1 Introduction

An emerging crisis all over the world is a large number of antimicrobial drugs becoming ineffective against most microbes due to emergence of resistance. It has been observed that most of the microbes are exhibiting insensitivity for more than one drug, a condition termed as Multi Drug Resistance (MDR). Studies have concluded that occurrence of resistance, mainly in bacteria, is mostly because of one or combination of the two mechanisms. First, expression of multiple genes responsible for resistance to a single or multiple drugs in a cell. Second, increased activity or overexpression of efflux proteins (or exporters) [1,2,3]. The efflux proteins are classified into six major classes namely small multidrug resistance (SMR), proteobacterial antimicrobial compound efflux (PACE), major facilitator superfamily (MFS), multidrug and toxin compound extrusion (MATE), resistance nodulation division (RND) and ATP binding cassette (ABC) superfamily. Of these, ABC superfamily is the largest [4]. In order to solve the resistance issues, ABC exporters have always been the desired target protein. Great efforts have been taken in either inhibiting or modulating them. But the major obstacles in their success was their high level of conservedness resulting in toxicity and gradual ineffectiveness of inhibitors due to rapid rate of mutations [5, 6]. With the failure of first and second line of drugs against Multi drug resistant microbes, due to either emergence of resistance or adverse side effects, there has been a constant search for new therapeutics. Antimicrobial peptides (AMPs) provide a ray of hope as an alternative strategy. Being specific in nature, AMPs withhold advantage over small molecules by minimizing toxicity and increased specificity [7]. Herein, we have focused on ABC superfamily of efflux proteins with potent role of antimicrobial peptides for targeting them in combating with MDR microbes.

2 ABC Exporters

ABC Exporters are a wide variety of ATP dependent proteins that confer translocation from cytoplasmic leaflet to exoplasmic leaflet of cell membrane, thus, also known as Floppases [8, 9]. They are found in all living organisms and play key roles in many biological processes. They utilize energy obtained from ATP hydrolysis to efflux a diverse range of molecules like amino acids, sugars, ions, small organic or inorganic compounds, peptides or proteins, hydrophobic drugs and a variety of toxins. The ABC exporters are often associated with MDR and several human diseases. The structural composition of ABC exporters includes 2 transmembrane domains (TMDs) each having six transmembrane helices and 2 nucleotide binding domains (NBDs). ABC transporter class is a highly conserved class of proteins categorised further in 7 subfamilies (ABCA to ABCG). In humans, 51 ABC genes are known while in L. major 40 members have been reported till date [10, 11]. The subfamilies of Human ABC transporters as classified by HUGO Gene Nomenclature Committee (HGNC) are mentioned below [12].

2.1 Classification

2.1.1 ABCA

The subfamily A of ABC family also known as ABC1 constitutes of few of the largest proteins amongst ABC transporters. It consists of 14 genes which are further sub grouped into 7 and 5 genes. The first sub group is ABCA1 like genes consisting of 50 exons. These sub groups includes ABCA1, ABCA2, ABCA3, ABCA4, ABCA7, ABCA12 and ABCA13. Mutations in these genes are associated with various genetic diseases. For instance, mutations in ABCA1 cause Tangier’s disease. Mutations in ABCA4 may lead to Stargardt disease, age-related macular degeneration and retinitis pigmentosa (Table 1). The other sub group consists of ABCA5, ABCA6, ABCA8, ABCA9 and ABCA10 genes consisting of 37–38 exons [13, 14].

