Biotransformation of hydroxyphosphonates 275
Spirit blue agar plates
compounds 2a and 2b provided products enriched
in R or S- stereoisomers depending on the lipase
applied.
The lipolytic activity of microorganisms was tested
on Spirit blue agar plates. Several media were tested
and the best one was chosen in order to achieve the
most vigorous growth and the best lipolytic activity
visualization. 5 g of peptone from meat, 5 g of
peptone from casein, 5 g of yeast extract, 20 g of
agar-agar and 0.15 g of Methyl Blue (Spirit Blue)
were suspended in 1 l of distilled water and auto-
claved (15 min at 121°C). In a separate flask lipase
reagent was prepared by suspending of 6 ml tribu-
tyrin and 0.6 g gum arabic in 24 ml of distilled water.
Lipase reagent was also autoclaved (15 min at 121°C).
Both mixtures were cooled to at least 50°C (stabiliza-
tion of the emulsion), poured into sterile plates and
allowed to solidify. Inoculated plates were incubatad
at 30Ϯ2°C for 72 hours and after this time, clear
zones around colonies of lipolytic microorganisms
were observed.
Higher values of enantiomeric excess were
achieved for hydrolysis of 2a. Unfortunately, reac-
tions exhibited low to moderate stereoselectivities.
In most cases enzymatic hydrolysis of 2a provided
products enriched in the R-enantiomer of the alco-
hol. The only exceptions were lipases from Candida
antarctica and Aspergillus niger, which more selectively
hydrolyzed the substrate of opposite configuration,
thus providing the S-product. The reverse stereose-
lectivity of most of lipases was observer in case of
hydrolysis of 2b.
Enzymatic acylation
Despite the high excess of enzyme used and pro-
longed reaction time, we were not able to obtain
conversions higher than 8–15% (Table III).The only
enzyme which exhibited satisfactory activity was
lipase from Candida antarctica A, which yielded
52% conversion of substrate 1a within 4 days.
However, even in this case the reaction had poor
enantioselectivity (17%ee).
Results and Discussion
Screening
Screening was carried out using a wide variety of
available enzymatic preparations and microorgan-
isms. Reactions were carried out using both types of
biocatalyst for a standard time (6 days for hydrolysis
by whole cells and 7 days for hydrolysis and acylation
by lipases) at room temperature (Table I). Enzymatic
hydrolyses were carried out in a biphasic system
(phosphate buffer and the mixture of n-hexane and
diisopropyl ether) using 50 mg of biocatalyst. The
acylation reaction was carried out under anhydrous
conditions (diisopropyl ether) and therefore only iso-
lated enzymes were tested.
Microbial hydrolysis
The esters 2a and 2b were also hydrolyzed using
whole bacterial cells. These microorganisms were
previously tested for their lipolytic activity using
Spirit Blue agar plates. The kinetic resolution was
carried out in a buffered medium and reactions were
stopped after reaching the conversion range of ∼50%
or after 6 days (Table IV).
Microbial hydrolysis gave the most satisfactory
results. All the reactions carried out when 2a
was used as substrate provided S-1a with different
ee values, depending on the strain used. The best
results were obtained using B. subtilis – after 8 hours
a product of 95% ee was obtained with satisfactory
yield. Escherichia coli might also be considered as a
potentially useful catalyst; however, high ee % values
were only achieved at low conversions. In contrast,
hydrolysis of 2b did not give satisfactory results, with
both conversions and enantioselectivities being
exceptionally low.
In all the cases conversion of compound 2a
was significantly higher than that of 2b. Only bio-
catalysts giving more than 10% substrate conversion
in the screening reaction were considered for further
studies. Whereas hydrolysis of compounds 2 gave
satisfactory results the synthesis of these compounds
by acylation of hydroxyphosphonates 1 failed (only
one reaction occurred with a conversion higher
than 10%). Therefore, in the hope of achieving bet-
ter substrate/product ratio the enzymes showing any
activity were used in further studies.
We were unable to calculate ee% values for
the unreacted 2. Neither diethyl- or dibenzyl ace-
toxyphenylmethanephosphonates showed separation
of signals in their 31P NMR spectra in the presence
of α- cyclodextrin. A better method for determina-
tion of ee values of these esters may be their chemi-
cal conversion into hydroxyphosphonates and using
methods suitable for alcohols – 31P NMR spectra
Enzymatic hydrolysis
In order to optimize the reaction, the time of reac-
tion and amount of enzyme were adjusted to achieve
approximately 50% of substrate conversion (Table
II). This is typical when trying to set up conditions
for kinetic resolution of enantiomers. Hydrolysis of