Polyfluoropyridyl Glycosyl Donors - American Chemical Society

---

Chapter 19

Polyfluoropyridyl Glycosyl Donors 1

1,

Christopher A . Hargreaves , Graham Sandford *, and Benjamin G . Davis 2

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

1

Department of Chemistry, University of Durham, South Road, Durham, DH1 3LE, United Kingdom Chemistry Research Laboratory, University of Oxford, Mansfield Road, Oxford OX1 3TA, United Kingdom

2

Synthesis of stereochemically defined oligosaccharides by a series of glycosylation processes involving the reaction between a glycosyl donor and acceptor is essential to synthetic carbohydrate chemistry and glycobiology. However, despite the importance of glycosylation chemistry and the development of much sophisticated methodology, there remains no general and stereoselective procedure for the synthesis of oligo- and polysaccharides. Families of novel glycosyl donors have been conveniently synthesized from polyfluoropyridine derivatives and, in a short series of model glycosylation reactions, have shown that they offer great possibilities for the controlled stereoselective synthesis of oligosaccharide systems. In particular, their tunable reactivity offers the promise of broad ranging activity suitable for cascade synthesis.

© 2007 American Chemical Society

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

323

324 Carbohydrates are, of course, of tremendous importance in biological chemistry and play key roles in many natural processes including metabolism, as structural components of many biosynthetic pathways and as messengers in precise communication events. For example, monosaccharide galactolipids are important binding elements in the initial infective interaction of the outer proteins of the HTV-1 virion with the surface of leukocytes, an anchoring process that is severely disrupted simply by the alteration of only one stereocentre in the carbohydrate. Such factors have recently provided increased impetus for research into the rapidly growing field of glycobiology where, for example, the concise synthesis of stereochemically defined oligosaccharides and glycocconjugates, " consisting of complex glycan units attached to a peptide or protein system is of fundamental importance. However, the synthesis of such systems for use as both mechanistic probes or as therapeutic agents presents a formidable challenge because, unlike oligopeptide and nucleic acid synthesis, there are no general automated, synthetic methodologies yet available. 1

2

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

3

5

30, |3 GLUCOSIDIC LINK

3a. a GLUCOSIDIC LINK

Scheme 1. Glycosylation reactions

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

325 Oligosaccharides are synthesized by a series of key glycosylation reactions in which, for example, a glycosyl donor 1 reacts with a glycosyl acceptor 2 in the presence of a suitable activating system to give disaccharide 3. (Scheme 1). While regioselectivity may be controlled by protecting group strategies, a successful glycosylation reaction requires both chemoselectivity, that is, the desired leaving group must be displaced by the appropriate hydroxyl group to give the correct disaccharide, and stereoselectivity, attack of the nucleophilic hydroxyl group must occur specifically at either the a or P face to give a single diastereoisomer. Consequently, for the coupling of two monosaccharide residues 4 and 5 bearing single free hydroxyl groups, Scheme 2, eight possible products may, in principle, be obtained from a glycosylation reaction that has poor selectivity.

5 - 5,

a and (3

Scheme 2. Synthesis of a disaccharide with no reaction controlfromdual function donor-acceptor building blocks

The challenge for the synthetic chemist is, therefore, to devise strategies that overcome these difficulties and allow syntheses of simple disaccharides (e.g., cc-4-5) from two residues. Moreover, such dual-

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

326 function acceptor-donor systems have the potential, with correct selectivity, to not only allow access to di- but also oligosaccharides through iterative reaction cascades. Chemoselectivity can be achieved by the use of glycosyl donors that bear substituents at the anomeric position which have different leaving group abilities and, consequently, may be selectively activated towards nucleophilic attack by specific activating species. This is illustrated in Scheme 3, where 4, bearing leaving group L G i which is activated specifically over L G by LG activator, is attacked by the hydroxyl group of the unactivated glycosyl donor 5-LG , to give the required disaccharide 4-5 only. This process is termed tuning and the reactivity of a glycosyl donor can be adjusted through the use of different protecting groups, activators and/or anomeric leaving groups. 2

