0:00
I will start by introducing the PolyPeptide group.
0:08
PolyPeptide is a global market leader for innovative clinical and commercial stage peptide drug substances.
0:13
Currently manufacturing around half of all approved peptide APIs with a strong custom project pipeline from biotech startups to large pharma.
0:24
It includes more than 1000 employees distributed over six production sites worldwide: three in Europe (Sweden, Belgium, France), two in the US, and one in India.
0:38
As an overview of the peptide market, since the first medical use of insulin, the development of peptide drugs has become a hot topic in pharmaceutical research because of their attractive pharmacological properties and safety profiles.
0:54
To date, more than 80 peptide APIs have been approved and marketed worldwide, with more than 170 in active clinical development and around 600 in preclinical evaluation.
1:06
Global sales of peptide APIs in the pharmaceutical industry exceeded 70 billion USD in 2019. We are witnessing an increase in the complexity of peptide sequences and the development of new formulation strategies and applications such as personalized therapy.
1:27
As part of our green chemistry program, PolyPeptide aims to reduce the environmental footprint based on four main pillars.
1:48
Reducing the amount of chemicals used in production, which will reduce the process mass intensity, recycling different solvents, reagents, and other waste, replacing hazardous chemicals whenever possible, and avoiding the use of hazardous chemical reagents by assessing new disruptive technologies starting with the first pillar: reduction.
2:12
To improve existing practices, it is important to identify the critical process steps of solid-phase peptide synthesis, which will guide us during the search for improvement opportunities that will be further implemented in industrial applications.
2:30
The performance of the manufacturing process is driven by the excellence of the chemical conditions applied to assemble the peptide sequence.
2:38
To this end, deep optimisation of the coupling and deprotection conditions was carried out to maximise the crude peptide purity and yield.
2:49
On the other hand, much less attention was paid to the optimisation of the washing steps, despite the fact that poor washing performance can be dramatic and detrimental to the obtained product quality, and that more than 75% of the solvent is typically consumed during these washing steps.
3:08
As the first improvement of the process, PolyPeptide implemented batch washing in the standard SPPS reactor used for large-scale manufacturing.
3:23
The concept of batch washing consists of the addition of DMF, followed by the mixing of the solid and liquid phases, then draining of the liquid phase.
3:33
This process is repeated several times until the amount of residual DMF present in the solution reaches acceptable levels.
3:44
Reduced solvent consumption was attained using a lower volume of solvent per batch wash, however, at the expense of the number of operations, which was significantly increased, leading to a detrimental impact on manufacturing time.
3:59
To overcome this and further improve washing, PolyPeptide implemented the percolation concept, which consists of a continuous addition of DMF while controlling the percolation velocity to allow continuous liquid phase draining.
4:18
Compared to batch wash, percolation allowed the reduction of solvent consumption by 57% and contributed to a significant reduction in washing time by more than 60%.
4:37
Percolation is now widely used in production for large-scale applications. Here is a comparison between a standard batch wash and percolation on a 600-liter reactor for washing following Fmoc deprotection.
4:52
While around 15,000 litres of DMF are required for batch washing, only 5000 litres were used per percolation, allowing a 67% reduction in volume.
5:05
Implementing percolation for the production of selected products and customer projects allowed DMF savings from 33% to 54%, with over 250 metric tons of DMF being saved in 2020.
5:23
The second pillar is recycling. PolyPeptide has a long experience in acetonitrile recycling in downstream processes.
5:33
Recycling is performed at an offsite disposal facility in Braine, where dedicated tanks are used to transport in-house waste resulting from purification.
5:43
90% of the waste is recycled to secure the supply of solvent.
5:48
This process is robust and FDA-approved. It is worth noting that other solvent recycling is currently under development at PPL, and significant improvements have been made.
6:02
The third pillar is the replacement of hazardous chemicals. As a peptide manufacturer, the main challenges we face are related to peptide chemistry itself, which involves the use of large amounts of highly hazardous reagents, solvents, and waste like dipolar aprotic DMF or excess amino acids and additives like piperidine.
