вход по аккаунту


Isolation of RNA Aptamers for Biological Cofactors by In Vitro Selection.

код для вставкиСкачать
Kenny. h i d . 1993. 4Y. 394: B. Mom, G De With, Acra Crvsttdhgr. S e ~ r B
1978. 34. 2785. S. V. Sereda, M Y. Antipin. T. V. Timofeeva. Y. T. Struchkov.
Sue. P/ij.c. Crj.s!ul/ugr. Ens/. Tunis/. 1988. 33. 66.
[13] E. C. K . Lai. D. Mackay, N. J. Taylor, K. N . Watson. c'un. J. Cheni. 1988, 66.
[14] It is not unusual to find more than one independent molecule in the unit cell.
In most cases these symmetry-independent molecules show the same conformation (for recent examples from our group. see R. Glaser. G. S. Chen. C. L.
Barnes. Angcw. Chrii?. 1992, 104. 749; A n g o t . Ciienr I n ! . E d Eii,q/. 1992, 31.
740: G. S. Chcn. R. Glaser. C. L. Barnes, J C'heni S o i . Chcn?.Coniiinni 1993,
1530: R Glaser, C. L. Mummert. C. J. Horan. C. L. Barnes. J P/JJ.Y.Org.
C/im. 1993. 6. 201.): but crystals containing two rotamers are rare.
[15] T. Ishida. M. Inoue. K . Nasu. T. Kurihara. A i t a Crj.sfu//ugr. SLY!.C 1983, 39.
470. In this paper the isomers of I1 are said to have the same NMR spectra in
contradiction to previous reports by the same group (T. Kurihara, Y. Sakamoto, M. Mori, T. Sakaki. Hrriwcyler 1978. 9, 1041).
[16] C. P Brock, G . L. Morelan. .I P/IVJ.Chon. 1986. YO. 5631; H . Grutzmacher.
H . Pritzkow, Aiigew. Cliew~.1992. 104, 92. A/igcw. C/ir/n.I n f . Ed. Engi. 1992.
31.99; A. Beck, R. Gompper, K Polhorn. H.-U. Wagner. ibid 1993. /U.i.l424
and 1993. 32. 1352.
[17] I.N . Levine, J. C h m . P / i n 1963. 38, 2326.
[l X] C. Standorfy in Gmtwlund Tliroriw(u/A s p t ~ tin
s rhc C/it~mi.s/i:ro f f h eCwhonNr!rogen Doirhli, Bond (Ed.: S. Patai). Wiley Interscience, New York. 1970,
[19] A Streitwieser. C. H . Heathcock. Orgunochc C/ieniii,, Verlag Chemie. Weinheim. 1980. p. 644; Intrui/ni~!mito Organic C/ienii.srrv. 3rd ed , Macinillan.
New York, 1985. p. 527.
[20] Conjugation might still be important in hetero-substituted azines. a) See literature cited in ref. [ l a ] : b) K . Hagen. K . Hedherg. J. Phrs. Cizeni. 1992. Y6,7976;
c) G. Kober, P. Rademacher. R. Boese. J Chow. So(..Prrkiri Trun, 2 1987,761.
d ) k t r r Chiwi. Scrmii. Ser. A 1988. 42. 571.
[21] K. B. Wiherg. P. R. Rahlen, M . Marque7. J A m . Clinii. So<. 1992. /14%8654.
and references cited i n ref. 11 a].
I221 W. J. Hehre, L. Radom, P. von R Schleyer. J. A. Pople. .4h h i ! m Molec,ukir
Orbiud Tlirorv, Wiley. New York. 1986.
[23] We found this same feature also in all other azines M e have studied; its origin
is currently under investigation.
[24] U. C. Sinha. Actu Crylallugr. Sect. B 1970, 26. 889.
[?5] a ) Positive chemical shifts reflect increased magnetic shielding relative to the
external standard. neat nitromethane. b) M . Witanowski, L. Stefaniak, G. A.
Webb. Nitrogen N M R Spi !r.o.vccip~( A n n u Rep. N M R Sp
tides, amino acids, metal ions, and antibiotics. Furthermore,
detailed knowledge of this kind could facilitate the development
of new ribozymes.
