Carratalá et al. Microb Cell Fact (2021) 20:30 https://doi.org/10.1186/s12934-021-01524-3 Microbial Cell Factories RESEARCH Open Access Title: insoluble proteins catch heterologous soluble proteins into inclusion bodies by intermolecular interaction of aggregating peptides Jose Vicente Carratalá1,2,3, Andrés Cisneros1,2, Elijah Hellman1,2, Antonio Villaverde1,2,3 and Neus Ferrer‑Miralles1,2,3* Abstract Background: Protein aggregation is a biological event observed in expression systems in which the recombinant protein is produced under stressful conditions surpassing the homeostasis of the protein quality control system. In addition, protein aggregation is also related to conformational diseases in animals as transmissible prion diseases or non-transmissible neurodegenerative diseases including Alzheimer, Parkinson’s disease, amyloidosis and multiple sys‑ tem atrophy among others. At the molecular level, the presence of aggregation-prone domains in protein molecules act as seeding igniters to induce the accumulation of protein molecules in protease-resistant clusters by intermolecu‑ lar interactions. Results: In this work we have studied the aggregating-prone performance of a small peptide (L6K2) with additional antimicrobial activity and we have elucidated the relevance of the accompanying scaffold protein to enhance the aggregating profile of the fusion protein. Furthermore, we demonstrated that the fusion of L6K2 to highly soluble recombinant proteins directs the protein to inclusion bodies (IBs) in E. coli through stereospecific interactions in the presence of an insoluble protein displaying the same aggregating-prone peptide (APP). Conclusions: These data suggest that the molecular bases of protein aggregation are related to the net balance of protein aggregation potential and not only to the presence of APPs. This is then presented as a generic plat‑ form to generate hybrid protein aggregates in microbial cell factories for biopharmaceutical and biotechnological applications. Keywords: Recombinant protein, Inclusion body formation, Protein aggregation, Intermolecular interaction, Antimicrobial peptides Background Protein aggregation is an event widespread distributed from bacteria to animals. In bacteria, it has been related to stress states with the deployment of a complex protein network to compensate for the reduction of the ability *Correspondence: neus.ferrer@uab.cat 3 of the cells to cope with the conformational stress [1]. Bioengineering, Biomaterials and Nanomedicine Networking Biomedical Research Centre (CIBER-BBN), 08193 Bellaterra, Barcelona, In contrast, in yeast, protein aggregation is an inher- Spain itable adaptive phenomenon [2]. In animals, protein Full list of author information is available at the end of the article © The Author(s) 2021. This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat​iveco​ mmons​.org/licen​ses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creat​iveco​mmons​.org/publi​cdoma​in/ zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Carratalá et al. Microb Cell Fact (2021) 20:30 Page 2 of 10 aggregation is also observed in pathological states related (CBDclos) [29], among others. Another peptide with to conformational diseases [3], but is also associated to high capacity to enhance protein aggregation is VP1, a the formation of hormone aggregates in secretory gran- peptide sequence present in the VP1 structural protein of ules [4]. the Foot-and-mouth disease virus [26]. Due to safety and During recombinant protein production, the detec- regulatory purposes, the use of viral protein domains, as tion of protein aggregates is a common outcome and is VP1, is not suitable for some applications. For that rea- observed in both eukaryotic and prokaryotic expression son, the development of novel APPs is of great interest. systems [5, 6]. Bioinformatic tools are available for pre- In this study, we have selected a small APP of 8 amino diction of protein and peptide solubility and to identify acids in length (L6K2) to study the potential of this type aggregation-prone hot spots in the amino acid sequence, of peptides to enhance the aggregation propensity of which can be modified during the design of recombinant soluble proteins [30]. In addition, we have amplified its genes [7–9]. However, changes in the primary