Stereocontrolled organocatalytic synthesis of prostaglandin PGF 2α in seven steps

Article abstract of DOI:10.1038/nature11411

Prostaglandins are hormone-like chemical messengers that regulate a broad range of physiological activities, including blood circulation, digestion and reproduction. Their biological activities and their complex molecular architectures have made prostaglandins popular targets for synthetic organic chemists for over 40 years. Prostaglandin analogues are widely used as pharmaceuticals and some, such as latanoprost, which is used to treat glaucoma, have become billion-dollar drugs. Previously reported syntheses of these compounds are quite lengthy, and every chemical step costs time and energy, generates waste and is accompanied by material losses. Using a new bond disconnection, here we report a concise synthesis of the most complex prostaglandin, PGF _2α, with high levels of control of relative and absolute stereochemistry, and fewer steps. The key step is an aldol cascade reaction of succinaldehyde using proline organocatalysis to create a bicyclic enal in one step and an enantiomeric excess of 98%. This intermediate bicyclic enal is fully primed with the appropriate functionality for attachment of the remaining groups. Access to this bicyclic enal will not only render existing prostaglandin-based drugs more affordable, but will also facilitate the rapid exploration of related chemical structures around the ubiquitous five-membered ring motif, such as potentially therapeutic prostaglandin analogues.

Full text of DOI:10.1038/nature11411

LETTER
doi:10.1038/nature11411
Stereocontrolled organocatalytic synthesis of
prostaglandin PGF2a in seven steps
Graeme Coulthard1, William Erb1 & Varinder K. Aggarwal1
Prostaglandins are hormone-like chemical messengers that prostaglandins was uncovered by Bergstro¨m et al. (see refs 2 and 10
regulate a broad range of physiological activities, including blood for reviews). The complex structures of prostaglandins, together with
circulation, digestion and reproduction1,2. Their biological their broad spectrum of biological activity, fuelled intense research
activities and their complex molecular architectures have made activity into their synthesis, comparable to that generated from
prostaglandins popular targets for synthetic organic chemists for b-lactam antibiotics and steroids. Woodward11, Corey12, Stork13,
over 40 years3,4. Prostaglandin analogues are widely used as Noyori14, Danishefsky15 and many others contributed ingenious
pharmaceuticals and some, such as latanoprost, which is used to strategies and developed new methodologies of general use in the con-
treat glaucoma5,6, have become billion-dollar drugs. Previously struction of these complex molecules3–5. But prostaglandins and their
reported syntheses of these compounds are quite lengthy, analogues are not only of academic interest, they have also found their
and every chemical step costs time and energy, generates waste way into a considerable number of pharmaceuticals. For example, an
and is accompanied by material losses. Using a new bond discon- analogue of PGF2a (1), latanoprost (2), is used in the treatment of
nection, here we report a concise synthesis of the most complex glaucoma16, and achieved sales of $1.75 billion in 2010 (ref. 17).
The manufacture of latanoprost6 requires 20 steps and uses the
prostaglandin, PGF2a, with high levels of control of relative and
absolute stereochemistry, and fewer steps. The key step is an aldol
original strategy developed by Corey et al.18 in the synthesis of the
cascade reaction of succinaldehyde using proline organocatalysis related prostaglandin, PGF2a. Corey’s synthesis involved the forma-
to create a bicyclic enal in one step and an enantiomeric excess of tion of a key intermediate, the Corey lactone (3), in nine steps from
98%. This intermediate bicyclic enal is fully primed with the appro- cyclopentadiene (4). From this lactone, they were able to assemble
priate functionality for attachment of the remaining groups7. the entire family of prostaglandins12, and specifically PGF2a, in eight
Access to this bicyclic enal will not only render existing further steps18.
prostaglandin-based drugs more affordable, but will also facilitate
the rapid exploration of related chemical structures around
the ubiquitous five-membered ring motif, such as potentially
therapeutic prostaglandin analogues.
