Abstract
In this paper we have researched two methods for peptide synthesis; solid phase peptide synthesis (SPPS) and recombinant protein production, hereafter called recombinant production. The goal was to create a decision support that can be of use when choosing a production method for a given peptide. Additionally, we wanted to find a way to tell if a given sequence might be difficult to synthesize with SPPS, and in those cases recommend recombinant production as an alternative. To accomplish this, we have investigated general problems that may occur for the two methods as well as amino acid and sequence specific issues. We have also researched if there are any known solutions to avoid these problems, and by evaluating the gravity and frequency of the problems, with these solutions in mind, the decision support was created.
A description of each considered issue is given, but the amount of each amino acid and the sequence of the amino acids in the peptide also needs to be considered when choosing the method. For 15 out of the 20 individual amino acids we have recommended the use of SPPS.
For the remaining five, three are dependent on the placement within the sequence and in two cases we recommend considering recombinant production. We believe we have created a decision support that fulfills its purpose and can be of use when choosing the production method for future biotherapeutics.
1 Abbreviations 7 2 A tool for efficient method selection to facilitate Affibody’s business 9 2.1 The Swedish biotech company and their revolutionary techniques 9 2.2 A decision support and more insight to the world of biotherapeutics as
mission 10
3 The decision support 11
3.1 Things to think about when choosing one of the methods 11
3.1.1 SPPS 11
3.1.2 Recombinant protein production 12
3.2 Specific amino acids 12
3.2.1 Alanine, Ala 12
3.2.2 Arginine, Arg 13
3.2.3 Asparagine, Asn 13
3.2.4 Aspartic acid, Asp 13
3.2.5 Cysteine, Cys 14
3.2.6 Glutamic acid, Glu 14
3.2.7 Glutamine, Gln 14
3.2.8 Glycine, Gly 15
3.2.9 Histidine, His 15
3.2.10 Isoleucine, Ile 15
3.2.11 Leucine, Leu 15
3.2.12 Lysine, Lys 16
3.2.13 Methionine, Met 16
3.2.14 Phenylalanine, Phe 16
3.2.15 Proline, Pro 16
3.2.16 Serine, Ser 17
3.2.17 Threonine, Thr 17
3.2.18 Tryptophan, Trp 17
3.2.19 Tyrosine, Tyr 17
3.2.20 Valine, Val 18
3.3 Sequences 18
3.3.1 “Difficult sequences” 18
3.3.2 Half-time 18
3.3.4 Self-assembling peptides 19
3.3.5 Sequences prone to form aspartimide 19
3.3.6 Sequences vulnerable to acidolysis 19
3.4 The peptide 19
3.4.1 C-terminus 19
3.4.2 Length 20
3.4.3 N-terminus 20
3.5 Conclusion of the decision support 20
4 Foundation of the support; the methods and potential manufacturing
problems 23
4.1 The two methods - environmental aspects and general problems 23
4.1.1 SPPS 23
4.1.1.1 General problems of SPPS 23
4.1.1.2 Environmental aspects and safety 24
4.1.2 Recombinant protein production 26
4.1.2.1 General problems of recombinant protein production 26
4.1.2.2 Environmental aspects 29
4.2 Manufacturing problems in SPPS due to peptide properties 30 4.2.1 The length of the peptide is a limiting factor 31 4.2.2 Aggregation causes problems during synthesis and purification 31
4.2.2.1 Aggregation and its problems 31
4.2.2.2 Sequences that can aggregate 32
4.2.2.3 Solutions to aggregation 34
4.2.3 Intramolecular cyclization 35
4.2.3.1 Pyroglutamate 35
4.2.3.2 δ-lactam 35
4.2.3.3 Aspartimide Formation can lead to many different problems 36 4.2.4 Racemization lowers the quality and biological activity 37
4.2.4.1 Mechanism 38
4.2.4.2 Problematic conditions 40
4.2.5 Diketopiperazine formation cleaves the peptide 41 4.2.6 Deamidation - spontaneously deciding the half time 42 4.2.7 Acylation of His may result in branched peptides 43
4.2.8 Alkylation - a problem for many amino acids 44
4.2.9 Deguanidination of Arg can lead to branched peptides 45
4.2.10 Oxidation causes dimerization and aggregation 46
4.2.11 Dimerization can cause aggregation 47
4.2.12 𝛽-elimination of Cys causes multiple impurities 48 4.2.13 Acidolysis - Breaking the peptide bond with acid 49 4.2.14 Dehydration is a risk with unprotected Asn and Gln residues during
coupling 49
4.2.15 Formylation as a PTM or an unwanted side reaction 50
4.2.16 Phosphorylation - PTM with problems 50
4.2.17 Solvent-induced side reactions 51
4.2.17.1 DCM can react with aliphatic amines 51
4.2.17.2 Acetone can be a contaminant for Pro and His 51
4.2.17.3 Methanol might induce reactions in the side chain of Asn and Gln 52 4.3 Manufacturing problems in recombinant protein production due to peptide
properties 52
4.3.1 Peptide length 52
4.3.2 Challenges with Self-Assembling peptides 52
4.3.3 Removal of N-terminus formylmethionine 52
4.3.4 Predicting solubility based on amino acid composition 53 5 Literature study and teamwork as main project methodology 55
5.1 Scouting for information and proper sources 55
5.2 Teamwork as the main driver for progress 55
6 Contribution statement 56
7 Acknowledgements 58
8 References 59
9 Appendices 70
9.1 Accessibility and production in biotherapeutics - An ethical analysis of the
industry 70
9.1.1 Does the interest of great profit outrun everyone’s right to proper
health care? 70
9.1.2 Does the benefit of clinical trials outweigh the risk to the subjects? 72 9.2 Solid phase peptide synthesis - Bruce Merrifield’s popular prodigy 73
9.2.1 How does it work? 73
9.2.2 Purification process 75
9.3.1 How does it work? 76
9.3.2 Purification process 77
1 Abbreviations
2-MeTHF 2-methyltetrahydrofuran ABD Albumin binding domain
Acm Acetamidomethyl
AIB 2-Aminoisobutyric acid
Ala Alanine
Asn Asparagine, Asparaginyl
Asp Aspartic acid, Aspartate, Aspartyl
Arg Arginine
Boc tert-Butyloxycarbonyl Bom Nπ-Butoxymethyl Bum Nπ-tert-Butoxymethyl
CSPS Classical solution-phase peptide synthesis
Cys Cysteine
Da Dalton
DCM Dichloromethane
DHA Dehydroalanin
DKD 1,4‐diazepine‐2,5‐diones (diketodiazepines) DKP Piperazine‐2,5‐diones (diketopiperazines) DMF N,N-dimethylformamide
DMPU N,N'-dimethyl propylene urea DMSO Dimethyl sulfoxide
DTT Dithiothreitol E. coli Escherichia coli EtOAc Ethyl acetate
Fmoc 9-Fluorenylmethoxycarbonyl
For Formyl
Fc Fragment crystallizable GCP Good Clinical Practice Gln Glutamine, Glutaminyl
Glu Glutamic acid, Glutamate, Glutamyl
Gly Glycine
Gu·HCl Guanidine·HCl
GVL Gamma-valerolactone
HF Hydroflouric acid
HPLC High performance liquid chromatography Hsps Heat shock proteins
IBs Inclusion bodies iso-Asp iso-Aspartate
Ile Isoleucine
IPTG Isopropyl thiogalatoside
Lys Lysine
mAbs Monoclonal antibodies MAP Methionine aminopeptidase MeCN Acetonitrile
Met Methionine
MetO Methionine sulfoxide MetO2 Methionine Sulfone NCL Native chemical ligation NBP N-butylpyrrolidone
Orn Ornithine
Pbf 2,2,4,6,7‐pentamethyldihydrobenzofuran‐5‐sulfonyl
PC Propylene carbonate
pH Potential of hydrogen, Power of hydrogen
Phe Phenylalanine
Pmc 2,2,5,7,8‐pentamethylchromanyl‐6‐sulfonyl
Pro Proline
PTM Post-translational modification
RP Reverse-phase
RP-HPLC Reverse-phase high performance liquid chromatography
Ser Serine
SpA Staphylococcus aureus protein A SPPS Solid-phase peptide synthesis
StBu t-butylthio
TFA Trifluoroacetic acid
Thr Threonine
Trp Tryptophan
Trt Trityl
Tyr Tyrosine
SAPs Self-assembling peptides
Val Valine
WHO World Health Organization
Xan 9-Xanthenyl
2 A tool for efficient method selection to facilitate Affibody’s business
In this bachelor thesis, we give a proposition for a decision support which can be used for method selection processes regarding the production of therapeutic peptides. Two methods are being investigated - solid phase peptide synthesis and recombinant protein production with Escherichia coli (E. coli), the former being the main focus while the latter is considered an alternative. In the decision support, the methods themselves are analyzed and compared, but the emphasis is on the properties of the molecule that is being produced. The aim of the decision support is that given a peptide, the most suitable method for production can be selected.