structure aggregating potential by protein engineering and demon- of the natural proteins may lead to secondary effects as strated the antimicrobial activity of this type of amphi- the appearance of immunogenic epitopes [10]. Therefore, pathic alpha-helices. We have also analyzed the role of the aggregation propensity of the produced protein may stereospecific interactions in the aggregation of heterolo- be reduced by lowering the transcription and transla- gous recombinant proteins in the presence of L6K2-con- tion rates of the gene, keeping intact the original amino taining peptides, providing a platform to obtain hybrid acid sequence [11]. The main variables include the media IBs inside the cells. These results have relevant implica- composition, incubation temperature, promoter strength tions in the biopharmaceutical and biotechnological and inductor concentration among others [12]. applications of IBs. In prokaryotes, the solubility of proteins is controlled by the protein quality control system, a complex network Results and discussion of protein factors involved in protein folding, unfolding Modulation of recombinant protein solubility in ClearColi and degradation [13]. In bacteria, aggregates, known as cells recombinant proteins in the endotoxin free ClearColi™ inclusion bodies (IBs), are dynamic protein clusters from In order to study the performance of APP fused to which solubilized active protein conformers are released under physiological conditions [14–16]. In fact, recent expression system, two model soluble proteins; iRFP experimental approaches have revealed the ability of (near-infrared fluorescent protein) and GFP (green fluo- active IBs to rescue enzymatic activities in cell cultures rescence protein) were selected as scaffolds and fused to and to target cancer stem cells in cancer animal mod- the surfactant-like peptide L6K2, previously described as els [17–21]. Therefore, IBs are envisioned as depots of APP (Fig. 1a) [14]. recombinant protein with the capacity to release the pro- In transformed ClearColi cells, recombinant H6iRFP tein of interest from a complex and stable scaffold that protein was equally distributed in both soluble and insol- protects the biological activity of the embed protein over uble cell fractions (Fig. 1b, c). As expected, upon L6K2 time. Therefore, the enhancement of protein aggregation fusion, a different distribution pattern was observed, in this type of nanostructures is gaining interest. In fact, where most of the protein was located in the insoluble bioprocess design during recombinant protein produc- cell fraction, suggesting an increased aggregation ten- tion has been shown to impact the size of the IBs and dency for this fusion protein (Fig. 1b, c). The change in the physicochemical quality of the recombinant protein solubility pattern was achieved within 1 h of induction achieving constant production of IBs [22–24]. In addi- and was maintained for up to 5 h (Fig. 1b). As a model tion, aggregation propensity of recombinant proteins in APP, with high ability to enhance aggregation tendency of expression systems may be enhanced by the addition of recombinant proteins, VP1 from the capsid protein of the aggregation-prone peptides (APPs) in the design of the Foot-and-mouth disease virus [26, 27] was fused to GFP coding DNA sequence of the gene [25]. APP promote (Fig. 1a). As expected, most of the protein signal in the the establishment of intermolecular interactions between sample was detected in the insoluble cell fraction (Fig. 1b, protein species enhancing the tendency of the resulting c). In contrast, the recombinant H6GFPL6K2 was mostly complexes to accumulate in the insoluble cell fraction detected in the soluble cell fraction (Fig. 1b, c). These [26]. Aggregation domains can be found in nature, in par- results indicated that the aggregation propensity of a ticular, several protein domains have been shown to pos- recombinant protein may be modulated by APP although sess such aggregation capacity, including a variant of the the solubility tendency of the scaffold protein may coun- human β-amyloid peptide (Aβ42 (F19D)) [27], a mutant teract this effect (compare solubility of iRFP and GFP of the maltose binding protein (MalE31) [28], and the when fused to L6K2 in Fig. 1b, c). In the case of VP1, the cellulose-binding