Table 1 Human ABC Transporter genes, functions, associated diseases and their tissue location
Full size table

2.1.2 ABCB

The subfamily B is the only human subfamily of ABC Transporters having half transporters as well as full transporters. If all the domains of an ABC exporter (i.e. 2TMDs and 2 NBDs) are present as a single polypeptide chain, then they are referred as full transporters. But if 2TMDs and 2NBDs are represented by more than one polypeptide chains, they are referred as half transporters. 1TMD and 1NBD comprises in a single polypeptide and such two polypeptides dimerize to form a complete functional unit. The dimers may either be homo- or hetero- in nature [11, 18]. The schematic arrangement of full transporters and half transporters are represented in Fig. 1. It constitutes of 11 unique genes of which, 4 are full transporters while the rest 7 are half transporters. Many members of this subfamily extrudes multiple drugs and imparts multi drug resistance, hence ABCB subfamily is aliased as MDR. ABCB1, commonly known as P-gp or P-glycoprotein or MDR1, is one of the major exporters responsible for MDR. P-gp is the first Human ABC to be cloned and also one of the best characterized and studied protein. Apart from MDR, mutations in ABCB genes are associated with diseases like spondylitis, diabetes type 2, coeliac disease, X-linked sideroblastic anaemia, and several cholestatic liver diseases (Table 1) [13, 14]. The ABCB2 and ABCB3 are half transporters and form a heterodimer to transport peptides into ER. These two transporters are associated with antigen processing (TAP) protein complex, hence, also named as TAP1 and TAP2 respectively [19].

Fig. 1
figure 1

The schematic arrangement of transmembrane domain (TMD) and nucleotide binding domain (NBD) in ABC exporters. In full transporters, 2TMDs and 2NBDs are expressed as single polypeptide. While in half transporters, more than one polypeptide are present either in homodimer or in heterodimer to form functional unit. The arrangement of domains is in head to tail fashion, i.e., C-terminal of first domain connects with N-terminal of the second

Full size image

2.1.3 ABCC

The subfamily C of ABC family of exporters constitutes of 13 gene members of which nine are involved in multidrug resistance. This family is aliased as Multidrug Resistant Proteins, MRPs. The members of this subfamily are involved in diverse functional activities ranging from extrusion of small molecules, drugs and toxins to ions and nucleoside transport as well as acting as receptors. The ABCC7 (CFTR, Cystic fibrosis transmembrane conductance regulator) functions as chloride ion channel and mutations in this gene results in cystic fibrosis, a genetic autosomal recessive disease. The sulfonylurea receptor (SUR) genes, ABCC8 and ABCC9, are associated with hypo- and hyper-glycaemia, dilated cardiomyopathy, familial atrial fibrillation and hypertrichotic osteochondrodysplasia. The other diseases associated with mutations in genes of ABCC subfamily are Dubin-Johnson Syndrome, Pseudoxanthoma Elasticum (PXE), diabetes mellitus 2, Borna disease, congenital bilateral aplasia, etc. (Table 1) [14, 19].

2.1.4 ABCD

ABCD subfamily consists of 4 genes, ABCD1-4. They encode half transporters and are expressed only in peroxisomes. They may function as either homodimer or heterodimer. Mutations in genes ABCD1 and ABCD2 are associated with X-linked adrenoleukodystrophy (ALD). Thus, this subfamily is also called as ALD. The other diseases associated are Bile Acid Synthesis Defect, Congenital and Zellweger syndrome, Methyl malonic Aciduria and Homocystinuria, cblj type (MAHCJ) and Disorders Of Intracellular Cobalamin Metabolism (Table 1) [14, 19].

2.1.5 ABCE

ABCE subfamily has only one member, ABCE1, also known as organic anion binding protein (OABP). ABCE1 is an ABC transporter protein with ATP-binding domain but no transmembrane domain and recognizes oligoadenylate for binding and help in promoting interferon activity during viral infections. Any defects in this gene results in Noonam Syndrome (Table 1) [14, 19].

2.1.6 ABCF

ABCF subfamily is also known as GCN20 after the most characterized GCN20 gene of S. cerevisiae. Similar to ABCE subfamily, it also has only ATP-binding domain and lacks transmembrane domains. It consists of 3 members, ABCF1-3. These genes play role in inflammatory processes as they were found to be upregulated by tumour necrosis factor-α. Autoimmune Pancreatitis, Hemophagocytic Lymphohistiocytosis and Intestinal Tuberculosis are the diseases associated with ABCF subfamily (Table 1) [14, 19].