In our analysis for the synthesis of prostaglandins and in common
with other syntheses of PGF2a,13,15,19 we recognized that lactol 5 was an
ideal late-stage intermediate because it enabled the incorporation of
the upper side-chain with control of double-bond geometry through a
Prostaglandins, such as PGF2a (Fig. 1; 1), are hormones that are Wittigreaction. Atthis point ourretrosynthetic analysis departed from
responsible for the control of a myriad of essential biological processes all previous syntheses. We were interested in disconnecting the C12–
from sleep to pain, fever to inflammation, menstruation to birth, and C13 bond20, because we recognized that the bicyclic lactol could
constriction of blood vessels to blood clotting1,2,8. They are derived from control the stereochemistry of the 1,4-addition of the lower side-chain
arachidonic acid and are transformed by prostaglandin synthetase into onto a suitable Michael acceptor 6. By selecting an aldehyde as the
a number of structurally related carbocyclic molecules. These sensitive electron withdrawing group of the Michael acceptor, the enal 7 could
and labile molecules are not stored in the body but are synthesized in be disconnected back to a simple aldol dimerization of succinaldehyde
response to stimuli. They were discovered in the early 1930s by von (8), a process that could be rendered asymmetric through proline (9)
Euler9 and by the mid-1960s the structures of the first family of catalysis21,22
.
O
a
b
HO
HO
HO
HO
O
Nine steps
Eight steps
CO2i-Pr
CO2H
Ph
OH
4
OH
Latanoprost (2)
OH
AcO
Corey lactone (3)
PGF2α (1)
c
O
OH
OP
OH
Wittig
Aldol
reaction
HO
O
1
9
O
O
O
Conjugate
addition
CO2H
6
5
=
12
11
20
15
OH
O
HO
13
EWG
HO
OH
O
O
FGI
PGF2α (1)
5
6
7
8
Figure 1
|
Prostaglandins in nature and medicine. a, General schematic of
analysis of PGF2a. The numbered carbon atoms are referred to in the main text.
FGI, functional groupinterconversion. Ph, phenyl. EWG, electronwithdrawing
group. i-Pr, isopropyl, CH(CH3)2.
Corey’s synthesis of PGF2a. b, The molecular structure of latanoprost (2), a
PGF2a analogue used for the treatment of glaucoma. c, Our retrosynthetic
1School of Chemistry, University of Bristol, Cantock’s Close, Bristol BS8 1TS, UK.
2 7 8
| N A T U R E | V O L 4 8 9 | 1 3 S E P T E M B E R 2 0 1 2
Macmillan Publishers Limited. All rights reserved
©2012

LETTER RESEARCH
Further
aldol
reactions
Further
aldol
reactions
Perhaps unsurprisingly, the desired aldol proved highly challenging
to realize in practice: treatment of succinaldehyde with proline in a
range of solvents and under a variety of conditions did not deliver any
of the desired bicyclic product but instead gave oligomeric material.
We therefore deconstructed the reaction cascade to determine
which of the two steps, the initial aldol step or the second aldol and
dehydration step, was causing the problem.
Model aldehyde 16 was treated with proline (9) and aldol product
17 was obtained as a 3.6:1 mixture of diastereoisomers in moderate
yield (Fig. 3a)23. This showed that aldol reactions of aldehydes bearing
a carbonyl group in the 4-position were suitable substrates for the
proline-catalysed aldol reaction.
To test the second aldol and dehydration step, model dialdehyde 18
was prepared by ozonolysis of the known lactone 19 (Fig. 3b)24.
However, treatment of this dialdehyde with proline (catalyst A)
provided only low conversion to the expected enal 20, clearly indi-
cating that the second step was the hurdle in the reaction cascade. We
therefore explored alternative catalysts and found that [Bn2NH2][OCOCF3]
(21, catalyst B)25 was much more effective, giving enal 20 in 51%
isolated yield (from 19).