2.1 The Swedish biotech company and their revolutionary techniques
”Our mission is to improve the lives of patients with serious diseases by being a
research-driven company with a long term commitment to development of protein based drugs”
That is how the innovative Swedish biotech company Affibody describes their main purpose and goal (Affibody Medical AB 2021). With their revolutionary Affibody® and Albumod™
technologies, they have taken some big leaps in the right direction on that journey. The company has been going for 23 years; ever since researchers from the Royal Institute of Technology and Karolinska institutet founded Affibody in 1998 (SwedenBIO 2021). Today, the company is situated in Stockholm, the capital of Sweden. In the upcoming paragraphs, we will introduce you to the company's two largest innovations with an emphasis on Affibody molecules whilst briefly describing the Albumod™ technology.
Monoclonal antibodies (mAbs) are commonly used as treatment for various diseases, but Affibody molecules can sometimes be a better alternative (Altai et al. 2018). Affibody molecules are chemically stable and small (6.5 kDa), which facilitates the penetration of tissues. The molecules can for instance block protein-protein interactions by binding to receptors or ligands, or conjugations to toxins and cytotoxic drugs (Ståhl et al. 2017).
Affibody molecules are specially engineered Staphylococcus aureus protein A (SpA) domain B, where a couple of the surface residues have been substituted (Gunneriusson et al. 1999).
SpA is used to create Affibody molecules because it binds to the Fragment crystallizable (Fc) region in mammalian antibodies (Rigi et al. 2019). Once domain B has been edited, it is called the Z-domain and is used in Affibodies as monomers or together with other Z-domains (Gunneriusson et al. 1999). Depending on how many Z-domains that are used in the Affibody molecules, the molecules will still be small, but vary in size.
The Affibody molecules can furthermore be fused with an albumin binding domain (ABD), which elongates the half-life of the molecule in the body and lowers the uptake by the
There are different ways to create Affibody molecules - solid phase peptide synthesis (SPPS) and recombinant protein production being the top candidates (see 9.2 and 9.3 for methods description). SPPS is a method to chemically synthesize molecules while recombinant protein production uses host cells as means of production. Which method that will be most suitable is case specific and depends to some extent on the method itself, but also on the properties of the molecule being synthesized. Since Affibody consistently creates a lot of new Affibody molecules, a decision support to optimize and streamline the method selection process for new molecules could potentially facilitate Affibody’s business.
2.2 A decision support and more insight to the world of biotherapeutics as mission
As described above, Affibody consistently creates new Affibody molecules. Biotherapeutics are, contrary to small molecule based drugs, often large and more complex and can hence be more challenging and costly to produce (Kabir et al. 2019). Thereby, any tool that can bring more efficiency to the process of producing biotherapeutics would contribute to a company involved in the biotherapeutics industry.
A possible tool of this kind is a decision support with the purpose to facilitate the selection process of a suitable method for production when a new Affibody molecule is created. A decision support like this will hopefully bring more efficiency, optimize the production and contribute to Affibody’s business. The goal with the decision support is that given a peptide, Affibody will be able to choose the most suitable method for production by analyzing influential properties of the molecule in question, but also based on the methods themselves.
In the decision support, we will solely compare SPPS and recombinant protein production with E. coli as potential methods of production. We will furthermore only look into peptides as the molecules of interest and analyze peptide and amino acid properties to see if there are any that could influence which method of production that will be more suitable. In this analysis, SPPS will be viewed as the primary option whilst recombinant protein production will be seen as an alternative when SPPS is not eligible. Another goal for the project is to implement an ethical analysis of the area of biotherapeutics, and also an environmental analysis of the two methods, since both are crucial parts when doing reviews and research in the area of medicine today.
By completing this project as ordered by the company Affibody, we hope to contribute to the development of the next generation's biotherapeutics.
3 The decision support
This decision support contains a collection of every amino acid and properties of the peptide that influence which method to choose. Depending on which amino acids the peptide consists of, different problems will arise when attempting SPPS or recombinant production in E. coli.
SPPS will be recommended as a primary method and recombinant production is seen as an alternative when SPPS is dissuaded. Instances where SPPS would be dissuaded is for example if SPPS would lead to low yield or be impossible to conduct. If a method recommendation is not explicitly stated, our recommendation is SPPS.
The problems being raised for each section are based on information gained from literature and relevant chapters are referred to as the problems are being presented. The details have been taken into consideration but ultimately the recommendation is based on our
interpretations of the information gained under the process.
3.1 Things to think about when choosing one of the methods
3.1.1 SPPS
SPPS was invented by Bruce Merrifield in the 1960’s and since then the method has
continued to grow in popularity. Due to the extensive research that has been conducted in this area, a lot of information regarding SPPS is available. This facilitates usage and
troubleshooting when utilizing the method. The extensive research during the past 60 years has also led to a lot of improvements of the method. SPPS is Affibody’s preferred method of production for several reasons. Compared to recombinant protein production, SPPS is a more standardized method which in this case entails reproducibility, an overall smooth process and high control. Furthermore, SPPS is conducted in a single vessel which contributes to the simplicity of the method. More about this can be found in chapter 9.2.
Another important positive aspect with SPPS is the attainability to incorporate non-canonical amino acids in the peptide. This enhances the possibility to more freely design the peptide to make it more suitable for specific purposes. A non-canonical component of this kind that is quite frequently used in SPPS is 2-Aminoisobutyric acid (AIB) (Johansson 2021).
