Eur. J. Biochem. /43, 613—620 (1984)
`© FEBS 1984
`
`Single-step elongation of oligodeoxynucleotides
`using terminal deoxynucleotidyl transferase
`
`Herbert SCHOTTand Herbert SCHRADE
`
`Institut fiir Organische Chemie, Universitat Tiibingen
`
`(Received March 26, 1984)
`
`— EJB 840317
`
`Procedures for the stepwise addition of one or more deoxyribonucleotide residues to the 3’ end of an
`oligodeoxyribonucleoside phosphate acceptor using commercially available terminal deoxynucleotidyl transferaseis
`described. 2—80 nmolofacceptors with a chain length offour, five or nine monomerunits were elongated with a
`single 2’-deoxyribonucleoside 5’-triphosphatein yields of 20 —30°,. The monomerscarried no protecting groups and
`were used both radioactively labelled and unlabelled.
`The elongated oligodeoxynucleoside phosphates were isolated by reverse-phase (Nucleosil C,,) high-
`performance liquid chromatography or paper chromatography. The isolated products were sequenced by the
`fingerprint method. Advantages and disadvantagesof this new methodologyfor the enzymatic synthesis of defined
`oligodeoxynucleotides are discussed.
`
`Besides the well-known chemical synthesis, two different
`approaches for
`the controlled addition of monomers or
`oligomers to the 3’ end of an oligodeoxyribonucleotide primer
`are known.
`In one approach, deoxyribonucleotide residues
`derived from 2’-deoxyribonucleoside 5’-diphosphates may be
`added to oligodeoxyribonucleotide primers with polynu-
`cleotide phosphorylase, purified from Escherichia coli B [1 — 3).
`This method allowsthe synthesis of oligodeoxyribonucleotides
`of defined sequences by repeated addition of single residues,
`applying conventional biochemical procedures. The other
`approach demonstrates conditions which allow the efficient
`addition of 2'-deoxyribonucleoside 3’,5’-bisphosphates
`to
`oligodeoxyribonucleotides and ofsingle-stranded oligodeoxy-
`ribonucleotides to one another using T4 RNA ligase [4].
`Both methods have been used to synthesize a considerable
`number of oligonucleotides. However, general application of
`these techniques is limited because the preparation of the pure
`polynucleotide phosphorylase requires many critical purifi-
`cation steps and the condensation time (several days for one
`reaction step) with T4 RNA ligase takes too much time in
`comparison to the rapid chemical synthesis which can be done
`by an automated synthesizer,
`However, an enzymatic synthesis should be a very attractive
`technique for the simple preparationof oligonucleotides in high
`purity, if the required enzymes are easily available and if the
`condensation time is short enough. A well known and com-
`mercially available enzyme is terminal deoxynucleotidyl trans-
`ferase which catalyses the polycondensation of deoxynu-
`
`cleoside triphosphates onto a suitable acceptor 3’-hydroxy
`group. The acceptor must have at
`least
`three phosphate
`residues. This enzyme has found wide application in DNA and
`RNA sequence analysis and is widely used for the purpose of
`construction of recombinant DNA invitro. Depending on the
`presence ofeither Mg** or Co?" ions, the enzyme may be used
`to add a number, preferentially 15—20 nucleotide residues, to
`the 3’ end ofeither single-stranded or double-stranded DNA
`fragments [5—9].
`Several series of homologous oligomers were examined for
`initiator utilization and product distribution in the homopoly-
`condensations and the copolycondensations, but the stepwise
`elongation of oligonucleotides by the repeated addition of
`monomers using terminal deoxynucleotidyl transferase has not
`yet been reported.
`Usually this enzyme condenses a varying number of
`monomersonto the 3’ end of an acceptor in a polycondensation
`reaction yielding oligonucleotides of different chain length.
`Even if the elongation could be shifted to a monoaddition, a
`complex reaction mixture would be obtained.
`Wehaveinvestigated the enzymatic reaction conditions and
`the chromatographic separation procedure which allows the
`stepwise synthesis of defined oligodeoxynucleotides using
`terminal deoxynucleotidyl transferase.
`
`MATERIALS AND METHODS
`
`Acceptors
`
`triethylammonium hydrogen car-
`Abbreviations. EtsNHCO,,
`bonate; HPLC, high-performance liquid chromatography.
`Enzymes. Terminal deoxynucleotidyl transferase (EC 2.7.7.31):
`polynucleotidyl phosphorylase (EC 2.7.7.8); T4 RNA ligase (EC
`6.5.1.3): alkaline phosphatase (EC 3.1.3.1); snake venom phos-
`phodiesterase (EC 3.1.4.1); spleen phosphodiesterase (EC 3.1.4.18):
`T, polynucleotide kinase (EC 2.7.1.78).