r

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

2

910

OR

OR

OR

S

Q

R

4-5

Scheme 3. Controlled chemoselective glycosylation by selective activation ofLGj Given the importance of glycosylation reactions to glycobiology, many classes of glycosyl donors have been synthesized and their use in oligosaccharide synthesis assessed. Whilst systems such as thioglycosides 6, trichloroacetamidates 7, alkenyl glycosides 8 and glycosyl fluorides 9 (Scheme 4) have been exploited in some very elegant oligosaccharide syntheses, there is still no general solution to the glycosylation problem. Typically, each donor system displays its own peculiar selectivities in specific carbohydrate systems. Furthermore, in many cases, such as the trichloroacetamidate and glycosyl fluoride systems, it is often difficult to synthesise a related family of donors which bear similar but slightly different substituents at the anomeric position that have different leaving group abilities that would make it possible for their reactivity to be subtly 'tuned* as required. 6 - 8

Scheme 4. Glycosyl donors in general use

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

327 Consequently, we have initiated a research programme that aims to devise synthetic approaches to families of related glycosyl donor systems, in which the leaving groups may be readily tuned, that provide a generally applicable solution to the glycosylation problem. We have pursued a glycosylation strategy whereby a range of electronically varied electron deficient pyridyl systems 10 bearing a range of functionality on the heterocyclic ring are utilized as leaving groups which may be activated towards glycosylation by a Lewis acid. (Scheme 5).

(RO) ^S^O^N^

+ HQ-^pgyLGz

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

n

N

Lewis Acid (LA) ^

(R0) ^^u9^ ^ n

(OR)

n

+

r

o/

*0L < °'^ ^ N

OH

LG2

(OR)n

Scheme 5. Novel glycosyl donor systems

We opted to use pentafluoropyridine 11 as our polyfunctional pyridine scaffold because this system is highly susceptible towards nucleophilic attack and could, in principle, be readily attached to a sugar residue and/or functionalized by a sequence of nucleophilic aromatic substitution reactions. The potential use of such systems has been noted previously. Using the strategy postulated in Scheme 6, families of related glycosyl donors could, potentially, be synthesized by reaction of pentafluoropyridine 11 with a suitable carbohydrate derivative and a series of nucleophiles to give several related classes of glycosyl donors 12, 13 and 14, depending simply on the order of nucleophilic substitution. Class 12 is formed by reaction of pentafluoropyridine 11 directly with a saccharide system, Class 13 by reaction of a nucleophile with PFP followed by a saccharide, and so on. Whilst donor 12 has been synthesised before, the full opportunities for using highly fluorinated pyridine systems for reactivity tuning have not been exploited. For all of these glycosyl donors, leaving group ability could, we hoped, be controlled by both the substituents on the pyridine rings (Nuc!, Nuc , etc.) and/or the Lewis acid activator. 11

11

12

2

Reactions of Polyfluoropyridine derivatives The chemistry of pentafluoropyridine 11 has been developing over many years since the its first viable synthesis and is dominated by nucleophilic aromatic substitution reactions. A wide variety of nucleophiles (Nuc Scheme 7) react with pentafluoropyridine to give, in 11,13

14

1}

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

328

Scheme 6. Synthetic strategy for the synthesis of related families of polyfluoropyridyl glycosyl donors

the vast majority of cases, products 15 arising from the selective displacement of the fluorine atom located at the 4-position. Many tetrafluorinated pyridine derivatives 15 have been synthesized by this methodology but the further reaction of such systems with subsequent nucleophiles to, potentially, give 16 have not been explored to any great extent, although some initial results have been reported. In principle, reaction of a second nucleophlile (Nuc , Scheme 7) could give rise to three products, 16, 17 and 18, arising from the displacement of the ortho and meta fluorine atoms or the substituent, Nuci, located at the 4position, respectively. However, the factors that determine this potential regioselectivity have not been established. Synthesis of glycosyl donors could, therefore, potentially be achieved by reaction of a suitable carbohydrate nucleophile and pentafluoropyridine to give donors of type 12. Suitable model reactions involving pentafluoropyridine have been recorded. In contrast, however, little is known of 13 or 14. Before embarking on the synthesis of glycosyl donor types 13 and 14 we decided to carry out model reactions between a series of tetrafluoropyridine derivatives (Nuc! = O C H , NH/Pr, C N , N 0 , Scheme 7) and sodium ethoxide to establish the regioselectivity of nucleophilic substitution on these systems which may be dependent upon the nature of the substituent located at the 4-position (Scheme 8). 1 5