domain of Clostridium cellulovorans aggregation tendency overcome the high solubility of the Carratalá et al. Microb Cell Fact (2021) 20:30 Page 3 of 10 Soluble a b c 100 Insoluble Insoluble fraction Soluble fraction Protein distribution (%) 0h 1h 3h 5h 0h 1h 3h 5h H6 iRFP 80 H6iRFP H6 iRFP PT L6K2 H6iRFPL6K2 60 H6 GFP PT L6K2 H6GFPL6K2 VP1GFPH6 VP1 GFP H6 40 PT PTPPTTPTPPTTPTPTP 20 0 K2 K2 FP FP L6 L6 1G iR FP FP VP G iR Fig. 1 Effect of the addition of L6K2 in the aggregation propensity of fluorescent proteins. a Schematic representation of the recombinant genes is shown displaying protein domains by separate squares. H6 indicates the presence of a Hisx6 tag. PT corresponds to a linker with the indicated amino acid sequence. b Comparative aggregation propensity of iRFP and GFP in the presence of L6K2. VP1GFPH6 corresponds to a GFP fused to VP1 considered as positive control of the experiment. c Relative solubility (%) of recombinant proteins detailed in panel a analyzed by Western Blotting. Equivalent number of transformed ClearColi cells were lysed and soluble and insoluble cells fractions were separated. All proteins were detected with anti-his antibody GFP, while GFP solubility was not affected by the L6K2 containing longer Leucine/Lysine tracks (L12K4 and addition (compare solubility of GFP when fused to VP1 L18K6) were detected as fluorescent cellular pole aggre- or L6K2 in Fig. 1b, c). gates. Therefore, the serial L6K2 repeats acted both as APP and periplasm localization signals since the fluo- Impact of APP length on protein solubility in ClearColi cells rescence pattern revealed the clustering of signal in dis- As the ability of L6K2 to reduce solubility of GFP was not crete aggregates on the periphery of the cell cytoplasm. significant while it was effective in iRFP, we decided to In addition, the aggregation tendency in L6K2 repeats evaluate the effect of peptide length in the solubility of increased with the number of repeats while L12K4 pre- GFP. For that, we redesigned the H6GFPL6K2 recombi- sented an aggregation pattern like the observed in cells nant gene to add at the C-terminus of the GFP sequence, expressing the positive control VP1GFP. This aggrega- different versions of the surfactant-like peptide L6K2 tion tendency was not recorded in Western Blot analy- (Fig. 2a). The aggregating potential of L6K2 peptide was sis of the soluble and insoluble cell fractions (Fig. 2b, c), amplified by reiteration of leucine and lysine repeats indicating that the aggregation tendency of the L6K2- in different positions (see Table 1) and analyzed by derived peptides may be sensitive to the tested experi- AGGRESCAN software [8]. Selected peptides displayed mental conditions of protein extraction. This was not the higher hot spot area (HSA) than the original L6K2 pep- case of the aggregation pattern of VP1GFP construct that tide. However, only L12K4 and L18K6 showed increased was perfectly replicated under confocal laser scanning normalized hot spot area (NHSA) and increased average microscopy and SDS-PAGE (Compare VP1GFP data in aggregation-propensity hot spot ­(a4vAHS). Fig. 2b–d). As previously observed, H6GFPL6K2 was detected in the soluble cell fraction of transformed ClearColi cells Antimicrobial activity of L6K2‑containing recombinant (Fig. 2b, c) and consequently, the emitted fluorescence proteins was homogenously distributed in the cytosol (Fig. 2d). During recombinant gene expression experiments of The addition of the L6K2 derived peptides had a positive L6K2-containing constructs, the growth of transformed impact in protein aggregation tendency. As expected, the E. coli cells was compromised, especially in the case of distribution of fluorescence in the transformed cells was L18K6 (data not shown). At that point, we reasoned detected in protein clusters (IBs; Fig. 2d and Additional whether the peptides were toxic to the cell by displaying file 1: Fig S1a). In fact, we detected two different aggre- antimicrobial activity. The modeling of the L6K2 derived gation patterns in the L6K2 derived constructs. On the peptides with PEP-FOLD 3 [31–33] displayed amphip- one hand, the proteins containing serial L6K2 repeats athic alpha helices in all cases (Additional file 2: Figure ((L6K2)x2 and (L6K2)x3) were preferentially detected in S2a). The preferred