`
`The acceptors (dT)4, (dC), and d(A-A-A-C) were synthe-
`sized in our laboratory (H. Schott, unpublished); d(A-A-A-G)
`[12], d(G-A-A-A) [12] and d(C,4,T,) [13] were isolated from
`partial hydrolysates of chemically degraded herring sperm
`DNA. dTTP, dATP and dCTP were obtained from Boehringer,
`Mannheim;
`[«#-*?P]dATP (specific activity = 2 Ci/mmol),
`
`CUREVAC EX2040
`CUREVAC EX2040
`Page 1
`Page 1
`
`

`

`614
`
`[y-*?PJATP and [8-7HJdGTP(specific activity = 9 Ci/mmol)
`were purchased from Amersham Buchler, Braunschweig.
`
`Enzymes
`
`Alkaline phosphatase (calf intestine), snake venom phos-
`phodiesterase and spleen phosphodiesterase were obtained
`from Boehringer, Mannheim; T, polynucleotide-kinase was
`purchased from Amersham Buchler and terminal deoxynu-
`cleotidyl transferase from Pharmacia P-L Biochemicals, Frei-
`burg.
`
`Chromatography
`
`DEAE-cellullose thin-layer plates: Polygram, Cel 300
`DEAE HR-2/15, were obtained from Macherey & Nagel,
`Diiren, and cellulose paper no. 2316 from Schleicher & Schiill,
`Dassel.
`
`Condensation reactions
`
`The elongation of acceptors was carried out at 37°C in the
`following buffer: 50 1! H,O, 70 p12 M Tris/HCl pH 7.6, 70 ul
`2M sodium cacodylate, 10 ul 0.1 M CoCl,,
`1 ul 0.1 M di-
`thioerythritol. The conditions of the condensation reactions are
`summarised in Table 1.
`1 U (condensations IT] —VI, 1X) or 5 U
`(condensations VII, VIII) of alkaline phosphatase were added
`at the end of the reaction time and the incubation continued for
`1h.
`
`Separations of reaction mixtures from condensation | and
`1! were performed on DEAE-cellulose thin-layer plates by
`homochromatography using homomix VI according to Jay et
`al. [14]. Elongated products were detected autoradiographi-
`cally, eluted with Et;NHCO, buffer and identified by the
`nearest-neighbour analysis [11, 15] (see Fig. 1).
`Separation of the reaction mixture from condensation VII
`was achieved by descending paper chromatography. First the
`mixture was chromatographed in solvent A: ethanol/iM
`ammonium acetate (pH 7.5) (7:3, v/v). Unreacted monomer.
`which migrates closely to the solvent front, was detected by
`liquid-scintillation counting of paper segments. The segment of
`the paper chromatogram which contained the spot of the
`monomer wascut off. The rest of the paper chromatogram was
`dried and then rechromatographed in solvent B: 1-propanol
`/concentrated ammonia/water (55:10:35, v/v/v). Elongated
`oligonucleotides (Fig. 4) were detected by liquid scintillation
`counting of paper segments, eluted from the paperand finally
`lyophilised.
`Separations of reaction mixtures from condensations II —
`V1, VII1 and LX were carried out by HPLC (Fig. 2). Nucleosil
`Cyx, particle size 7.0 1m (Macherey & Nagel, Diiren) was used
`as the stationary phase. The column (250 x 4.6mm internal
`diameter) was eluted under isocratic conditions with a mixture
`of 72°, A and 28°, B, except condensation mixture [X which
`was separated using 68", A and 32° B.A = 0.1 Mammonium
`acetate (pH 7.5); B = methanol/water (60:40, v/v).
`A high-pressure liquid chromatograph (Waters 6000 A) was
`equipped with an ultraviolet detector (Du Pont Instruments
`model 836) and a high-pressure sampling valve (Waters U6K).
`The monitoring wavelength was 254nm,
`the flow rate
`it ml/min. The flow detector for radioactivity consisted of a
`detector unit (BF 5026) with a measuring cell (BF 5029) and a
`measuring unit
`(BF 2240: ratemeter BF 2305, Berthold,
`Wildbad). The time constant used was 2s,
`
`RESULTS
`
`At first, chemically synthesised tetranucleoside trisphos-
`phates of different sequences, as well as these isolated from
`partial DNA hydrolysates, were used as acceptors and elon-
`gated with purine or pyrimidine mononucleotides. Two re-
`actions were evaluated qualitatively and four quantitatively. In
`condensation VII, all components of the mixture of sequence-
`isomeric nonanucleoside octakisphosphates d(C,, T;), isolated
`from the partial hydrolysate of depurinated DNA, were
`elongated together with a dG residue. This reaction was also
`investigated quantitatively.
`In condensation VILL the syn-
`thetically obtained tetranucleoside trisphosphate d(A-A-A-C)
`was clongated to d(A-A-A-C-A) which was then converted to
`d(A-A-A-C-A[*?P]pT) (condensation [X).
