Phosphonamide pyrabactin analogues as abscisic acid agonists

Article abstract of DOI:10.1039/c5ob00137d

A four step synthesis towards novel phosphonic pyrabactin analogues is presented. Via a stomatal closure and germination assay, the ability of the analogues to selectively induce the ABA-signaling pathway was demonstrated.

Full text of DOI:10.1039/c5ob00137d

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
Phosphonamide pyrabactin analogues as abscisic
acid agonists†
Cite this: DOI: 10.1039/c5ob00137d
M. Van Overtveldt,a T. S. A. Heugebaert,a I. Verstraeten,‡b D. Geelenb and
C. V. Stevens*a
Received 23rd January 2015,
Accepted 16th March 2015
A four step synthesis towards novel phosphonic pyrabactin analogues is presented. Via a stomatal closure
and germination assay, the ability of the analogues to selectively induce the ABA-signaling pathway was
demonstrated.
DOI: 10.1039/c5ob00137d
www.rsc.org/obc
The phytohormone abscisic acid (ABA, Fig. 1) 1 is a crucial structural resemblance to ABA 1, pyrabactin 2 is able to elicit
factor in the adaptive response of plants to abiotic environ- ABA-responses, such as the promotion of guard cell closure
mental stresses. ABA 1 is known to induce stomatal closure, and the inhibition of seed germination, but fails to trigger sub-
thereby reducing transpiration during periods of heat, drought stantial vegetative responses.4,5 The fact that not all ABA-path-
or salt stress.1 In view of the ongoing climate change, causing ways are affected made pyrabactin 2 an indispensable tool for
an intensification of earth’s hydrological cycle and thus entail- genetic studies and the identification of the PYR1/PYL/RCAR
ing more droughts, manipulation of the ABA-content and -sig- START-domain receptor family.6 Furthermore, because of its
naling is a promising method to improve the plant’s tolerance inherent specificity and high stability, pyrabactin 2 can be
to water scarcity and the crop yield in suboptimal growth used as a lead compound to synthesize agricultural applicable
conditions.
ABA-agonists that act as anti-transpirant without side effects
While treatment of crops with exogenous ABA 1 resulted in on plant growth and development.7
an increased drought resistance,2 the potential for its appli-
This work presents the synthesis and biological evaluation
cation in agricultural settings is limited. In addition to the of a small library of pyrabactin analogues, containing phos-
unwanted hormonal effect on seed germination and plant phonate and phosphonamide linkers instead of the originally
growth, ABA 1 is light sensitive and degrades rapidly when in reported sulfonamide bridge (2). This may result in an
contact with plants.3 Recently, a synthetic, more stable, ABA- enhanced biological compatibility as sulfonamides tend to
agonist named pyrabactin 2 (Fig. 1) was identified through have a negative effect on plant growth and are regularly used
chemical genetic analysis.4 Although there is no apparent as herbicides.8 Phosphate, on the other hand, is an essential
plant nutrient to sustain optimal growth and the plant enzy-
matic system is well equipped to handle phosphates, phospho-
nates and phosphonic acids. Furthermore, phosphorus and
sulfur have a similar geometry and atomic weight. As such,
similar receptor binding can be expected. The proposed syn-
thetic pathway is in analogy with the first reported pyrabactin
synthesis route9 and comprises the phosphonylation of aryl
halides (step 1), activation of the phosphonate function (step 2
and 3), and coupling with an amine or alcohol (step 4,
Scheme 1).
Fig. 1 Plant growth regulators (S)-(+)-abscisic acid ((+)-ABA) 1 and
pyrabactin 2.
The first step in the synthetic sequence, the coupling reac-
tion of aryl halides with diethyl phosphate or phosphite, was
performed in two distinct ways: lithium–halogen exchange and
transition metal catalysis (Table 1).
