D'Ercole, Annunziata. "Development and scale-up of synthetic strategies for exotic macrocyclisation to increase druggability of peptides as active pharmaceutical ingredients of industrial interest". Doctoral thesis, 2022. http://hdl.handle.net/2158/1264636.
Resumen
In the framework of the PhD project of industrial interest, I have been involved
in the development and optimization of synthetic procedures to obtain peptides of
pharmaceutical interest, both on the laboratory scale and for the industrial
production, in the context of the University-Industry Joint Laboratory PeptFarm
of the University of Florence.
This research develops through two parallel lines. The former is related to the
development of a multigram, scalable cGMP-compliant MW-SP synthetic
approach for the manufacture of the cyclic peptide Active Pharmaceutical
Ingredient Eptifibatide acetate. Additionally, an alternative, patentable synthetic
approach has been developed to overcome patent restrictions.
On the other hand, the development of an efficient synthetic strategy for the
preparation of a library of stapled peptides of pre-clinical interest derived from
relaxin hormone was performed, aiming to investigate their biological role.
Moreover, the feasibility of an oral administration of serelaxin gastroprotected
formulations were investigated.
According to the industrial perspective on research needs and opportunities in
manufacturing, automation of as many steps as possible within an industrial
production frame is pivotal to guarantee safety requirements. Solid-phase
strategies are considered methods of election for medium-length peptide
syntheses not only at the research scale but for large-scale production, as well.
The possibility to use microwave-assisted technology on the large scale recently
introduced, prompted us to evaluate the possibility to conveniently set-up a safe
and fully cGMP-compliant pilot process to produce Eptifibatide acetate Active
Pharmaceutical Ingredients (API), a generic hexapeptide, characterized by a
single disulfide bridge. We investigated strategies based on the use of the
microwave-assisted solid-phase peptide synthesis (MW-SPPS), by the use of a
DIC/Oxyma Pure coupling protocol at 90 °C. This fully automated technology,
previously accessible only at R&D level, has been recently made available also
for the large-scale manufacturing of peptide APIs, taking constantly into account
6
the cost-effectiveness and dangerousness of each procedure. Accordingly, we
developed an optimized process at the laboratory scale (1-5 mmol), which was
subsequently successfully scaled-up to 70 mmol, obtaining all the information
required by regulatory agencies to validate the process and qualify the pilot-scale
plant. The process consists of 5 steps: 1) automated microwave-assisted solid phase
synthesis of Eptifibatide linear precursor; 2) cleavage from the resin with
concomitant amino acid side-chains deprotection; 3) disulfide-bond formation in
solution; 4) purification by flash column chromatography; 5) ion-exchange solid phase
extraction. Since the direct scale-up of a kg-scale, cGMP compliant
peptide API production procedure is a challenge that requires an accurate
understanding of each involved step, we preliminary performed a quality
management risk assessment, which enabled a smooth and effective achievement
of a successful final result. Moreover, in our optimization process, a reduction in
time, solvents and waste have been obtained, ensuring compliance with the
quality specifications, according to regulatory agencies requirements (FDA and
EMA). Satisfactory results were obtained in terms of Eptifibatide acetate HPLC
purity (99.6%) and Yield (22.1%).
Additionally, the investigation of an alternative on-resin cyclization strategy for
therapeutic peptide industrial production of Eptifibatide acetate has been carried
out in parallel with the aim to develop a robust and economically competitive
production process avoiding intermediate steps of isolation to preserve the
recovery guaranteeing a GMPs quality product, to overcome patent restrictions.
A scalable, fully automated approach performed entirely in the same reactor has
been developed. We explored and compared four solid-phase disulfide formation
approaches (A, B, C, D) between the C-terminal Cys and the N-terminal 3-
Mercaptopropionic acid (MPA). These mainly differ one from each other for the
final cyclization step, obtained by direct formation of an S-S disulfide bridge
(strategies A-B) or via side-chain-to-tail amide bond formation (strategies C-D).
Strategy D resulted the best one, thanks to the concomitant reduction of the Stert-
butylthio (StBu) Cys protecting group (PG) and disulfide formation with the
MPA reducing agent, enjoying the advantage of using an already qualified
starting material. This strategy (D) represents an inventive (non-obvious)
7
strategy, (since we were the first to propose MPA to deprotect StBu on cysteine),
which proposes for the first time to perform all the processes including disulfide
bond formation in a single reactor (novelty), scalable on multigram-scale by
Liberty Pro synthesizer (Industrial applicability). Therefore, according to the
three patentability criteria required for a new production process: Novelty;
inventive and industrial applicability, the present PhD work identifies a new
patentable production process.1
In line with the synthesis of conformationally constraint relaxin derivatives, the
present work describes the development of an innovative, efficient and
reproducible MW-assisted Copper-Catalyzed Azide-Alkyne Cycloaddition (SP
MW-CuAAC) performed on solid phase to prepare side-chain-to-side-chain
clicked H1-relaxin single B-chain analogues, overcoming the several synthetic
drawbacks (aggregation tendency and poor solubility) which hamper
relaxins syntheses.
All the relevant parameters, that are, resin (PEG-PS vs PS), solvent mixtures
(H2O:t-BuOH:DCM 1:1:1, DMSO:DMF 1:2), catalytic system (CuBr vs CuSO4),
microwave energy and reaction time were optimized using a systematic
approach.2 Two generations of H1-relaxin single B-chain stapled analogues were
obtained. First-generation (VR and VIR) and second-generation H1-relaxin
single B-chain peptides (VII and VIIR; VIII and VIIIR; IX and IXR) were
characterized by different lengths and different positions and orientations of the
triazolyl ring, and were designed with the aim to stabilize the α-helix
conformation and to expose the binding cassette motif. The α-helicity induced by
the side-chain to side-chain stapling obtained was demonstrated by the circular
dichroism (CD) performed both in phosphate buffer and in SDS micelle, thanks
to the collaboration with Prof. A. Carotenuto (University of Naples Federico II).
Moreover, in the frame of the collaboration with Prof. D. Bani (University of
Florence) and Prof. A. Hossein (Institute of Neuroscience and Mental Health,
University of Melbourne, Australia), H1-relaxin analogues were biologically
tested, to verify binding to cells expressing the receptor RXFP1 and activity
8
through cAMP signaling pathway in HEK-293T cells stably expressing the
RXFP1 receptor.
Moreover, since the major challenge in the development of peptide drugs is to
improve their oral bioavailability, we investigated the relative bio-potency of the
intact serelaxin molecule (the recombinant form of human H2-relaxin) and the
purified porcine one, in comparison with their proteolytic fragments, obtained
after treatment with Simulated Intestinal Digestion Fluid (SIF). Signalling events
downstream receptor activation in THP-1 human monocytic cells was measured.