`
`Elongation of homologous pyrimidine nucleoside phosphates
`
`For qualitative investigations, the two chemically synthe-
`sised homologous pyrimidine tetranucleoside trisphosphates,
`(dC), and (dT),. were condensed with [z-**P]JdATPto obtain
`radioactively labelled elongated products. The acceptor was
`applied ina 40-fold molar excess (with respect to the monomer)
`favouring the formation of the monoaddition products d(T-T-
`T-T[*?P]pA) and d(C-C-C-C[*?P]pA), respectively. The con-
`densation wascarried out according to the conditions givenin
`Table 1 with 1 U terminal deoxynucleotidyl transferase/nmol
`monomer. After incubation at 37 °Cfor 4h, the crude reaction
`mixtures were separated on DEAE-cellulose precoated sheets
`by homochromatography [10] according to the chain length.
`Because of the incorporation of [7*P}pdA units,all elongated
`oligonucleotide phosphates were detected by the autoradiog-
`raphy of the chromatograms. In the autoradiograms of each
`reaction mixtures, only one spot occurred which corresponded
`to the monoaddition products d(T-T-T-T[??P]pA) and d(C-C-
`C-C[>?P\pA) respectively. This assumption was confirmed by
`the ‘nearest-neighbour analysis’ [11, 15].
`To determine the number of added adenylic acid units, the
`spot of the corresponding radioactively labelled oligonu-
`cleoside phosphate was scraped from the DEAE-cellulose
`plate, eluted with the volatile Et, NHCO, buffer and sub-
`sequently lyophilised. Analiquot of the freeze-dried substances
`was totally hydrolysed with spleen phosphodiesterase so that
`nucleoside 3’-monophosphates were obtained. The hydrolysate
`was separated by paper chromatography and the radioactively
`labelled mononucleotides wereidentified. Only dT[**P]p could
`be found in the totally hydrolysed d(T-T-T-T[?*P]pA) and only
`dC{??P]p in the hydrolysate of d(C-C-C-C[*?P]pA) (Fig. 1).
`After the total hydrolysis of both elongation products with
`snake venom phosphodiesterase which liberates nucleoside
`5’-monophosphates, only [°?P]pdA was found, as expected.
`These results demonstrate that in the enzymatic reaction both
`acceptors are elongated, each by only a single adenylic acid
`unit.
`
`Elongation ofpurine nucleoside phosphates
`
`In the following two quantitatively evaluated reactions, the
`sequence-isomeric purine tetranucleoside trisphosphates d(G-
`A-A-A) and d(A-A-A-G) were used as acceptors which had
`been isolated from depyrimidinated herring sperm DNA [12].
`Both acceptors were condensed with dCTP and dTTP re-
`spectively. In contrast to the preceding examples, the excess of
`acceptor relative to the amount of monomer was now less than
`20°; however, again about 1 U enzyme/nmol monomer was
`
`CUREVAC EX2040
`CUREVAC EX2040
`Page 2
`Page 2
`
`

`

`Table 1. Conditions of the enzymatic elongation of oligodeoxyribonucleoside phosphates using terminal deoxynucleotidyl transferase and
`mononucleoside triphosphates
`The monomerwas obtained as a lyophilisate. The enzyme concentration was 10 U/ul; the reaction was carried out at 37°C
`
`
`
`
`
`
`Conden- Amount Enzyme Buffer—ReactionAcceptor Amount Monomer Reaction
`sation
`volume
`time
`
`
`615
`
`I
`Ul
`Ml
`IV
`Vv
`Vi
`VIL
`Vill
`IX
`
`d(T-T-T-T)
`d(C-C-C-C)
`d(A-A-A-G)
`d(A-A-A-G)
`d(G-A-A-A)
`d(G-A-A-A)
`d(C,, Ts)
`d(A-A-A-C)
`d(A-A-A-C-A)
`
`nmol
`(Ax6o units)
`8.6 (0.08)
`8.1 (0.06)
`5.2 (0.3)
`5.2 (0.3)
`2.6 (0.15)
`2.6 (0.15)
`= 50 (4.0)
`80 (4.3)
`5 (0.35)
`
`[x*°P]GATP
`[o°°P]dATP
`dTTP
`dCTP
`dTTP
`dCTP
`[8-SH]dGTP
`dATP
`[o-32P]dTTP
`
`nmol
`
`units
`
`0.2
`0.2
`4.4
`44
`2.2
`a2
`40
`60
`33
`
`0.2
`0.2
`4.5
`4.5
`2.0
`2.0
`20
`20
`3.0
`
`pl
`——<—<—S ree
`I
`25
`5
`25
`5
`7
`5
`7
`5
`7
`5
`7
`16
`50
`5
`8
`5
`7
`
`h
`
`4
`+
`4
`4
`3
`3
`4
`4
`3.5
`
`pdC pdA pdG
`
`pdT
`
`MIGRATION [cm]
`
`Fig. 1. Results ofthe nearest-neighbour analysis of ** P-labelled penta-
`nucleoside tetrakisphosphates. Paper electrophoresis (a) of the four
`mononucleotides, (b) of the total enzymatic hydrolysate using spleen
`phosphodiesterase of pentanucleotide tetrakisphosphate d(T-T-T-T
`{°?P|pA) and (c) of d(C-C-C-C[**P]pA).