2.1.7 ABCG

ABCG subfamily or White consists of 5 genes that encode half transporters. The domain orientation is opposite to other ABC transporters with ATP-binding site at N’ and trans-membrane domain at C’ end, and hence is also known as ‘reverse half transporters’. ABCG1 is involved in cholesterol transport and any defects are linked with cardio-metabolic diseases. ABCG5 and ABCG8 helps in sterol transport and are linked with sitosterolemia disease. ABCG2 is the most studied member and its function is efflux of toxins and drugs thereby causing multi drug resistance (Table 1) [14, 19].

2.2 General Architecture

ABC Transporters are one of the largest families of known proteins that are highly conserved and share common architecture which includes 2 TMDs and 2 NBDs arranged in a head to tail fashion (Fig. 1). Each TMD consist of 6 helices and are capable of recognizing a wide variety of substrates. TMDs are arranged in the form of two wings such that the helices TM1 and TM2 of a subunit and TM3-6 of other domain form one wing in a domain swapped arrangement. ABC transporter proteins transport a wide variety of substrates through TMDs and hence compared to NBDs, TMDs are much more diverse. Being diverse in nature is supported by varying number of trans-membrane helices in different subfamilies. The other domains of ABC proteins, i.e. NBDs, are more conserved in nature. NBDs consist of a larger and a smaller domain. Larger domain, also known as catalytic sub domain, consist of 2 β sheets and 6 α helices. The Walker A motif (also called P-loop; GXXGXGKS/T) and Walker B motif (ΦΦΦΦD) are part of this domain. The Φ in Walker B motif represents any hydrophobic residue. The smaller helical domain has 3–4 helices and the conserved ABC signature motif or C-motif (LSGGQ). Apart from this, ABC domain has a glutamine residue also called Q-loop and is thought to be connecting TMD and NBD. Another conserved region known as H motif or switch region having highly conserved histidine residue play an important role in the interaction of ATP with ABC domain. The Walker A motif of NBD1 and signature motif of NBD2 interacts with ϒ-Phosphate of ATP and vice versa, thereby sandwiching it between two NBDs [11, 20, 21]. The interaction of ATP with NBDs in ABCB1 are shown in Fig. 2.

Fig. 2
figure 2

The interacting partners of ATP with NBD of ABC Transporter Protein ABCB1

Full size image

2.3 Mechanism of Action

The transport of substrate through ABC transporter proteins requires binding of two ATP molecules followed by their hydrolysis. Since, an ATP molecule interacts with both NBDs, thus binding of ATP brings both NBDs closer to each other in a dimeric form. This is known as closed-dimer conformation. The 12 transmembrane helices (TMH) are arranged in a manner to form a large cavity which can either be in open-inwards conformation (open towards cytosol) or in open-outwards conformation. The rearrangement between these two conformations is guided by ATP hydrolysis. The open-inward conformation has a high affinity binding site for substrate. Upon ATP binding, the NBDs dimerizes and results in rearrangement of membrane cavity to form open-outward conformation. This conformation has low affinity for substrate. Thus, ATP hydrolysis results in the release of substrate and conformational change back to open-inward form. In this manner, they alternate between these two conformations, hence called alternating access model. Apart from this, two models have been proposed explaining the mechanism of catalytic cycle of ABC exporters: (i) The ATP Switch Model and (ii) The Constant Contact Model.

As per ATP Switch Model, the two NBDs fully dissociate from each other at the end of catalytic cycle. The binding of ATPs followed by dimerization of NBDs is the power stroke driving the conformational changes from open-inwards to open-outwards form. Due to this rearrangement, the high affinity binding site converts into low affinity resulting in the release of substrate in the extracellular space. Finally, hydrolysis of either or both ATPs occurs which disrupts the closed NBD dimer and results the conformation back to open-inwards form. This is the most accepted model for the catalytic cycle mechanism. This model is being supported by various structures obtained in inward-facing form like ABCB1, ABCC7, Sav1866, MsbA, MetNI, MalFGK2, ModBC, etc. [22,23,24].