If SPPS is chosen as the method for production, it is important to keep in mind that SPPS has two major problems - solubility and the length of the peptide. A higher amount of
hydrophobic amino acids will consequently lead to aggregation of the peptide which results in solubility issues. Homooligopeptides of hydrophobic amino acids may in the same way lead to aggregation and thus solubility issues. SPPS furthermore has issues with the synthesis of peptides longer than about 50 amino acids. Some solutions to this problem, like ligation methods, have been presented in literature but the problem with longer peptides still persists and is thus something to keep in mind if choosing SPPS. More about these issues can be
SPPS also has some drawbacks regarding both environmental aspects and safety. The safety issues can be related to the use of unsafe and hazardous chemicals, such as
N,N-dimethylformamide (DMF). DMF is connected to several health issues and is thus dangerous to humans. DMF may also lead to unwanted reactions involving the side chains of some of the amino acids. DMF is unfortunately one of the most frequently used solvents in SPPS, but research is consistently made to find a replacement. Additionally, great amounts of chemicals are utilized in SPPS in the washing steps. This is not very environmentally friendly and thus another thing to keep in mind if choosing SPPS. More about environmental aspects and safety regarding SPPS can be found in 4.1.1.1.
Lastly, some amino acids are particularly troublesome for SPPS. Asp and His may lead to problems that are important to keep in mind if the target peptide contains notable amounts of these specific amino acids. The formation of aspartimide is a consequence of Asp in
combination with amino acids with low steric hindrance. Thus, the issues due to Asp also depend on the placement in the peptide, and not only the amount of the amino acid. Amino acid specific problems are evaluated in 3.2 and more thoroughly explained in 4.2.
3.1.2 Recombinant protein production
Recombinant protein production in E. coli requires less hazardous chemicals than SPPS, and is overall bio-safe (4.1.2.2). Another advantage is that it can be used to produce peptides with more than 50 residues in length (4.3.1). The peptide is produced by a cell host, therefore it is synthesised in a natural environment with the protection measurements of the cell’s central dogma. However, due to this there is less freedom for alterations like non-canonical amino acids, since they are not part of the cell's genetic code. Also, the process requires a lot of trial and error. To achieve maximum yield, the expression system and growth condition needs to be optimized for the specific targeted protein (4.1.2.1). Recombinant protein production also has solubility issues, where the peptide can aggregate into inclusion bodies (4.1.2.1). This is dependent on the amino acid composition, and computational tools have been developed to predict solubility of the peptide based on the primary structure (4.3.4).
3.2 Specific amino acids
3.2.1 Alanine, Ala
SPPS
Ala may cause problems with aggregation, especially if in a homooligopeptide (4.2.2.2).
Recombinant production
No found problems.
3.2.2 Arginine, Arg
SPPS
δ-lactam formation occurs at the guanidino group of Arg during coupling steps (4.2.3.2). The formation of δ-lactam may lead to deletion in the peptide (4.2.3.2). Arg might undergo deguanidination which can lead to formation of ornithine (4.2.9). When DCM is used as a solvent, it might react with Arg (4.2.17.1).
Recombinant production
Arg codons, AGA and AGG, are not commonly found in E. coli genes and overexpression of rare Arg codons can result in poor synthesis (4.1.2.1). N-terminal Met will not be removed if Arg is the penultimate residue as Arg is a non-cleavable amino acid (4.3.3). High frequencies of Arg can improve solubility of the protein (4.3.4).
3.2.3 Asparagine, Asn
SPPS
Asn can form aspartimide through a nucleophilic attack under basic conditions (4.2.3.3).
Aspartimide formation is problematic since it might lead to racemization and other by-products (4.2.3.3 and 4.2.4.1). Asn can be deamidated under physiological conditions leading to it being converted into Asp (4.2.6). Asn can undergo dehydration at the coupling step under basic conditions (4.2.14). When DCM is used as a solvent, it might react with Asn (4.2.17.1).
Recombinant production
N-terminal Met will not be removed if Asn is the penultimate residue as Asn is a non-cleavable amino acid (4.3.3).
3.2.4 Aspartic acid, Asp
SPPS
Asp can form aspartimide, leading to side-products (4.2.3.3). If the aspartimide forms DKP or DKD, the peptide may be terminated (4.2.3.3) or obtain deletion sequences (4.2.5).
Aspartimide intermediates are very difficult to separate (4.2.3.3). Aspartimide formation can also lead to racemization and other by-products (4.2.3.3 and 4.2.4.1). The formation of aspartimide is sequence dependent (3.2.5). In the presence of methanol solvent, Asp might undergo methyl esterification side reactions (4.2.17.3).
If the sequence contains many Asp, recombinant production might be a better choice. The formation of aspartimide is however sequence dependent so a high number of Asp does not necessarily correlate with high degrees of formation of aspartimide.
Recombinant production
N-terminal Met will not be removed if Asp is the penultimate residue as Asp is a
3.2.5 Cysteine, Cys
SPPS
Racemization of Cys causes problems (4.2.4.2). Alkylation of Cys leads to impurities in the final product (4.2.8). If cations released from resin linker cleavage react with Cys the peptide could be irreversibly immobilized to the solid support (4.2.8). Cys can form disulfide bridges and other oxidized forms when oxidized (4.2.10). Formation of disulfide bridges can cause aggregation (4.2.10). Disulfide bridges can be disulfide scrambled if Cys residues are free (4.2.10). Cys can be subject to dimerization via intermolecular disulfide bridges when oxidized (4.2.11). β-elimination of disulfide bridges creates impurities (4.2.12). Resin-bound Cys can form piperidinylalanine (4.2.12).
If the C-terminal amino acid is Cys, SPPS is not recommended (4.2.4.2). Purification of Cys-rich sequences may be difficult and if the peptide contains multiple Cys, disulfides have to be handled properly.
Recombinant production
Cys can affect the solubility of the protein (4.3.4). Cys can lead to incorrect formation of disulfide bridges that may cause aggregation into inclusion bodies (4.3.4).
3.2.6 Glutamic acid, Glu
SPPS
Glu can undergo a cyclization reaction and form pyroglutamate (4.2.3.1). In the presence of methanol solvent, Glu might undergo methyl esterification side reactions (4.2.17.3).
Recombinant production
N-terminal Met will not be removed if Glu is the penultimate residue as Glu is a
non-cleavable amino acid (4.3.3). Low frequencies of Glu may lead to a higher probability of insoluble product (4.3.4).
3.2.7 Glutamine, Gln
SPPS
Gln can undergo a cyclization reaction and form pyroglutamate (4.2.3.1). Gln can be
deamidated under physiological conditions leading to formation of Glu (4.2.6). Dehydration may occur when Gln is coupled (4.2.14). When DCM is used as a solvent, it might react with Gln (4.2.17.1).
Recombinant production
N-terminal Met will not be removed if Gln is the penultimate residue as Gln is a
non-cleavable amino acid (4.3.3). High frequencies of Gln can improve solubility of the protein (4.3.4).
3.2.8 Glycine, Gly
SPPS
Gly may induce β-sheet packing leading to aggregation (4.2.2.2). It may also aggregate if in a homooligopeptide (4.2.2.2). Gly-containing peptides are prone to form DKP due to minimal sterical hindrance between side chains (4.2.5). Gly should be avoided at the C-terminus (4.2.5). If the C-terminal amino acid is a Gly, recombinant production should be considered.
Recombinant production
No found problems.
3.2.9 Histidine, His
SPPS
Racemization of His causes problems (4.2.4.2). The imidazole of His can be acylated which may lead to branched peptides (4.2.7). His can gain a formyl group if certain protecting groups are used (4.2.15). Acetone contaminants can make N-terminal His undergo enamination modification (4.2.17.2). If the sequence contains many His recombinant production should be considered.
Recombinant production
High frequencies of His can improve solubility of the protein (4.3.4).
3.2.10 Isoleucine, Ile
SPPS
Ile may cause problems with aggregation, especially if in a homooligopeptide or in a sequence with many other β-branched hydrophobic amino acids (4.2.2.2).
