Remote control by protecting groups in the diastereoselective addition of acetylides to 2-(4-quinolyl)propanol

Article abstract of DOI:10.1016/S0957-4166(98)00418-2

Differently O-protected 2-(4-quinolyl)propanols undergo, in the presence of chloroformate esters, a regio- and stereoselective addition of acetylides on the carbon α to nitrogen with moderate to good diastereoselectivity. The bulkiness of the OH protecting group, which has a 1,7 relationship with the newly created stereocentre, appears to be mainly responsible for this remote stereocontrol.

Full text of DOI:10.1016/S0957-4166(98)00418-2

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 9 (1998) 3923–3927
Remote control by protecting groups in the diastereoselective
addition of acetylides to 2-(4-quinolyl)propanol
Giuseppe Guanti,∗ Sara Perrozzi and Renata Riva ∗
Dipartimento di Chimica e Chimica Industriale and C.N.R., Centro di Studio per la Chimica dei Composti Cicloalifatici ed
Aromatici,† via Dodecaneso 31, I-16146 Genova, Italy
Received 21 September 1998; accepted 14 October 1998
Abstract
Differently O-protected 2-(4-quinolyl)propanols undergo, in the presence of chloroformate esters, a regio- and
stereoselective addition of acetylides on the carbon α to nitrogen with moderate to good diastereoselectivity. The
bulkiness of the OH protecting group, which has a 1,7 relationship with the newly created stereocentre, appears to
be mainly responsible for this remote stereocontrol. © 1998 Elsevier Science Ltd. All rights reserved.
Quinolines, together with their hydrogenated derivatives, are components in many natural products.1
These compounds generally have complex structures and usually only one of the possible stereoisomers
is noteworthy for biological activity.
The stereoselective functionalization of quinolines, having at least one stereogenic centre as a side
chain, by nucleophilic addition to an activated heterocyclic system, is still under explored. Only a few
examples are known in the literature and they usually involve the nucleophilic addition to a partially rigid
ring fused quinoline system, bearing at least two stereogenic centres.2a,c,d However, recently Isobe’s
work, on the synthesis of some Dynemicin A model compounds,2b has shown that in the addition to
quinolyl derivatives containing only one stereocentre, it is possible to get a good diastereoselection if a
suitable combination of various factors affecting the stereochemistry of the reaction is realized.
∗
†
Corresponding authors. E-mail: guanti@chimica.unige.it; riva@chimica.unige.it
Associated to the National Institute of C.N.R. for the Chemistry of Biological Systems.
0957-4166/98/$ - see front matter © 1998 Elsevier Science Ltd. All rights reserved.
PII: S0957-4166(98)00418-2

3924
G. Guanti et al. / Tetrahedron: Asymmetry 9 (1998) 3923–3927
In connection with our interest in the chemistry of enediynes,3 we recently turned our attention to the
synthesis of some Dynemicin A analogues, using 1 and 2 as precursors. For this purpose, following a
chemoenzymatic procedure, we recently prepared 2 in both enantiomeric forms.4 Now, we report our
results on the synthesis of racemic 1 and the effect of different factors on the regio- and stereoselectivity
with the addition of some acetylides to this substrate. The employed quinolines were synthesized as
reported in Scheme 1, starting from 4-ethylquinoline 3, which was prepared by a modified Skraup
procedure.5,6 Treatment of 3 with formaldehyde gave 1a,7 and with the protection of the alcoholic
function under varying conditions gave the derivatives 1b–h.8
Scheme 1.
On these compounds we studied the addition of trimethylsilylacetylides in the presence of suitable
chloroformate esters (Scheme 2), which enhance the electrophilicity of the heterocyclic ring.9 We chose
first quinoline 1b and tested the influence of the following variables both on yield and diastereomeric ratio
(Table 1): (a) various chloroformates (entries 1–3), (b) various experimental details (order of addition
of the reagents, temperature, etc.; entries 3–10), and (c) various counterions of the acetylide (entries
11–12). Although C-4 is also susceptible to nucleophilic attack,2b we usually found (except in entry 11)
a very regioselective reaction in which the nucleophile attacked 1b only at position 2 to give a mixture
of the stereoisomers 4 and 5 with a moderate diastereomeric ratio. Whatever the protecting group or
nucleophile used, it is noteworthy that in all the performed reactions (Tables 1 and 2) diastereomer 4
always prevailed.10
Scheme 2.
With the exception of entry 1, for each change in experimental conditions, we did not notice any
appreciable influence on the d.r., which was usually in the range 64–68%:32–36%. However, by changing
the order of addition of the reagents (see note c in Table 1), only the yield of the reaction was substantially
influenced. The best conditions required the sequential addition of nucleophile and chloroformate to a
quinoline solution (entry 3) at −78°C, followed by slow warming of the reaction mixture; alternatively,
a solution of nucleophile and chloroformate, warmed from −78 to 0°C, can be treated with the quinoline
at −78°C (entry 9). The nature of chloroformate (entries 1–3) has an important influence on the ideal
reaction temperature, PhCO2Cl being the best also in terms of diastereomeric ratio. Changing the
nucleophile from RMgBr to RMgCl produced an appreciable lowering of the yield (entries 3, 12), while