6:28
For example, the production of one kilogram of GLP-1 agonist exenatide generates up to 34 tons of waste and 118 tons of CO2.
6:37
The need for adopting greener practices has been widely recognized and communicated.
6:42
While greening is considered important, quality, cost, and timelines cannot be compromised.
6:48
On the other hand, greening of registered manufacturing processes is difficult due to regulatory constraints.
6:55
Any future changes in manufacturing practices will be primarily driven by safety, regulatory, customer, and market constraints, as well as better process performance, which will lead to cost reduction.
7:16
In November 2021, Annex 17 of the REACH Regulation was amended to regulate and restrict the manufacture and use of DMF on the European market starting at the end of 2023.
7:28
Despite DMF being an excellent universal solvent for SPPS, it is reprotoxic.
7:33
It is very important now to coordinate our efforts and work closely to find green alternatives.
7:41
However, the main bottlenecks for the implementation of these alternatives are associated with the properties and characteristics of DMF, which is considered a cheap, non-flammable, readily available solvent.
7:53
DMF has an optimal polarity and viscosity for resin swelling, solubilisation of different amino acids, reagents, and by-products, permitting the synthesis of long and short peptides in high yields and purities.
8:12
Several green alternatives, including NBP, GVL, EtOAc, DMSO, and others enabling the synthesis of short peptides, were published.
8:22
However, none of them were able to compete with DMF.
8:27
Current direction in finding potential alternatives to DMF involves the use of binary solvent mixtures.
8:34
Besides the Tolomelli-Cabri Lab, who used a mixture of DCM/Anisole, PolyPeptide group was among the first to use binary solvent mixtures such as DMSO-ethyl acetate and NBP-ethyl acetate for the synthesis of short peptides.
8:51
Furthermore, the synthesis of longer peptides up to 28-mer residues using binary solvent mixtures was recently achieved by Bachem and Novo Nordisk using three main solvent mixtures: DMSO/2-Me-THF, DMSO/DOL, and NBP/DOL.
9:09
Regardless of the choice of binary solvent mixture, the ratio of both solvents needs to be further tuned based on additional considerations that I will shortly discuss.
9:19
Besides the kinetics of couplings and Fmoc deprotection and the crude peptide purity, the four parameters that we decided to add in the consideration for large-scale green SPPS are the kinetics of pre-activation, solubility of amino acids and other additives in green solvents, precipitation of diisopropylurea, and critical impurities assessment.
9:45
Before starting with the kinetics of pre-activation, I will give an overview of peptide bond formation, which consists of three steps: activation of the amino acid using DIC, followed by the nucleophilic addition of oxyma to provide the active ester, which will be further coupled with the amino acid loaded on the resin to provide the dipeptide.
10:09
To date, the coupling kinetics were evaluated as a block, so these three steps were considered as one step, and no study has integrated the step of pre-activation independently from the step of coupling.
10:23
To further emphasize this, I reproduce two graphs from a recent publication by Bachem and Novo Nordisk, where they studied the kinetics of coupling of Fmoc-glycine with a tripeptide loaded on ACTC resin using DIC oxyma in neat solvent as well as in binary solvent mixtures.
10:42
As you can see, I highlighted the binary mixtures that were chosen as potential alternatives to DMF.
10:49
The choice of these solvent mixtures was oriented toward global overall kinetics that were either comparable or superior to those of DMF.
11:04
At PolyPeptide, we decided to look for the contribution of pre-activation to the global coupling kinetics.
11:10
Surprisingly, having similar global coupling kinetics doesn't mean that the three steps of peptide bond formation have the same reaction rate.
11:22
Further studies and investigations proved that the rate of formation of the amino acid-DIC oxyma complex was much slower than the rate of formation of the active ester.