Recombinant production
N-terminal Met will not be removed if Ile is the penultimate residue as Ile is a non-cleavable amino acid (4.3.3).
3.2.11 Leucine, Leu
SPPS
Leu may cause problems with aggregation, especially if in a homooligopeptide or in a sequence with many other β-branched hydrophobic amino acids (4.2.2.2).
Recombinant production
N-terminal Met will not be removed if Leu is the penultimate residue as Leu is a non-cleavable amino acid (4.3.3).
3.2.12 Lysine, Lys
SPPS
DHA formation in combination with a pH over 10 can lead to DHA reacting with Lys, resulting in formation of lysinoalanine (4.2.12). The side chain of Lys can be formylated (4.2.15). When DCM is used as a solvent, it might react with Lys (4.2.17.1).
Recombinant production
N-terminal Met will not be removed if Lys is the penultimate residue as Lys is a
non-cleavable amino acid (4.3.3). High frequencies of Lys can improve solubility of the protein (4.3.4).
3.2.13 Methionine, Met
SPPS
Met may cause problems with aggregation, especially if in a homooligopeptide (4.2.2.2). Met may be alkylated if located at the C-terminus leading to formation of homoserine lactone (4.2.8). Met can oxidize if it gets in contact with air or an oxidizing agent, leading to a higher risk of aggregation (4.2.10). Usage of strong oxidants can irreversibly oxidize Met to MetO2
(4.2.10).
Recombinant production
Non-cleavable Met residues may cause problems when removing the N-terminal Met (4.3.3).
3.2.14 Phenylalanine, Phe
SPPS
Phe may cause problems with aggregation, especially if in a homooligopeptide or in a sequence with many other β-branched hydrophobic amino acids (4.2.2.2).
Recombinant production
No found problems.
3.2.15 Proline, Pro
SPPS
Although hydrophobic, Pro should not cause problems with aggregation (4.2.2.2). Pro should be avoided at the C-terminus (4.2.5). Pro-containing peptides are prone to form DKP due to minimal sterical hindrance between side chains (4.2.5). Acetone contamination in
combination with p-nitrobenzaldehyde might cause an asymmetric aldol reaction if the N-terminal residue is a Pro (4.2.17.2). If the C-terminal amino acid is a Pro, recombinant production should be considered.
Recombinant production
No found problems.
3.2.16 Serine, Ser
SPPS
A problem with the hydroxyl group is that it can undergo acylation (4.2.7). Cations released from resin cleavage can attack the hydroxyl group and form impurities (4.2.8). β-elimination can lead to the hydroxyl moiety being removed (4.2.12). The hydroxyl group in Ser can undergo dehydration (4.2.14). Usage of formic acid can induce formylation at the hydroxyl group (4.2.15). The hydroxyl group is at risk for phosphorylation (4.2.16).
Recombinant production
No found problems.
3.2.17 Threonine, Thr
SPPS
A problem with the hydroxyl group is that it can undergo acylation (4.2.7). Cations released from resin cleavage can attack the hydroxyl group and form impurities (4.2.8). β-elimination can lead to the hydroxyl moiety being removed (4.2.12). The hydroxyl group in Thr can undergo dehydration (4.2.14). Usage of formic acid can induce formylation at the hydroxyl group (4.2.15). The hydroxyl group is at risk for phosphorylation (4.2.16).
Recombinant production
No found problems.
3.2.18 Tryptophan, Trp
SPPS
Trp may cause problems with aggregation, especially if in a homooligopeptide (4.2.2.2).
Alkylation of Trp leads to impurities in the final product (4.2.8). If cations released from resin linker cleavage react with Trp the peptide could be irreversibly immobilized to the solid support (4.2.8). Trp can be subject to oxidation, which increases the risk of disulfide
scrambling (4.2.10). Trp can undergo a three step dimerization mechanism caused by
alkylation (4.2.11). The likelihood of cross-link formation increases proportionately with the number of Trp (4.2.11). Formylation can occur in basic conditions (4.2.15).
Recombinant production
No found problems.
3.2.19 Tyrosine, Tyr
SPPS
Cations released from resin cleavage can attack the hydroxyl group and form impurities (4.2.8). Tyr can be subject to dimerization through oxidation, which might lead to aggregation (4.2.11). The hydroxyl group of Tyr is at risk of phosphorylation (4.2.16).
Recombinant production
No found problems.
3.2.20 Valine, Val
SPPS
Val may cause problems with aggregation, especially if in a homooligopeptide or in a sequence with many other β-branched hydrophobic amino acids (4.2.2.2).
Recombinant production
No found problems.
3.3 Sequences
3.3.1 “Difficult sequences”
SPPS
“Difficult sequences'' have a tendency to aggregate (4.2.2.2). Though a common problem with SPPS, aggregation has many solutions, some of which are described in (4.2.2.3).
Recombinant production
Solubility is dependent on the amino acid sequence, and some have a higher risk of aggregation (4.3.4).
3.3.2 Half-time
SPPS
The half-time of the protein can be prolonged by Asn-Pro coupling with carboxylic side Pro (4.2.6).
Recombinant production
No found problems.
3.3.3 Homooligopeptides
SPPS
Homooligopeptides have a tendency to aggregate when they comprise hydrophobic amino acids (4.2.2.2).
Recombinant production
No found problems.
3.3.4 Self-assembling peptides
SPPS
Self-assembling peptides may pose problems with aggregation as they can spontaneously assemble into micelles or β-sheets while in solution (4.2.2.2).
Recombinant production
Self-assembling peptides may lead to toxicity for the host cell, degradation or formation of inclusion bodies (4.3.2).
3.3.5 Sequences prone to form aspartimide
SPPS
Gly, Asn, Asp, Arg, Thr and Cys are prone to form aspartimide with Asp. Sequences with Asp coupled to these amino acids should be carefully synthesized. Coupling of Asp or Asn to non-bulky amino acids like Ser and Ala may also increase likelihood of aspartimide
formation. Gly is the worst amino acid to have coupled with Asp as it does not offer any steric hindrance. For this reason, one should use backbone protection if the target sequence contains Gly-Asp or Asp-Gly (4.2.3.3).
If the target sequence is prone to form aspartimide one may need to consider recombinant production as an alternative to SPPS since aspartimide formation causes many problems.
Recombinant production
No found problems.
3.3.6 Sequences vulnerable to acidolysis
SPPS
Acidolysis leads to unwanted fragmentation. One of the most vulnerable sequences is Asp-Pro (4.2.13).
Recombinant production
No found problems.
3.4 The peptide
3.4.1 C-terminus
SPPS
DHA formation followed by hydrolysis may lead to C-terminal peptide fragments with a pyrovoyl group (4.2.12). Inefficient loading may lead to C-terminal deletion (4.1.1.1). DKP formation mainly takes place on the C-terminus during the dipeptide step (4.2.5). In the presence of methanol solvent, the C-terminus might undergo methyl esterification side
reactions (4.2.17.3). One should consider using recombinant production as an alternative to SPPS if the sequence has a Cys, Gly or Pro at the C-terminus.
Recombinant production
No found problems.
3.4.2 Length
SPPS
SPPS is not suitable for peptides over 40-50 amino acids, unless NCL or other ligation method is used (4.2.1).
Recombinant production
In recombinant production, too short peptides are broken down by proteases and too long peptides have lower solubility (4.3.1).
3.4.3 N-terminus
SPPS
DHA formation followed by hydrolysis may lead to N-terminal aminated peptide products (4.2.12). In a formylation reaction a formyl functional group can be added to the N-terminus (4.2.15). Gln or Glu residue at the N-terminus will likely undergo a cyclization reaction and form pyroglutamate (4.2.3.1).