The only observed by-product was 4-fluoronaphthalene, due
to early quenching of the reactive species. In order to avoid
this side reaction, the amounts of both reagents were elevated
from 1.1 to 1.5 equivalents and the reaction time after chloro-
phosphate addition was extended from 30 minutes to 1 h.
aSynBioC Research Group, Department of Sustainable Organic Chemistry and
Technology, Faculty of Bioscience Engineering, Ghent University, Coupure Links 653,
9000 Ghent, Belgium. E-mail: Chris.Stevens@UGent.be
bDepartment of Plant Production, Faculty of Bioscience Engineering, Ghent
University, Coupure Links 653, 9000 Ghent, Belgium
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c5ob00137d
‡Current address: Botany and Plant Sciences, University of California, Riverside,
Genomics Building, 900 University Avenue, CA-92521 Riverside, USA.
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
halogen exchange approach. The synthesis of diethyl 4-fluoro-
naphth-1-ylphosphonate 9 proceeded without difficulties as
expected, since lithium exchange selectively occurs with the
bromide (Table 1, entry 1).
The synthesis of the corresponding 4-bromonaphth-1-
ylphosphonate 10 entailed more complications as both
bromine atoms are susceptible to lithium–halogen exchange.
Indeed, the reaction of 1,4-dibromonaphthalene 4 with butyl
lithium and diethyl chlorophosphate, using the optimized pro-
cedure for the synthesis of diethyl 4-fluoronaphth-1-ylphos-
phonate 9, generated multiple side products. The most
common by-products were either formed by double lithiation
of the starting material or via reaction of lithiated naphthalene
Scheme 1 Proposed synthetic route towards phosphorus containing
pyrabactin analogues.
All three halogen containing phosphonates 9, 10, and 12 with the desired end product. Several adjustments were made
were prepared by treating bromoarenes 3, 4, and 6 with butyl to improve the procedure and minimize inefficient side
lithium and diethyl chlorophosphate, using the lithium– product formation. First of all, as the solubility of 1,4-dibromo-
Table 1 Synthesis of diethyl arylphosphonates
Yielda
(%)
Entry
1
Aryl bromide
Aryl phosphonate
Reagents (equivalents)
Reaction conditions
BuLi (1.5)
ClP(O)(OEt)2 (1.5)
−78 °C, 1 h
−78 °C (add phosphate)
20 °C, 1 h
88
2
BuLi (1.1)
ClP(O)(OEt)2 (10)
−78 °C, 0.5 h
74
−40 °C, 0.5 h
−78 °C (add to phosphate)
20 °C, 1 h
3
4
Pd(OAc)2 (0.02)
PPh3 (0.06)
DIPEA (1.5)
78 °C, 16 h
96
64
HP(O)(OEt)2 (1.2)
BuLi (1.1)
ClP(O)(OEt)2 (10)
−78 °C, 0.5 h
−40 °C, 0.5 h
−78 °C (add to phosphate)
20 °C, 1 h
5
6
Pd(OAc)2 (0.02)
PPh3 (0.06)
DIPEA (1.5)
78 °C, 16 h
78 °C, 16 h
82
85
HP(O)(OEt)2 (1.2)
Pd(OAc)2 (0.02)
PPh3 (0.06)
DIPEA (1.5)
HP(O)(OEt)2 (1.2)
a After purification via column chromatography with a mixture of ethyl acetate and hexanes as eluent and silica as adsorbent. b 1,4-
Dibromonaphthalene 4 was purchased from TCI Europe or synthesized from naphthalene (Y = 87%) based on a modified procedure of
Cakmak et al.10
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
naphthalene 4 in Et2O appeared to be poor at −78 °C, lithia-
tion was performed for an additional 30 minutes at −40 °C in
order to obtain complete dissolution and lithiation of the
starting material. To avoid double lithiation at the elevated
temperature, the amount of added butyl lithium was reduced
to 1.1 equivalents. Self-condensation was circumvented by
ensuring a constant excess of diethyl chlorophosphate via
increasing the amount to 10 equivalents. Moreover, lithiated
4-bromonaphthalene was added to the excess of diethyl chloro-
phosphate in Et2O. Prior to this inversed addition, both solu-
tions were cooled to −78 °C to minimize temperature rise
during transfer via the cannula. As such, acceptable conver-
sion and yield were obtained (Table 1, entry 2). The same pro-
cedure was applied for the synthesis of diethyl 4-bromophen-1-
ylphosphonate 12 (Table 1, entry 4).