11:37
This means that the reaction between the amino acid and DIC was the limiting step or the rate-determining step, and that the pre-activation and coupling steps had different contributions to the overall kinetics.
11:51
While in DMF, pre-activation was slow and coupling was fast, we wondered whether pre-activation kinetics in alternative solvents would have similar or different behaviour.
12:04
We carried out a study of the pre-activation kinetics of Fmoc-phenylalanine in DMF as well as in a series of different green solvents reported in the literature.
12:16
As we can see, pre-activation in DMF was slow, while kinetics were much faster in the alternative solvents.
12:26
Pre-activation kinetics in binary solvent mixtures were also studied.
12:32
The difference in pre-activation kinetics between DMF and other alternative solvents will guide our choice of binary solvent mixtures, as it will have a direct impact on synthesis time, possible formation of side products, and implementation at a larger scale.
12:52
The second parameter we added is the solubility of amino acids, which is critical if we want to attain fast kinetics.
13:01
We studied the solubility of Fmoc-protected amino acids, the 20 proteinogenic ones, in the three solvent mixtures chosen as alternatives to DMF: DMSO-ethyl acetate, DMSO/2-Me-THF, and NBP-ethyl acetate were all capbale of solubilising all amino acids at 0.5M.
13:40
The third parameter we were interested in studying was the precipitation of diisopropylurea.
13:48
Diisopropylurea is a crystalline, colourless side product formed during the reaction between the complex of Fmoc-amino acid-DIC and oxyma to give the active ester.
14:01
Despite the presence of this byproduct not causing significant problems on a development or research scale, its formation becomes much more problematic when scaling up to production manufacturing, as it can clog tubing during transfers and the frit of the reactor, leading to inefficient resin washing and requiring larger volumes of solvents to eliminate it.
14:36
Diisopropylurea is soluble in DMF at concentrations lower than 0.18 molar, while no data regarding its solubility in alternative solvents had been reported.
14:49
We carried out a comprehensive study regarding the solubility of diisopropylurea in several neat and binary solvent mixtures at room temperature.
14:59
The three solvent mixtures we chose were DMSO-ethyl acetate, DMSO-2-Me-THF, and NBP-ethyl acetate.
15:09
The solubility of diisopropylurea in DMSO-ethyl acetate and DMSO-2-Me-THF was similar and moderate, comparable to that of DMF.
15:20
Its solubility in NBP-ethyl acetate was very poor, implicating larger washing volumes required to eliminate it, which would lead to increased solvent consumption, against the principles of green chemistry.
15:41
If NBP-ethyl acetate is to be used as a binary solvent mixture in SPPS, it should not be used at room temperature.
15:54
PolyPeptide is currently addressing this problem, and the path forward will be communicated shortly.
16:03
The fourth parameter we added in our consideration for large-scale SPPS was the critical impurities assessment.
16:12
Studies and publications have primarily focused on the crude peptide purity following synthesis by SPPS.
16:20
Binary solvent mixtures were chosen based on their overall purity profile.
16:29
Several articles have addressed the impact of the choice of alternative solvents on the overall purity of the crude peptide and the formation of SPPS-related side products.
16:39
Critical impurity assessment was less investigated. To secure overall process efficiency, it is indispensable to detect the presence of these critical impurities in the downstream and minimize them as much as possible to achieve better purification with higher yields and lower costs.
17:09
Regarding our green program to reduce the environmental footprint, we were able to reduce the amount of solvent used to rinse the resin following Fmoc deprotection by 70% through the development of new percolation practices.
17:28
Recycling of acetonitrile is well established at Braine. Other solvent recycling is in progress.
17:34
The replacement of DMF by green solvents is ongoing.
17:39
Four new parameters were added to guide our choice: pre-activation kinetics, solubility of amino acids, precipitation of diisopropylurea, and critical impurity assessment.
18:11
PolyPeptide is currently assessing the use of mechanochemistry for solvent-free peptide synthesis.