G. Guanti et al. / Tetrahedron: Asymmetry 9 (1998) 3923–3927
3925
Table 1
Addition of acetylides to 1b using different experimental conditions (R1=SiMe2t-Bu; R3=SiMe3)
Table 2
Addition of various nucleophiles to differently protected quinolines (R2=Ph, nucleophile=
R3–C≡CMgBr)
using the corresponding cerous acetylide (entry 11) we obtained not only a very low yield in addition
products, but we observed also a certain quantity of the regioisomers derived from nucleophile addition
on C-4.
In order to test the scope of this reaction we examined the influence of the hybridization of the attacking
carbon, using sp2 and sp3 nucleophiles, such as vinylmagnesium bromide and ethylmagnesium bromide.
In both cases the reaction was very sluggish and we obtained only small amounts of addition products.

3926
G. Guanti et al. / Tetrahedron: Asymmetry 9 (1998) 3923–3927
Moreover, in the second case, we isolated also small amounts of products arising from addition to C-4,
which slowly decomposed to give the starting compound 1b.
Since we were not satisfied by the extent of the chiral induction, we studied under experimental
conditions equivalent to the ones reported in entry 3 (Table 1), the influence of different protecting groups
on the primary alcoholic function. The results are reported in Table 2. Except for 1e (entry 5), which gave
a decreased d.r. compared to 1b, or for 1d (entry 4), in all other cases (entries 1–3, 6–11) we found an
increase in stereoselectivity, with the triphenylmethyl as protecting group giving the best results.
It is not easy to rationalize the influence of the protecting group: while in the case of TBDPS or
trityl groups the increase of diastereoselection can be attributed to the greater bulkiness, the induction
observed with 1a (in this case the reaction must occur on the alcoholate, due to the presence of an excess
of acetylide) or 1d can hardly be explained in terms of just steric hindrance. Moreover, anisyl or p-
nitrophenyl groups gave an enhanced induction. Thus, it is likely that stereoelectronic effects are also
operating. The stereochemical outcome of the addition reaction, together with forcefield calculations
performed with Chem 3D Plus from CSC on a simplified analogue of 1b (R=SiMe3), suggest that the
‘H-inside’ conformation 6 (with the benzylic hydrogen directed through the peri hydrogen) with the
nucleophile attacking preferentially from the opposite side to CH2OR could represent a reasonable model
of the transition state (Scheme 3). The bulkiness of the alcohol protecting group is probably responsible
for the enhanced induction observed in entries 8–9 and the use of different acetylides (entries 10–11) do
not seem to affect in an appreciable way the yield or diastereoselection. Thus, the two stereocentres in 4
being in a 1,4 relationship and the O-protecting group in the side chain and the new created stereocentre
in a 1,7 relationship, this reaction represents a further interesting example of long-range stereocontrol.11
It is noteworthy that although the control of diastereoselectivity achieved in this study is not excellent
(up to ca. 7:1), simple crystallization of the diastereomeric mixture derived from 1h (entry 9), allowed
us to increase this ratio to 99:1. This result is particularly interesting in view of a possible utilization of
these intermediates in the synthesis of Dinemycin A analogues. Studies in this field are now in progress
in our laboratories.
Scheme 3.
Acknowledgements
We wish to thank M.U.R.S.T. and C.N.R. for financial support of this project.
References
1. (a) Alvarez, M.; Salas, M.; Joule, J. A. Heterocycles 1991, 32, 759–794. (b) Michael, J. P. J. Nat. Prod. Rep. 1995, 12,
77–89 and 465–475.
2. See for example: (a) Nicolaou, K. C.; Dai, W.-M. J. Am. Chem. Soc. 1992, 114, 8908–8921; (b) Nishikawa, T.; Yoshikai,
M.; Obi, K.; Kawai, T.; Unno, R.; Jomori, T.; Isobe, M. Tetrahedron 1995, 51, 9339–9352 and references therein; (c) Shair,

