Deoxygenated and alkylated furanoses: Thorpe—Ingold effects on tautomeric equilibria and rates of anomerization

Carhohydrare

Elsevier

21

210 (1991) 21-38

Research,

Science Publishers

B.V., Amsterdam

Deoxygenated and alkylated furanoses: Thorpe-Ingold effects on tautomeric equilibria and rates of anomerization Joseph R. Snyder and Anthony S. Serianni* Depariment

of Chemistry

und Biochrmbtry.

University

ofNotre

Dame.

Notre

Dame,

IN 46556

(U.S.A.)

(Received May 3rd, 1990; accepted July 12th, 1990)

ABSTRACT

2-Deoxy-D-gl~c,c,ra-tetrose,

3-deoxy-uL-glycero-tetrose,

rose, 3-C-methyl-oL-erythrose,

3-C’-methyl-oL-threose,

oxy-5-O-methyl-u-erythro-pentose carbon,

and characterized

have been prepared,

and ring-opening

saturation-transfer substitution ization.

n.m.r.

was found

Increased

kc,,,\,. In contrast.

2-deoxyfurdnose

and acyclic (k,,,)

spectroscopy

to significantly

substitution furanose

enhances

k,,,

anomerization

(aldehyde

and ring-closing

forms

(k,,,,,) rate constants

at pZH 5.0 (acetate

at the anomeric

(75 MHz) spectroscopy.

and hydrate) buffer)

and 60

affect both the thermodynamics the proportion

and 3-de-

in some cases with “C-substitution

by ‘H- (300 and 620 MHz) and “C-n.m.r.

tions of cyclic (2 and /j furanoses) (‘H20) solution,

3-deoxy-3,3-di-C-methyl-uL-y!vcero-tet-

2-deoxy-5.O-methyl-u-crythro-pentose

The propor-

were determined were measured The degree

and kinetics

of cyclic forms in solution

in aqueous by ‘H and “C

of furanose

of furanose

by stimulating

ring

;anomerfuranose

was less affected by the degree of substitution;

implicate furanose ring conformation

however, kinetic studies of as a potential determinant of kopen.

INTRODUCTION

The spontaneous ring-opening and ring-closing reactions that are characteristic of reducing sugars in solution’ provide a unique biologically important system to study the mechanistic features of reversible intramolecular hemiacetal and hemiketal formation. Comparisons of intramolecular reactions to their intermolecular counterparts can be helpful in identifying and quantifying the factors that determine the efficiency of chemical processes. These comparisons serve as a means to better understand catalysis, the chemical basis of which lies partly on the proper juxtaposition of the reactive centers, either imposed internally (intramolecular reactions) or mediated by an external inorganic or organic agent (e.g., a metal or an enzyme). This juxtaposition, which is embodied in the concept of “effective molarity”‘, minimizes the trial-and-error search time required in the uncatalyzed reaction and results in enhanced reaction rates. In “neutral” furanoses (i.e., furanoses that do not contain an ionizable substituent), rate constants of ring-opening (kopcn) and -closing (k,,,,,) depend highly on ring structure and configuration”-‘. Within an anomeric pair, the anomer having O-l and O-2 cis opens at a similar or greater rate than that having these atoms tram, and a mechanistic model’ has been proposed to explain this behavior. In contrast, x anomers of the pentose 5-phosphates ring open more readily than B anomers at pH 4.2 (ref. S),

0008.6215/91/903.50

@ 199

I ~ Elsevier Science Publishers

B.V

‘2

.i.K.

regardless

of the relative

ring-opening

mechanism

configuration is involved

greater

solutions

than

pentoses

that

found

es’). Furthermore, similari. aqueous

these data

solutions

aldotetrofuI~;~~lost‘\

tnore quantitatively

for the reduced

thermodynamics

To examine

and alk} latcd furanoses the eflkt

of furanosc,

ofring

anomeri;latinn.

Q4H

OH 19

HO

OH 10

OH

SI’KIANNI

;t dilTcrent

he cnl~an~~d

amount

form found is 5 10 fold

(~“!g . 5Cl-methyl(id+. apiofur-ano+ V,TI’Efound to be in alk-l-substitut~ci

of .icyc~lic aldch~ Jc form in

these etYec:h i’urthcr, ~SEL‘ ~yntlwkcd

(2. 5. 7, IO. 11. 15. and 19). and euamincd

deoxygenation

H,COH,C

0 q@’

and apiofuranoses

tha: /c~,,~,,.must

suggest

and u-threose.

of the Hldupentofuranosc~

and cubstltuted

of these compounds.

scvernl dcoxygenated

H&OH&

in solutions

in order to account

that

of acyclic aldehyde

n-erythrose

k,,,,,, for the ~lldotetl-afuranoses

Hence.

aldofuranoses

that the percentage

of the ~{ldotetrofuranoses,

or 5-dcoxypentc~ses”)

and C.-Z, suggesting

al C-l

.A.S.

with these compounds.