2

12

2

5

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

2

329

11

15

16

17

18

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

Scheme 7. Nucleophilic substitution in penta- and tetra-fluoropyridine systems

Model reactions for: ( R O ) - ^ S ^ ^ /

O H

(RO)n

n

OC H 2

2

5

CH ONa, CH OH 3

13

OC H F^ A . / F 5

Nu^

3

(75%)

reflux F

F

F

3

(81%) F"

CF(CF ) /F 3

F

N 23

3

EtOH, reflux

21

/L

"OCH

CH CH20Na

"xr F^

N 20

19

2

N ^ X)CH CH 22 2

CF(CF ) /F 3

CH CH ONa 3

2

F\

3

2

A .

(81%)

EtOH, reflux F

F

CN

F

F

i5c

CH CH ONa 3

25

F F

2

CN F^ A . / F

+

3

CN F ^ O E t

2

MeCN

F"S*S N0

N "OCH CH 24

V

F "OEt 26a,b 87%, 1 :3

^

F

OEt 2

lY N ^ h 27

CH CH ONa 3

2

EtOH, reflux

(43%) + Others F^N/^OEt 28

Scheme 8. Reactions of tetrafluoropyridine derivatives with alkoxides as models for preparation of donor type 13

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

330 Reaction of 4-methoxy-tetrafluoropyridine gave a complex mixture of products arising from substitution of both the fluorine atom at position 2 and the methoxy substituent itself. In contrast, ethoxy substituents, which are poorer leaving groups than methoxy, were not displaced by sodium methoxide and the 2-methoxy-4-ethoxy pyridine system 20 was obtained in high yield from 19. The 4-amino- and 4-perfluoroisopropyl derivatives, 22 and 24, gave products from substitution of the ortho fluorine selectively. In contrast, the cyano system 25 led to a mixture of ortho and meta substituted products 26a,b due to the activating influence of the electron withdrawing cyano substituent on the adjacent 3-position which competes effectively with the activating influence of ring nitrogen. Nitro substrate 27 gave product 28 arising from displacement of the labile nitro group only. These simple model reactions indicated that 11,19, 21 and 23 could be useful substrates for the preparation of glycosyl donors whereas 25 and 27 are less suitable. Synthesis of Polyfluoropyridyl Glycosyl Donors Using the methodology from the model studies described above, a short series of glycosyl donors was synthesized by reaction of the tetrabenzyl glucose salt 29 and either pentafluoropyridine 11 or an appropriate tetrafluoropyridyl system such as 22. (Scheme 9)

F

Scheme 9. Synthesis of polyfluoropyridyl donors

Other trisubstituted donors 35 and 36 were also synthesized by a three stage strategy (Scheme 10) adapted from earlier work involving the chemistry of perfluoro-4-isopropylpyridine, thus extending the number of possible donors that may be accessed by the general strategy outlined in Scheme 6. 15

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

331

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

Scheme 10. Synthesis ofpolyfluoropyridyl donors

A l l donors were found to be hydrolytically stable and, in some cases, the individual anomers could be separated and purified by high pressure flash chromatography. As well as making use of the usual characterization methods, we observed that the a and P anomers showed distinct F nmr resonances, making the identification and measurement of anomeric ratios a ready process. 1 9