We report here on the isolation of R N A motifs that bind to
the flavin portion of flavin adenine dinucleotide (FAD) and
flavin mononucleotide (FMN) . In addition, selection experiments were performed in order to isolate RNA aptamers for
nicotinamide adenine dinucleotide ( N A D + ) and nicotinamide
mononucleotide f N M N ') (Scheme 1).
Isolation of RNA Aptamers for
Biological Cofactors by In Vitro Selection**
Petra Burgstaller and Michael Famulok"
In vitro selection enables the simultaneous screening of a
large number ( 2 lo1') of different D N A or R N A sequences for
certain functionalities. In vitro selection experiments encompass
a number of sequential steps, in which the first is always the
synthesis of a library of random D N A sequences. After amplification of the D N A by polymerase chain reaction (PCR), an
R N A pool is produced by transcription in vitro. Those RNAs
(aptamers) that specifically bind to the target molecule are selected from this R N A pool, for example, by affinity chromatography.['. 21 Such specific ligand-binding nucleic acids not only
can have use as potential lead structures for protein inhibitor^.[^'
but can also help to increase our understanding of the biochemically important recognition processes that involve the interaction of R N A with substrates such as nucleotides, proteins, pep[*] Dr. M. Famulok, DipLChem. P Burgstaller
Fachberelch Chemie, lnstitut fur Biochemie der Universitit Munchen
Am Klopferspitz 1Xa. D-82152 Martinsried (FRG)
Telefix: Int. code (89)8578-2470
This work was supported b) the Deutsche Forschungsgemeinschaft and the
European Union (project No. Biot-CT93-0345). We thank E.-L. Winnacker
for his support and F. Michel, D. Faulhammer. and T. Luchterhandt for helpful
Scheme 1 The FMN. NAD'. and N M N + hgands employed in the selection.
and 7.8-dimethylalloxa~inewhich was employed for investigating the binding specificity.
For the selections we used a pool of 32P-labeled RNA 113-mers
with a complexity of 10'' different sequences which consisted of
a random sequence of 14 nucleotides flanked by two defined
primer binding sites. By means of affinity chromatography with
agarose derivatized with the given cofactor, those R N A sequences were enriched that were bound to agarose and eluted
with a solution containing the appropriate hgdnd (Scheme 2.
Table I ) .
After six cycles significant amounts of the RNA were bound
to the F M N and N A D + columns. Binding to the FAD matrix
could be detected after four cycles, whereas enrichment of
primer sequences:
0.5 mL adipic acid
dihydrazide agarose
1 mL cofactor
affinity elution of
bound RNAs
elution of
unbound RNAs
with buffer
NMN was not evident after eight cycles of selection/amplification. The FMN-. FAD-. and NAD+-selected RNA pools from
the last of the respective cycles listed in Table 1 were treated with
Table 1 . Bound fraction [%I of KNA used in the selection cycle. The values represent the fraction of bound RNA that was eluted from the column with the corresponding ligand in the buffer solution; the column was previously washed with five
column volumes of buffer solution.
Selection cbcle
21 38
reverse transcriptase and the cDNA amplified with PCR and
cloned. We sequenced 17 clones from the FMN-, 14 from the
FAD-. and 16 from the NAD+-selected RNA pools (Scheme 3).
Scheme2. a) The design of the DNA Iibrary for the synthesis of the KNA pools
employed. The framed sequence is the T7promoter, the underlined sequences the restriction sites for the subsequent cloning.
b) A schematic representation of the in vitro
selection cycle. The RNA was loaded onto
the preselection column (0.5 mL adipic acid
dihydraride agarose) and eluted from the
column with 1.0 mL of a buffer solution
( 2 5 0 m ~NaCI; 5 0 m ~tris-HCL, pH 7.6;
5 mM MgCI,). Nonbinding RNAs were separated by washing with five column volumes
of the buffer solution. Bound RNAs were
eluted with a buffer solution of the ligand
and then treated with reverse transcriptase
(50 mM Tns-HCI. pH 8.3; 75 mM KCI; 3 mM
MgCI,; 10 mM dithiothreitol (DTT). 0.3%
TWEEN 100; 0 . 4 m ~of one of the deoxynucleotide triphosphates dATP, TTP.
dGTP, dCTP; 1 WM primer M20.106; 40 U
Superscript reverse transcriptase: 45 min reaction at 4 2 ' C ) . The cDNA was amplified
with PCR (10 mM tris-HCL, pH 8.3; 50 mM
KCI; 0.01 % gelatine: 1.5 mM MgCI2: 0.3%
TWEEN 100; 0.2 mM one of the deoxynucleotide triphosphates dATP, TTP, dGTP.