According to the Constant Contact Model, the two NBDs do not dissociate from each other during catalysis, rather each ATP hydrolyses and binding site open alternately. This means that in every cycle, one NBD will hydrolyse a bound ATP and another NBD will release ADP and Pi hydrolysed in the previous cycle. Thus, the NBDs always exists in a dimer form. Example of this model is explained by heterodimeric ABC transporter TM287/TM288 from T. maritima [25]. Similar activity was also observed in ABCC7 as well as ABCB1 suggesting that there may be more than one mechanism for transport of substrates in a single transporter or a transporter might transport different substrates with different mechanisms [11, 23, 24, 26].

3 Antimicrobial Peptides

Antimicrobial Peptides (AMPs) are small peptides ranging from 12 to 100 amino acid residue length and can be linear or combination of alpha helices or beta sheets or loops or mixed structures. AMPs have a broad spectrum of target organisms ranging from viruses to bacteria to parasites and have been fighting against various microbes including antibiotic-resistant microbes for the last two decades. A database, linking AMPs (LAMP2; https://biotechlab.fudan.edu.cn/database/lamp/index.php), currently have 23,253 sequences of which 7824 are natural while 15,429 are synthetic AMPs [27, 28]. Apart from this, other databases of AMPs are dbAMP (https://140.138.77.240/~dbamp/), DRAMP (https://dramp.cpu-bioinfor.org/), SATPdb (https://crdd.osdd.net/raghava/satpdb/), DBAASP (https://dbaasp.org/), CAMP (https://www.camp3.bicnirrh.res.in/index.php), APD (https://aps.unmc.edu/AP/main.php), etc. The first discovery of AMP dates back to 1939, where a mixture of AMPs, gramicidins, was obtained from soil Bacillus brevis [29,30,31]. Most of the natural AMPs are transcribed and expresses as pro-peptides and later are post translationally modified in their active form at their respective target cells. For instance, a human cathelicidin gene of 18Kda weight, known as hCAP18, is processed at C-terminus end to release a 37 amino acid residue peptide beginning with two leucines, thus, named LL37. This processing occurs in neutrophils. In case of skin, LL37 is further cleaved by serine proteases to produce diffeerent variants with varied degree of activity (like RK-31, KS-30, K20). The LL37 peptide was later found to exhibit potent and broad spectrum antimicrobial activity [32,33,34]. Similarly, defensins are also stored as unprocessed pro-peptides form in the paneth cells of small intestine. They subsequently undergo proteolytic cleavage resulting in the activation of mature peptides. This proteolytic cleavage is done by trypsin in humans and matrix metalloproteinase 7 (MMP7) in mouse [35]. AMPs have been discovered in both prokaryotes and eukaryotes. In higher order animals, they are mostly found in tissues and organs that are more susceptible for airborne pathogens and hence act as first line of innate immune defence [36, 37].

3.1 Mechanism of Action

AMPs target their cell either by disrupting membrane integrity or by penetrating inside cell and targeting intracellular proteins, DNA or RNA synthesis. Thus, on the basis of their mechanism of action, AMPs can be classified as either Membrane—Active or Intracellularly—Active. The membrane active AMPs initially interact with the membrane and then either aggregate or form pores in the membrane thereby permeablising it. Since the AMPs that interact with the membrane, are mostly amphiphilic in nature (having both ionic and hydrophobic residues). The mechanism of action of membrane actives can further be classified as Carpet like, Membrane thinning, aggregate formation, Barrel stave or Toroidal pore formation as represented in Fig. 3. In the barrel stave model, the AMPs insert into the membrane bilayer perpendicularly by forming pores. The hydrophobic and hydrophilic associations of phospholipid chains line the lumen of this pore thereby allowing AMP to penetrate and disrupt the membrane bilayer. In the toroidal pore model, the AMPs enter perpendicular to the membrane with inward folding of the membrane lipids resulting in formation of continuous channels between inner and outer leaflet of bilayer membrane. The pore is lined by both peptides and lipids head group. The basic differecne between toroidal pore and barrel-stave pore is that in the former peptides are always in contact with phospholipid head even when they enter the lipid bilayer. Less regular pore structures and peptides oriented parallel to membrane are characteristic features of disordered toroidal pore model [38, 39]. As the name suggests, in the carpet mechanism, the AMPs bind parallel to the surface of membrane and upon reaching above a threshold concentration, they permeate and disperse the membrane in a detergent-like manner without pore formation [39, 40]. The aggregate formation mechanism is similar to carpet formation, the difference being that in carpet model peptides interact only with phospholipid head while in aggregate form they may interact with complete lipid or rather no fixed pattern is followed [41]. Sometimes AMPs insert themselves only in one leaflet of membrane bilayer causing gaps in the membrane. To fill it, neighbouring lipid molecules are pulled in and gradually may also lead to replacement of complete monolayer with AMP. Such phenomenon is known as membrane thinning [42, 43]. As for intracellular active AMPs, they can either directly penetrate inside cell or through endocytosis. The AMPs falling in this category like PR-39, indocilin, tPMP-1, aHNP-1, etc. targets and inhibit DNA and protein synthesis. They may also inhibit enzymatic activities or may mediate immunomodulatory effects in hosts [7, 44,45,46].