O
HO
O
O
Oligomer
O
O
13
O
12
–H2O
(ii)
(i)
O
OH
OH
9
Intramolecular
aldol
reaction
O
O
OH
O
CO2H
N
H
O
O
Dehydration
O
O
8
O
O
7
11
10
(iii)
(iii)
8 + 9
10 + 9
O
O
O
OH
O
OH
Further
Further
aldol
reactions
O
O
aldol
reactions
O
HO
Oligomer
O
O
OH
OH
15
14
We therefore tested the aldol reaction cascade of succinaldehyde
with a combination of proline and [Bn2NH2][OCOCF3] in a range of
solvents and under a variety of conditions (see the Supplementary
Information for full details). This time the reaction was successful;
selected data, illustrative of the process, is summarized in Table 1.
The reaction could be conducted at (unusually) low loadings of both
catalysts and at relatively high concentration (2 M) (entries 1–3 in
Table 1). The sequenced addition of the two catalysts and the timing
was critical to the success of the reaction: if [Bn2NH2][OCOCF3] was
present at the outset (entry 4 in Table 1), or not added at all (see earlier
discussion), the reaction failed. We spent some time investigating this
facet of the reaction because the low yield is a consequence of
oligomerization of the intermediate trialdehyde. As the concentration
of the trialdehyde builds up it is prone to undergo further aldol
reactions with succinaldehyde, leading to oligomers. The second
catalyst seemed to inhibit the initial proline-catalysed aldol reaction,
so consecutive addition of the two catalysts was required. The
optimum time to add the second catalyst was found to be dependent
on the amount of proline catalyst used: using just 2% proline, the yield
of the lactol 7 increased with increasing time before the addition of the
second catalyst (entries 5–8 in Table 1), peaking at around 10 h. The
reaction could also be conducted with just 1% proline but considerably
longer reaction times were required (entry 9 in Table 1).
Figure 2
|
Potential reaction pathways of the proline-catalysed aldol
reaction of succinaldehyde. The desired pathway is shown in blue; others are
labelled (i) to (iii) and described in the main text.
However, the ‘simple’ concept shown in the retrosynthesis belies a
highly complex reaction cascade (blue, Fig. 2) with many potential
pitfalls (black, Fig. 2): (i) the aldol product 10 is required to form
the less favoured hemiacetal 11, bearing cis substituents in the
5-membered ring, but not the trans hemiacetal 12; (ii) the hemiacetal
11 is required to undergoan intramolecularsecondaldol and eliminate
to give 7 but aldol 10 itself should not eliminate to give 13; (iii) aldol 10
is a reactive trialdehyde which will be prone to undergo further aldol
reactions with succinaldehyde (8) or with itself, leading to 14 and 15
and ultimately oligomers.
(10 mol.%)
a
CO2H
N
O
H
O
OH
THF, room temperature
O
O
61%, 3.6:1
diastereomeric
ratio
O
O
O
16
O
17
Table 1
|
Effect of catalyst loading and time delay on yield
b
O3
O
Catalyst (20 mol.%)
THF, room temperature
14 h
MeOH
–78° C
Me2S
O
O
CO2H
OH
N
O
(i)
O
O
H
O
O
THF (2 M), room temperature
O
Catalyst A: (S)-proline: 5%
Catalyst B: [Bn2NH2][OCOCF3]
(21): 51%
O
(ii) [Bn2NH2][OCOCF3]
THF (1 M), room
8
O
O
19
temperature,14 h
O
18
(not purified)
20
7
Entry
(S)-proline (mol.%)
Time (h)
Yield (%)
c
O
1
2
3
4
5
6
7
8
9
10
5
5
2
2
2
2
2
1
2
2
4
0
4
14
16
19
,2
10
16
20
20
18
O
OH
O
N
O
H
=
O
O
H
OH
O
O
O
O
22
10
O
6
10
10
24
24
Figure 3
|
Model studies. a, Model studies to investigate the intermolecular
proline-catalysed aldol reaction of aldehydes bearing an ester at the 4-position.
b, Model studies to investigate an intramolecular aldol reaction and
dehydration. c, A transition state structure accounting for the observed
enantioselectivity.