dCTP: 3 . 0 ~primer
M38.27: 3 . 0 ~ ~
primer M20.106; 2.5 U 1 0 0 p L - I TAQ
D N A polymerase; conditions of cycle:
94 C - 1 min. 55°C - 2min. 72 'C - 2inin).
the PCR-DNA is transcribed (40 mM trisHCL, pH 7.9; 6.8 mM spermidine; 22 mM
MgCI,: 0.01 % Triton X-100: 5 mM of the
nucleotide triphosphates ATP. UTP. GTP.
CTP; 10mM DTT. SOU T7 RNA polymerase; 1-2 pCi x-"P-GTP: reaction for
16h at 37 ' C ) . The selection cycle can now
be repeated with the enriched KNA pool.
Thirteen of the FMN-binding RNAs contained two conserved
regions with the sequences AGGNUAU and AGAAGG (the sequence of the clone FMN-2 occurred three times). Both consensus sequences are flanked by variable nucleotides, which pair
together to form a defined secondary structure. The monoclonal
RNAs of the FMN-2, FMN-7, FMN-12, and FMN-15 sequences
all have affinity for FMN. By means of analytical affinity chrom a t o g r a p h ~ [we
~ ' determined the dissociation constant K,, for the
FMN-2 aptamer/FMN complex. The specificity of this RNA
was quantified by elution with 7,s-dimethylalloxazine ( I ) , FAD,
ATP, and GTP. Based on the conserved secondary structure we
constructed an RNA sequence of 35 bases (35FMN-2 = 5'-
around 0.5 p~ was measured. The ribose moiety of FMN was
not recognized by the aptamers, as both FMN-2 and 35FMN-2
bind to 1. Both of these aptamers show very low affinity for
adenosine triphosphate (ATP) and d o not bind to guanosine
triphosphate (GTP) at all. As expected, they bind to FAD with
similar values of K, as for binding to FMN (Table 2).
Four of the fourteen FAD-binding RNAs were identical to
the FMN-2 clone. The FAD-3 clone can form the same secondary structure. In comparison to all the other sequences that
form this motif. the two consensus sequences are in reversed
FMN Selection
based on the secondary structure of FAD-1 (27FAD-1 =
containing the consensus sequence and the stem region exhibits
binding constants that are somewhat weaker but on the same
order of magnitude as those of the full-length aptamer FAD1 .['] It is, therefore. one of the smallest specific ligand-binding
RNAs known to date. The corresponding D N A sequence of this
aptamer has no measurable affinity for FAD.
Why was only one binding motif with affinity for the flavin
portion obtained in the F M N selection, whereas in the FAD
selection a second motif of significantly lower affinity could also
be isolated? This may well be explained by the different concentrations of the ligand in the F M N and the FAD agaroses (FMN:
0.5 mM; FAD: 3.0 mM). A comparison ofthe elution volumes V,
of FAD-1 for the two derivatized agarose types, V, = 2 mL for
the F M N agarose and V, = 13 mL for the FAD agarose, proFAD Selection
vides values of Kd of a similar magnitude. This implies that all
Clone (Clones 2, 5, 16, and 22 are identical to FMN-2.)
FAD-I -type binding species were removed from the F M N
rAvj 3
N, , ~ ~ G A A G G ~ G ~ G A -, - - - u
- cNm- A G G u U A ~ u ~
agarose with the 5.0 mL washing employed during selection and
A ~ A G
as a result were not enriched.
All sixteen NAD'-binding sequences contained the base seA
quence GGAAGAAACUG. This same consensus sequence is a
component of an ATP-binding RNA motif isolated recently by
, :-i
N~~-uGcGGGcAAAAGGAAGUGUA~~CUCCC~-N. Sassanfar and Szostak.[ldlIndeed, the aptamers that we isolated
can fold into the identical motif (Scheme 4). This aptamer binds
T F CI '.