Recent studies have indicated in aqueous

SSYI)i:K.

and :tlkyi:rt~on on lhe kinetics and

THORPE-INGOLD

EFFECTS

23

EXPERIMENTAL

Materials. -

Acrolein, acetol, 2,2-dimethyl-3-hydroxypropionaidehyde, 2-deoxy-D-erythro-pentose, and anhydrous sodium thiosulfate were purchased from Aldrich Chemical Co. p-Xylose and 5% palladium-barium sulfate (PddBaSO,) were purchased from Sigma Chemical Co. Potassium (“C)cyanide (K13CN, 99 atom-% 13C) was purchased from Cambridge Isotope Laboratories. Standard iodine solutions (0.1~) were purchased from Fisher Scientific Co. Instrumentation. - ‘H- (300 MHz) and ‘“C-n.m.r. (75 MHz) spectra were obtained on a Nicolet NT-300 Fourier-transform n.m.r. spectrometer equipped with a 293B pulse programmer and quadrature-phase detection. 2D ‘H-‘H COSY and 13C-‘H chemical shift-correlation spectroscopy’” were conducted on the same spectrometer using software supplied by G. E. NMR Systems. F.t.-n.m.r. spectra (‘H) at 620 MHz were obtained at the NM R Facility for Biomedical Studies, Departement of Chemistry, Carnegie Mellon University. ‘H- and ‘3C-saturation-transfer n.m.r. (s.t.-n.m.r.) spectroscopy was performed on the NT-300 300 MHz n.m.r. spectrometer as previously described4. S.t.-n.m.r. spectra were recorded with saturation times ranging from 0.1 to 20 s and relaxation delays of > 20 s. At least ten (10) saturation times were employed in each experiment, and signal intensities were plotted semi-logarithmically as previously described’ in order to obtain 5, from the slope. Using T, (spin-lattice relaxation times) and r, values, ring-opening rate constants (kopen)in s- ’ were determined from the relationship3. 1/r, = k open+ l/T,. ‘3C-S.t.-n.m.r. was used to obtain kopenin 5, 7, 10 and 11, while ‘H-s.t.n.m.r. was used to measure kopcnin 2, 15 and 19. Ring-closing rate constants (k,,,,,) were determined from Kegvalues and kopn. In addition to high-resolution ‘H- and 13C-n.m.r. spectroscopy, gas-liquid chromatography-mass spectrometry (g.l.c.-ms.) was used to characterize 2, 5, 7 and 15. Mass spectra of their respective peracetylated alditol acetate derivatives were obtained on a DuPont DP-102 g.l.c.-m.s. instrument operated in the direct chemical-ionization mode with isobutane and a source temperature of 200”. The peracetylated alditol acetate derivatives were prepared as described by Blakeney et ~1.~~A gas-liquid chromatographic column (2 mm id. x 1.83 m) containing SP-2340 (3% on Chromosorb W-AW) was used with a temperature program of 190-230” at 2”/min. The quasimolecular ion (M + 1)’ was used to characterize each compound: 2, m/z 233; ( l-13C)5, m/z 234; 7, m/z 261; 15, m/z 277. Spectra in all cases showed an intense signal at m/z (M - 59) corresponding to the loss of a single OAc group from the structure. 2-Deoxy-D-glycero-tetrose 2 (Scheme 1). - To a solution of 3-deoxy-D-erythropentose” l(2.2 g, 16.4 mmol in 1 mL of distilled H20) was added with stirring a solution of Pb(OAc), (8.9 g, 20 mmol) in glacial acetic acid (225 mL). After 40 min at room temperature, solid oxalic acid (5 g, 40 mmol) was added with stirring. After 15 min, the suspension was filtered (vacuum) through Celite to remove lead oxalate, and the filtrate was evaporated at 30” in uacuo to _ 5 mL and added to 100 mL of 0.1% (v/v) H,SO,. After 6 h at 37”, the acidic solution was neutralized with batchwise addition of Dowex-1

OH

3

Schcmc

x

5

4

2

X (HCO:

) resin.

chromatographed

and after concentration

on a column (4.4

x

to - 10 m I_ at 30

124 cm) containing

Dowes-50

mesh) (Ba’ ’ ) ion-exchange resin. using distilled water as the cluent” were collected at a flow rate trf 1 mL~min

ij! ~‘cI(‘L(o,it was y 8 (700~-100

Fractions

f 10 mtI

and were assayed with phenol sulfuric

acid“

l-he ! ield c>f‘2 45~~56 were pooled and concentrated to - 10 nrL at 30 in I’OCUO.