Table 2 Monohydrolysis of diethyl arylphosphonates
Diethylphosphonates 11, 13, and 14 were prepared via the
Hirao cross-coupling reaction, an efficient method for organo-
phosphorus syntheses developed in the early 1980’s.11 Several
Hirao-based procedures were evaluated, including the use of
microwave technology and acetate additives.12 Despite the
good yields reported in these papers, we only managed to
achieve a limited conversion and/or poor purity. However, the
method developed by Gooβen and Dezfuli, which employs an
alcoholic solvent and a sterically demanding tertiary amine as
base, generated satisfactory results and all products were
obtained in very good yields (Table 1, entries 3, 5 and 6).13
Although the yields using this method were generally higher
than the yields obtained via the lithium–halogen exchange
method, the procedure could not be applied to improve the
phosphonylation of dibromoarenes as the Hirao-coupling reac-
tion is not selective towards one bromine atom.
Scheme 2 Activation and coupling reaction of monoethyl arylphos-
phonates 15–20 with an amine or alcohol using thionyl chloride.
Prior to the coupling reaction of the synthesized phospho- up was generally followed by purification via column chrom-
atography. In some cases, additional purification steps were
nates (9–14) with an amine or alcohol, activation of the phos-
required (see ESI, Table S-1†). The isolated yields range from
phonate group for nucleophilic addition is required. The
activation was carried out in two steps: hydrolysis with NaOH 11% to 78%.
Two types of by-products, both formed during the activation
step, were frequently detected (see ESI, Scheme S-2, Fig. S-1
to obtain the monoethyl arylphosphonate and subsequent con-
version to the corresponding phosphonyl chloride. The mono-
hydrolysis reaction proceeded efficiently in a mixture of 1,4- and S-2, Table S-2†). Most commonly, self-condensation of the
activated phosphonate with the inactivated equivalent took
place. The resulting dimeric anhydride (22, Scheme 2),
dioxane and water containing 1.5 M NaOH. Both solvents were
proven to be required because reactions in 2.5 M NaOH as well
as in 1,4-dioxane with six equivalents NaOH, both at reflux, however, can react during the coupling step, which results in
at most 50% of the desired analogue and 50% of starting
product. The second occurring phenomenon was overactiva-
yielded only trace amounts of hydrolyzed product. The opti-
mized reaction conditions are a 1 : 1 to 2 : 3 ratio dioxane–
water and a 4 h reflux period. All products (15–20) were tion of the phosphonic acid and hence, double addition–elim-
ination of the nucleophile (21, Scheme 2). The dual activated
phosphonic dichloride is formed via a SOCl2-driven dealkyla-
obtained in good to excellent yield and purity after acid–base
extraction (Table 2).
A first approach to the synthesis of activated phosphonyl tion mechanism, similar to the cleavage of dialkylphospho-
nates with trimethylsilyl halide.15
chlorides involves the use of thionyl chloride as chlorinating
agent (Scheme 2).14 SOCl2 also served as solvent as dilution of
In an attempt to avoid these side reactions, a second chlori-
the reaction mixture with dry Et2O entailed a severe drop in nation method, using TMSNEt2 and oxalyl chloride, was evalu-
ated (Scheme 3).16 This alternative activation method proved
yield. The monoethyl arylphosphonates (15–20) were dissolved
to be successful for the synthesis of pyrabactin analogues VI
in 10 equivalents SOCl2 and heated to reflux. After completion
of the reaction, SOCl2 was evaporated and the unpurified and VII starting from ethyl 4-bromonaphth-1-ylphosphonic
acid 16, and enhanced the yield of the activation and coupling
step from 11–13% to 52–59%. In contrast, the synthesis of ana-
residue was redissolved in dry CH2Cl2. One equivalent of the
appropriate amine or alcohol and triethyl amine were added
and the mixture was stirred for 16 h at 20 °C. Extractive work logue XIV using oxalyl chloride and DMF resulted in only
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 3 Activation and coupling reaction of monoethyl arylphos-
phonates 16 and 20 with an amine using oxalyl chloride.