N ,, - ~ G Y s ~ z c G A A A G G A A G G G I J A A ~ - N
to the adenosine porN,?-YGGUGhCGLACUC
tion of ATP with a Kd
Scheme 3. F M N - and FAD-binding RNA sequences. Selected RNAs were treated
of 4.0 pM and also has
with reverse transcriptase and amplified by PCR with the M27.39 and M20.106
affinity for NAD'. To
primers (see Scheme ?a). The M27.39 primer was employed to provide the DNA
enrich aptamers, which
with E ( o R / restriction sites. After digestion with EmRI and BurwHI. the excised
may be present in small
fragment was cloned into the vector pGem3r (Promega) [7] and sequenced following
the dideoxy method [S]. Only those bases are shown which are relevant in forming
concentrations in the
the binding motrfs shown above each set of sequences. The consensus sequences and
Scheme4. RNA-binding motif [Id] for the
a schematic representation of the btem region are also shown.
which require the
adenosine portlon of ATP or NAD+.
NMN' portion for
binding to N A D + , we performed three additional selection
Table 2. Dissociation constants Kd for the binding of FMN-2. 35FMN-2. FAD-1.
cycles, in which a negative selection with ATP preceded the
and 27FAD-1 to various ligands in solution [a].
elution with NAD'. However, no further binding species were
K d [W]
obtained. This result was confirmed by an NMN' selection
(Table I ) , which likewise did not lead to an enrichment of binding species over eight cycles. The simplest explanation for these
results would be that the pool does not contain a single sequence
capable of binding to NMN'
4 x 10'
2 x 10'
What i s remarkable about this result with NAD' and ATP
3 103
selections['d1is the fact that two completely independent selections employed for the structurally related, but yet different
[a] The K , values were determined by elution of "P-labeled RNA from a column.
which contained 1.0 mL of a 0.5 insf F M N or 3 mM FAD agarose, and were calcuagaroses and hgands (and nucleic acid pools) led to identical
li:,)) [If]. [L] is the
lated according to the equation K,, = [L] x i(V-, - V,) x (V,
aptamers, and indeed exclusively to those with affinity for the
concentration of the ligdnd in the solution. Vc,the elution volume in the prccence of
adenosine portion of the ligand. The fact that this motif can be
the ligand in solution. V, the elution volume with buffer solution in the absence of
isolated repeatedly from independent pools with a complexity of
the ligdnd. Ve the void volume of the column (0.7 mL)
1014-101'sequences illustrates the high certainty and precision
of this selection technique. Furthermore, this result indicates
that the adenosine-specific aptamer is the "optimal sequence
order. This finding in particular strongly supports the proposed
solution" for the molecular recognition of adenosine by RNA.
secondary structure. The FAD selection provided another bindIn conclusion, we were able to select new RNA motifs for the
ing motif, the conserved 13-mer 5'-PuAAAGGAAGUGUA-3'.
specific recognition of biological cofactors from randomized
which is flanked by pairing base pairs, leading to formation of
RNA pools. The aptamers described recognize preferentially
the stem-loop structure shown in Scheme 3. This motif was
certain regions of a ligand. Other regions were only poorly
found in five sequences (FAD-I. FAD-4, FAD-9, FAD-I 1. and
bound or not at all. The RNA motifs 35FMN-2 and 27FAF-1
FAD-20). The affinity of FAD-I for FAD or F M N is signifirepresent a new class of supramolecular systems for the highly
cantly lower than the affinity of the FMN-2 motif for these
specific recognition of flavins based on "irrational receptor deligands (Table 2). FAD-I also recognizes exclusively the flavin
sign-,19. 101
portion of FAD, since binding to 1 ( K , = 23 p ~is )significantly
better than binding to F M N (84 p ~ ) and
, there is virtually no
Received: November 25, 1993
affinity for the adenosine portion of FAD ( K t T P2 4.0 mM).
Revised version: January 18. 1994 [Z6511 IE]
German version: A n p i . Ch<vn. 1994. 106. 1163
FAD-1 likewise does not bind to GTP. A minimal 27-mer RNA
i "
A. D. Ellington. J. W. Szostak. N u r u r ~1990. 346. 818: b ) rhid. 1992, 359.
850: c ) M . Famulok. J. W. SLostak. J. An,. Chim. Sor. 1992, 114. 3990:
d ) M S,issanT;ir. .I.W. Szostak. Nutcrrv 1993. 364. 550: e) D. P. Bartel. J. W.
Smstnk. .%wiiw 1993. 261. 1411 : f) G . J. Connell, M. Illangesekare. M. Yarus.
B f o ~ ~ h w i i 1993.