Fractions

Wil\ 70”% ( I .? g. I I .5 mmol) hascd on 1. S-Dcci.~-_;r-l,r.-glqcero-i~//.c~”\~~ 5 { ,sc~hrn~c~ -71. added a solution

of distilled

hydroxypropionaldehydc. distilled

(Kugelrohr

apparatus)

After

I5 min. the reaction mixture

was adjusted tri ptl 4.3 and aerated

hood to remove excess HC’N. The solution acid. and then further glass-tiber

batchwise

filter

After

1tac~199911

treatment

with excess Dowcu-50

a column

(4.4

x

x

was

treared

with

BaC’O,

through ii C“eJitcpad IO rcmo\c the butchwise

8 (14 ’ ) and DOMW- I x 8 (WCC.); I rains.

I JO cm) containing

Row rate of I mL;min

The

(X5(/b

Dowes-50

as theeluent.

x

S (X0

Fractions

100 inesh) (C’s” 1 IOIIt 10 mL.)u~erecoJlcctcd at ;t

and asayed for aldose using phenoi ~sulfuricb aci#‘. Fractions

containing 3-deoxy-l,r.-y!,.i,rJ~~)-tetrose IYIC’LIO

ttlrc?ugtl

was concentrated at .!(I i/7~MUO to _ 5 mL and chromatugraph~d on

exchange resin”‘. usingdistilledwater

ed at !O in

acetic

and wduc~d tvlrh Jld-HaSO,

The miuturc V,YIS :~acun9ll-tit~eleJ filtration

yidtf

5 (fractions

64 73) we‘rept~led arId concentrat-

based on 4). The (I-“C)-substituted

derivative of5 was

prepared and purified as described above for the unenriched c~~~~poundI-I! substituting K”C’N

5

N, in ;I well-wntrd

BaSO, precipitate (and tzxcess BaC’O,). the filtrate was deionized by qarutc, deionized solution

j(3.

IIT 4 nnL of’

was readJusted to pH 4.3 uith gIGal

to t’emokc spent catalyst, and the filtrate

to ren~ove SO,-

Wil?;

~~13 as;aycd tq g.1.c.“.

for 3 h with

acidified to pH I .7 with 18~1ti,SO,

and H, according to puhlishcd prucedura”. a

of KCh

~-dcox~gl~~crrald~ll~d~

prepared from acrolcin (311” 4I. 1 1 g. / 5 rnnltrl.

t-J,<_))with stirring.

and the solution

To an aqueous solution

I-J,(>) adjusted to p1-I 7. with glacial acetic iiCiCi

( I .i g. 20 mmol. in 60 mL. ofdistilled

for- KC’Y in the protocoi.

THORPE-INGOLD

25

EFFECTS

0

CHO

K’kN

H, ,

Pd-B.&O,

6 Scheme

7 3

3-Deoxy-3,3-di-C-methyl-DL-glycero-tetrose 7 (Scheme 3). - To an aqueous solution of KCN (2.0 g, 31 mmol, in 90 mL of distilled H,O) adjusted to pH 7.7 was added with stirring an aqueous solution of 2.2-dimethyl-3-hydroxypropionaldehyde 6 (2.0 g, 20 mmol, in 5 mL of distilled H?O) previously distilled on a Kugelrohr apparatus. After 15 min, the reaction mixture was assayed by g.l.c.14, and the solution was readjusted to pH 4.3 and aerated as described above to remove excess HCN. The solution was reduced at pH 1.7 with Pd-BaSO, and Hz according to published procedures”, deionized, and chromatographed as described above. Fractions 7&75 were pooled and concentrated at 30” in IXZCUO (70% yield). 3-Deoxy-3,3-di-C-methyl-or-(l“C)glq’cero-tetrose was prepared and chromatographed as described above for the preparation of 7 by substituting K13CN for KCN in the reaction scheme.

% L-=0

H, . Pd-E&O,

:H,OH

(HO):

(W

CH,OH

8 Scheme

CHO I

K CN

9

10

11

4

3-C-Methyl-DL-erythrose 10 and 3-C-methyl-DL-threo.w 11 (Scheme 4). --To an aqueous solution of KCN (32.5 g, 500 mmol, in 100 mL of distilled H,O) adjusted to pH 7.5 with glacial acetic acid was added acetol8 (6.85 mL, 100 mmol, glass-distilled twice). The pH was maintained at pH 7.5-7.6 with additions of dilute acetic acid. After 20 min, the solution was adjusted to pH 5.5, and an aliquot was removed for g.1.c. analysis14. The solution was readjusted to pH 4.3 and aerated for 6 h with Nz in a well-vented hood, trapping the excess HCN as previously described15h,‘7. The nitriles were hydrogenated over Pd-BaSO, at pH 1.7 as described above. After reduction was complete ( - 6 h), the solution was filtered through a glass fiber filter to remove spent catalyst, and the filtrate was treated batchwise with BaCO, to remove SO,‘-. The white suspension was filtered through a Celite pad, the filtrate was deionized by batchwise and separate treatment with Dowex-50 x 8 (20-50 mesh) (H+) and Dowex-1 x 8 (2&50 mesh) (HCO,-) ion-exchange resins, and the solution was evaporated to - 30 mL at 30” in vucuo, giving 2-C-methyl-or_-glyceraldehyde 9 in 90% yield. To an aqueous solution of 9 (2.8 g, 27 mmol, 30 mL HzO) was added a solution of

13

14

Scheme 5

7‘0 ;I solution

,7-Dco.~~~-.5-O-r??eth~~/-l,-erythro-pc~nt~).s~~ 15 t SU’I~WW3 ,. (v’v) HCI in methanol

(140 mL) was added

After 6 min at 25 . the acidic solution

mmol).