Recombinant production
Removal of N-terminal Met/N-formyl Met is often essential for the protein to function correctly. Retention of N-terminal Met is dependent on the second amino acid in the
sequence. However, over-expression of the target protein may lead to retention regardless of the following amino acid residue (4.3.3).
3.5 Conclusion of the decision support
In order to decide if SPPS will be eligible for the production of a therapeutic peptide or if recombinant production will be a better choice, we have created a conceivable decision process. We want to highlight that this is our own perception based on the information gathered in the literature study.
The predominant parameter in the decision process and hence the first feature to look into is the length of the peptide. If the peptide is longer than ~50 amino acids, SPPS may be more troublesome than advantageous due to low yield and impurities. However, there are ways to prevail this issue but this entails additional steps in the manufacturing process and thus loss of simplicity.
The second step is to see which amino acid that is situated in the C-terminus of the peptide.
This is important to pay attention to since the amino acids Pro, Gly and Cys will be
inconvenient in this position. Pro and Gly are prone to cause unwanted side reactions and Cys may undergo racemization, leading to a decrease in chiral purity.
The following step is to look into the hydrophobic amino acid content in the peptide, since this correlates to one of the most common problems in SPPS - aggregation. Aggregation due to hydrophobic amino acids may cause troubles with solubility, which in turn results in purification issues. Though aggregation is the most common problem for SPPS, there are several ways to overcome it. However, a very high content of hydrophobic amino acids may still be problematic, so this is nevertheless an important feature to pay attention to.
The next step is to examine how frequently the amino acids Asp and His occur in the target peptide. If there are high amounts of His the utilization of SPPS may for instance result in low chiral purity due to racemization, which is not suitable in the industry of biotherapeutics.
The problems related to Asp are not only dependent on frequency, since the neighbouring amino acids will be the determining factor in the most severe problem connected to Asp;
aspartimide formation. If the neighbouring amino acids offer low steric hindrance, the risk of aspartimide formation rises which consequently may lead to termination of the peptide.
Even though the majority of the amino acids do not correspond to problems so severe that SPPS is not eligible, some still have problems that are important to keep in mind if utilizing SPPS. For instance, if an amino acid occurs in high frequency as for example in a
homooligopeptide, the risk of an increase in unwanted side reactions rises. In the diagrams below, figure 1a and 1b, a summary of how many problems the amino acids are involved in is displayed followed by a conclusion of the ratio of which method of production that is
recommended.
Figure 1a. An overall view of the number of identified problems per amino acid. The amino acids are on the x-axis and the number of problems identified are on the y-axis. Amino acids marked with * have the recommendation to consider production via recombinant production. Amino acids marked
with ** have a position specific recommendation.
Figure 1b. Number of amino acids per type of recommendation. 15 amino acids have the recommendation that SPPS should work. Two amino acids have the recommendation to consider
recombinant production. Three amino acids have the recommendation that you should consider recombinant production if they are positioned in the C-terminal.
If one pays attention to the features described above when choosing the method of production for a therapeutic peptide, we believe that an informed and fair decision can be made.
4 Foundation of the support; the methods and potential manufacturing problems
The decision support in its entirety was presented in chapter 3, but in order to fully understand the reasoning displayed there we will now go into more detail and explain the problems further.
4.1 The two methods - environmental aspects and general problems Although SPPS in this thesis is viewed as the primary option, the most fundamental part of the method selection process is to compare the methods themselves. The reason for this is that they may differ tremendously in for example environmental impact, reproducibility and amount of general problems. This chapter will go more into detail and showcase different aspects of SPPS and recombinant protein production. Here general problems will be presented, meaning problems that can occur during the process for any peptide, with no regard for specific amino acids or the sequence. For description of the methods and their purification processes, see appendix chapter 9.
4.1.1 SPPS
4.1.1.1 General problems of SPPS
The most common, general problems in SPPS today are problems connected to solubility.
Another problem of this kind is connected to the synthesis of longer peptides. There are some suitable candidates in the search for a solution to this problem, where one example is native chemical ligation of peptides, abbreviated NCL. These two problems can be derived from properties of the peptide and will thus not be investigated in this particular section of the report. A problem which, in contrast to the problems mentioned above, is general for SPPS and therefore described here is deletion and insertion into the sequences as well as
termination of the elongation of the peptides that are being synthesized.
Deletion sequences, or failure sequences as they are also called, are peptides that are missing one or several of the amino acids in the target peptide (D’Hondt et al. 2014). The deletion of amino acids is due to incomplete removal of protecting groups or insufficient activation of an incoming amino acid, leading to the incoming amino acid not being able to incorporate into the resin bound, growing peptide (Sanz-Nebot et al. 1999, D’Hondt et al. 2014). Inefficient loading of resin may also lead to C-terminal deletion. In most cases, deprotection is effective but for longer peptides, incomplete deprotection leads to deletion peptide impurities even at high concentrations of the effective removal group piperidine and repeated trials (Luna et al.
2016 s. 2, Wu L 2019). Due to the deletion sequence being similar to the target peptide (Sanz-Nebot et al. 1999), the formation leads to purification problems (Benoiton 2016).
To prevent the formation of deletion sequences, a Kaiser test and capping can be applied. By doing a Kaiser test during the peptide synthesis, the completeness of the coupling and cleavage will be known. If the test shows a positive result, in other words that the reactions are completed, the synthesis can move to the next cycle. If the test shows a negative value the capping procedure will be applied, terminating the failed sequences into truncated sequences.
This is done by permanently blocking the N-terminus of the free and unreacted amino acid during the whole synthesis. Alternatively, the N-terminus can be acylated (Verma et al.
2020). The resulting, truncated peptides are therefore shorter, less similar to the targeted peptide and much easier to separate (Wu LC et al. 2017). This method is not a risk free approach and can still give deletion impurities (Benoiton 2016).
In Fmoc-SPPS, extra Fmoc is added during coupling in order to have an efficient reaction.
Fmoc stands for 9-Fluorenylmethoxycarbonyl. However, if the washing is not completed, the amino acid with the Fmoc attached can acylate an amide group in the peptide, resulting in incorporation of this amino acid in the next coupling reactions (D’Hondt et al. 2014, Wu L 2019). Insertion sequences are much more common in classical solution-phase peptide
synthesis (CSPS) (Eggen et al. 2005), but the risk will increase in SPPS if there is easy access to the position of the acylation. In other words, the risk increases if there is a lack of bulky side chains (D’Hondt et al. 2014). The insertion mechanism can be avoided by washing to remove the piperidine, decreasing the risk of premature deprotection of the amino acid (Wu L 2019).
Termination of the elongation of the peptide, leading to a truncation, may be caused by several different side reactions. Several cyclization reactions, such as aspartimide (see 4.2.3.3) and DKP formation (see 4.2.5) can cause the termination of a peptide. A truncation impurity will affect the yield of the desired product, but due to the difference in mass from the desired product, it is possible to purify the peptide. This can be done with the capping method, as previously described.
4.1.1.2 Environmental aspects and safety
An important topic of today’s industrial operations is the environmental aspect, and SPPS unfortunately suffers from a few drawbacks in this particular area. In an article by Rasmussen et al. from 2019 the aim is to investigate eco-friendly alternatives in peptide chemistry
involved in the therapeutics industry (Rasmussen et al. 2019). They describe attainable ways of how to decrease the amount of manufacturing steps and also look into some greener replacements of hazardous chemicals. The authors of the article state that decreasing the amount of manufacturing steps will likely save energy, and this is good from an
environmental point of view.