`In (b) can be seen the
`migration of dT[*?P]p and in (c) of dC[*?P]p
`
`used (Table 1). After 3-4h the reactions were stopped by
`addition of alkaline phosphatase, which dephosphorylates the
`applied monomers to nucleosides. The crude reaction mixtures
`were separated by reverse-phase HPLC (Nucleosil C,,) under
`isocratic elution conditions. The peaks were monitored by
`ultraviolet detection; the resulting elution profiles are shown in
`Fig. 2. To simplify the interpretation of the elution profiles,
`so-called ‘blanks’ were chromatographed underidentical con-
`ditions. These “blanks’ consisted of all the reaction components
`except the transferase and weretreated identically: only one of
`the ‘blank’ runs is shown (Fig. 2a). By comparing the ‘blank’
`(Fig. 2a) and the reaction mixture (Fig. 2b), the new product
`
`(peak 3 in Fig. 2b) can be identified easily. The elution profiles
`in Fig.2b-—e demonstrate that
`the enzymatic elongation
`yielded different reaction products which are well separated.
`After
`the reaction of d(A-A-A-G) + dTTP,
`the elution
`profile (Fig. 2b) shows that only one additional main peak
`appears, but
`in the other three elution profiles two or three
`new peaks were present. The fractions of the numbered peaks
`were lyophilised and their products identified by the fingerprint
`method [10].
`Aliquots of the freeze-dried products were therefore first
`phosphorylated enzymatically at the 5’-hydroxyl groups. The
`oligonucleotides were labelled, using [}-3?P]JATP and polynu-
`cleotide kinase, and were then partially hydrolysed with snake
`venom phosphodiesterase. The resulting hydrolysate was chro-
`matographed two-dimensionally. From the autoradiograms
`(fingerprints), which were obtained after chromatography,the
`purity as well as the sequence of the components could be
`determined. The fingerprints obtained are shown in Fig.3.
`Fromthe interpretation ofthe fingerprints of the enzymatically
`elongated tetranucleoside trisphosphates, the following results
`are obtained, which are summarised in Table 2. In all cases, the
`unreacted monomers were degradedwith alkaline phosphatase
`to the corresponding nucleosides and inorganic phosphate.
`They are eluted in peaks 1 and la—c.
`In the reations [1] and V dTTP was used, and each ofthe
`acceptors d(A-A-A-G) and d(G-A-A-A) were elongated with
`only one monomerunit to d(A-A-A-G-T) and d(G-A-A-A-T)
`respectively. If dCTP was used, however, besides the monoad-
`dition products d(A-A-A-G-C) or d(G-A-A-A-C), di- and
`triaddition products were also obtained (reactions [V, VI). In
`reaction VI [d(G-A-A-A) + dCTP], besides the monoaddi-
`tion product, the nucleotide d(G-A-A-A-C-C) was obtained
`(peak 11, Fig. 2e), and in reaction IV |d(A-A-A-G) + dCTP],
`the nucleotides d(A-A-A-G-C-C)(peak 5, Fig. 2c) and d(A-A-
`A-G-C-C-C) (peak 6, Fig. 2c) were found. By reaction of d(G-
`A-A-A) with either dTTP or dCTP (reaction V and VI res-
`pectively), not only elongated products were found but also
`d(A-A-A) (peak 9, Fig. 2d, e). Apparently the acceptor d(G-A-
`A-A) lost its 5’-terminal dG residue prior to the elongation step
`because neither d(A-A-A-T) nor d(A-A-A-C) could be found in
`the reaction mixture, although these should have been present
`if the already clongated acceptor had been degraded. Un-
`expectedly, after
`the elongation of the sequence-isomeric
`
`CUREVAC EX2040
`CUREVAC EX2040