Ambcrlite

IRA-68

?-deoxv-r~-c~l,ti?ro-pet~t~)sc was neutralized

. X0 g). The neutral

resin (OH

solution

with batchwire

at a flow rate of I mL/min

furanoside

eluted first (fractions

and assayed

33 43). followed

addition

U’X conct’ntrated

with phenol

sulfuric

by the /Fifuranoside

of

to a syrup

1 ~* 7 I 100 400 [ IO rnL,~ were

at 30’ in L’CI~IIO. and the tnixture ofglycosides was separated on ;I Dowermesh) (OH ) column, using distilled water as the ~‘lucnt“. Fractions collected

01‘0.5’,0

I2 (5.0 g. 37.3

acid’ ‘, The x

(fractions

46 58)

(yield, 94%). Either glycoside in a three-neck ping funnel.

13 (3.1 g. 19 mmol) was dissolved

flask equipped

andp-toluenesulfonyl

CH2Clz (31 mL) was added -5

with a drying

and (I- during

chloride

tube.

in anhydrous

thermometer.

(3.9 g, ?(I.6 mmol) dissolved

slowly at -- 5 . maintaining

the addition.

pyridine

The temperature

the reaction

was mainrained

(X0 mL)

nnd equalizing

drop-

in h.p.l.c.-grade mixture

betwreen

for I !I ;I[ 0

atic~

which time the solution was allowed to warm to room tcmpcrat urc. After stirring overnight. ice-cold water (200 mL) was added, the mixture was extracted with CIlc’II (3 x 40 mL). and the organic After evaporation. A solution

extracts

4.3 g (yield.

were pooled

and dried

iivt’r anhydr~ous

M$W,.

7 1 ‘1%)of 14 was obtained.

of 14 (8 2. 55.2 mmol)

in dry methanol

( 10 ml_) was m~xeci \vith 3.6~

NaOCH, (35 mL.. 161 mmol) in a Pyrex screw-capped (Teflon-lined) test tube. and the red reaction mixture was heated for 0.5 h in an oil bath at 100 Afta- cnoling. water (5 mL) was added to dissolve the precipitate, and the aqueous solution \&as cxtractcd with CHCil (7 x 30 mL). The organic extracts were pooled and dried over an hydrous I\lgSO, and evaporated

at 30 in PCI~uo (yield, 3.1 g. 84%).

THORPE-INGOLD

H,COH,C

27

EFFECTS

H,COH,C

o

y_$

SNNaOH.CS,.CH,I

0

ys 0

DMSO

Me

0

HO

o

H,CS(S)CO

%

Me

nBw$nH loiuene, retlux

a..

Me

0

ir

Me

16

17

H&OH+2

O OH

H,O’ 0

-

Me

w-

OH

-1;;

Me 18

19

Scheme 6

The glycoside was hydrolyzed with Dowex-50 x 8 (H’) (3 g, 20-50 mesh) ion-exchange resin for 1 h at 37” in CHCI,-H20 (15 mL) to generate 15 (yield, 50%, based on 12). 3-Deo?cy-S-O-methyl-D-erythro-penlusc 19 f Schelne 6j .-To a solution of 1.2-Oisopropylidene-5-~-methyl-~-o-xylofuranose (16) (5.4 g, 26.5 mmo1)‘9 in methyl sulfoxide (53 mL) was added 5M aqueous sodium hydroxide (9 mL, 39.8 mmol) and carbon disulfide (2.1 mL, 34.5 mmol). The resulting burgundy solution was stirred for 30 min, and methyl iodide (2 mL, 32.3 mmol) was added. The yellow solution was stirred for an additional 30 min and poured into 100 mL of H,O. The mixture was extracted with hexane (3 x 30 mL), and the organic extracts were pooled and dried over anhydrous MgSO,, giving 3.4 g of 17. To a solution of 17 (1.4 g, 5 mmol) under reflux in dry toluene (10 mL) was added via a dropping funnel, over a period of 30 min, a solution of tri-n-butyl stannane (2.08 g, 7.1 mmol) in dry toluene (71 mL). Refluxing was continued under N, for 60 h, and the reaction was monitored periodically by t.1.c. (silica gel; 32 ether-he~ane solvent: R, 1’7, 0.93; R, 18, 0.71). The toluene was removed if? WKUQto give a crude, white product. which was dissolved in 150 mL. of hot acetonitrile and extracted with hexane (3 x 50 mL) to remove tin-containing products The acetonitrile phase was evaporated to a syrup, giving 0.6 g (6~O~ yield) of 18, Hydrolysis of the isopropylidene ketal moiety in 18 to yield 19 was achieved by reAux in 0.1 % H,SO, (+ 60% yield based on 18). RESULTS