Fig. 3
figure 3

Mechanism of action of AMPs. a Carpet model; b aggregate formation; c barrel-stave pore formation; d toroidal pore formation; e membrane thinning

Full size image

3.2 Nature of AMPs

The widely explored and most studied AMPs are against bacteria. The majority of these AMPs are either cationic or amphipathic and target bacterial cell membrane. The examples of antibacterial peptides are Magainin, Cecropin A, Mellitin, Buforin II, Protegrin, Polyphemusin, Indolicidin, PR-39, LL-37, α/β Defensins, Drosocin,etc. An AMP, Nisin, was found to be effective against methicillin resistant Staphylococcus aureus thus opening a new door for fighting against resistant microbes [45, 47].

Antifungal peptides are also mostly amphipathic in nature and kill by targeting cell wall or binding to chitin or targeting intracellular components. D-V13K, P18, Indolicin, Defensins are a few examples of AMPs possessing antifungal property [4, 7]. As for the antiviral activity of AMPs, they either integrate into the envelope of both DNA/RNA viruses and cause membrane instability or disruption or by preventing viral particles from entering host cell by blocking their receptors on cell membrane or preventing cell to cell spread of viral particles. Magainin, Cecropin, Mellitin, LL-37, Defensin, Brevinin-1, Dermaseptin, Tachyplesin, Protegrin, Polyphemusin, Lactoferricin, Indolicidin, etc. are a few antiviral AMPs targeting a wide spectra of viruses like HIV, Herpes Simplex Virus (HSV), Influenza A Virus (IAV), Jumin Virus, Adenovirus, Human Papillomavirus (HPV), Human Cytomegalovirus (HCMV), Vesicular Stomatitis Virus (VSV), etc.[7, 45]. The AMPs with antiparasitic properties form a small group when compared with those AMPs having antibacterial, antifungal and antiviral properties. The first AMP presenting antiparasitic activity was Magainin 2 which can kill Paramecium caudautum by swelling and eventual bursting of cells. Analogues of Defensins (synthetic peptides) were found to be effective against Leishmania major and Trypanosoma brucei. Apart from antiprotozoal, antinematodal peptides have also been reported against Caenorhabditis elegans like PMAP-23, cathelicidin, etc. An acylated synthetic AMP, Oct-CA(1–7)-M(2–9), was found to be effective against canine leishmaniasis. Dermaseptin S1-S5 AMPs were found to kill Leishmania mexicana by disturbing their lipid membrane composition. Cecropin A and their derivatives were more effective on the intracellular amastigote form compared to the extracellular promastigote form as tested on Leishmania aethiopica. Another AMP, Tachyplesin, is effective on Leishmania braziliensis as well as Trypanosoma cruzi. Similar to multidrug chemotherapy, combinatorial effect of AMPs have also been studied. For example, a combination of Mellitin and Cecropin A derivative shows greater effectiveness against Leishmania donovani than administered alone [7, 45, 48, 49].