Dimethylformamide (DMF) is a chemical that is commonly used as a solvent in SPPS for the coupling steps and deprotection of the N-terminal protecting group (Chan & White 1999).
However, this chemical might not be the best option as DMF is related to several health dangers for humans, including cardiac damage and dysfunctions in liver and kidneys (Hu et
al. 2020). The search for greener and safer alternatives for DMF has increased over the years and will now be further explored.
Rasmussen et al. found suitable exchanges for DMF, and in their article they mention for example gamma-valerolactone (GVL) and 2-methyltetrahydrofuran (2-MeTHF) (Rasmussen et al. 2019). Both of these potential replacement solvents were also mentioned in an article about sustainability challenges in peptide synthesis by Isidro-Llobet et al. from 2009, where specific examples of SPPS using these chemicals also are described (Isidro-Llobet et al.
2009). Even though frequently mentioned, GVL in particular might not be a good option since it, according to an article from 2020 by Martin and co-workers, leads to unacceptably low chiral purity which makes it inappropriate for the biotherapeutics industry (Martin et al.
2020). In the same article by Martin et al. it was also stated that propylene carbonate (PC) performed well as solvent and decreased side reactions and increased chiral purity - but it was also said that this particular solvent might not be a good replacement for DMF in industrial large scale synthesis.
In Rasmussens and co-workers elaborated examples of SPPS processes with greener solvents, they more than once mention ethyl acetate (EtOAc) mixed with other chemicals such as N-butylpyrrolidone (NBP) or acetonitrile (MeCN) (Rasmussen et al. 2019). Furthermore, they state that the elaborated examples of greener versions of SPPS were completed with not so much of a rise in the expenses of raw materials. This is a very positive result since the economic aspect also is important in the industry of biotherapeutics. Something that is important to mention is that DMF is a commonly used solvent for a reason; EtOAc might be cost friendly and obtainable from renewable sources, but it does not work as well as DMF according to a written work (Pawlas et al.). To work around this issue, the authors display a few other alternatives to EtOAc, but also find that EtOAc in combination with DMF
enhances coupling rates and decreases some unwanted reactions. Therefore, they state that a greener chemical that could replace DMF and achieve the same results as the combination of EtOAc and DMF would be good. Their suggestions regarding a chemical of this kind is N,N'-dimethyl propylene urea (DMPU) and dimethyl sulfoxide (DMSO). In a previously mentioned article by Martin et al. they discuss many greener options for DMF but in
accordance with Pawlas et al. also says that mixtures of different solvents could be a potential solution (Martin et al. 2020, Pawlas et al.). For example, Martin and his colleagues mention that EtOAc mixed with PC resulted in a slightly higher yield than EtOAc on its own.
Furthermore, they describe an example of SPPS using mixtures of EtOAc with either NBP or MeCN depending on the present step in the process, which led to good results. However, they later state, in accordance with Pawlas et al., that DMSO might be a better alternative than NBP and MeCN. Also, Martin et al. in a comparison for costs of the synthesis found that if the solvents were recycled, the cheapest alternative per amino acid cycle would be
DMSO/EtOAc diluted to a 1:9 ratio. This specific mixture was cheaper than both neat DMF and NBP/EtOAc.
authors state that this degradation leads to dimethylamine (DMA) impurities, which in turn can cause troubles when utilizing SPPS since it can impact the deprotection of Fmoc groups.
In the article they also mention possible solutions to be able to regenerate the DMF but this consequently leads to an increase in preprocessing steps which is contradictory to what Rasmussen et al. tried to achieve.
Another way to greenify SPPS is through recycled resins, which is mentioned in a previously noted article by Isidro-Llobet et al. (Isidro-Llobet et al. 2009). In the article, the authors discuss that resin recycling still is a quite small field of research since the recycling most commonly is applied to only one type of resin; the 2-chlorotrityl resin. Furthermore, they state that the greener resins of today are not ideal in some important areas. Resins based on for example polyethylene glycol tend to swell more than other resins and thus a greater amount of solvent is needed. They furthermore discuss that some of the greener resins vary in how compatible they are with green solvents. Other environmentally friendly properties the authors mention regarding resins are biodegradability and the possibility to obtain the resin’s start-up components from renewable sources.
Based on the information above, it is important to mention that even if a chemical may be hazardous and not considered “green”, they can still be used for manufacturing purposes if handled appropriately. If the handling of the chemical is done correctly, there should not be any safety issues for either employees or the environment. An example of this is that MeCN is not dangerous if kept in closed systems and disposed of properly, but if not handled
correctly it is an environmental burden. One thing that is not to be overlooked on this topic is however that careful and proper handling often leads to an increase of expenses, and this is something that companies in general take closely into consideration. In conclusion, it is quite safe to say that this issue is a matter of finding a balance between the use of well-performing but quite hazardous chemicals and the responsibility of adequate handling. Because certainly, one will not refrain from manufacturing medicine that will help a lot of people just because the production process may potentially be non-optimal in some areas.
4.1.2 Recombinant protein production
4.1.2.1 General problems of recombinant protein production
E. coli is the most applied cell host for recombinant proteins, making up approximately 30 % of all approved recombinant proteins for biotherapeutics (Huang et al. 2012). There are several reasons for this including low cost, fast growth, high yield and extensive knowledge of the genome of the species (Correa & Oppezzo 2015). Several protocols and strategies for production of heterologous proteins in E. coli are available (Rosano & Ceccarelli 2014). The lack of need for post-translational modifications and E. coli’s extensive use, limits this study's research to only include production of polypeptides using E. coli as the expression host.
There is no one-fits-all procedure for production of recombinant proteins. In order to maximize the yield of expressed and purified active protein, the expression system and
growth condition needs to be optimized for the specific targeted protein (Ferrer-Miralles et al.
2015). There are two main challenges for recombinant production in E. coli, namely limited heterologous gene expression and aggregation (Ferrer-Miralles et al. 2015). This is in agreement with Overton (2014), claiming that even though the protein is successfully translated, there is no guarantee that it will fold correctly. What follows are common problems encountered in recombinant protein production using E. coli hosts.
Expression vectors and promoters
The first problem one might encounter is the design of the expression vector, including what promoter and selection marker to use, and if any tags are needed in downstream steps
(Rosano & Ceccarelli 2014). There are many available expression systems for use. Usually a system contains a strong promoter that is easily inducible (Tucci et al. 2016). The promoter is regulated with an inducer, since it allows controlled expression and the ability to separate between cell growth and protein production in downstream steps (Overton 2014). A common promoter is the pET system (Rosano & Ceccarelli 2014). The pET system uses medium copy number plasmids, where the target genes are placed downstream of a T7 promoter and T7 RNA polymerase is supplied in trans (Baneyx 1999). The pET plasmid is not effectively activated by E. coli RNA polymerase and requires addition of the inducer IPTG (Overton 2014). A disadvantage with using strong promoters systems is that they can lead to aggregation of the peptide (Baneyx 1999).
Plasmid instability
After successful transformation of the vector into a plasmid, the problem of plasmid instability and loss begins. If a high copy number is used, plasmid loss is more likely since they are a heavier metabolic burden (Overton 2014). The balance between a high enough copy number to generate as high yield as possible against a low copy number to reduce metabolic burden needs to be assessed for each case (Overton 2014). To ensure that bacteria containing the correct plasmid are grown, antibiotic resistance markers are used and cells are grown in medium with the antibiotic to kill off the cells without the plasmid (Baneyx 1999).