~ r r ~ 32,
~ 5497: g) M . Famulok. J. Am. Clren~.Soc., 1994. 116.
Revieirs. a ) M. F;imulok,J. W. Szostak. Ang~ir.Climi. 1992. 103, 1001 ;Arigcw.
C'/IIWI./ t i / . Ed. G i g / . 1992. 31, 979. b) J. W. Szostak, Eenil.\ Broi/iivn Sci. 1992.
17. X9: c ) M . Famulok. J. W. Szostak in .Yuilric k i d s rrriii Moi~~culur
(Eds: F. Eckstein. D. M J. Lilley). Springer. Berlin. 1993, p. 271.
Tuerk. L. Gold. S c w i w 1990, B Y . 505; b) L. C. Bock. L. C . Griftin. J. A.
1111. E. H . Vermaas. J. J. Toole. Nuruw 1992. 355. 564: c) C. Tuerk. S.
M:icIhug~tl. L. Gold. Pro<. :Z;u//. 4 < d . .%/. US2 1992. 89. 69x8: d) D.
Schneider. c'. Tuerk, L. Gold. J. M d . B i d . 1992, 228. 862; e ) M . L. Riordan.
1. c'. Martin. , \ ' i i r i w 1991. 350, 442.
1986. 3.59. 1.
F ti Arnold. S A Scolield. H. W Blanch. J Cli~-oti?ufup~.
It i \ striking that FAD-I and 27FAD-1 bind to I more strongly than to FAD
o r F M N . An explanation for this could be that I does not have a negaticely
charped phosphate group and thus r-epulsion of the R N A phosphatc backbone
IS no1 possible. Similar explanations have been XiLen for comparable observation\ (bee ref. [Id]). Other possible explanations include gain or loss of binding
ii rewlt o f differences in the hydrate shell of 1 and F M N or FAD.
n~ interiictions. and steric phenomena. DetAiled structural investigntion, w!ould provide a greater insight into such observations.
I t I S possihle that the stringency of the selection cond~tionidid not allow
i s ~ l i i t i ~ofoptamers
that bind N M N ' or the N M N portion of NAD'. We
iirc currently examining different buffer conditions. metal ion?. and ligand
concentrations as well as elution volumes (as with FAD-1). which could lead
t o isolation of specific motifs. Houever. the pool certainly did not contain
motifs specific for NMN' with K , values comparable to those of the ATPbinding speciea ( I 10 pM)
a ) J. Sambrook. E. F. Fritscli. T. Maniatis. .Molimilm C l ~ n i n g2nd
ed.. Cold
Spring Harbor Laboratory Press, Cold Spring Harbor, 1989: b) F. Asubel. R.
Brent. R E. Kingston, D. D. Moore. J. G . Seidman, J. A. Smith. K . Struhl.
C ' f i r . i . r ~ f l /Pr-oIlXY/ir i~r1A4i~/i~<iduf'
B;o/opi.. Wiley. New York. 1987.
F. Sanger. S. Nicklen. A. R. Coulson. P~YJC.
,Vur/. .4cud. Sci. USA 1977. 74.
Examples of ILivln receptora based on rational receptor design: Y Aoyama. Y
Tm,ik;i. H. To]. H. Ogoshi. J Ain. Clion. SO<.1988. 110. 634: b) Y. Yano. N .
Taniurd. K. Mitsui. T. Nabeshima. Chiw. Lei/. 1989. 1655; c) Y. Aoyama. K.
Mirohdmi. H. ToL ;hid. 1990. 651.
See rel. 121 i n G. \on Kiedrowski. Arrpeii.. C ~ I W1991.
J . 103. 839; ,411g~w.
/ii/ Ed. D ? , q / . 1991. SO. 822.
Functionalization of C,, Buckminsterfullerene
by [8 + 21 Cycloaddition: Spectroscopic and
Electron-Transfer Properties of a
E r n s t Beer, Michaela Feuerer, Andreas K n o r r ,
Albert Mirlach, and Jorg Daub*
8-Methoxyheptafulvene (1) has pronounced x-donor properties"' and reacts with acceptor-substituted alkenes to form tetrahydroazulenes in a regiospecific [8 21 cycloaddition.''] I n
contrast, buckminsterfullerene C,, (2) has a high electron affinity (E,&= 2.6-2.8 eV).r31This is the reason for the electrophilic
nature of the x-system: reactions normally occur at the double
Prof. Dr. J. Daub. DipLChem. E. Beer. M. Feuerer. Dip1.-Chem. A. Knorr.