The various studies conducted by diverse research groups imply that AMPs possessing antiprotozoal activity might have different/multiple mechanistic actions which depend on certain peptide motifs and these differ from those having anti-bacterial/fungal/viral activity [45]. The emerging resistance in microbes renders a constant pressure on the need of new drugs to fight against antibiotic resistance and multi-drug resistance (MDR). Nagarajan D. and group have tried to fight against MDR bacteria’s by using machine learning technique to design a library of synthetic AMPs. Of these, two AMPs were quite successful against MDR clinical isolates of Escherichia coli, Acinetobacter baumannii, Klebsiella pneumoniae, Pseudomonas aeruginosa, Staphylococcus aureus, and coagulase-negative staphylococci (CoNS) strains [50]. Their work has upheld the importance of computationally designed antimicrobial peptides which is the need of the hour. Other limitations of natural occurring AMPs are their long translating sequences resulting in high production cost, limited availability and low potency, their susceptibility to enzymatic degradation and cytotoxicities. With these limitations and avaialble information withheld by the sequence and structures, natural AMPs have led to the design and development of a new generation of synthetic AMPs with more potency, less toxicity apparently being more useful for therapeutics [7, 51].

In the initial era of AMPs discovery, it was believed that AMPs only target membranes and by disrupting them causes antimicobial effect. Thus, the stereospecific orientation of amino-acids in an AMP doesn’t affect their activity. Although later it was revealed that AMPs do have intracellular targets with specific stereochemical orientation like proteins, DNA or RNA and the presence of d- or l-amino acids in an AMP makes a difference. The AMPs with intracellular targets do not permeabilize membrane but penetrates them and causes antimicrobial activity [52]. Some examples of AMPs having stereochemical targets are buforin II, human α defensin 5, human α defensin 1, human β defensin 4, indolicidin, PR-39, etc. [53,54,55,56,57]. The effect of stereospecificity is not only limited to amino acids. The d- or l-enantiomeric changes in sugar (post-translational modification of AMPs) have drastically affected the antimicrobial activity of Drosocin [58]. The stereochemical isomers of MP196, a hexapeptide, showed improved activity against penicillin-binding protein of E. coli [59].

3.3 Designing of New Synthetic AMPs

While designing synthetic AMPs, few properties that need to be considered includes length of peptide, net charge, helicity, hydrophobicity, amphipathicity and last but not the least, their solubility in aqueous solution. Depending on the design and structure activity relationship, designing of synthetic AMPs can be categorised in three different classes, i.e., by modification of sequences of naturally occurring antimicrobial peptides, de novo designing of peptides and using computational assistance to design combinatorial libraries to help in identification of new synthetic AMPs.

3.3.1 Modification of Existing AMPs

This strategy starts with choosing an appropriate natural occurring active AMP as a starting step. A large number of databases containing details of all the naturally occurring AMPs from diverse species like plants, insects, mammals, etc., serve as grand reservoir for selecting the initial template sequence for designing new AMP. This template sequence can further be used for sequence truncation, amino acid substitution, cyclization which may result in improvising its structural activity, reducing size and hence improving antimicrobial activity with reduced toxicity [7, 60].

As most natural AMPs are longer in length, thus, truncating them to short 20–25 residues length not only decreases the cost facilitating large scale production, but also increases stability, decreases the chances of proteolytic cleavage hence making them more active. Removal of few residues may result in increase in the activity as well as specificity while decreasing the toxicity. Zasloff et al. have demonstrated through their study that truncating N-terminus of Magainin 2 by three residues have increased its antimicrobial activity drastically [61]. Similarly, truncating C-terminus of human cationic antimicrobial peptide (hCAP-18) has resulted in shorter and more functional α-helical peptide [62]. Many cationic host defense peptides have been modified using this strategy, for e.g., α-helical cecropins, lactoferrin, cathelicidins and magainins and β-sheet or β-hairpin like AMPs Defensins, gramicidins, protegrins, etc.[60].