`Page 3
`Page 3
`
`

`

`
`
`
`
`
`012
`
`- 008
`3
`x«
`
`OoL
`
`e
`S
`é&
`
`13
`
`|
`
`i
`
`|mL |
`
`012
`
`E
`st 0.08
`os
`—
`
`0.04
`
`0
`
`2
`
`‘
`
`616
`
`a)
`
`0.12
`
`0.08
`
`E2
`
`Ea
`
`004
`
`0
`
`40
`TIME [ min ]
`
`20
`
`0
`
`40
`
`20
`TIME [ min ]
`
`0
`
`E 0.08
`Cc
`~t
`wurs
`a
`
`0.04
`
`0
`
`e
`
`20
`40
`TIME
`[ min ]
`
`
`9
`
`7
`
`40
`TIME [ min ]
`
`
`20
`
`0
`
`12
`|
`
`c
`
`2p"
`
`e
`£
`
`4
`
`TIME [ min }
`
`TIME
`
`[ min ]
`
`Fig. 2. Separation of mixtures of a deoxynucleoside and oligonucleoside phosphates on a Nucteosil Cy, (7 um) column (250 « 4.6mm internal
`diameter) by isocratic elution at room temperature. Flow rate:
`1 ml/min. The mixtures were obtained after the enzymatic elongation of
`tetranucleoside trisphosphates (a—f} or pentanucleoside tetrakisphosphates (g) with nucleoside triphosphates using the terminal de-
`oxynucleotidyl transferase (except a) and were subsequently treated with alkaline phosphatase. The column waseluted with 72°, A(A = 0.1M
`ammonium acetate) and 28°, B (B = methanol/water, 60:40). Reaction mixtures: (a) = d(A-A-A-G) + dTTP treated only with alkaline
`phosphatase: (b) = d(A-A-A-G) + dTTP: (c) = d(A-A-A-G) + dCTP: (d) = d(G-A-A-A) + dTTP; (e) = d(G-A-A-A) + dCTP; (D = d(A-A-
`A-C) + dATP: (g) = d(A-A-A-C-A) + [x-**P]dTTP, the columneluted with 68°, A and 32° B. Peaks: 1 = dT; fa = dC; lb = dA,
`te = [22PJPO}> .2 = d(A-A-A-G); 3 = d(A-A-A-G-T):4 = d(A-A-A-G-C); 5 = d(A-A-A-G-C-C); 6 = d(A-A-A-G-C-C-C): 7 = d(G-A-A-A);
`8 = d(G-A-A-A-T); 9 = d(A-A-A); 10 = d(G-A-A-A-C): 11 = d(G-A-A-A-C-C); 12 = d(A-A-A-C): 13 = d(A-A-A-C-A); 14 = d(A-A-A-C-
`A-A); 15 = d(A-A-A-C-A-A-A); 16 = d(A-A-C-A); 17 = d(A-A-A-C-A[??P]pT); 18 = d(A-A-A-C-AP?P]pTP?P]pT)
`
`CUREVAC EX2040
`CUREVAC EX2040
`Page 4
`Page 4
`
`

`

`Table 2. Results of the enzymatic elangation ofoligodeoxyribonucleoside phosphates using terminal deoxynucleotidyl transferase
`HPLCof the reaction mixture was carried out on a 7-{1m Nucleosil C,, column (250 x 4.6 mm internal diameter) eluted with 72°, A and 28 °, Bor,
`for reaction LX, 68°. A and 32°B; A = 0.1M ammonium acetate; B = 60%, CH,OH/40% H,O
`
`617
`
`HPLCof the reaction mixture
`
`
` Identified by fingerprint Reaction Acceptor + Monomer
`
`
`
`III
`
`IV
`
`Vv
`
`VI
`
`d(A-A-A-G)+ dTTP
`
`d(A-A-A-G)+ dCTP
`
`d(G-A-A-A)+ dTTP
`
`d(G-A-A-A)+dCTP
`
`Vill
`
`d(A-A-A-C)+ dATP
`
`IX
`
`d(A-A-A-C-A) + [a-**P]dTTP
`
`peak
`
`Fig. 2
`
`retention
`time
`
`yield
`
`Fig. 3
`
`sequence
`
`Z
`3
`
`4
`3
`2
`6
`
`7
`$
`9
`
`10
`7
`1
`9
`
`12
`16
`13
`14
`15
`
`13
`17
`18
`
`b
`b
`
`c
`c
`c
`c
`
`d
`d
`d
`
`e
`e
`e
`
`t
`f
`f
`t
`f
`
`g
`2
`g
`
`min
`22.7
`31.8
`
`14.7
`20.5
`22.7
`32.0
`
`21.8
`29.8
`36.0
`
`19.0
`28
`24.8
`36.0
`
`26.8
`55.0
`62.1
`81.0
`95.0
`29,3
`40,3
`78.5
`
`a
`68.5
`21.7
`
`26.0
`20.7
`47.0
`4.0
`
`51.4
`31.1
`12.0
`
`25.8
`45.8
`13.8
`8.9
`
`58.1
`2.9
`21.4
`8.7
`2.1
`58.3
`31.9
`94
`
`a
`b
`
`c
`d
`a
`e
`
`f
`2g
`=
`
`h
`f
`i
`=
`
`j
`k
`|
`m
`n
`|
`°
`Pp
`
`d(A-A-A-G)
`d(A-A-A-G-T)
`
`d(A-A-A-G-C)
`d(A-A-A-G-C)
`d(A-A-A-G)
`d(A-A-A-G-C-C-C)
`
`d(G-A-A-A)
`d(G-A-A-A-T)
`d(A-A-A)
`
`d(G-A-A-A-C)
`d(G-A-A-A)
`d(G-A-A-A-C-C)
`d(A-A-A)
`
`d(A-A-A-C)
`d(A-A-C-A)
`d(A-A-A-C-A)
`d(A-A-A-C-A-A)
`d(A-A-A-C-A-A-A)
`d(A-A-A-C-A)
`d(A-A-A-C-A[2P]pT)
`d(A-A-A-C-A[??P]pTE33P]pT)
`
`acceptor d(A-A-A-G), a corresponding degradation product
`could not be detected (Table 2).
