Propylphosphonic anhydride (T3P) mediated synthesis of 3-oxoisoindoline-1-carboxamides from 2-formylbenzoic acid, amines, and isocyanides. Preparation of isoindolinone alkaloids

Article abstract of DOI:10.1016/j.tetlet.2018.08.063

Propylphosphonic anhydride (T3P) was successfully applied to the synthesis of an isoindolinone library by the utilization of an Ugi four-center, three-component reaction (Ugi-4C-3CR). The use of T3P significantly shortened the required reaction time and t

Full text of DOI:10.1016/j.tetlet.2018.08.063

Accepted Manuscript
Propylphosphonic anhydride (T3P® ) mediated synthesis of 3-oxoisoindoline-1-
carboxamides from 2-formylbenzoic acid, amines, and isocyanides. Preparation
of isoindolinone alkaloids
Valentina Varga, Mátyás Milen, Péter Ábrányi-Balogh
PII:
S0040-4039(18)31070-0
https://doi.org/10.1016/j.tetlet.2018.08.063
TETL 50238
DOI:
Reference:
Tetrahedron Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
10 July 2018
24 August 2018
29 August 2018
Please cite this article as: Varga, V., Milen, M., Ábrányi-Balogh, P., Propylphosphonic anhydride (T3P® ) mediated
synthesis of 3-oxoisoindoline-1-carboxamides from 2-formylbenzoic acid, amines, and isocyanides. Preparation of
isoindolinone alkaloids, Tetrahedron Letters (2018), doi: https://doi.org/10.1016/j.tetlet.2018.08.063
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Propylphosphonic anhydride (T3P®) mediated synthesis of 3-oxoisoindoline-1-
carboxamides from 2-formylbenzoic acid, amines, and isocyanides. Preparation of
isoindolinone alkaloids
Valentina Varga,1 Mátyás Milen,2 Péter Ábrányi-Balogh1,*
1Hungarian Academy of Sciences, Research Centre for Natural Sciences, 1519 Budapest,
POB 286, Hungary
2Egis Pharmaceuticals Plc., Directorate of Drug Substance Development, 1475 Budapest,
P.O.B. 100, Hungary
*To whom correspondence should be addressed. Tel.: +36 1 3826961; e-mail: abranyi-
balogh.peter@ttk.mta.hu
Graphical abstract
Abstract
Propylphosphonic anhydride (T3P®) was successfully applied to the synthesis of an
isoindolinone library by the utilization of an Ugi four-center, three-component reaction (Ugi-
4C-3CR). The use of T3P® significantly shortened the required reaction time and the
corresponding products were obtained in good to high yields. Moreover, a side-reaction was
observed when phenylethylamine derivatives and tryptamine were used as the amine
component. The latter reaction was applied to the microwave-assisted, one-pot synthesis of
the isoquinoline alkaloid (±)-nuevamine. Surprisingly, the traditional Ugi four-component
reaction (Ugi-4CR) was unsuccessful in the presence of T3P®. In this case an α-amino amide
was produced excluding the carboxylic acid from the multicomponent reaction.
Keywords: isoindolinones, Ugi-reaction, T3P®, natural product synthesis

Introduction
Isocyanide-based multicomponent reactions, especially the Ugi four-component reaction
(Ugi-4CR), have received significant attention in synthetic organic and medicinal
chemistry.1,2 The Ugi-4CR offers an efficient and atom-economical way to construct
differently functionalized molecules from readily available aldehydes, amines, carboxylic
acids, and isocyanides.3 This reaction usually takes place under mild conditions with wide
functional group tolerance, but mostly with long reaction times. Multiple variations of the
Ugi-reaction, such as the Ugi-4CR, several three-component Ugi reactions (Ugi-3CR), Ugi-
tetrazole synthesis, oxidative Ugi reaction, and Ugi-azide repetitive processes have been
applied in the generation of diverse heterocyclic libraries.4–7
Oxoisoindole (2,6-dihydro-2H-isoindole-1-one, 1, Fig. 1) is a prevalent and important
heterocyclic motif which has been widely used as a valuable building block in the
pharmaceutical industry and in the synthesis of natural products.8 For example, lenalidomide
(2, trade name: Revlimid®) is a commercial drug for the treatment of multiple myeloma,
myelodysplastic syndrome, and lymphoma.9,10 (±)-Lennoxamine (3) and (±)-nuevamine (4)
are two representative isoindolinone alkaloids which were isolated from the Chilean barberry,
Berberis darwinii.11
Figure 1. Structures of oxoisoindole (1) and selected oxoisoindole derivatives
Due to widespread interest in the isoindolinone ring system, several approaches have been
developed for the synthesis of these heterocycles, but mostly utilise multistep protocols,
expensive metal catalysts, halogenated solvents, or harsh reaction conditions.12–15 The
1