Another strategy to increase stability against protease cleavage is by cyclization of AMPs. The cyclization can be done either by amide linkage between backbone in head-to-tail fashion, or disulphide bridging or native chemical ligation. The disulphide linkage cyclization of an indolicidin analogue CP-11 has led to a more rigid backbone as being demonstrated by Hancock and group [63,64,65]. One more way of modification of an existing AMP is the creation of hybrid peptides from truncated fragments of multiple naturally occurring peptides with antimicrobial activity. In such a hybrid peptide, usually one component is having more antimicrobial activity with higher toxicity and the other component is less toxic but less active too. Such a combination leads to a novel chimeric peptide having higher activity and less toxicity. The examples of hybrid peptides are a series of combinations of cecropin A (CA) and melittin; protegrin-1 (PG-1) with CA or PG-LB (bovine lactoferricin) [60, 66, 67].

The most widely used strategy is modifying an AMP by changing its amino acid content. Numerous studies have been reported where researchers have tried varied number of permutations and combinations to study the effect of different amino acids on their antimicrobial activity. It has been observed that higher proline content reduces the ability of peptide to penetrate the cell membrane. It was also professed that swapping neutral amino acids like Asn and Gln with positively charged residues have resulted in reduction of cytotoxic effects on eukaryotic cells. Introduction of d-amino acids in a peptide sequence have shown more stability against proteases [68, 69]. Addition of an amide group at the end of sequence has also shown to increase the efficiency of a peptide. For instance, amidation at C-terminal end of PMAP-23 increases the uptake by 10 folds, better and faster interaction with cell membrane of gram negative bacteria, better membrane permeabilisation, etc.[70]. The use of an unnatural AMP like β-didehydrophenylalanine have shown higher stability against proteolytic cleavages and hence been widely used in medicinal chemistry to alter a native bioactive AMPs [71]. Another classic example of modifying natural amino acids by substitution is represented by Jiang et al. group. They have used 2 natural AMPs, Piscidin 1 (fish origin) and dermaseptin S4 (frog origin), and substituted hydrophobic positions with one or two Lys residues. This subsitution not only improved antimicrobial activity against A. baumannii and P. aeruginosa but also decreased hemolysis of human red blood cells. I9K substitution in Piscidin 1 imrpoved activity by 55-fold and 32-fold respectively, against A. baumannii and P. aeruginosa. While L7K and A14K substitutions in sequence of dermaseptin S4 increased activity by 730-fold and 980-fold respectively, against A. baumannii and P. aeruginosa [72]. Apart from these different post-translational modifications can be incorporated like phosphorylation, methylation, glycosylation, amidation, formation of disulphide linkage, etc.[7].

3.3.2 De novo Design

This method is based on the intrinsic characteristics features of the naturally occurring antimicrobial peptides. These features include cationicity, amphipathicity, residue length, or structural features like formation of α-helical or β-sheet conformations or having mixed extended secondary structures. These features are responsible for stability as well as strong microbial activity. This method combines a few basic residues with non-polar residues. The basic residues like Arg, Lys or His assists in initial electrostatic interactions with microbial membrane surfaces which is negatively charged. The non-polar residues like Ala, Val, Leu, Phe, Tyr and Trp mediates the peptide’s insertion into microbial membrane lipid bilayers. This stratagem is widely used in drug development area as it leads to the identification of motifs which are smaller in size and can be further used to develop pharmacophore essential for antimicrobial effects. This, in turn, provides an opportunity to increase the ease of optimization through systematic modifications to the short peptide sequence. It also reduces the production cost as well as toxicities and immunogenicity’s which are the demerits associated with peptides with longer length of amino acid residues. As mentioned earlier, the combination of Arg and Lys residues have been found to possess strong antimicrobial activity. Similarly, combinations of Arg with a hydrophobic residue Trp also have strong effect on microbes and hence have been investigated vigorously by researchers. A handful of studies have revealed that combination of Arg and Trp along with a critical length of peptides (5–6 residue) not only balances amphipathicity but also accomplishes minimal requirement for showing antibacterial activities [73,74,75].