`The desired monoaddition product was obtained in yields
`between 20°and 30°, in all four reactions. Moreover, the
`unchanged acceptor could be recovered in chromatographi-
`cally pure form and could be used again for another elongation
`reaction, Using this possibility,
`the yield of the enzymatic
`elongation could be substantially increased.
`
`Elongation of sequence-isomeric pyrimidine nonanucleoside
`octakisphosphates
`
`In reaction VII, the mixture of sequence-isomeric pyrim-
`idine nonanucleoside octakisphosphates, d(C,, T;), isolated
`from partially hydrolysed depurinated herring sperm DNA
`[13], was elongated enzymatically with [8-7 H]dGTP under the
`conditions given in Table | by a single dG residue. The acceptor
`was used in about 20%, molar excess with respect
`to the
`monomers.
`1 U of terminal deoxynucleotide transferase was
`apphed for 2 nmol monomer in contrast to the preceeding
`reactions,i.e. only half as much enzyme was used with respect
`to the monomer. Again the reaction was stopped after 4h by
`addition of alkaline phosphatase, which dephosphorylates
`[8-"H]JdGTP to [8-"H]dG.
`The reaction mixture could not be separated by HPLC
`however, because the sequence isomersare partially separated
`by chromatography on Nucleosil C,, and it
`is not generally
`possible to separate the elongated products from the original
`acceptors. Therefore the separation was performed by paper
`chromatographyin twodifferent systems. First the mixture was
`chromatographedin the solvent system A wherethe nucleoside
`
`[8--H]dG runs within the front, whereas the unconverted
`sequence isomers and their elongated products remain at the
`origin. To determine the yield of the reaction, a strip of the
`paper chromatogram, containing an aliquot of the reaction
`mixture, was first cut in the direction of the run, and then cut
`perpendicularly into strips of 2cm width. Then the radioac-
`tivity of the strips was determined. Usually about 10°, of the
`applied radioactivity was located in the region of the elongated
`sequence isomers whereas about 90° was recovered as
`[8-*H]dG. For preparative isolation ofthe elongated sequence
`isomers, first
`the upper region of the paper chromatogram
`containing [8-* H]dG wascut off. The paper chromatogram was
`then chromatographed in solvent system B in which the
`unchanged sequence-isomeric nonanucleoside octakisphos-
`phates form a single spot. This spot moves faster than the
`mixture of the nucleotides elongated by one dG residue which
`also migrate as one spot. The positions of the unchanged
`oligonucleoside phosphates could be detected by their ul-
`traviolet absorption, whereas all the elongated sequence iso-
`mers were detected by their radioactivity (Fig. 4). Only one
`radioactive spot appeared in the paper chromatogram which
`contained the monoaddition products of the sequence-isomeric
`oligonucleoside phosphates d(C4,T.-[8-*H]G). It can therefore
`be concluded that only one dG residue was condensed onto the
`sequence isomers.
`
`Stepwise elongation ofa tetranucleoside trisphosphate
`to a hexanucleoside pentakisphosphate
`
`In the two quantitatively evaluated examples, reactions
`VIII and [X (Table 2), the synthetic tetranucleoside trisphos-
`
`CUREVAC EX2040
`CUREVAC EX2040
`Page 5
`Page 5
`
`

`

`oyoOo
`
`Ss
`
`a2H
`
`e
`
`0
`
`30
`20
`10
`MIGRATION [cm]
`
`Fig. 4. Paper-chromatographie separation ofthe reaction mixture d(C4,
`Ts) + [8?HJ]dGTP catalysed by terminal deoxynucleotidyl trans-
`ferase.
`The paper was
`chromatographed first
`in
`system A
`= ethanol/1 M ammonium acetate pH 7.5 (7:3, v/v) and after drying
`in system B = 1-propanol/NH,OH conce./water (55:10:35; v/v/v).
`(a) Migration of d(Cy, T,) detected by ultraviolet absorption: (b)
`migration of d(C,, T;-[8-4H]G) detected by the radioactivity
`
`impurities of the chemically synthesised acceptor. Peak 13
`contained the desired monoaddition product d(A-A-A-C-A).
`The yield (+ 21°) was about twice as high as the one obtained
`by the elongation of the sequence isomers although less enzyme
`was applied. The peaks 14 and {5 contained the diaddition
`product d(A-A-A-C-A-A) and the triaddition product d(A-A-
`A-C-A-A-A), respectively, in yields of 9% and 2". In peak 16
`the tetranucleoside trisphosphate d(A-A-C-A) was eluted,
`which was formed from the elongated acceptor d(A-A-A-C-A)
`by loss of a 5’-terminal adenylic acid residue. The degradation
`must have taken place after the elongation because a degraded
`acceptor, d(A-A-C), could not be elongated. This finding is
`remarkable because prior to the elongation the acceptor, d(G-
`A-A-A), was partially degraded to d(A-A-A). These results
`indicate different kinetics which are presumably influenced by
`the sequence of the acceptor.