Reactions were carried out on 200 mg of succinaldehyde 8. ‘Time’ refers to the time before
[Bn2NH2][OCOCF3] (2 mol.%) was added and the reaction was diluted to 1 M. The yield is the NMR
(nuclear magnetic resonance) yield based on an internal standard (1,3,5-trimethoxybenzene); results
are an average of five reactions.
1 3 S E P T E M B E R 2 0 1 2
| V O L 4 8 9 | N A T U R E | 2 7 9
Macmillan Publishers Limited. All rights reserved
©2012

RESEARCH LETTER
OH
O
(iii) MeOH
amberlyst 15
MgSO4
O
O
(i) H2O
69%
(ii) (S)-proline
O
O
O
O
O
then
14% (two steps)
99:1 enantiomeric
ratio
8
23
[Bn2NH2][OCOCF3]
7
24
O
O
15.0 g
prepared
S
(iv)
25
O
(v) O3
then
NaBH4
O
O
Li2(CN)Cu
O
OTBDMS
Me3SiCl / NEt3
HO
OTBDMS
OTBDMS
27
OSiMe3
60% (two steps)—1.4 g
49% (two steps)—4.6 g
26
OH
(vii)
HO
CO2H
O
BrPh3P
(vi) HClaq. / THF
29
CO2H
57% (two steps)—152 mg
47% (two steps)—1.9 g
HO
OH
PGF2α (1)
HO
OH
28
Figure 4
|
A concise asymmetric synthesis of PGF2a (1). Reaction conditions (iv), 25 (1.1 equivalents), THF, then Me3SiCl, Et3N. (v), O3, CH2Cl2/MeOH
are as follows (see the Supplementary Information for full details). (i), H2O, at (3:1), 278 uC, then NaBH4 (3 equivalents), at 278 uC to room temperature.
75 uC for 4 h, then at 115uC to distil MeOH and H2O. (ii), (S)-proline (vi), 1.5% aqueous HCl/THF (3:2), at room temperature for 16 h. (vii), (4-
(2 mol.%), THF(2 M), at roomtemperature for20 h, then [Bn2NH2][OCOCF3] carboxybutyl)(triphenyl)phosphonium bromide (6 equivalents), potassium
(2 mol.%), THF (1 M), at room temperature for 14 h. (iii), MeOH tert-butoxide (12 equivalents), THF, at 0 uC to room temperature. TBDMS,
(2.0 equivalents), amberlyst 15, MgSO4, CH2Cl2, at room temperature for 14 h. tert-butyldimethylsilyl.
The reaction was conducted at relatively high concentration, which acetal and silyl ether withaqueous HCl followed byWittig reactionwith
greatly assisted the scale-up of the reaction. Despite the low yield,
work-up and purification of the enal was straightforward as the
oligomeric side products could be largelyremoved byfiltration, leaving
a relatively pure crude material. Partial purification through a plug of
silicagavethedesired enal 7inabout 16% yieldandwith anenantiomeric
ratio of 99:1, on the multigram scale. The absolute stereochemistry of
the product was ultimately established through the synthesis of PGF2a
(1) and follows from the List–Houk model for this type of reaction
(transition-state structure 22, Fig. 3c)26. A mixture of diastereoisomers
of trialdehyde 10 was expected to be formed (a diastereoisomeric
ratio of about 3.6:1 was expected from results with the model com-
pound 16) butas the minor diastereoisomer cannot give the alternative
diastereomericenalitmustbeconsumed bytheformationofoligomers.