Along with maintaining amphipathic nature, the design of synthetic AMPs in terms of secondary structure should also be considered. Synthetic AMPs which can self-assemble into stabilized α-helical, β-sheet and/or hairpin structures are favoured. Due to charge repulsion between cationic residues, AMPs remain unstructured in an aqueous solution. This unstructured form helps them in interacting with cell membrane components of microbes like lipopolysaccharides, lipoteichoic acids and phospholipid head groups which neutralizes the charges on peptides leading to their self-assembly in a specified secondary structure via hydrogen bonding and hydrophobic interactions with the phospholipid bilayer causing membrane damage and perturbation. α-helix is the most studied class closely followed by β-sheet and hairpin structures. Decades of study have revealed that the degree of helicity is correlated with the length of peptide. Increase in the length of the peptide up to a certain extent increases the antimicrobial activity of the peptide. If the length is increased beyond a certain threshold then it results in loss of activity, selectivity or even stability. Synthetic α-helical AMPs like (FFRR)3, (LLRR)3 and (LLKK)3 displays the most optimal balance between α-helicity, antimicrobial activities and hemolytic potentials. Another important factor is the arrangement of hydrophobic and hydrophilic residues. Interestingly, arranging these hydrophobic and hydrophilic residues facing opposite to each other in the helical wheel projection has found to achieve ideal amphiphilicity. Yet another study have focused on the positioning of Trp residues at the hydrophobic–hydrophilic interface in helical wheel projection to enhance the stability as well as membrane penetration ability of the AMP [60, 76,77,78].

Another advantage of using de novo design is the liberty to use amino acids of D-conformation. Mant et. al. have successfully designed de novo new AMPs with selective activity against gram-negative bacteria while having no toxicity against human red blood cells under the most stringent conditions (18 h at 37 °C and up to 1000 μg/ml or > 350 μM of AMP). They have also shown that the unusual use of all D-conformation amino acids in peptides, diaminopropionic acid (Dap) and diaminobutyric acid (Dab), on the polar surface of amphipathic alpha-helix AMP have resulted in increased specificity for prokaryotes (bacteria) over eukaryotic cells. Aditionally, when tesetd on over 27 strains of A. baumannii resistant against atleast 20 antibiotics, the AMPs showed identical antimicrobial activity with MIC in range of 0.5 to 1.2 μM. Thus introduction of D-amino acids have significant advantage on antimicrobial activity with minimized hemolytic effect on red blood cells [79].

A new wave in direction of designing synthetic peptides is the use of machine learning approach to study sequences and nature of natural occurring peptides and on that basis predicting a whole new library of primary sequences. The structure of these newly designed peptides and then their function is predicted with computer assistance. Various databases like LAMP (linking antimicrobial peptides) have been created to store and compare various AMPs.

4 Discussion

ABC transporters being the largest family of efflux transport proteins and playing key role in MDR have always been an attractive target protein. But their highly conserved nature resulted in targeting them to be quite challenging. Most of the ABC inhibitors available currently leads to toxicity. Even the presence of modulators have helped in minimizing their toxic effects to some extent but failed to improve their effectiveness. Antimicrobial peptides with specificity as their characteristic feature have given a ray of hope in fighting against resistant microbes. Both natural and synthetic AMPs have shown great potential with their antimicrobial activity. For instance, LTX-109, a synthetic AMP, and their stereoisomers have proven quite effective against S. aureus as well as methicillin resistant S. aureus (MRSA) infections [80, 81]. In our recent work we have demonstrated that small peptides can either be used to modulate activity of transporter proteins leading to reversal of Miltefosine resistance in Leishmania major or can directly act as AMPs wildtype L. major strains [82]. In nutshell, AMPs provide an effective solution against resistant microbes and can be used either directly or in combination with existing antibiotics to maximize the antimicrobial effect (Fig. 4).

Fig. 4
figure 4

Schematic representation of the role of AMPs when administered alone or in combination with existing antibiotics

Full size image