‘H- and ‘3C-N.m.r. chemical sh$is. - The ‘H chemical shifts of the deoxy- and alkyd-substituted furanoses in “HO at 25” are given in Table I. The r- and p-anomeric proton signals of5,7, 10, 11, and 19 were assigned via their correlation with anomeric

Spin coupling

II

constants”

” Values arc reported

5.9

5.6 4.2 to.5

4.0

5.1

0.5

4.0

2.4

5.5

4.1 4.8 5.0 1.9 4.0

4.4

1.8 4.2

12

5.1 5.7

1,2

I.1

5.7

7.1

2’,3

4.8

1.6

-13.9

- ~ 14.2

3.3’

6.6

4.4

4.5

9.3

3,4

in ‘HZ0 at 25”.

8.6

3.4’

9.5

3.1

3’,4

values which were not determined

2,3’

furanoses

denotes

6.5

3.7

> 1.9 5.9

to *O. 1 Hz. “Nd”

- L4.0

-14.2

5.2

- 14.4 - 5.9 -14.7 1.8

2,2’

2.3

and alkyl-substituted

Coupliny constant (Hz)

for deoxyfuranoses

in hertz (Hz) and are accurate

2-Deoxy-a-D-gl~crro-tetrose 2a 2-Deoxy-/j’-o-glycero-tetrose 2b 2-Deoxy-u-y!vcero-tetrose hydrate 2c 3-Deoxy-r*-DL-g!)lcero-tetrose 5a 3.Deoxy-B-Dr-yl~~ero-tetrose 5b 3-Deoxy-3,3-di-C-methyl-Dt_-&crro-tetrose 7a 3-Deoxy-3,3-di-C-methyl-m-g/yeera-tetrose 7b 3-C-Methyl-r-oL-erythrose 10a 3-C’-Methyl-/?-uL-erythrose lob 3-C-Methyl-a-ut-threose lla 3-C-Methyl-p-oL-threose lib 2-Deoxy-5.O-methyl-a-D-er).tAropcntose 15a 2-Deoxy-5-O-methyl-B-D-~r.~~~zr~pentose 15b 3-Deoxy-5-U-methyl-r-D-~r~f~r~~pentose 19a 3-Deoxy-5-O-methyl-P-u-er~thropentose 19b

Compound

‘H-‘H

TABLE

1.4

3’,4’

nd nd

- IO.0

-8.7

-8.3

5 -8.4

-11.8

4,4’

7.8

1.2

4.1

3.2

6.0

4.5’

3.2

4,5

- 10.8

- 10.9

-11.0

5.5

G

B

.!

2

III

chemical

2c

42.4 72.2 77. I

99.7 98.3 103.7 to +O.l

42.5

99.5

p.p.m.

34.4

33.7

72.X

72.5

79.9 80.9

83.6 7x.7

42.8

76.8 77.9

85.7

104.2

42.5

31.4 32.1 35.0

71.5 12.6 70.3

c-3

76.3 82.7

19.3

99.2

91.1 91.9 103.1 91.2 104.9 99.6 90.9

72.4 76.9 72.5

42.6 43.5 42.0

c-2

91.2 103.2 93.4

99.13 99.67 90.4

C-l

20.1 20.2

23.0 21.9

24.2

26.1

C-3’

20.6

obsh

C-3”

79.4

76.7

85.2

84.8

78.6 76.9

77.2 77.4

79.7

78.6

66.4 68.1 60.0

15.7 14.9 67.1

c-4

in ‘Hz0

are referenced

furanoses

Spectra

(p.p.m.)

and alkyl-substituted Chemicul Shft

’ Values are reported in p.p.m. and are accurate * “Obs” denotes obscured resonance.