The complete synthesis of PGF2a from commercially available
materials is shown in Fig. 4. Heating 2,5-dimethoxytetrahydrofuran
23 (41 g) in water followed by evaporation and extraction gave crude
dialdehyde27 which was directly subjected to the aldol reaction as
described above. The hemi-acetal 7 was converted into a 2:1 incon-
sequential diastereoisomeric mixture of methoxy acetals 24 which
were carried through the subsequent reaction sequence. On a large
scale (285 g of 23), further modifications were required owing to the
partial decomposition of the dialdehyde during the extended time
needed to remove the larger volume of solvent used for extraction.
the phosphonium salt 29 (ref. 29) gave 1.9 g of the target molecule
PGF2a (1), which was identical in all respects to the natural product30.
Thus, we have developed a short (seven steps) synthesis of
prostaglandin PGF2a from inexpensive 2,5-dimethoxytetrahydrofuran
23. The key step is an organocatalytic aldol dimerization reaction
of succinaldehyde, which generates the bicyclic enal 7 in high
enantiomeric ratio and fully primed with functionality suitable to
introduce the required side chains directly. The aldol cascade required
proline to perform the first aldol reaction and a second catalyst
([Bn2NH2][OCOCF3]) to induce an intramolecular aldol reaction
and elimination. Although the aldol reaction was low-yielding, the
enantioselectivity was very high, isolation and purification was
straightforward and the reaction could be conducted on the multi-
gram scale. Its application in a short synthesis of the most complex
of prostaglandins, PGF2a, has been demonstrated. Indeed, the bicyclic
enal 7 is an ideal building block not just for the cost-effective synthesis
of the whole family of prostaglandins, but for also exploring the
chemical space around the ubiquitous five-membered carbocyclic ring
motif, where other biologically active molecules undoubtedly lie.
METHODS SUMMARY
Procedure for formation of methoxyacetal 24. A solution of succinaldehyde 8
(109.5 g, 1.27 mol) in 2-MeTHF (650 ml) was stirred at room temperature (20 uC)
Instead, we found that after distillation of the methanol (MeOH) and (S)-proline (2.93 g, 25.4 mmol, 0.02 equivalents) was added and the reaction
stirred at room temperature for 20 h. THF (650 ml) was added, followed by
[Bn2NH2][OCOCF3] (8.43 g, 25.4 mmol, 0.02 equivalents). The reaction was
stirred for a further 14 h. Celite (60g) was added to the reaction and the volume
ofthereaction mixture wasreduced by threequartersunder reduced pressure. tert-
Butyl methyl ether (TBME) (975 ml) was added slowly and with vigorous stirring
of the mixture. The mixture was stirred for 20 min before filtration of the resulting
solids. The solids were washed with TBME (33 150ml) and the filtrate was
concentrated under reduced pressure. The material was purified by column chro-
matography (about 600g silica), eluting with petrol/EtOAc (6:4 to 5:5), to give the
lactol 7 (as an approximately 2:1 mixture of diastereoisomers), as a brown oil.
This residue, containing 7, was dissolved in CH2Cl2 (190 ml) and stirred at
room temperature. MeOH (5.47 g, 6.90 ml, 170.6 mmol, 2.0 equivalents based
most of the water, azeotropic removal of the remaining water using
2-MeTHF (2-methyltetrahydrofuran) provided a solution of the dia-
ldehyde in a solvent (2-MeTHF) that could be used directly in the aldol
reaction. Using this improved and simplified protocol, from 109.5 g of
inexpensive succinaldehyde we were able to obtain 15.0 g of pure
methoxy acetal 24 in 14% yield over two steps. Conjugate addition of
the mixed vinyl cuprate28 25 to 4.0g of methoxy acetal 24 followed by
trapping with TMSCl (trimethylsilyl chloride) furnished the silyl enol
ether 26. Subsequent controlled ozonolysis followed by treatment with
NaBH4 gave the alcohol 27 (4.6g, 49% over two steps). As planned,
these two steps occurred with complete stereocontrol at the newly
created stereogenic centres. Finally, simultaneous deprotection of the on about 13.4% of 7 detected by internal standard in the previous reaction) was
2 8 0
| N A T U R E | V O L 4 8 9 | 1 3 S E P T E M B E R 2 0 1 2
Macmillan Publishers Limited. All rights reserved
©2012

LETTER RESEARCH
18. Corey, E. J., Weinshenker, N. M., Schaaf, T. K. & Huber, W. Stereo-controlled
synthesis of dl-prostaglandins F2a and E2. J. Am. Chem. Soc. 91, 5675–5677
(1969).