isoindolinone scaffold can be constructed using an Ugi four-center, three-component reaction
(Ugi-4C-3CR) from 2-formylbenzoic acid (5a), amines (6) and isocyanides (7) (Scheme 1).16
This Ugi-4C-3CR proceeds smoothly in methanol or ethanol at room temperature and leads to
the products in good yields. The only drawback of this process is the prolonged reaction time
(12–24 h).16–18 As part of our continuing interest in the development of propylphosphonic
anhydride (T3P®) promoted reactions,19–22 herein, we report a T3P®-assisted synthesis of 3-
oxoisoindoline-1-carboxamides applying an Ugi-4C-3CR.
Scheme 1. Ugi-4C-3CR reaction for the synthesis of oxoisoindole derivatives
Results and Discussion
The T3P®-mediated model reaction of 2-formylbenzoic acid (5a), aniline (6a), and tert-butyl
isocyanide (7a) (1 equiv. of each) was optimized by varying different experimental
parameters. Initially, the effects of the reaction time and addition order of the components
were investigated. It was found, that after adding all reactants, 15 min is required for the
optimal yield (Table 1, entry 2). Increasing the reaction time did not improve the conversion
(Table 1, entries 3–5). When tert-butyl isocyanide (7a) or T3P® was added 10 min after the
rest of the reactants, the yields slightly decreased to 68% and 67%, respectively (Table 1,
entries 6, 7).
Table 1. Optimization of the Ugi-4C-3CR reaction leading to oxoisoindole 8a.
2

T3P®
Yielda
Entry
Reaction time
(equiv.)
8a [%]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
5 min
15 min
1 h
2 h
4 h
15 min
15 min
15 min
15 min
15 min
15 min
2 h
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
1.0
2.0
0.5
0.1
0
71
74
73
70
75
68b
67c
69d
65e
66f
76
73
48 h
48 h
39
traceg
Reagents and conditions: 5a (1 mmol), 6a (1 mmol), 7a (1 mmol), T3P® (see the appropriate column) EtOAc (3
mL), room temperature (unless otherwise indicated).
a Isolated yield (8a).
b tert-Butyl isocyanide (7a) added 10 min after the rest of the reactants.
c T3P® added 10 min after the rest of the reactants.
d Reaction was carried out at 70 °C.
e Aniline (6a, 1.3 equiv.).
f tert-Butyl isocyanide (7a, 1.3 equiv.).
g Detected by HPLC-MS.
Next, the temperature and the amount of components were screened. The desired product (8a)
was obtained in 69% yield at 70 °C for 15 min. (Table 1, entry 8). This was followed by
increasing the amount of aniline (6a) or tert-butyl isocyanide (7a) to 1.3 equivalents, resulting
in decreased yields of 65% and 66%, respectively (Table 1, entries 9, 10). In the case of
changing the amount of T3P® to 2 equivalents, the yield was slightly increased to 76% (Table
1, entry 11). Reducing the amount of T3P® to 0.5 equivalents resulted in the formation of 8a
in 73% yield, but 2 h was necessary for full conversion (Table 1, entry 12). Further reducing
the amount of T3P® to 0.1 equivalents gave 8a in only 39% yield after 48 h (Table 1, entry
3

13). In the absence of the T3P® the product could be detected only in trace amounts by
HPLC-MS (Table 1, entry 14).
With the optimized reaction conditions in hand, the scope and limitations of the T3P®-assisted
synthesis of 3-oxoisoindoline-1-carboxamides (8) was examined (Scheme 2). The reaction of
2-formylbenzoic acid (5a) with various primary aromatic amines (6a–j) and isocyanides (7a–
c) was studied.
First, an isoindolinone library was synthesized using tert-butylisocyanide (7a), 1-
pentylisocyanide (7b) or 2,6-dimethylphenylisocyanide (7c), and five different amines (6a–e)
resulting in products 8a–o. It should be noted that in the case of 7b and 7c the amount of
T3P® could be reduced to 1.5 equivalents. Overall, the products were obtained in good to high
yields (47–85%, Scheme 2).
After extension of the reaction with the three isocyanides, 7c was chosen for testing a wider
scope and limitation. Using 4-methoxyaniline (6f) as the amine component, product (8p) was
obtained in a reasonable yield (64%). Compound 8q was produced in similar yield (60%) with
the bulkier fluorene-2-amine (6g). Aryl amines containing electron withdrawing groups
strongly hindered the reaction. Particularly, the use of 3,5-bistrifluoromethyl aniline (6h) gave
product 8r in 40% yield. Furthermore, no product formation was observed with 4-nitroaniline.
Benzylamine (6i) showed low reactivity and the corresponding product (8s) was isolated in
25% yield. When the reaction was performed with 1,4-phenylenediamine (6j), two products
were isolated, the mono- (8t) and the bis-isoindole derivative (8u) in 61% and 20% yield,
respectively. Compounds 8t and 8u could only be separated using reverse-phase flash
chromatography.
4

