P. Burtscher et al.
optimization was balancing the incorporation with the specific Lower than expected from experiments with C-13 was the
incorporation.
specific activity at 795 MBq/mmol. This number was close to
the lower limit of the requested range (20–40 mCi/mmol). But as
the titer was unusually high, this specific activity was considered
to be minimal. It is interesting to note that the D-[14C]Trp could
(partially) be recovered during work-up.
Adaption of the fermentation process
SANDOZ Kundl provided us with a Streptomyces longisporoflavus
high-producing strain and adapted the existing fermentation
protocol to our needs. Without definite optimization, the impact
of the amount of precursor added, of the time point of feeding
and harvesting were investigated.
Second run with radiolabeled precursor
Although not robust, these results confirmed the general
Attempts to modify the added precursor to L-Trp-OMe10 in applicability of the process and underlined its suitability for
order to prevent inhibition of the system were not successful. the production of C-14-labeled staurosporine. A repetition
Therefore, L-Trp appeared to be the building block of choice. (30 mg/50 ml) on a preparative scale (i.e. using more fermenta-
First experiments with [13C6]L-Trp indicated efficient incorpora- tion flasks) appeared to be reasonable. Entries 2 and 3 were
tion (450%, determined by 13C-NMR) at a level of 1 g/kg. Taking performed in the same run. In order to minimize the risk of a too
into consideration that two L-Trp moieties contribute to one low specific activity, the addition was raised to 40 mg/50 ml for
molecule of staurosporine provided evidence that the method five flasks. In this run (entries 2 and 3), the titer was 50% lower.
has a distinct chance of realization.
The specific activities, however, were significantly higher. Under
identical reaction conditions raising the addition from 30 mg/
50 ml to 40 mg/50 ml increased the average specific activity from
1000 to 1397 MBq/mmol. In contrast to entry 1, the radio-
chemical purity of the crude material was less (74–78%)
resulting in 123 and 141 MBq after HPLC-purification, respec-
tively, correlating to a radiochemical yield of 7.6 and 6.5%. In this
run, the volatile activity amounted to 2% of the total activity
employed. Finally, 4520 MBq of [3-14C]D/L-Trp provided 342
MBq of [4c,7a-14C2]staurosporine 1b.
Adaption to the radiofermentation
In contrast to the pre-experiments, a racemic mixture of D/L-Trp11
was used for the radiofermentation. Owing to its commercial
availability synthetically produced [3-14C]D/L-Trp was selected as
building block resulting in the labeling of the positions 4c and 7a
of the staurosporine moiety.
Owing to differences in the fermentation equipment, the
transfer of the protocol to the isotope lab revealed some
difficulties. Lower shaking intensity (due to lower extension) and
no capability to control the humidity of our shaking device
resulted in a generally lower titer and a considerably higher
inter-experimental variability. Despite experimental attempts to
compensate for these effects (i.e. by increasing the shaking
velocity and by saturation of the circulating air stream with
water at 351C), the quality of the above experimental data could
not be reached.12
Formation of [14C2]PKC412
In a final synthetic step, the [4c,7a-14C2]staurosporine 1b was
N-benzoylated to obtain [14C2]PKC412 2b (Figure 2).
Experimental
Materials
Orientating experiments under these conditions gave evi-
dence that an increasing addition of D/L-Trp (15, 30 and 60 mg/
50 ml) raised the incorporation significantly. Addition of 30 mg
D/L-[13C]Trp ( = 15 mg L-Trp)/50 ml produced about 30% of
doubly labeled staurosporine. Based on the specific activity of
the [3-14C]D/L-Trp (50 mCi/mmol) employed, this would be
sufficient to meet the customer’s requirement (20 mCi/mmol).
Although significantly less productive, up to 50% of the
potential appeared to be sufficient and prompted us for a first
radiolabeled run.
The goals of the first radiolabeled run were to verify the above
range of data, to test the organism for radiosensitivity and to
determine the amount of volatile activity (i.e. [14C]CO2]
produced during the fermentation process. For this reason,
the air was continuously circulated through 1 N aqueous NaOH
to capture any [14C]CO2.
Strain: Streptomyces longisporoflavus GLS7 high-producing
mutant of Streptomyces longisporoflavus R19.13
Procedure
(a)Pre-culture medium containing sucrose and soybean flour, pH
adjusted to 6.8–7.5, standard sterilization conditions. Inoculated
with 2–6% frozen culture, cultivation time 48–56 h.
(b)Fermentation broth containing glucose, sucrose, soybean
flour, ammonium sulfate calcium sulfate and some antifoam.
pH adjusted to 6.8–7.5.
(c)Fermentation conditions
Inoculated with 10% seed culture.
Addition of precursor after 24 h, amount of precursor added
(see Table 1), cultivation time 140–150 h.
(d)Work-up: The combined phases basified with 6 M NaOH to
pH 9.2 were combined with ethyl acetate and filtered over
Super Cel. The organic phases contained the crude C-14
labeled staurosporine, which was purified by using flash
chromatography.
First run with radiolabeled precursor
Surprisingly, this run (entry 1) showed very high yields of
staurosporine providing no indication of radiation sensitivity at
the applied radio concentration. The radiochemical purity of the
crude material was 490% resulting in 78 MBq of 1b after HPLC-
purification. Even when the radiochemical yield amounted to
only 10%, only 4% of the total radioactivity was found in the aq.
NaOH impinger solution giving evidence for low volatile activity.
The identity and radiochemical purity were determined by using
HPLC, MS and 1H-NMR. The identification of the labeling
position by depletion of the corresponding 13C-NMR was not
possible due to the insufficient signal intensity.
Copyright r 2010 John Wiley & Sons, Ltd.
J. Label Compd. Radiopharm 2010, 53 613–615
Burtscher