19. Howard, C. C. et al. Total synthesis of prostaglandin-F2a involving stereocontrolled
and photo-induced reactions of bicyclo[3.2.0]heptanones. J. Chem. Soc. Perkin
Trans. I 852–857 (1980).
20. Corey, E. J., Nicolaou, K. C. & Beames, D. J. A shortsyntheticroute toprostaglandins
utilizing position-selective epoxide opening by thevinyl Gilman reagent. Tetrahedr.
Lett. 15, 2439–2440 (1974).
21. Mukherjee, S., Yang, J. W., Hoffmann, S. & List, B. Asymmetric enamine catalysis.
Chem. Rev. 107, 5471–5569 (2007).
added via syringe. Amberlyst 15 (1.22 g) and MgSO4 (25.2 g) were added in one
portion and the reaction mixture was stirred at room temperature for 14 h. The
reaction mixture was filtered through a sinter funnel and the solids washed with
CH2Cl2 (33 60 ml). The filtrate was concentrated under reduced pressure and
purified by column chromatography (about 300g silica), eluting with petrol/
EtOAc (9:1 to 4:1), to give the methyl acetal 24 (as an approximately 2:1 mixture
of diastereoisomers, 15.0 g, 14.0% (over two steps from succinaldehyde)) as a
yellow oil. See the Supplementary Information for characterization data.
Received 9 February; accepted 11 July 2012.
Published online 15 August 2012.
22. List, B., Lerner, R. A. & Barbas, C. F. Proline-catalyzed direct asymmetric aldol
reactions. J. Am. Chem. Soc. 122, 2395–2396 (2000).
23. Northrup, A. B. & MacMillan, D. W. C. The first direct and enantioselective cross-
aldol reaction of aldehydes. J. Am. Chem. Soc. 124, 6798–6799 (2002).
24. Corey, E. J. & Ravindranathan, T. A simple route to a key intermediate for the
synthesis of 11-desoxyprostaglandins. Tetrahedr. Lett. 12, 4753–4755 (1971).
25. Corey, E. J. et al. Stereospecific total synthesis of gibberellic acid. A key tricyclic
intermediate. J. Am. Chem. Soc. 100, 8031–8034 (1978).
26. Bahmanyar, S., Houk, K. N., Martin, H. J. & List, B. Quantum mechanical predictions
of the stereoselectivities of proline-catalyzed asymmetric intermolecular aldol
reactions. J. Am. Chem. Soc. 125, 2475–2479 (2003).
27. Gourlay, B. S., Molesworth, P. P., Ryan, J. H. & Smith, J. A. A new and high yielding
synthesis of unstable pyrroles via a modified Clauson-Kaas reaction. Tetrahedr.
Lett. 47, 799–801 (2006).
28. Lipshutz, B. H., Kozlowski, J. A., Parker, D. A., Nguyen, S. L. & McCarthy, K. E. More
highly mixed, higher order cyanocuprates ‘‘RT(2-thieny)Cu(CN)Li2’’. Efficient
reagents which promote selective ligand transfer. J. Organomet. Chem. 285,
437–447 (1985).
1. Funk, C. D. Prostaglandins and leukotrienes: advances in eicosanoid biology.
Science 294, 1871–1875 (2001).
2. Bindra, J. S. & Bindra, R. Prostaglandin Synthesis (Academic Press, 1977).
3. Bindra, J. S. in The Total Synthesis of Natural Products (ed. ApSimon, J.) Vol. 4
353–449 (Wiley, 1981).
4. Das, S., Chandrasekhar, S., Yadav, J. S. & Gree, R. Recent developments in the
synthesis of prostaglandins and analogues. Chem. Rev. 107, 3286–3337 (2007).