However, the author points out that using antibiotics can lead to contamination of the product unacceptable for use in medicals. The spread of antibiotic resistance are concerns for large scale cultures (Rosano & Ceccarelli 2014). Therefore, alternative approaches have been developed in order to avoid the use of antibiotics. A method to ensure that plasmid-free cells are unable to grow is by using vectors that carry some gene leading to cell death upon plasmid loss, according to the authors. Other metabolic selection markers used together with an appropriate host strain are also being implemented (Overton 2014).
Host strain and growth medium
Further problems involve which host strain and growth medium to use. This is dependent on the expression plasmid system and regulatory requirements (Overton 2014). There are many modified E. coli strains adapted for expression vectors and other challenges of the production process (Tucci et al. 2016). Strains are often specialized for specific situations (Rosano &
(Tucci et al. 2016). In regard to the growth medium, it needs to contain all nutrients required for rapid cell growth (Overton 2014).
mRNA stability
The half-life of E. coli mRNAs is relatively short, ranging between 30s to 20 min (Baneyx 1999). This can cause mRNA degradation. Stable secondary structures in the 5′ UTR and removal of features triggering degradation can increase mRNA stability (Overton 2014), (Baneyx 1999).
Codon bias
Organisms of eukaryotic and prokaryotic species have different codon usage, which causes problems when expressing heterologous proteins in E. coli (Tucci et al. 2016). Those problems are frameshifts, misreading, truncated peptides and low expression. For example, the human Arg codons AGA and AGG are rarely found in E. coli genes (Baneyx 1999).
Overexpression of genes with high contents of rare Arg codons may result in defective synthesis of the corresponding protein, since they lack the corresponding tRNAs (Overton 2014). Overton and Tucci et al. suggest two solutions to this problem; optimization of the recombinant gene to contain codons common to E. coli or to use strains that overexpress the corresponding tRNAs. Successful codon optimization can increase expression level
significantly to result in higher yield and lower cost (Tucci et al. 2016). Nevertheless, this is not true for all cases, as it has been shown that codon optimization for an E. coli strain can result in aggregation when changing the rare codons (Correa & Oppezzo 2015).
Solubility and Inclusion body formation
Biological activity is linked to solubility (Ferrer-Miralles et al. 2015). However, many recombinant proteins are unable to be expressed as soluble (Habibi et al. 2014).
Ferrer-Miralles et al. explain that, when overexpressing a recombinant gene in E. coli, achieving maximum yield may be hindered by the folding machinery being saturated. This will lead to misfolding of the protein and cause aggregation. Protein aggregates in bacteria are known as Inclusion bodies (IBs) (Ferrer-Miralles et al. 2015), and are composed of unfolded and partially folded proteins (Overton 2014).
There are several solutions for solving the solubility problem and misfolding, categorized into modifying host strain, expression vector or the growth condition (Overton 2014). As previously mentioned, recombinant proteins misfold due to a higher rate of producing polypeptides than the rate of the heat shock response, meaning the production of Heat shock proteins (Hsps) (Ferrer-Miralles et al. 2015). Hsps consist of proteins involved in folding machinery, such as chaperons. The authors suggest the solution to introduce a higher amount of Hsps to the system. However, a specific limiting Hsps might still be lacking, leading to various successes for this approach, according to the article. Decreasing the growth temperature has been shown to improve solubility and activity (Arya et al. 2015).
Hydrophobic interactions are favored in high temperatures and Hsps expression is induced in lower temperatures (Ferrer-Miralles et al. 2015).
Another solution is to attach a signal sequence on the N-terminus to transport the synthesized polypeptide from the cytoplasm into the periplasm (Overton 2014). The article states that the cytoplasm contains more proteases, making it an unfavorable environment for protein
folding. Transporting the polypeptide to the periplasm is especially important if it requires several disulfide bonds (Tucci et al. 2016). This is because the cytoplasm of E. coli contains several disulfide bond reductases, which might cause disulfide bonds formed between two Cys residues to be reduced. Also, cell lysis has to be performed when extracting the protein, mixing it with other cellular components (Overton 2014). However, if produced in periplasm, the protein is easier released by removing the outer membrane. Here the target protein will be mixed with fewer other components, meaning less purification steps needed, according to the author.
Nonetheless, formation of inclusion bodies is not always a bad thing. IBs can protect proteins against proteolysis and permits accumulation of proteins toxic to the host cell (Wingfield 2015). Inclusion bodies mostly consist of the target protein of interest. Hence, expressing the protein as inclusion bodies can aid in purification, as it can reduce the need for extensive chromatographic purification steps (Singh et al. 2015).
Fusion Tags
There are three types of fusion tags used to improve solubility and aid in purification or downstream processes; co-expression, fusion partner and affinity tags (Correa & Oppezzo 2015). Additionally, tags allow for detection of the protein along the expression and
purification process (Rosano & Ceccarelli 2014). One way of assuring a high enough content of Hsps is to co-express them along with the target sequence (Correa & Oppezzo 2015). By adding a fusion partner, usually some extremely soluble protein that drives the overall expression to soluble form, the yield of target protein can be improved (Correa & Oppezzo 2015). Affinity tags can be used to protect the peptide and improve purification, by adding a short sequence at the N- or C-terminus which is recognized by certain molecules during purification steps (Correa & Oppezzo 2015). Usually, the fusion tags later need to be removed during downstream purification, since they can inhibit protein activity (Rosano &
Ceccarelli 2014). Tag removal is performed by adding a cleavage site between the tag and fusion partner, meaning the protein of interest, for cleavage by a specific protease (Tucci et al. 2016).
4.1.2.2 Environmental aspects
Recombinant protein production is seen as more environmentally friendly than SPPS
(Funfrock 2019). In general recombinant production is a bio-safe process (Tucci et al. 2016).
This is in agreement with the undertaken analysis, where no hazardous components were identified. The following environment analysis was performed by looking at a typical
protocol for recombinant production in E. coli, and examining environmental precautions for the mentioned chemicals and cultivation additives using safety data sheets from
tuberosum Epoxide hydrolase I Purification and characterization” 2020 by the Department of Chemistry - BMC (Uppsala University).
Chemicals
NaH2PO4(Disodium phosphate) Imidazole
NaH2PO4is not considered to be toxic or to accumulate in organisms (MSDS No. 255793 Sigma-Aldrich Sweden AB 2019). Waste should be disposed of according to local
requirements. Furthermore, the safety data sheet states that the environmental protection measures are to avoid leakage to sewage.
Imidazole is not classified as hazardous to aquatic environments and the substance is biodegradable (Imidazole No. X998 Carl Roth 2020). Neither does it accumulate in organisms to a high degree. Waste should be disposed of as hazardous waste according to regional regulations, according to the safety data sheet.
Cultivation additives Ampicillin
Isopropyl thiogalactoside (IPTG)
Ampicillin does not accumulate significantly in organisms (MSDS No. A9393 Sigma-Aldrich Sweden AB 2019). Environmental protection measures are to avoid leakage to sewage and ground water, according to the safety data sheet. Furthermore, waste should be handled according to regional requirements to companies with permission for waste management.
Isopropyl thiogalactoside (IPTG) is not classified as hazardous to aquatic environments (IPTG No. 2316 Carl Roth 2020). Also, it does not accumulate significantly in organisms.
Environmental protection measures are to avoid leakage to sewage and ground water, according to the safety data sheet. Waste should be disposed of according to local requirements (IPTG No. 2316 Carl Roth 2020).