Scheme 2. Scope and limitations for the Ugi-4C-3CR synthesis of an isoindolinone library.
1
Notably, in the H-NMR spectrum of 8s, the two benzylic protons gave doublets at different
chemical shifts (4.05 ppm and 5.34 ppm), that could be explained by their diastereotopic
5

character and also by the proximity of the isoindolinone ring oxygen atom. This phenomenon
was investigated with quantum chemical calculations and is presented in detail in the ESI.
A plausible mechanism for the T3P®-mediated synthesis of 3-oxoisoindoline-1-carboxamides
(8) is outlined in Scheme 3. The reaction steps leading to heterocycle 8 presumably begins
with the formation of previously reported intermediate 9,21 followed by addition of the
isocyanide resulting in 10. This positively charged species (10) is than attacked by the ring-
opened anionic residue of the T3P®-water adduct (QO-) leading to the imidic phosphonic
anhydride 11. Under aqueous work-up conditions anhydride 11 is hydrolysed resulting in the
1
desired amide 8 after tautomerization. In order to provide evidence for this mechanism, H
and 31P NMR measurements were performed in the presence and also in the absence of T3P®.
These results support that T3P® plays a role in the scavenging of water, forming intermediate
9, and reacting with isocyanide 10. A detailed mechanistic study is under preparation.
Scheme 3. Proposed mechanism for the T3P®-promoted formation of isoindolinones 8.
6

Next, tryptamine (6k) or 3,4-dimethoxyphenyl-ethylamine (6l) were investigated in the Ugi-
4C-3CR. Besides the desired products 8v and 8w, formation of side-products 13a and 13b
were also observed in 18% and 12% yield, respectively (Scheme 4, pathway A). The structure
of these heterocycles was elucidated by NMR spectroscopy, and it was found that the applied
isocyanide was not necessary for their formation. Thus the reaction of 2-formylbenzoic acid
(5a) and the corresponding amine in the presence of T3P® gave compounds 13. The reaction
conditions were optimized for 13a (Scheme 4, pathway B, Table 2). It turned out, that for the
highest yield in 2 h (Table 2, entries 3–6), it was necessary to increase the amount of the
amine and T3P®. The application of two equivalents of tryptamine and five equivalents of
T3P® at room temperature gave product 13a in 68% yield (Table 2, entry 5). When the
reaction mixture was heated under microwave (MW) conditions,23 the corresponding product
(13a) was obtained in 74% yield (Table 2, entry 6). Notably, previous syntheses using
sulphuric acid gave 13a in 68% yield,24 the application of boric acid and acetic acid gave 13a
in 96% yield,25 and a solvent- and additive-free MW synthesis gave 13a in 95% yield.26 Using
homoveratrylamine (6l) as the starting material, compound 13b was produced in 45% yield
under the same reaction conditions. Moreover, in order to utilize the above Pictet-Spengler
acylation cascade, 6-formyl-2,3-dimethoxybenzoic acid (5b) and 2 equivalents of 2-[3,4-
(methylenedioxy)phenyl]ethylamine (6m) were reacted under the optimized conditions,
resulting in formation of the alkaloid (±)-nuevamine (13c) in 35% yield. This moderate yield
might be explained by the decreased reactivity of the carbonyl and carboxyl groups of 5b
caused by the electron donating methoxy groups. Notably, the nuevamine skeleton has been
previously synthesized from analogues of 5, but via a two-step procedure.27 To the best of our
knowledge, our method represents the first one-pot synthesis of natural product (±)-
nuevamine (13c).
7