conformation of the L6K2-contain- periplasmic areas around the cells, while the constructs ing peptides was maintained in the presence of the PT Carratalá et al. Microb Cell Fact (2021) 20:30 Page 4 of 10 a b c Soluble GFP H6 Insoluble fraction Soluble fraction 100 Insoluble Protein distribution (%) 0h 1h 3h 5h 0h 1h 3h 5h H6 GFP PT L6K2 GFPH6 80 H6GFPL6K2 H6 GFP PT (L6K2)x2 60 H6GFP(L6K2)x2 H6GFP(L6K2)x3 H6 GFP PT (L6K2)x3 40 H6GFPL12K4 H6 GFP PT L12K4 H6GFPL18K6 20 H6 GFP PT L18K6 VP1GFP 0 4 K2 x2 x3 FP FP VP1 GFP 2K 2) 2) L6 1G G PT PTPPTTPTPPTTPTPTP L1 6K 6K FP VP FP (L (L G FP FP G G G d Fig. 2 Production of GFP-containing recombinant proteins fused to aggregation prone peptides in ClearColi. a Schematic representation of L6K2-containing GFP constructs. H6 indicates the presence of a Hisx6 tag. PT corresponds to a linker with the indicated amino acid sequence. VP1 corresponds to the VP1 structural protein of Foot-and-mouth disease virus with high tendency to aggregate. b Detection of GFP in soluble and insoluble cell fractions of ClearColi transformed with expression plasmids containing the corresponding recombinant genes. c Relative solubility (%) of recombinant proteins detailed in panel a analized by Western Blotting. d Confocal analysis of ClearColi cultures producing GFP with several aggregation-prone peptides. Arrows indicate the distribution of protein aggregates in the expressing cells Table 1 Predictions of “hot spots (HS)” of aggregation in aggregating polypeptides by AGGRESCAN [8] Name HS region HS size Sequence HSA NHSA a4vAHS Ref. L6K2 1–6 6 LLLLLLKK 6.211 1.035 0.949 [22] (L6K2)x2 1–14 14 LLLLLLKKLLLLLLKK 12.789 0.913 0.865 This study (L6K2)x3 1–22 22 LLLLLLKKLLLLLLKKLLLLLLKK 19.367 0.880 0.842 This study L12K4 1–13 13 LLLLLLLLLLLLKKKK 14.625 1.125 1.074 This study L18K6 1–19 19 LLLLLLLLLLLLLLLLLLKKKKKK 23.025 1.212 1.171 This study HS hot spot, HSA hot spot area, NHSA normalized HSA, a4vAHS average aggregation-propensity in each HS linker, which has been described as a flexible peptide for of the peptide in expression systems [37]. However, the separating protein domains (Additional file 2: Fig. S2b) study of their activity when fused to reporter proteins [34]. This configuration has been described in naturally by genetic engineering has not been explored in depth. produced or synthetic cationic antimicrobial peptides Examples of this strategy include the fusion between (AMP) which have been proposed as a potential new GWH1 [38] and GFPH6 [14, 39] and the secretory pro- class of antimicrobial drugs [35]. The production of small duction of AMP-containing fusion partners [40]. In those peptides is difficult to be reached by recombinant tech- studies, the fusion of the AMP to the N-terminus of nologies due to reduced stability, and alternative strate- recombinant protein preserved the bactericidal activity gies have been taken to overcome such a main bottleneck of the AMP even though with its C-terminus anchored [36]. One possibility is the fusion of AMP to partner by the fusion. Therefore, we analyzed the putative antimi- proteins for a potential dual effect on the final product. crobial activity of the purified soluble versions of H6GF- First, the reduction of the toxicity of the AMP over the PL6K2 and H6GFP(L6K2)X2 proteins and compared expressing host, and the improvement in the stability these data with that obtained for purified GWH1GFPH6. Carratalá et al. Microb Cell Fact (2021) 20:30 Page 5 of 10 The results indicated that the antimicrobial activity of though their mechanism of action is nonspecific, it has L6K2-containing recombinant proteins is strain specific been described a differential efficacy of the same anti- (Fig. 3), being comparable to the antimicrobial activity of microbial peptide between Gram-negative and Gram- GWH1 peptide fused to GFP in E. coli cultures (Fig. 3b). positive bacteria [45, 46]. In the case of Gram-positive In addition, the position of the peptide at each end of the bacteria, apart from membrane disruption, the reaction scaffold protein did not appear to be relevant to the anti- requires further interactions with the cell wall [45]. microbial activity. On the other hand, the incubation of S. aureus with the proteins containing amphipathic alpha- Pull‑down effect on aggregation tendency of H6GFPL6K2 helices had only