Fig. 3 A. Average stomatal aperture in time, relative to the aperture
measured at the start of the experiment, following 10 µM treatment with
ABA 1, pyrabactin 2 and pyrabactin analogues I–XII (stock solutions in
DMSO). B. Untreated epidermal stoma. C. Epidermal stoma 30 minutes
after treatment with 10 µM ABA in DMSO.
Fig. 2 Overview of the prepared phosphonamide and phosphonate
pyrabactin analogues and their overall yield (4 steps).
In the absence of the test compounds, a reduction of the
stomatal aperture of approximately 10% was observed after
slightly improved yield compared to the thionyl chloride method, 15 minutes, with no further change up to 30 minutes (see ESI,
as formation of dimeric anhydride remained significant.
Fig. S-4†). Therefore, 10% stomatal closure was considered a
Despite these difficulties in the final synthetic step, a total result of the incubation conditions.
of 14 phosphonic pyrabactin analogues were synthesized in
ABA 1 is highly effective in reducing the stomatal opening
acceptable overall yields. A summary of their structures is and functions within 5 minutes. A relative closure of 64% was
given in Fig. 2.
reached 30 minutes after incubating the epidermal peel in
The ability of the novel phosphonamide and phosphonate 10 µM ABA 1 solution. Pyrabactin 2 also induced stomatal closure
pyrabactin analogues to selectively elicit certain ABA-related rapidly, but to a much lower degree (21%) than ABA 1. Several
responses in plants was assessed via a stomatal closure and compounds of the small library of phosphonic pyrabactin ana-
germination assay. In case of abiotic environmental stress, logues (I, II, III, V, VII, IX and XI) were found to reduce the sto-
such as decreased water availability or increased salinity, ABA matal aperture to a greater extent than pyrabactin 2 (Fig. 3).
1 accumulates in the leaves where it mediates stomatal closure While compounds II, V, VII, IX and XI are moderate inducers
by changing the osmotic potential of the guard cells through of stomatal closure (31–38% reduction in pore opening area),
the regulation of Ca2+ levels.17 As a measure for drought analogues I and III strongly affected the stomatal aperture,
response, ABA-induced stomatal closure has been extensively reducing the opening with more than 50% after 30 minutes.
studied in for example Vicia faba, Nicotiana tabacum and Arabi- The speed at which the stomata close, indicated by the incline
dopsis thaliana.18 Implementing this biological assay, the of the graph (Fig. 3), is lower than for ABA 1. A few analogues
effect of the phosphonic pyrabactin analogues on the stomatal (VI, VIII, X and XII) did not significantly affect the size of the
aperture of detached abaxial tobacco (N. tabacum) leaf epider- stomatal aperture.
mal strips was measured and compared to pyrabactin 2 and
Remarkably, unlike the other phosphonamide pyrabactin
ABA 1 (Fig. 3). As both isomers of abscisic acid show different analogues described, compound IV did not cause closure
biological activity, the optically pure natural enantiomer of the stomata, but induced opening when the intrinsic
((+)-ABA) was used as reference compound.19 More detailed closure of 10% in untreated stomata is taken into account.
graphs for each treatment, specifying the standard error of the This observation is intriguing, as a high structural resem-
mean and the significance of the difference between the aper- blance is present between analogue IV and the newly
ture averages, can be found in Fig. S-5 of the ESI.†
identified ABA-agonists. Moreover, analogue IV is not a true
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2015

View Article Online
Organic & Biomolecular Chemistry
Paper
ABA-antagonist,§ suggesting that this molecule does not
induce the typical ABA-signaling process in the stomata.