Scheme 4. Reactions of compounds 5a,b with tryptamine (6k) or phenyl-ethylamines (6l,m).
Table 2. Optimization for the reaction of 2-formylbenzoic acid (5a) with tryptamine (6k)
T3P®
[equiv.]
Entry
6k
[equiv.]
T
t
Yield 13a
[%]b
1.5
3
5
3
5
1
1
1
2
2
2
rt
rt
rt
rt
rt
24 h
3 h
2 h
2 h
2 h
23
41
51
48
68
74
1
2
3
4
5
5
110 °Ca
10 min
6
aMW irradiation.
bIsolated yield
The mechanism for the formation of compounds 13 could be envisaged two ways (Scheme 5).
In the first case, the primary amine (6) and 2-formylbenzoic acid (5a) presumably forms
imine 14 and the heterocyclic ring is closed via a Pictet-Spengler cyclization leading to
carboxylic acid 15 (Scheme 5, pathway A, red arrows). This might be followed by
8

intramolecular acylation resulting in heterocycle 13. In the second case, acylation of the
amine (6) by 2-formylbenzoic acid (5a) and subsequent reaction of the aldehyde functional
group with the amidic NH (16) might form an iminium intermediate (17) that leads to 13
through a Pictet-Spengler ring closure (Scheme 5, pathway B, blue arrows). However, in
parallel with this pathway, the Bischler-Napieralski ring closure of 16 cannot be excluded,
which is also supported by the observation that 6,7-dimethoxy-1-(2-formylphenyl)-3,4-
dihydroisoquinoline (18) was detected by HPLC-MS and its structure was determined by
1HNMR in the crude reaction mixture (Scheme 5, pathway C, black arrow). Moreover, when
monitoring the reactions of 5a with 6k and 6l, imine 14 could be distinguished by HPLC-MS
from amide 16, which was formed only in 20% and 6%, respectively, supporting the
dominance of pathway A.
Scheme 5. Proposed T3P®-mediated reaction mechanisms for the formation of heterocycles
13.
9

In order to provide evidence for the above mechanisms, additional experiments were
performed (Scheme 6). Imine 14 could not be isolated from the reaction of 5a and 6k without
any additive or in the presence of MgSO4 or T3P®, which can be explained by its low
stability.28 However, in the reaction of 5a and 6k in ethanol at reflux,29 2’-carboxyphenyl-
1,2,3,4-tetrahydro-β-carboline (15) was isolated in 96% yield. When 15 was reacted in the
presence of 5 equivalents of T3P® under MW conditions at 110 °C for 10 min, the desired
product 13a was formed in 30% yield. Considering pathway B, firstly, amide 16 was
synthesized via the acylation of tryptamine with 2-formylbenzoyl chloride. Notably, in the 1H
and APT NMR spectra the cyclized 3-hydroxyisoindolinone was observed as the stable
isomer. It was also observed that the reaction proceeded in solution already at room
temperature without any additive. Adding 5 equivalents of T3P® over 10 min resulted in full
conversion, the HPLC chromatogram showed 13a and 18 in a 13:1 ratio. Heterocycle 13a was
isolated in 36% yield.
Scheme 6. Reactions supporting the proposed reaction mechanisms.
Based on this preliminary experimental investigation of the proposed mechanistic pathways
one can see the dominance of pathway A starting with imine formation following by Pictet-
10

Spengler cyclization. Nevertheless, pathway B is also conceivable, but in a much smaller
amount.
Finally, attempts were made to extend the reaction to 2-acetylbenzoic acid. However, in this
case we could not obtain the desired product, and only the formation of 2-formylbenzanilide
was observed by HPLC-MS, which might be explained by the decreased reactivity of the
electron rich acetyl oxo group than that of the formyl derivative.
Following the heterocycle synthesis, we were motivated to investigate the role of T3P® in a
traditional Ugi-4CR of benzaldehyde (19), aniline (6a), benzoic acid (20), and 2,5-
dimethylphenyl isocyanide (7c). Reacting these four components in ethyl acetate gave the
desired α-aminoacyl amide (21) in 30% yield after 48 h (Scheme 7, pathway A). In the
presence of T3P® the formation of 21 was not observed, but instead, β-aminoamide 22 was
isolated in 65% yield after 5 min (Scheme 7, pathway B).22 The reaction was monitored by
TLC and HPLC-MS and the conversion did not increase even after 24 h. From this result it
could be concluded that T3P® promotes formation of the imine from aniline 6a and
benzaldehyde 19, but hinders the role of the carboxylic acid in the Ugi-reaction protonating
the imine intermediate before attack of the isocyanide.22
11

Scheme 7. Investigation of the classical Ugi-4CR
Conclusion
A novel T3P®-promoted synthetic procedure was developed for the synthesis of
isoindolinones (8) in a one-pot, multicomponent Ugi-reaction from 2-formylbenzoic acid
(5a). The scope and limitations of the reaction was demonstrated by the application of various
amines (6) and isocyanides (7). A side reaction was observed in the reaction of tryptamine
(6k) and phenylethylamine derivatives (6l,m) that was optimized and applied for the first one-
pot synthesis of the alkaloid (±)-nuevamine (13c). Proposals have been presented for possible
reaction mechanisms supported by additional experiments. Finally, it was observed that the
Ugi-4CR was not successful in the presence of T3P®.
Acknowledgement
P. Ábrányi-Balogh is grateful for funding by the Hungarian Science Foundation OTKA
(project PD124598).
References
12