Model Glycosylation reactions We began our studies to establish the feasibility of utilizing these novel glycosyl donors by carrying out some model glycosylation reactions using cyclohexanol as the glycosyl acceptor, to mimic a sugar residue bearing a secondary hydroxyl group, and activation by a range of Lewis acids including boron trifluoride, copper (II) triflate, titanium tetrachloride and aluminium trichloride, amongst others. A l l reactions were carried out in acetonitrile at room temperature, monitored by F nmr and the product disaccharide 37 was isolated in each case. Reactions are shown in Scheme 11 which gives the yields of disaccharide 37 after 24 h. The data from Scheme 11 reveal that donor 31 is activated selectively by copper triflate whereas donors 34 and 30 are not. Titanium tetrachloride activates donors 31 and 30 but not 34 whilst all three donors are activated by boron trifluoride and aluminium trichloride at various rates. Although only initial, these results clearly demonstrate that glycosyl donor families of this type have variable reactivity that may be tuned via the substituents attached to the pyridine ring and the Lewis acid activator - an exciting prospect. Furthermore, activation of anomerically pure P-glycosyl donor 3ip by copper triflate gave the corresponding disaccharide 37 as a mixture of a and P anomers in the ratio 98 : 2., indicating an almost complete stereochemical inversion of configuration at the anomeric site upon 1 9

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

332 OBn 0

Lewis acid BnO'^\^^ \ cydohexanol , fenoX^^O^ stir, rt n

Donor OBn

37

F

31

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

OBn

CF(CF )

M

3

2

Donor

Lewis Acid

31 34 30

Cu(S0 CF ) Cu(S0 CF ) Cu(S0 CF )

3 1

T i C

3

3

3

2

3

3

2

3

3

2

75 0 0

4

5 5

TiCI

34 OBn

,

Yield after 24 h

0

4

0

T i C ,

1

4

2

F I

NH

30

Y

Scheme 11. Glycosylation reactions activated by a variety of Lewis Acids glycosylation. (Scheme 12) This indicates that the mechanism of this particular process may be S 2 in character which is especially unusual in a potentially coordinating solvent such as acetonitrile. In future, this may allow us to control the stereochemistry of the glycosidic bond formation step in oligosaccharide synthesis. N

OBn

OBn OH

CXj(SObCF3) cyctohexanol , 2>

, T. 31& 100% pure

I r

—=i

-

X

l

*

^

37,78% 98%a,2%panomer

Scheme 12. Stereochemistry of the glycosylation process

The presence of fluorine atoms within the glycosyl donors allow all the glycosylation reactions to be directly monitored by F nmr spectroscopy very accurately and conveniently. Scheme 13 shows reaction between donor 31p with cyclohexanol initiated by copper(H) triflate over time. Resonances attributed to the donor decrease over time with a corresponding increase in the resonances arising from the pyridinol side product formed after glycosylation has occurred. This 'real time' reaction monitoring is a great benefit for convenient analysis of these processes and an advantage of using fluorinated systems. 19

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

333

6 (ppm) 19

Scheme 13. Monitoring a glycosylation reaction over time by F nmr

The possibility of measuring the conversion of glycosyl donor to products very conveniently and non-intrusively by F nmr over time has allowed us to measure rates of reactivity of each donor system upon activation by a variety of Lewis acids, adapting a process analogous to that carried out on thioglycosde donors by L e y and Wong. Analysis of data concerning the conversion of glycosyl donor 33 to products using secondorder kinetic analysis, in line with our previous observation that these processes may be S 2 in character, allows the tentative determination of the initial rate constant by utilizing a plot of l/[Donor] - l/[Donor] versus time where the gradient is the rate of glycosylation for the acceptor upon activation by a range of Lewis acids. (Scheme 14) For glycosylation of 33 with cyclohexanol, we see that varying the Lewis acid can have a profound effect on the rate of reaction. For example, reaction using boron trifluoride etherate as activator is 120 times faster than when titanium tetrachloride is used. Wong has suggested that, for an ideal set of glycosyl donors, it would be desirable for a series of related donors to have relative glycosylation rates of between 1 and 1000. Even 1 9

10

16

N

0

16

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

334 OBn

OBn 0C H 2

B

N

F

^ o 5 ^ < ^ ^ 0 ^ ^ BnO \\ T

5

CydQtoxanol (1 gguiv.)» ^ o ^ ^ ^ Q Lewis acid(1.5equiv.) ^^^^^^BnO^ OH

OC H 2

300 -,

5

BF (C H ) 0 k = 11.9x10 dm molV 3

2

3

5

2

3

1

1 I FeCI 1 [!k = 11.3x10' dm mor s ! 3

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

250 -

3

200

150

50

1

i /! Aici jk = 5.5x10" d r n W V j

t 1

I 100

1

[ i f 1 f i i i

o

-

3

3

3

1

A

I1

u

j

Cu(S0 CF )