5. Collins, P. W. & Djuric, S. W. Synthesis of therapeutically useful prostaglandin and
prostacyclin analogs. Chem. Rev. 93, 1533–1564 (1993).
6. Nair, S. K. &Henegar, K. E. inModernDrugSynthesis(edsLi, J. J.& Johnson, D. S.) Ch.
21 329–338 (Wiley, 2010).
7. Ungrin, M. D. et al. Key structural features of prostaglandin E2 and prostanoid
analogs involved in binding and activation of the human EP1 prostanoid receptor.
Mol. Pharmacol. 59, 1446–1456 (2001).
8. Stix, G. Betterwaysto targetpain. Sci. Am. 68–71, http://www.scientificamerican.com/
article.cfm?id5better-ways-to-target-pain (16 December 2006).
9. Von Euler, U. S. Information on the pharmacological effect of natural secretions
and extracts from male accessory sexual glands. Arch. Exp. Pathol. Pharm. 175,
78–84 (1934).
10. Flower, R. J. Prostaglandins, bioassay and inflammation. Br. J. Pharmacol. 147,
S182–S192 (2006).
11. Woodward, R. B. et al. Novel synthesis of prostaglandin F2a. J. Am. Chem. Soc. 95,
6853–6855 (1973).
12. Corey, E. J. & Cheng, X.-M. The Logic of Chemical Synthesis (Wiley, 1995).
13. Stork, G., Sher, P. M. & Chen, H. L. Radical cyclization-trapping in the synthesis of
natural products. A simple, stereocontrolled route to prostaglandin F2a. J. Am.
Chem. Soc. 108, 6384–6385 (1986).
14. Suzuki, M., Yanagisawa, A. & Noyori, R. Prostaglandin synthesis. 16. The three-
component coupling synthesis of prostaglandins. J. Am. Chem. Soc. 110,
4718–4726 (1988).
15. Danishefsky, S. J., PazCabal, M. & Chow, K. Novel stereospecific silyl group transfer
reactions: practical routes to the prostaglandins. J. Am. Chem. Soc. 111,
3456–3457 (1989).
16. Resul, B. et al. Phenyl-substituted prostaglandins: potent and selective
antiglaucoma agents. J. Med. Chem. 36, 243–248 (1993).
29. de los Angeles Rey, M. et al. New synthetic strategies to vitamin D analogues
modified at the side chain and D ring. Synthesis of 1a,25-dihydroxy-16-ene-
vitamin D3 and C-20 analogues. J. Org. Chem. 64, 3196–3206 (1999).
30. Sheddan, N. A. & Mulzer, J. Cross metathesis as a general strategy for the
synthesis of prostacyclin and prostaglandin analogues. Org. Lett. 8, 3101–3104
(2006).
Supplementary Information is available in the online version of the paper.
Acknowledgements We thank EPSRC and the European Research Council (FP7/
2007-2013, ERC grant no. 246785) for financial support. V.K.A. thanks the Royal
Society for a Wolfson Research Merit Award and EPSRC for a Senior Research
Fellowship.
Author Contributions G.C. and W.E. were involved in the discovery and subsequent
development of the aldol reaction and G.C. applied it to PGF2a. V.K.A. conceived and
directed the investigations and composed the manuscript with revisions provided by
G.C. and W.E.
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare no competing financial interests.
Readers are welcome to comment on the online version of the paper. Correspondence
and requests for materials should be addressed to V.K.A. (v.aggarwal@bristol.ac.uk).
17. Pfizer reports fourth-quarter and full-year 2010 results; provides 2011 financial
guidance and updates 2012 financial targets. http://www.pfizer.com/files/
investors/presentations/q4performance_020111.pdf (2011).
1 3 S E P T E M B E R 2 0 1 2
| V O L 4 8 9 | N A T U R E | 2 8 1
Macmillan Publishers Limited. All rights reserved
©2012