A second well known hormonal effect of ABA 1 is the inhi-
bition of seed germination.20 Also pyrabactin 2 was proven to
be an effective germination inhibitor.4 The inhibitory effect on
seed germination of the novel phosphonic pyrabactin ana-
logues was investigated via an Arabidopsis thaliana germina-
tion assay. The seeds were incubated on medium containing
10 µM of (+)-ABA 1, pyrabactin 2 or the different pyrabactin
analogues and the germination ratio was documented 10 days
post growth at 25 °C (see ESI, Fig. S-6†). The assay revealed
that the analogues were poor inhibitors of germination (less
than 14% inhibition). Only one compound (II) markedly sup-
pressed seed germination (60% inhibition) compared to ABA 1
and pyrabactin 2 (38% and 34% inhibition respectively). These
results indicate enhanced specificity of the novel pyrabactin
analogues towards vegetative responses and confirm their
selectivity in inducing certain ABA-signaling pathways.
2 (a) C. J. Mittelheuser and R. F. M. Van Steveninck, Nature,
1969, 221, 281–282; (b) R. J. Jones and T. A. Mansfield,
J. Exp. Bot., 1970, 21, 714–719.
3 S. R. Cutler, P. L. Rodriguez, R. R. Finkelstein and
S. R. Abrams, Annu. Rev. Plant Biol., 2010, 61, 651–679.
4 S. Y. Park, P. Fung, N. Nishimura, D. R. Jensen, H. Fujii, Y. Zhao,
S. Lumba, J. Santiago, A. Rodrigues, T. F. F. Chow, S. E. Alfred,
D. Bonetta, R. Finkelstein, N. J. Provart, D. Desveaux,
P. L. Rodriguez, P. McCourt, J. K. Zhu, J. I. Schroeder,
B. F. Volkman and S. R. Cutler, Science, 2009, 324, 1068–1071.
5 M. R. Puli and A. S. Raghavendra, J. Exp. Bot., 2012, 63,
1349–1356.
6 N. Kitahata and T. Asami, J. Plant Res., 2011, 124, 549–557.
7 Q. Hao, P. Yin, C. Y. Yan, X. Q. Yuan, W. Q. Li, Z. P. Zhang,
L. Liu, J. W. Wang and N. E. Yan, J. Biol. Chem., 2010, 285,
28946–28952.
8 (a) L. J. Andus and J. H. Quastel, Ann. Bot., 1948, 12, 27–34;
(b) S. H. Crowdy and D. R. Jones, Nature, 1956, 178, 1165–1167.
9 (a) L. Duan, F. Zhou, W. Tan and Z. Li, Method for prepar-
ing 4-bromo-N-(2-pyridinylmethyl)naphthalene-1-sulfonamide,
China Pat., CN101781244A, 2010; (b) F. Zhou and L. Duan,
Agrochemicals, 2010, 7, 484–485, 496.
In conclusion, 14 novel phosphonamide and phosphonate
pyrabactin analogues were prepared in 4 steps. In general, the
pyrabactin analogues induce ABA-like stomatal closure
responses, indicating that they likely act as ABA-agonists. In
addition, the analogues do not exert a general influence on the 10 O. Cakmak, I. Demirtas and H. T. Balaydin, Tetrahedron,
ABA-signaling pathway, as was confirmed by the absence of 2002, 58, 5603–5609.