4.2 Manufacturing problems in SPPS due to peptide properties Besides the methods themselves, something that might have an impact on the selection process and what makes the decision case-specific is the peptide itself. What properties of the targeted peptide will be too troublesome for SPPS and thus make recombinant protein
production an option to consider? In the coming pages, we will analyze these properties and display their impact on for example efficiency, purification, yield and result when utilizing the two methods.
4.2.1 The length of the peptide is a limiting factor
An issue in chemical synthesis of peptides is that with increasing peptide length chemical synthesis becomes increasingly difficult (Kulkarni et al. 2018). Longer sequences tend to have problems with aggregation and steric crowding, leading to truncation, racemization and unwanted side products (Kulkarni et al. 2018). Additionally, incomplete deprotection and coupling becomes more pronounced with increasing length (Al-Warhi et al. 2012).
Accumulation of by-products, including truncated sequences and side products, leads to low purity and yield (Kulkarni et al. 2018).
The use of SPPS in industrial settings produces acceptable yields for peptides up to 40 or 50 amino acids, although some problems may arise when the peptide length exceeds 20 amino acids (Funfrock 2019). In these cases, recombinant production can be used as an alternative synthesis method (Rodríguez et al. 2014). Another potential solution making it possible to synthesize longer sequences is the use of NCL, or native chemical ligation (Mueller et al.
2020). NCL is a method where a native peptide bond is formed between two peptide fragments, where one of the fragments has a Cys residue with a peptide thioester at the
N-terminus (Kulkarni et al. 2018). This method leads to a higher yield and purity of the target protein, compared to regular SPPS of longer sequences (Lindgren et al. 2012).
4.2.2 Aggregation causes problems during synthesis and purification
4.2.2.1 Aggregation and its problems
Aggregation is a problem that can occur during any step of the production chain (Mueller et al. 2020). Due to inter- or intramolecular interactions peptides precipitate rather than stay in solution, see figure 2 (Paradís-Bas et al. 2016, Kent). In contrast to incomplete coupling or deprotection, aggregation problems cannot always be solved by repeated or prolonged reaction times (Amblard et al. 2006). The reason for this is that aggregation leads to reagents being unable to access the N-terminal amino acid meaning the synthesis cannot continue as normal. This reduces coupling and deprotection yields significantly (Isidro-Llobet et al.
2009). Apart from the issues during the synthesis the aggregates can cause problems during purification and characterization of the peptide (Mueller et al. 2020). The capability to induce structural associations is sequence dependent (Paradís-Bas et al. 2016), although sequences containing high proportions of certain amino acids have a tendency of being difficult to synthesize (Amblard et al. 2006).
Figure 2. Aggregation of a peptide, where the yellow amino acids are hydrophobic and the blue amino acids are hydrophilic.
4.2.2.2 Sequences that can aggregate
In a review article Paradís-Bas et al. (2016) discuss different types of sequences that can aggregate, either during the synthesis or after the peptide is removed from the solid surface.
Sequences that aggregate during synthesis are called “difficult sequences” while sequences that aggregate after removal from the solid support are called self-assembling peptides, or SAPs (Paradís-Bas et al. 2016). It has been found that “difficult sequences” typically contain high numbers of β-branched hydrophobic amino acids like Leu, Ile, Val and Phe (Mueller et al. 2020). This causes the sequences to have a tendency to aggregate due to formation of β-sheets or α-helices, leading to lower solubility in aqueous and organic solvents.
Furthermore, Gly is known to contribute to induction of β-sheet packing. Peptides that form α-helices are generally soluble in water while peptides that form β-sheets are insoluble due to their ability to interact by hydrophobic interactions (Paradís-Bas et al. 2016).
A sequence type known to aggregate either during synthesis or after removal from the surface mentioned in the review article are homooligopeptides, or oligopeptides containing only one type of amino acid (Paradís-Bas et al. 2016). When these comprise hydrophobic amino acids they show a high tendency to aggregate, leading to difficulties. While the only hydrophobic amino acids mentioned by the article are Ala, Gly, Ile, Leu and Val, the full list of
hydrophobic amino acids should also include Phe, Met, Trp and Pro (Aftabuddin & Kundu 2007). Though hydrophobic, a homooligopeptide of Pro should not cause significant problems with aggregation since Pro lacks a hydrogen atom on their α-amino group. This is however not explicitly backed up by the literature. The assembly of homooligopeptides is driven by hydrogen bond formation between amides on the backbone of the peptide, see figure 3 (Paradís-Bas et al. 2016). For this reason, the aggregation may be prevented using backbone protecting groups.
Figure 3. Hydrogen bond formation between amides on the backbone of the peptide, where R = side chain, R’ = N-terminus or continuation of the peptide, and R” = C-terminus or continuation of the peptide.
Another type of sequence discussed by Paradís-Bas et al. that may pose problems with aggregation is self-assembling peptides, or SAPs. These peptides typically assemble
spontaneously in solution after the peptide is removed from the solid support, although some may form β-sheet interactions during synthesis. These peptides mainly cause problems during purification as many can be synthesised without difficulties. SAPs consist of a hydrophobic and a hydrophilic part, where the hydrophobic part consists of more than three non-polar amino acids and the hydrophilic part consists of one or two charged or polar amino acids.
Depending on the charges in the hydrophilic part, SAPs can be divided into amphiphilic or ionic SAPs. The amphiphilic SAPs hold only one type of charge while ionic SAPs hold both positive and negative charges. This difference leads to different means of aggregation. Ionic SAPs form β-sheets while amphiphilic SAPs form micelles, see figure 4.
Figure 4. Micelle formation of amphiphilic self-assembling peptides. Here the C-terminus has the charged amino acid, but it could also be located at the N-terminus. In the figure R is a hydrophobic side chain and R’ is
the continuation of the peptide.
4.2.2.3 Solutions to aggregation
Solubility is of key importance when it comes to synthesis, purification and handling of peptides (Abdel-Aal et al. 2020) and the main issue with aggregating sequences (Paradís-Bas et al. 2016). While there are no unique protocols for handling aggregating hydrophobic peptides, there are a couple different modifications you can do to improve the production (Mueller et al. 2020). These can be divided into internal and external modification. Internal modifications are modifications to the amino acids while external modifications are
modifications to the peptides’ environment.
Two examples of internal modifications are solubilizing tags and backbone protecting groups (Mueller et al. 2020). Solubilizing tags usually consist of multiple hydrophilic amino acids that are attached to the N-terminus, C-terminus or a side chain of a hydrophobic sequence (Mueller et al. 2020). Backbone protecting groups may help with solubility where hydrogen bond formation between backbones of peptides is the cause of the aggregation. These backbone protecting groups work by substituting the nitrogen atom of the peptide bond, preventing the formation of interchain associations (Abdel-Aal et al. 2020). This leads to better synthesis of “difficult peptides” and increased peptide solubility.
An example of a backbone protecting group is pseudoproline. Oxazolidine, or pseudoproline, is usually derived from Ser, Thr or Cys (Haack & Mutter 1992). They work by linking its side chains to the N-terminal nitrogen atoms within the pseudoproline residue. This disrupts the backbone of the peptide and its secondary structure. The originally intended secondary structure can be restored by cleaving the bond under standard deprotection with
Trifluoroacetic acid (TFA) (Abedini & Raleigh 2005). Pro itself has shown to help with some of these issues in difficult sequences, thanks to its lack of a hydrogen atom on their α-amino