G. Guanti et al. / Tetrahedron: Asymmetry 9 (1998) 3923–3927
3927
M. D.; Yoon, T. Y.; Mosny, K. K.; Chou, T. C.; Danishefsky, S. J. J. Am. Chem. Soc. 1996, 118, 9509–9525; (d) Myers, A.
G.; Tom, N. J.; Fraley, M. E.; Cohen, S. B.; Madar, D. J. J. Am. Chem. Soc. 1997, 119, 6072–6094.
3. (a) Banfi, L.; Guanti, G. Angew. Chem., Int. Ed. Engl. 1995, 34, 2393–2395; (b) Banfi, L.; Basso, A.; Guanti, G.
Tetrahedron 1997, 53, 3249–3268; (c) Banfi, L.; Guanti, G. Eur. J. Org. Chem. 1998, 1543–1548.
4. Banfi, L.; Guanti, G.; Mugnoli, A.; Riva, R. Tetrahedron: Asymmetry 1998, 9, 2481–2492.
5. (a) Schering-Kahlbaum, A.-G. British Patent 283,577, 14 January 1927; C.A. 1928, 22, 4132; (b) Zöllner, C. US Patent
1,804,045, 12 May 1930; C.A. 1931, 25, 3668.
1
6. All new compounds quoted in this communication were fully characterized by H-NMR, 13C-NMR (in most cases), IR,
GC–MS (when feasible).
7. Guanti, G.; Narisano, E.; Riva, R. Tetrahedron: Asymmetry 1997, 8, 2175–2187. Emiacetalic compounds of formaldehyde
contaminating 1a were hydrolyzed with K2CO3.
8. 1b,c: R2t-BuSiCl, imidazole, DMF, r.t.; 1d: NaH, MeI, THF, r.t.; 1e: Ac2O, Py, r.t.; 1f,g: ArOH, PPh3, DEAD, CH2Cl2,
r.t.; 1h: TrCl, Py, CH2Cl2, r.t.
9. Fraenkel, G.; Cooper, J. W.; Fink, C. M. Angew. Chem., Int. Ed. Engl. 1970, 9, 523.
10. Compounds 4 and 5 were transformed into lactones 7 and 8 respectively. [a. m-CPBA, CH2Cl2, r.t., 92 and 90%; b. p-
TSA, MeOH, r.t., 60 and 51%; c. TPAPP, NMO, CH2Cl2, r.t., 62 and 61%] This route has most likely maintained the
configuration of the two original stereogenic centres. The trans relationship between H2 and H3 was unambiguously
established by J values in the proton spectrum of the diols that are precursors of 7, 8 (9.4 and 7.7 Hz respectively).
1
Comparing the H-NMR spectra of 7 and 8 we observed in 7 a strong upfield shift (2.52 vs. 3.15) for H α to C_O and
in 8 a strong upfield shift (0.77 vs. 1.22) for CH3 α to C_O. These facts can be explained by the shielding effect of the
aromatic ring on the groups positioned over it (H in 7, CH3 in 8). Moreover NOEDIF experiments showed a positive
NOE effect between H3 and CH3 α to C_O in 7 and between H3 and H α to C_O in 8 while a strong NOE effect was
observed between OCH3 and H3 in both stereoisomers. These data agree also with the conformation predicted by force
field calculations and allowed us to assign unambiguously the relative configuration of all the stereocentres in 7 and 8.
Further details will be given in a forthcoming full paper.
11. As a recent review in this field see: Thomas, E. J. J. Chem. Soc., Chem. Commun. 1997, 411–418.