the typical ABA germination inhibition capacity. The ability to 11 (a) T. Hirao, T. Masunaga, Y. Ohshiro and T. Agawa, Synthesis,
protect leaves from excess evaporation via the stomatal pores,
without interfering with germination, makes these analogues
1981, 56–57; (b) T. Hirao, T. Masunaga, N. Yamada, Y. Ohshiro
and T. Agawa, Bull. Chem. Soc. Jpn., 1982, 55, 909–913.
very promising leads for agro-pharmaceuticals. Indeed, 12 (a) M. Kalek, A. Ziadi and J. Stawinski, Org. Lett., 2008, 10,
drought is one of the main factors limiting crop productivity,
causing more than 50% yield loss worldwide.21 Furthermore,
the synthesized pyrabactin analogues may be helpful to
unravel ABA-perception and –signaling involved in stomatal
4637–4640; (b) M. Kalek, M. Jezowska and J. Stawinski, Adv.
Synth. Catal., 2009, 351, 3207–3216; (c) Y. Belabassi,
S. Alzghari and J.-L. Montchamp, J. Organomet. Chem.,
2008, 693, 3171–3178.
regulation. Complementary studies regarding compound stabi- 13 L. J. Goossen and M. K. Dezfuli, Synlett, 2005, 445–448.
lity and the effect on rooting processes of different species are 14 W. Tang and Y.-X. Ding, J. Org. Chem., 2006, 71, 8489–8492.
ongoing to assess their applicability in agriculture.
15 (a) C. E. McKenna, M. T. Higa, N. H. Cheung and
M.-C. McKenna, Tetrahedron Lett., 1977, 18, 155–158;
(b) G. M. Blackburn and D. Ingleson, J. Chem. Soc., Perkin
Trans. 1, 1980, 1150–1153.
16 S. Park, B. Seo, S. Shin, J.-Y. Son and P. H. Lee, Chem.
Commun., 2013, 49, 8671–8673.
17 J. I. Schroeder, G. J. Allen, V. Hugouvieux, J. M. Kwak and
D. Waner, Ann. Rev. Plant Physiol. Plant Mol. Biol., 2001, 52,
627–658.
18 (a) R. A. Fischer and T. C. Hsiao, Plant Physiol., 1968, 43,
1953–1958; (b) D. A. Thomas, Aust. J. Biol. Sci., 1970, 23,
961–980; (c) M. R. G. Roelfsema and H. B. A. Prins, Physiol.
Plant., 1995, 95, 373–378.
Acknowledgements
Financial support for this research from the Research Foun-
dation – Flanders (FWO Vlaanderen, M.V.O. and T.H.) is grate-
fully acknowledged. I.V. was supported by the Agency for
Innovation by Science and Technology in Flanders (IWT) and a
fellowship from Ghent University (BOF).
Notes and references
19 (a) W. R. Cummins and E. Sondheim, Planta, 1973, 111,
365–369; (b) M. K. Walker-Simmons, R. J. Anderberg,
P. A. Rose and S. R. Abrams, Plant Physiol., 1992, 99, 501–507.
20 (a) A. Garciarrubio, J. P. Legaria and A. A. Covarrubias,
Planta, 1997, 203, 182–187; (b) E. Nambara, M. Okamoto,
K. Tatematsu, R. Yano, M. Seo and Y. Kamiya, Seed Sci.
Res., 2010, 20, 55–67.
1 J. Leung and J. Giraudat, Annu. Rev. Plant Physiol. Plant
Mol. Biol., 1998, 49, 199–222.
§The absence of an antagonistic action of pyrabactin analogue IV was proven by
measuring the pore area of 100 random stomata before and after treatment with
10 µM ABA 1 and a combination of 10 µM ABA 1 and 10 µM analogue IV. Upon
combinational treatment, a pore area reduction of 39% was observed after
60 minutes, indicating that the effect of ABA 1 (pore area reduction of 31%) was
not counteracted.
21 A. Skirycz and D. Inze, Curr. Opin. Biotechnol., 2010, 21,
197–203.
This journal is © The Royal Society of Chemistry 2015
Org. Biomol. Chem.