a slight effect on cell viability under the Aggregation of different proteins may be enhanced by tested conditions (Fig. 3a). Interestingly, the antimicro- the stereospecific interaction of APP in bacteria [26]. In bial activity of the recombinant proteins was completely that context, we reasoned that the aggregation ability of different when Micrococcus luteus cells were challenged. a recombinant protein with the same APP may enhance The addition of the purified proteins had a positive effect the aggregation tendency of H6GFPL6K2 when produced on cell viability at lower concentrations while at the high- simultaneously in cells. For that purpose, we generated a est protein concentration (8 µmol/L) the cell viability dual expression vector including the gene encoding for dropped drastically (Fig. 3c). H6iRFPL6K2, which displayed a high tendency to aggre- As observed in Fig. 3b, the antimicrobial activity of the gate beside the gene coding for H6GFPL6K2 to be simul- L6K2-containing constructs was detected in E. coli cul- taneously expressed. In cells expressing at the same time tures at low protein concentrations. This mechanism may the aggregation prone H6iRFPL6K2 construct and the explain the cell growth inhibition observed in ClearColi soluble H6GFPL6K2 construct, the fluorescence of the cultures transformed with expression vectors with cloned GFP shifted from the cytoplasm to polar protein aggre- L6K2-derived genes. Small cationic or amphipathic mol- gates (IBs) (Fig. 4). The green fluorescence distribution in ecules, similar to the ones described in this work, have expressing cells was similar to the pattern observed when been described as produced by prokaryotes and eukary- co-expressing VP1GFPH6 and H6iRFPL6K2. otic organisms as defense against infectious agents. These The change in the aggregation propensity of the H6GF- molecules belong to a non-specific ancient system of PL6K2 seemed to be directed by the pull-down ability of innate immunity and they perform their activity through the L6K2 peptide present in the H6iRFPL6K2 construct. direct interaction with membranes, nucleic acids, protein The intermolecular interactions between L6K2 present or even activate autolysins [41–44]. In the case of inter- in the two proteins enhances the aggregation tendency acting with membranes, they cause the destabilization of of GFP. In the expressing cells, the newly formed H6GF- the cytoplasmic membrane by forming pores or by their PL6K2, when interacting with H6iRFPL6K2 with a high arrangement parallel to the membrane surface, disrupt- tendency to aggregate was dragged to the insoluble cell ing the proton motive force and provoking the leakage of fraction. Therefore, it may be hypothesized that when two vital molecules which lead to cell death. However, even different proteins share aggregation prone domains, even a 120 b 100 c 180 160 100 80 140 Cell viability (%) Cell viability (%) Cell viability (%) 80 120 60 100 60 80 40 40 60 20 40 20 20 0 0 0 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8 Protein amount (µM) Protein amount (µM) Protein amount (µM) Fig. 3 Survival curves of microbial cells in the presence of putative antimicrobial peptides fused to GFP protein. a Staphylococcus aureus, b Escherichia coli and c. Micrococcus luteus. The results are presented as mean of two replicas for each analyzed point with corresponding standard error bar. Similar CFU where seeded on 96-well plates and the indicated protein concentrations were added to evaluate its effect on cell metabolism Carratalá et al. Microb Cell Fact (2021) 20:30 Page 6 of 10 a H6 iRFP PT L6K2 H6 GFP PT L6K2 b H6 iRFP PT L6K2 VP1 GFP H6 Fig. 4 Confocal microscopy images of recombinant GFP in co-expression experiments. a Detection of H6GFPL6K2 in expressing H6iRFPL6K2 cells. b Detection of VP1GFPH6 in expressing H6iRFPL6K2 cells. A schematic representation of the corresponding constructs is depicted beside the confocal Microscopy images if one of the proteins is still soluble, the protein with the role of the scaffold protein in this process. Therefore, highly aggregation propensity may lead the accompany- this result has a direct application for biopharmaceutical ing soluble protein to the insoluble cell fraction through and biotechnological applications through protein engi- coexpression. However, although