3-Deoxy-cc-ILL-gl~~era-tetrose 5a 3-Deoxy-P-DL-g+ra-tetrose .5b 3-Deoxy-DL-g~~,Lrro-tetrose hydrate 5c 3-Deoxy-3,3-di-C-methyl-a-rrL-glycerotetrose 7a 3-Deoxy-3,3-di-C-methyl-P-r)L-g~~~er(~tetrose 7b 3-Deoxy-3,3-di-C-methyl-oL-g~~~~~u-tctrose hydrate 7c 3-C-Methyl-?-oL-erythrose 1Oa 3-C-Methyl-fi-w-erythrose 10b 3-C-Methyl-DL-erythrose hydrate 10~ 3-C-Methyl-a-DL-threose lla 3-C-Methyl-$or-threose Ilb 3-C-Methyl-or-threose hydrate llc 2.Deoxy-5-O-methyl-sc-D-er~thro-pentose I5a 2.Deoxy-S-O-methyl-8-o-er~rhro-pentose 15b 3-Deoxy-5-O-methyl-a-twrythro-pentose 19a 3-Deoxy-5-0-methyl-/Gn-erythro-pentose 19b

2a 2b hydrate

shifts” of deoxyfuranoses

2-Deoxy-r-D-g/lwro-tetrose 2-Deoxy-~-~-gl~~cfro-tetrose 2-Deoxy-wglyceru-tctrose

Compound

“C-N.m.r.

TABLE

externally

77.6

76.1

75.1

73.7

C-5

at 25”.

to the C-l signal of,&D-(1 -“C)glucopyranose

59.96

60.03

60.1

60.1

CH,

(97.4 p.p.m.) w

(-/.(‘-’

c-1

THORPE-INGOLD

aldehyde

EFFECTS

33 of 20 and 15 shows an 1 l-fold

at 30” and a 2.9-fold increase at 60“. Comparison

increase

in aldehyde

content

upon C-2 deoxygenation

increase

in aldehyde

content

may result, in part, from the more electrophilic

of aldopentofuranose

20 and 21 (compared

to that in 2 and 15) caused by the presence

at C-2 of the former.

Increasing

form in aqueous

solutions

temperature

of all compounds

has a smaller effect on the amount nyl] ratio consistently yl] (Table

studied

of hydrate

decreases

enhances

with increasing

of acyclic aldehyde

(Table V). Increasing however,

temperature

in

of an oxygen substituent

the amount

in solution;

rings. This carbonyl

temperature

the [hydrate]:[carbo-

due to increasing

[carbon-

V).

Increasing

alkyl substitution

of acyclic forms in solution. hydrate) decreases aldotetrofuranoses

of the aldofuranose

ring decreases

Thus, at 30” the percentage

in the series aldotetrofuranoses 21:22 (1O:ll (- 0.7%) --f 3-dimethyl-glq,cvro-tetroses

The quantity

of acyclic

hydrate

the total percent

of acyclic forms (aldehyde

seems to be more

11%)

+

--f 3-C-methyl-

7 (- 0.1%).

affected

by furanose

ring

alkylation than that of the acyclic aldehyde. For example, at 30” aqueous solutions of D-erythrose 21 contain 250-fold more acyclic hydrate than those of 7, but only 20-fold more acyclic aldehyde. the gem-diol

oxygens

The hydrate

of 7 contains

unfavorable

at C-l and the gem-dimethyl

1,3-interactions

between

groups at C-3; these interactions

are

absent in the acyclic aldehyde of 7. Thus, the [hydrate]:[carbonyl] ratio in aldofuranoses will depend, among other factors, on the state of alkylation of the ring. reduces

As previously noted14, alkyl substitution at C-4 of the furanose ring significantly the percentage of acyclic forms in solution. This effect may be observed by

comparing

the amounts

Ring-opening

of acyclic forms in solutions

und ring-closing

rate constants.

of the homologs -

2:15, and 20:21.

The ring-opening

(k,,,,,)

ring-closing (k,,,,,) rate constants for the deoxyfuranoses and alkyl-substituted es (p2H 5.0 in 50mM acetate buffer, 60”) are given in Table VI. In compounds an oxygen substituent at rates greater than,

at C-2 (5,7,10,11, or approximately

and O-2 trans. This behavior tetrofuranoses3

5-deoxy-

is consistent

and

furanoscontaining

and 22) anomers having 0- 1 and O-2 cis open equal to, corresponding anomers having O-l with that observed

and 5-O-methylpentose@,

previously

in the aldo-

and the D-pentuloses’4.

The 3-C-methyl-tetroses

10 and 11 are the deoxy analogs of the apiofuranoses configurations in both previously studied’. A comparison of kopen of corresponding and smallest for the series shows that kopcn is greatest for the /I-threo configuration ,4-erythro configuration,

with koprnof the a-rr.ythro and a- threo configurations

ate in magnitude. Thus, the conversion of a -CH20H apiofuranose ring to -CH, found in 3-C-methyl-tetroses

intermedi-

substituent at C-3 of the does not affect the relative

ring-opening reactivities of the four isomers. A comparison of kopen(Table VI) for 2 and 15 suggests that alkylation (in this case at C-4) reduces ring-opening reactivity. Likewise, a comparison of 21 and 7 leads to the same conclusion.

However,

and 11) and further

the effect is small and not uniform

(for example,

compare

22

studies will be required to examine this effect in more detail. kcloseis enhanced when the furanose ring is alkylated. For example,

In contrast, the alkylation of 2 to give

15 causes

a -3-fold

enhancement

in k,,,,,.