`In the second step (IX) 5 nmol of the isolated monoaddition
`product d(A-A-A-C-A) was used as acceptor which was
`converted with 3.3 nmol [«-??P]dTTP and 3U terminal de-
`oxynucleotidyl transferase during 3.5h. After the reaction had
`been interrupted by addition of alkaline phosphatase, which
`dephosphorylated [«-*?P]dTTP, the crude mixture was sepa-
`rated likewise by HPLC on Nucleosil C,, (Fig. 2g). To detect
`the radioactivity,
`in addition to the ultraviolet detector, a
`radioactivity measuring detector was used and both the
`ultraviolet absorbance and the radioactivity of the eluate were
`monitored. For practical reasons,
`the corresponding peaks
`were shifted to a minor extent. The interpretation of the elution
`profile was simplified considerably, because nowall elongated
`products were labelled radioactively whereas the unlabelled
`acceptor showedonly ultraviolet absorbance.
`The degradation products of the monomer were eluted in
`pek as 1 and Ie, which were formed bythe action ofthe alkaline
`
`CUREVAC EX2040
`CUREVAC EX2040
`Page 6
`Page 6
`
`618
`
`a) diA-A-A-G) (2)
`e
`SB
`
`e
`
`bl dlA-A-AG-T) (3)
`!
`e
`
`q®
`
`cl glA-A-AG-C} [4 }
`
`e
`
`.
`
`.
`
`[7}
`
`d@ di4-A-A-GC-C)
`
`[5]
`
`. 6
`
`e
`
`e
`
`.
`
`e] dA-A-A-G-C-C-C}
`°
`
`e
`
`e
`
`e
`e
`6
`&
`
`(6)
`
`f) dIG-A-A-A)
`e
`’
`
`e
`e
`
`g) diIG-A-A-A-T)
`
`(8)
`
`h) dIG-A-A-AC)
`
`(10)
`
`i) dIG-A-A-A-C-C}
`
`(11)
`
`=
`-
`@
`e
`e
`
`°
`.
`.
`e
`e
`
`-
`
`-e
`
`-
`
`jf) dtA-A-A-C)
`
`(12)
`
`k] qiA-A-C-A)
`
`(16)
`
`1) diA-A-A-C-A}
`
`113)
`
`o-
`
`@
`
`*
`
`+
`
`e
`
`0)
`
`dIA-A-A-C-A-T)
`(17)
`- .
`e
`z
`
`$$
`
`3
`2
`mw
`
`ELECTROPHORESIS
`
`4
`
`ab GIA-A-A-C-A-A-A} [15]
`=
`°
`oe
`e
`e
`e
`
`(14)
`*¢
`
`m) dIA-A-A-C-A-A)
`oe
`
`*
`@
`e
`
`p) dlA-A-A-C-AT-T} 18}
`’
`
`2
`
`- e
`
`el
`
`Fig. 3. Determination of the sequences ofoligonucleoside phosphates
`isolated by reversed-phase HPLC (Fig.2) after the enzymatic elon-
`gation of different acceptors using terminal deoxynucleotidy] trans-
`ferase. The isolated products were first labelled with **P and then
`partially hydrolysed with snake venom phosphodiesterase. The result-
`ing partial hydrolysates were two-dimensionally chromatographed.
`The first dimension consisted of an electrophoretic separation on
`cellullose acetate strips at pH 3.5, the second a homochromatography
`on DEAE-cellulose thin-layer plates. (a—p) The resulting autoradio-
`grams. The numbers following the determined sequence of the
`fingerprints correspond to the peak numbers of Fig. 2
`
`phate d(A-A-A-C) was converted (according to the conditions
`in Table 1) enzymatically into the hexanucleoside pentakis-
`phosphate d(A-A-A-C-A[*?P]pT) in twosteps on a preparative
`scale. In the first step (VII), 80 nmol of the acceptor d(A-A-
`A-C) were treated with 60nmol dATP and 20U terminal
`deoxynucleotidyl transferase. After 4h, the reaction was stop-
`ped by addition of alkaline phosphatase and the reaction
`mixture was separated by reverse-phase HPLC on Nucleosil
`Cyg (Fig.2f). The fractions of the numbered peaks were
`lyophilised and identified by the fingerprint method. These
`fingerprints are shown 1n Fig. 3. The quantitative evaluation of
`the reaction based on the comparison of the different peak
`areas is summarised in Table2. About 58° of the starting
`acceptor d(A-A-A-C) remained unchanged and was elutedin
`peak 12 of Fig. 2f. Both side peaks of peak 12 are probably
`
`

`

`619
`
`DISCUSSION
`
`phosphatase. Peak 1 contains dT and peak 1c inorganic
`degraded products were found. The sequence of the oligonu-
`phosphate. The unchanged acceptor d(A-A-A-C-A) could be
`cleotide acceptoris probably important in determiningif and
`when a 5’-nucleotide is enzymatically removed. For example,
`recovered from peak 13 and amounted to about 58°, of the
`d(A-A-A-A-C) looses tts 5’ monomerunit after elongationat
`reaction mixture. Peak 17 contained the desired elongation
`product d(A-A-A-C-A[*?P]pT) with a yield of 32%. The
`the 3’ terminus, but d(G-A-A-A) loosesits 5’-nucleotide before
`diaddition product d(A-A-A-C-A[*?P]pT[**P]pT) was found
`enzymatic elongation. The different composition of a reaction
`under peak 18 which amounted to 9°, of the reaction mixture.