1

3

3

2

k =0.2x10" dm mor s" 3

' Ii \i

3

1

1

TiCI k =0.1 x10' dm^or's' 4

3

50000

100000

150000

1

200000

Time (s) Scheme 14. Kinetic data for the reaction of 33 with cyclohexanol

with this single donor, a range of relative rates from 1 to 120 is already available by varying the Lewis Acid activator alone. A larger range of donors, based on fluoropyridine scaffolds, would, we believe span a much wider range. Finally, we have used this family of glycosyl donors for the synthesis of disaccharide 38 to demonstrate the potential of using these systems in oligosaccharide synthesis.(Scheme 15)

Conclusion We have developed general routes for the synthesis of families of related polyfluoro-pyridyl glycosyl donors from reactions of highly

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

335 OOH

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

Scheme 15. Synthesis of disaccharides

fluorinated pyridine derivatives and appropriately protected monosaccharide systems using nucleophilic aromatic substitution processes. The applicability of the glycosyl donors has been demonstrated in several glycosylation reactions and we find that the activation of donors towards attack by glycosyl acceptors and the rates of glycosylation depends on both the substituents that are located on the pyridine ring and the Lewis acid activator present. The demonstrated possibility of 'tuning the reactivity of the polyfluoro-pyridyl glycosyl donors provides a new approach to the important 'glycosylation problem'. Several early advantages of using fluorinated donors have been observed, such as analysis by F nmr spectroscopy and, in particular, improved stereochemical control for this very difficult transformation. We hope, in future, to be able to utilise our donor systems to address key synthetic challenges in the field of glycobiology.

1

1 9

References 1. 2. 3. 4.

Davis, B . G . Chem. Ind. 2000, 134. Bhat, S. Proc. Natl. Acad. Sci. USA 1991, 88, 7131. Davis, B . G . J. Chem. Soc., Perkin Trans 1 1999, 3215. Gridley, L . J.; Osborn, H . M. I. J. Chem. Soc., Perkin Trans 1 2000, 1471. 5. Pratt, M. R.; Bertozzi, C. R. Chem. Soc. Rev. 2005, 34, 58. 6. Paulsen, H . Angew. Chem. Intl. Ed. Engl. 1982, 21, 155. 7. Boons, G . J. Contemp. Org. Synth. 1996, 3, 173. 8. Davis, B. G. J. Chem. Soc., Perkin Trans 1 2000, 2137. 9. Mootoo, D . R.; Konradsson, P.; Udodong, U . ; Fraser-Reid, B . J. Am. Chem. Soc. 1988, 110, 5583. 10. Grice, P.; Ley, S. V . ; Pietruszka, J.; Osborn, M. I.; Henning, W. M.; Priepke, H . W. M.; Warriner, S. L . Chem. Eur. J. 1997, 3, 431.

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.

336

Downloaded by CORNELL UNIV on August 18, 2016 | http://pubs.acs.org Publication Date: January 11, 2007 | doi: 10.1021/bk-2007-0949.ch019

11. Brooke, G . M. J. Fluorine Chem. 1997, 86, 1. 12. Huchel, U.; Schmidt, C.; Schmidt, R. R. Eur. J. Org. Chem. 1998, 1353. 13. Chambers, R. D.; Sargent, C. R. Adv. Heterocycl. Chem. 1981, 28, 1. 14. Chambers, R. D.; Hutchinson, J.; Musgrave, W. K . R. J. Chem. Soc. 1964, 3573. 15. Chambers, R. D.; Hassan, M . A . ; Hoskin, P. R.; Kenwright, A . ; Richmond, P.; Sandford, G . J. Fluorine Chem. 2001, 111, 135. 16. Zhang, Z. Y.; Rollman, I.; Ye, X. S.; Wischnat, R.; Baasov, T.; Wong, C. H . J. Am. Chem. Soc. 1999, 121, 734.

Soloshonok et al.; Current Fluoroorganic Chemistry ACS Symposium Series; American Chemical Society: Washington, DC, 2007.