Likewise,

34

.I R. SNYDER.

ring-closing aldotetroses genation and,:or

reactivity is greater for the 3-C’-methyl-tetrc-,ses 21 and 22. The reduced ring-closing reactivity

‘(‘.{q., compare a reduction

Interestingly.

2 with 21 and 22) is probably

in the elrctrophilicity the ring-closing

of --C‘H, for

the state of substitution The enhanced cis ~vas previously either directly ring-opening. nevertheless

carbon.

reactiv-ities of the 3-C ‘-melhyJ-let roses

10 and 11 are

7; this i4 not unexpcctcd.

OF1 at (‘-3 of 10 and 11 constitutes

a relatively

as the

small change

in

of the ring. ring-opening

explained””

reactivity

hy invoking

or ricz intervening EJowevcr, although ring-opens slightly

of the anchimeric

validity

11 than for the b! ring deox>-

due to the loss ~~l‘a ring
of the carbonql

to those of the 3-deoxS-3..i-di-i’-methyltetroces

similar

substitution

10 and caused

4.S. SFRIANKI

of furanose anchimeric

water moiecules.

anotners assistance

the hydras\

having

O- 1 and O-2

by O-1 tn abstracting. i proton

on 0 i during

2 lacks a hydroxyl substituent at C -2. it> x anonter 2a faster than 2b. and thus ~mitrgfy ch;tllenpcs the

mechanism.

‘H

‘H Coupling

constanls

(Yable JJ) shout that.

with the ‘I;‘. in 2a. ‘J,, ,.,,_:.> :b ‘.fb,_,.,, ~‘,?. and is~,,.I,5 .,,_i ) iJl,_7_5 .,,_{. This pattern isconsistent El, “E, and E4 conformations delertnined to be preferred by Za in Ihe gas J:hasc t’rom 011 initio molecular orbital calcu!:ttions with the 3-2 1G basis set”. trr these cottformers. the dihedral angles made by JJ-?.S to H-l and F-J-3 arc 0 ~30 . while those made by i-J-Z/? to Ii-l and H-3 are X0 120 and 00 150 , rrspectivelq, and ihub larger c
for the former.

More importantly,

and O-3 quasi-axial. or nt‘ar quasi-axial. ing the hydroxyl proton on 0-i In contrast.

the 7-dcoxypent

the preferred

allowing

ofuranose

conformer:,

O-3 to substitute

anomers,

l)f 2a orienl

for 0-11 in

1% and 1Sb. rtng-open

0-I

abstract-

at stmilar

rates (Table VI). ‘H- ‘J-1 spin-couplings in 1Sa (Table II) diifsr from corresponding values in 2a and aggest a ~restercontribution of”north” conl‘ormcr~ ((,.!I.. ‘F,*)in u.hich O-3 is quasi-equatorial or near quasi-equl9torial and thus trnal7le 11 t p;trticipa:e in

35

THORPEEINGOLD EFFECTS TABLE VII Thermodynamic Compound

parameters for the ring-closing reactions of 21, 10, and 7 at 30” AG:”

u-Erythrose 21 - 1.6 3-C-Methyl-DL-erythrose 10 -2.6 3-Deoxy-3,3-di-C-methyl-uL-g~~~~r~tetrose 7 -2.8

AG/;”

AH:’

dH$

AS,”

4s;

-2.1 -3.0

-4.0 -4.3

-5.1 -5.8

-7.8 -5.8

-9.8 -92

-3.3

- 5.2

-6.2

-8.0

-9.6

” In kcal/mol. h In kcal/mol k 1.5.’ In caliK/mol k 3.

proton abstraction at O-l. Thus, this limited set of data suggests that furanose ring conformation may influence the relative ring-opening reactivities of some furanose anomers. A conformational effect may also be operating in 21 and 22, and 10, and 11. Within these pairs, the threo isomers show a greater diference in kopenbetween anomers than is shown between erythro isomers. This effect was observed previously in the 5-deoxy- and 5-O-methyl pentose@, where the smaller difference in erythro isomers was attributed to the lower reactivity of erythro isomers having O-l and O-2 cis. This reduced reactivity may be caused by intramolecular hydrogen bonding between the cis O-2 and O-3 which limits the ability of O-2 to participate in proton abstraction’. It is possible, however, that conformational properties of 0- 1, O-2-& erythro isomers do not allow, on average, O-l and O-2 to orient as well at those of O-l, O-2-& threo isomers. Ring alkylation and thermodynamic parameters. - Free-energies of conversion (AC”) for the ring closure reactions of 7, 10 and 21 at 30” are given in Table V!I. These values are expected to become more negative as ring alkylation increases, since the [cyclic]:[carbonyl] ratio increases with increased alkylation (see above). By determining equilibrium constants as a function of temperature, enthalpies of conversion (AH”) were determined, thus allowing AC” values to be divided into their two components, AH” and entropies of conversion, As” (Table VII). While the data set is limited and subject to inherently large errors due to the experimental methodology, these results suggest that changes in ring alkylation have only a small and random effect on AS” as ring alkylation increases. In contrast, a small but consistent decrease is observed in AH”. Thus, it appears that enthalpic factors are primarily responsible for the observed effect of ring alkylation on the free-energies of conversion for furanose ring-closure.

