similar previous experiment with only one incubation with
`DNA, an enrichment by a factor of 21 with a 40% yield was
`achieved.)
`It should be noted that several very successful procedures
`for isolating rDNA or tDNA from short DNA strands by
`hybridization with rRNA or tRNA and separation of RNA-
`DNA hybrids from the. remaining single-strand DNA have
`already been described (Colli et al., 1971; Kohne, 1968;
`Marks and Spencer, 1970; Brenner et al., 1970). Our final
`objective was the isolation of rDNA sequence in long DNA
`strands. Several hybridization experiments with DNA with an
`average single-strand molecular weight of 19 x 106 were tried,
`but very little enrichment was achieved.
`
`Discussion
`We have described a procedure for the formation of hydra-
`zone bonds between terminally oxidized RNA and hydrazide
`groups on a resin. For high molecular weight RNA, the rate
`of the coupling reaction is greatly accelerated if the RNA is
`nonspecifically bound to the resin. In the present instance, the
`nonspecific binding seems to be due to an electrostatic inter-
`action between the RNA and some unidentified basic groups
`introduced into the resin by the carbodiimide-glycinamide
`step.
`It is quite likely that there are steric barriers to reaction of a
`3'4erminal aldehyde group on an RNA with a hydrazide on
`the surface of a resin. In general, for a random coil molecule,
`the topological end is buried on the physical inside of the
`random coil particle, with only a low probability of being
`available for reaction with a reagent that cannot penetrate
`the coil. If the nucleic acid is adsorbed as a mobile two-di-
`mensional random coil on the surface of the resin, the reaction
`rate of the 3'4erminal aldehyde wtih a hydrazide would
`be accelerated by the reduction in dimensionality effect
`discussed by Adam and Delbriick (1968).
`We do not know the nature of the putative positively
`charged basic group introduced by the carbodiimide-glycin-
`amide step. The titration curve of Figure 1 suggests that the
`pK. is less than 7.5, since the dissociation of a complex be-
`tween the BH+ group (BH+ signifies the protonated basic
`group) and the negative RNA site should occur at a higher
`pH than that at which BH+ dissociates to B and H+. An
`obvious first guess as to the identity of the basic group would
`be the dirnethylaminopropyl group of the carbodiimide, but
`this should have a pK, greater than 8.
`Because of other commitments, we were unable to pursue
`the study of the enrichment of rDNA genes by hybridization
`to resin coupled RNA beyond the few experiments reported
`here. One point that we were unaware of at the time should
`
`be mentioned. Hunt (1969) reports that the RNA hydrazone
`bond is sensitive to cleavage by @ elimination in the presence
`of amine buffers. Therefore, the Tris used in our DNA experi-
`ments should not have been used.
`Our few experiments suggest that the complementary
`rDNA sequence in a high molecular weight DNA strand are
`less available for reaction with rRNA on the resin surface
`than are the same sequences on a shorter DNA strand. This
`we believe is an example of the excluded volume effect in
`nucleic acid renaturation (Wetmur and Davidson, 1968 ;
`Wetmur, 1971).
`The coupling procedure, or some improved modification, is
`potentially useful for many other applications, so we have
`accordingly described it here.
`
`Acknowledgment
`Many of the specific ideas for the chemical coupling
`described here were developed in the course of discussion
`with Professor Michael A. Raftery, to whom we are deeply
`grateful.
`
`References
`Adam, G., and Delbriick, M. (1968), in Structural Chemistry
`and Molecular Biology, Rich, A., Davidson, N., Ed.,
`San Francisco, Calif., W. H. Freeman.
`Alberts, B. M., Amodio, F. J., Jenkins, M., Gutmann, E. D.,
`and Ferris, F. J. (1968), Cold Spring Harbor Symp. Quant.
`Biol. 33,289.
`Alberts, B., and Herrick, G. (1971), Methods Enrymol. 21, 198.
`Axen, R., Porath, J., and Ernback, S. (1967), Nature (London)
`214,1302.
`Bonavida, B., Fuchs, S., and Sela, M. (1970), Biochem.
`Biophys. Res. Commun. 41,1335.
`Brenner, D. J., Fournier, M. J., and Doctor, B. P. (1970),
`Nature (London) 227, 448.
`Colli, W., Smith, I., and Oishi, M. (1971), J. Mol. Biol. 56,117.
`Cuatrecasas, P., Wilchek, M., and Anfinsen, C. G. (1968),
`Proc. Nut. Acad. Sci. U. S. 61,636.
`Gilham, P. T. (1968), Biochemistry 7,2809.
`Gilham, P. T. (1971), Methods Enzymol. 21,191.
`Hoare, D. G., and Koshland, D. E., Jr. (1967), J. Biol. Chem.
`242,2447.
`Hunt, J. A. (1965), Biochem. J. 95,542.
`Hunt, J. A. (1969), Biochem. J . 116,119.
`Jeanteur, P., Amaldi, F., and Attardi, G. (1968), J. Mol. Biol.
`33,757.
`Jovin, T. M., and Kornberg, A. (1968), J. Biol. Chem. 243,
`250.
`Kohne, D. E. (1968), Biophys. J. 8,1104.
`Litman, R. M. (1968), J. Biol. Chem. 243,6222.
`Marks, A., and Spencer, J. H. (1970),J. Mol. Biol. 51,115.
`McIlreavey, D. J., and Midgely, J. E. M. (1967), Biochim.
`Biophys. Acta 142,47.
`Nelidova, 0. K., and Kiselev, L. L. (1968), Mol. Biol. 2,47.
`Poonian, M. S. Schlabzch, A. J., and Weissbach, A. (1971),
`Biochemistry 10,424.
`Torriani, A. (1966), in Procedures in Nucleic Acid Research,
`Cantoni, G. L., and Davies, D. R., Ed., New York, N. Y.,
`Harper & Row, p 224.
`Wetmur, J. G. (1971), Biopolymers 10,601.
`Wetmur, J. G., and Davidson, N. (1968), J. Mol. Biol. 31, 349.
`
`B I O C H E M I S T R Y , V O L . 1 1 , N O . 4, 1 9 7 2 537
`
`Exhibit 2103 Page 5