Dr. A. Mirlach
Institut fiir Organische Chemie der Universitit
Uinvcrsititstrasse 31
D-93040 Regensburg (FRG)
Tcief;ix. Int. code + (941)943-4984
'This work was supported by Hoechst AG. A. K. thanks the Fonds der Chemischen lndustrie for a doctoral fellowship. The following colleagues also contributed to this paper: Dr. K . Mayer, J. Kiermeier, Dr. ter Meer(MS). Prof. K .
Mullen. Dip].-Chcm. F. Bcer (GPC). Dr. E Eibler. DiplL-Chem. R. Vasold
(HPLC). Dr T. Burgermeister, and F. Kastner ( N M R )
bond common to two annulated six-membered rings (6,6
bonds), and 2 can be reduced to the h e x a a n i ~ n . '51
~ .Owing to
these properties, the [8 + 21 cycloaddition of 1 with 2 should
lead to aniiulated fullerenes,r61which would open up a potential
route to polyfunctional fullerene derivatives with opto-electronic
properties. We report here for the first time on an [X 21 cycloaddition of 2 (Scheme 1).
Schemc 1. Reaction of 8-methoxyheptafulvene ( 1 ) with C,,] 12)
Treatment of 8-methoxyheptafulvene (1) with C,, (2) (in
toluene, under a nitrogen atmosphere, at room temperature)
gave a mixture of products whose HPLC trace shows more than
90 % of one main product. The main fraction from gel permeation chromatography (GPC) on polystyrene gel (eluent: chloroform) was further purified by column chromatography ( S O 2 ,
eluent: toluene/hexane, 6: 4) .['I The structure and stereochemistry of the main product were determined by mass spectrometry,
and NMR and electron spectroscopy. A peak at nz;: 854 in the
mass spectrum (negative ion, secondary ion mass spectrometry)
proves the formation of a monoadduct. In addition, the C,,
signal at mi= 720 indicates that a thermal retro-[8 + 21 cycloaddition occurs. The 'H NMR spectrum displays the typical resonance signals of a tetrahydroazulene moiety: [*I the signals for
1'-H. 8'-H, 3'a-H in 3 are shifted downfield in comparison to
those of structurally related tetrahydroazulenes, reflecting the
magnetic anisotropy of the spherical C,, unit. In contrast to the
C,, derivatives studied up to now, the reduced symmetry of 3
(C, rather than Ih) results in the separation of all signals in the
13CN M R spectrum. The two signals at 6 =75.13 and 71.58 due
to the quaternary aliphatic carbon atoms show conclusively that
no bonds within the carbon sphere were broken. Further proof
supporting the proposed structure comes from 13C NMR polarization transfer experiments (I3C DEPT). The seven methine
carbon atoms and the carbon atom of the methoxy group give
positive signals. The UV/VIS spectrum shows a band at 2 =
432 nm, which is also observed in other C,, cycloadducts bridged
at a 6,6 bond.[" The [8 21 cycloaddition can take place at either
the 6,6 or the 5,6 double bond. Reaction at the 6.6 linkage would
produce two c'isltrans diastereomers and at the 5,6 bond up to
four diastereomers. Because of the slowness of the [8 + 21 cycloaddition and by analogy with known examples, we assign
structure 3 to the major product. In this case the 6,6 double
bond in 2 reacts with the sterically less hindered side of 1, giving
3 with hydrogen atoms C-3'a and C-I' in the ci.s configuration.
The spatial proximity of 3'a-H and 1'-H was confirmed by NOE
experiments. The regioselectivity of the reaction is also supported by semiempirical calculations (MNDO, AM1, PM3) .l'"l The
calculated heats of formation ff, of the possible isomers as well
as the length ofthe bridged single bond length in the C,, unit are
shown in Table 1 .
Cyclovoltammetric and spectroelectrochemical studies''']
show that the cyclopentane-annulated fullerene 3 is reduced in
four reversible or quasi-reversible steps (Table 2, Fig. 1). The
radical anion Y-,dianion 3 2 - , and radical trianion 3'3- were
Без категории
Размер файла
466 Кб
cofactor, selection, isolation, rna, aptamer, biological, vitro
Пожаловаться на содержимое документа