1. Váradi A, Palmer TC, Dardashti RN, Majumdar S. Molecules. 2016;21:19/1–
19/22.
2. Kaim LE, Grimaud L. Tetrahedron. 2009;65:2153–2171.
3. Ugi I, Werner B, Dömling A. Molecules. 2003;8:53–66.
4. Tanaka Y, Hasui T, Suginome M. Org Lett. 2007;9:4407–4410.
5. Barreto AFS, dos Santos VA, Andrade CKZ. Beilstein J Org Chem.
2017;13:2596–2602.
6. Singh K, Kaur A, Mithu VS, Sharma S. J Org Chem. 2017;82:5285–5293.
7. Cárdenas-Galindo LE, Islas-Jácome A, Colmenero-Martínez KM, Martínez-Richa
A, Gámez-Montaño R. Molecules. 2015;20:1519–1526.
8. Speck K, Magauer T. Beilstein J Org Chem. 2013;9:2048–2078.
9. McCormack PL. Drugs Aging. 2015;32:409–418.
10. Cruz MP. Pharm Ther. 2016;41:308–313.
11. Valencia E, Freyer AJ, Shamma M, Fajardo V. Tetrahedron Lett. 1984;25:599–
602.
12. Zhu C, Flack JR. Tetrahedron. 2012;68:9192–9199.
13. Liu B, Wang Y, Liao B, Zhang C, Zhou X. Tetrahedron Lett. 2015;56:5776–5780.
14. Tyagi V, Khan S, Chauhan PMS. Tetrahedron Lett. 2013;54:1279–1284.
15. Liu L, Bai S-H, Li Y, Wang L-X, Hu Y, Sung H-L, Li J. J Org Chem.
2017;82:11084–11090.
16. Ley SV, Taylor SJ. Bioorg Med Chem Lett. 2002;12:1813–1816.
17. Emaeili-Marandi F, Yavari I, Saeedi M, Mahdavi M, Shafiee A. Helv Chim Acta.
2016;99:87–190.
18. Akrami S, Firoozpour L, Goli-Garmroodi F, Moghimi S, Mahdavi M, Zonouzi A.
Synth Commun. 2016;46:1708–1712.
19. Milen M, Ábrányi-Balogh P, Dancsó A, Frigyes D, Pongó L, Keglevich G.
Tetrahedron Lett. 2013;54:5430–5433.
20. Ábrányi-Balogh P, Földesi T.; Grün A, Volk B, Keglevich G, Milen M.
Tetrahedron Lett. 2016;57:1953–1957.
21. Milen M, Dancsó A, Földesi T, Slégel P, Volk B. Tetrahedron. 2016;72:5091–
5099.
22. Milen M, Dancsó A, Földesi T, Volk B. Tetrahedron. 2017;73:70–77.
13

23. Desroses M, Koolmeister T, Jacques S, Llona-Minguez S, Jacques-Cordonnier M-
C, Cázares-Körner A, Helleday T, Scobie M. Tetrahedron Lett. 2013;54:3554–
3557.
24. Hoefgen B, Decker M, Mohr P, Schramm AM, Rostom SAF, El-Subbagh H,
Schweikert PM, Rudolf DR, Kassack MU, Lehmann J. J. Med. Chem.
2006;49:760–769.
25. Fodor KJ, Kocsi VL, Kiss K, Károlyi BI, Szabolcs Á, Silaghi-Dumitrescu L,
Csámpai A. Monatsh. Chem. 2013;144:1381–1387.
26. Jida M, Soueidan OM, Deprez B, Lacondea G, Deprez-Poulain R. Green Chem.
2012;4:909–911.
27. Heaney H, Shuhaibar KF. Synlett. 1995;1:47–48.
28. Tian Y, Sun J, Zhang K, Li G, Xu F. Synthesis. 2015;50:2255–2265.
29. Berti F. Annali di Chimica (Rome, Italy). 1962;52:1101–1121.
14

Highlights
 T3P® was applied successfully in an Ugi four-center three-component reaction
 An isoindolinone library was synthesized in a short reaction time with good yields
 The scope and limitations was shown applying of various amines and isonitriles
 A side reaction was extended for the synthesis of the alkaloid (±)-nuevamine
30.
15