`mixture shows also that the condensation catalysed by terminal
`The sequences ofthe isolated oligonucleotide phosphates were
`deoxynucleotidyl
`transferase is influenced by the acceptor
`confirmed by the fingerprint method (Fig. 3). Peak 17 was
`sequence and the monomer unit to be added to it. Diaddition
`isolated and lyophilized and a white powder was obtained.It is
`products, which are found in the examined reaction mixtures
`up to 21 °,, can be very useful in cases where the oligonucleotide
`remarkable that, in contrast to the first step of the synthesis,
`desired contains homologous segments. Polyaddition can be
`after the second step no degradedoligonucleoside phosphates
`were found.
`favoured by varying the reaction conditions, so that inasingle
`reaction cycle two or three monomer units may be added to a
`given acceptor.
`The elongation of acceptors with monomer units as de-
`scribed here may be regarded as a new way to the stepwise
`enzymatic synthesis of defined oligonucleotides. The time
`needed for one condensation cycle, including the isolation of
`the reaction products,
`is about 5h. A comparison with
`automated chemical synthesis of defined oligonucleotides
`shows that the machines can do asingle addition step in a
`shorter time. However, the resulting reaction product is less
`well defined andstill carries protecting groups which have to be
`removed. The limiting factor in both procedures of synthesis is
`probably the separation of reaction products, which will be
`more and more difficult with increasing chain length. An
`important advantage of the enzymatic synthesis is the isolation
`of a pure, well defined product after each reaction cycle. In
`contrast,
`the product of an automated chemical synthesis is
`identified, without internal feed-back, only at the end of the
`synthesis. The difficulties in an isolation of the desired oligo-
`nucleotide are therefore considerably greater when the chemi-
`cal method has been used for its synthesis than with our
`enzymatic method.
`The amountof oligonucleotides which can be produced by
`enzymatic synthesis
`is
`sufficient
`for common molecular
`biological experimentslike hybridisation, use as primer, linker,
`adapter or point mutator. These proceduresare usually carried
`out with labelled oligonucleotides which have to be synthesized
`enzymatically from a given synthetic oligonucleotide. The
`enzymatic synthesis catalysed by terminal deoxynucleotidyl
`transferase leads directly to a labelled product. This must be
`regarded as a great simplification in comparison to other
`published procedures.
`
`The results show that oligodeoxyribonucleotides can be
`elongated enzymatically using terminal deoxynucleotidyl
`transferase. Acceptors must have at least four monomerunits
`and be obtainable by chemical synthesis or by isolation from
`partial hydrolysates of chemically degraded DNA.Thepurity
`of the commercially available transferaseis fully sufficient for
`this purpose. The addition of purine nucleotides to a pyri-
`midine acceptor and vice versa is of special
`interest, since
`acceptors with both purine and pyrimidine monomer units are
`not yet available from partial hydrolysates of chemically
`degraded DNA.
`The reaction mixtures, which contain 2—80 nmolof accep-
`tors yield 20—30°, of monoaddition products. Remarkably,
`the reaction is not affected by the chain length of the acceptors
`used. which have four, five or nine monomer units in the
`reported examples. The yield can be increased by recycling the
`unchanged acceptors, which can be recovered easily in chro-
`matographically pure form.
`The reaction mixtures are separated by different chroma-
`tographic procedures, of which thin-layer homochromatog-
`raphy on DEAE cellulose sheets is most suitable for analytic
`purposes. Very small quantities of addition products are
`detectable when radioactively labelled monomers are used in
`the elongation reaction,
`larger quantities of products are
`separated by paper chromatography or HPLC.
`The examples described show HPLC separations using
`reverse-phase chromatography on Nucleosil C,,. The elution
`profiles indicate that the capacity of the analytical column (250
`x 4.6mm internal diameter) used is not yet fully utilized and
`that even larger batches could be separated in one run.
`In
`comparison to anion-exchange chromatography, which can
`also be used for the separation of the reaction mixtures, reverse-
`phase
`chromatography
`has
`the
`following advantages.
`Separations can be carried out with a volatile buffer system;
`therefore,
`the products can be isolated without a desalting
`procedure. By-products having the same net charge as the
`desired elongation product can also be removed by reverse-
`phase but not by anion-exchange chromatography. On the
`other hand, the elution profile of the reaction mixture is not
`determined by the net charge of
`the oligonucleotides.
`Normally, the highest peak of the newly appearing peaksis the
`monoaddition product. The correct identification, however,
`has to be performed by the fingerprint method.
`The reaction mixtures [except d(A-A-A-G) + dTTP]
`contained not only the mono- but also di- and triaddition
`products although the acceptor con