DISCUSSION

The anomerization reaction of reducing sugars in solution has been known for over a century’, yet the effects of carbohydrate structure on the kinetics of this

THORPE-INGOLD

EFFECTS

37

stituents may act to stabilize pseudocyclic conformations of the open-chain forms. Alkylation at various sites in the furanose ring does not have an equal effect on the kinetics of furanose ring-closing. For example, while 2 and 5 have the same degree of substitution, their k,,,,, values differ (Table VI). The most extreme comparison involves the hydroxymethylation of the anomeric carbon of aldotetrofuranoses, generating 2-ketopentofuranoses. This kind of alkylation results in a significant reduction ofk,,osF. The present study shows that alkylation at non-anomeric furanose ring carbons erzhar?ces ring-closure, but the magnitude of this enhancement will probably depend on the specific site involved. For example, while gem-dialkyl effects may be invoked to explain the enhanced k,,,,, upon substitution at C-4 of aldofuranoses, substitution at this site is also likely to affect ring-closing rates by altering the electronic character of O-4. The observed difference in the ring-opening behavior of the anomers of 2 and 15 points to the potential importance of furanose conformation and dynamics in governing kopen.From the limited data available, it appears that O-3 may play a role in catalyzing the opening of aldofuranoses rings under the proper conformational circumstances. This effect, which deserves further scrutiny, serves to emphasize that numerous, competing structural factors influence the kinetics of furanose anomerization. ACKNOWLEDGMENTS

This work was supported by grants from the National Institutes of Health (GM 33791) and the Research Corporation (10028). We thank Dr. Mistra of the NMR Facility for Biomedical Studies, Department of Chemistry, Carnegie Mellon University, Pittsburgh, PA (partially supported by NIH grant PH00292) for his assistance in obtaining the 620 MHz ‘H-n.m.r spectra. REFERENCES 1 (a) H. S. lsbell and W. Pigman, Adu. Carbohydr. Chem. Biochem., 24 (1969) 13-65. (b) W. Pigman and H. S. Isbell, Adv. Curhohydr. Chrm., 23 (1968) I l-55. 2 (a) W. P. Jencks, Cafalysis in Chemistry und Enzymology, McGraw-Hill. Inc.. 1969. 741. (b) M. L. Bender, R. J. Bergeron, and M. Komiyama, The Bioorganic Chemisrry qf’Enzyma~ic Catnlysis. John Wiley and Sons, New York, 1984, pp. 216247. 3 A. S. Serianni. J. Pierce, S.-G. Huang, and R. Barker, J. Am. Chrm. Sot., 104 (1982) 40374044. 4 J. R. Snyder and A. S. Serianni, J. Org. Chem., 51 (1986) 2694-2702. 5 J. R. Snyder and A. S. Serianni, Curbohydr. Res., 166 (1987) 85~ 99. 6 J. R. Snyder and A. S. Serianni, Curhohydr. Res., 184 (1988) 13-25. 7 J. R. Snyder and A. S. Serianni, J. Am. Chem. Socc, I I I (1989) 2681-2687. 8 J. Pierce, A. S. Serianni and R. Barker, J. Am. (‘hem. SIC., 107 (1985) 2448--2456. 9 (a) G. A. Morris and L. D. Hall, J. Am. Chem. SM., 103 (1981) 4703-4711. (b) A. B. Blakeney, P. J. Harris, R. H. Henry, and B. A. Stone, Carhohydr. Res., I 13 (1983) 291-299. 10 Z. J. Witczak and R. L. Whistler, Carbohydr. Res., 110 (1982) 326-329. I1 J. K. N. Jones and R. A. Wall, Can. J. Chem., 38 (1960) 2290-2294. 12 J. E. Hodge and B. T. Hofreiter, Methods Carbohydr. Chem., 1 (1962) 38@394. 13 A. R. Gallop0 and W. W. Cleland, Arch. Biochem. Biophys., 195 (1979) 152- 154. 14 A. S. Serianni and R. Barker, J. Org. Chem., 45 (1980) 3329-3341. 15 (a) A. S. Serianni, E. L. Clark. and R. Barker, Carhnhydr. Res.. 72 (1979) 79-91. (b) A. S. Serianni and R. Barker, in E. Buncel and J. Jones (Eds.), 1sorope.v in lhe Physical and Biomedical Sciences, Elsevier, New York, 1987, pp. 21 l-236.

38

.I.K.

SNk’DliR.

A.S. SliRlAUNl