the secondary structure neering. In fact, these protein nanoclusters have been of the iRFP and GFP proteins is not similar (Additional described as a source of soluble active protein obtained file 3: Figure S3), the effect of the iRFP scaffold protein upon incubation in non-denaturing conditions [14–16] in the aggregation enhancement of H6GFPL6K2 may not and have also been administered as biocompatible depots be ruled out. For that reason, a spectral variant of GFP for tumor targeting of therapeutic proteins [17–21]. Fur- (EBFP2; highly similar in amino acid sequence and sec- thermore, protein aggregation seems to be a common ondary structure) was fused to VP1 domain generating mechanism described in most of the expression systems VP1EBFP2H6 construct (Additional file 4: Figure S4). [47–49] that opens up the possibility of expanding this Predictably, when produced recombinantly, this protein type of strategy to proteins that are difficult to produce was mainly accumulated in the insoluble cell fraction in prokaryotes. Therefore, the fusion of common APP to (Additional file 5: Figure S5). different therapeutic recombinant proteins can induce The distribution of the GFP fluorescence in cells the colocalization of two recombinant proteins in IBs, simultaneously transformed with plasmids coding obtaining protein formulations with potential synergic H6GFPL6K2 and VP1EBFP2H6 was homogeneously dis- activities. tributed in the cytoplasm of the cells, in agreement with the data obtained in the expression experiment of H6GF- Conclusions PL6K2 alone (compare the distribution of GFP fluores- Protein aggregation is a universal event which is associ- cence in Figs. 2c and 5a, and Additional file 1: Fig. S1b). ated to conformational diseases in eukaryotes. In bac- On the other hand, the fluorescence emitted by EBFP2 teria, although it has been described as a symptom of fused to VP1 in those cells was mainly detected in polar metabolic stress resistance, some studies suggest the IBs as expected. When VP1GFPH6 was expressed along relevance of protein aggregation in physiological adap- with VP1EBFP2H6, the GFP fluorescence was located tation to stress [1]. In most of the recombinant protein exclusively at the poles of the cells, as IBs (Fig. 5b and production experiments described so far, a variable por- Additional file 1: Fig. S1b). The colocalization analysis of tion of the protein accumulated in bacterial IBs. In recent the fluorescence emission from both proteins indicated years, the use of IBs as active protein deposits has begun the preference of H6GFPL6K2 to aggregate in the pres- to be explored for biopharmaceutical and biotechnologi- ence of the same APP (Fig. 5c) ruling out an aggregating cal applications [22, 50]. The current study highlighted Carratalá et al. Microb Cell Fact (2021) 20:30 Page 7 of 10 a VP1 EBFP2 H6 H6 GFP PT L6K2 c b VP1 EBFP2 H6 VP1 GFP H6 Fig. 5 Overlay fluorescence images of H6GFPL6K2 and VP1GFPH6 with VP1EBFP2H6. a Detection of fluorescence emission from H6GFPL6K2 and VP1EBFP2H6 coexpressing cell cultures. b Detection of fluorescence emission from VP1GFPL6H6 and VP1EBFP2H6 coexpressing cell cultures. c Quantitative colocalization analysis of EBFP2 and GFP fluorescence signals in coexpression experiments. Overlap coefficients between the different fluorescences emitted by the fusion proteins VP1EBFP2H6/VP1GFPH6 and VP1EBFP2H6/H6GFPL6K2 expressed in ClearColi cells. Analysis performed from images obtained by confocal microscopy. *p < 0.001, one-way analysis of variance (ANOVA) the ability to enhance protein aggregation by the fusion with NcoI and HindIII and inserted into the MCS1. For of APPs to recombinant proteins used as baits for the dual expression plasmids, H6iRFPL6K2 + H6GFPL6K2, capture of soluble proteins. This effect was even observed H6iRFPL6K2 + VP1GFPH6, VP1EBFP2H6 + VP1GFPH6 for highly soluble proteins as GFP. In addition, hybrid and VP1EBFP2H6 + H6GFPL6K2 a two-step cloning IBs, enriched in two different recombinant proteins, strategy was followed. After the generation of the MCS2 were formed through stereospecific interactions between cloning plasmids (pETDuet-H6GFPL6K2 and pETDuet- common APP. Therefore, the presented data described VP1GFPH6), H6iRFPL6K2 and VP1EBFP2H6 frag- the potential of APP in the control of the aggregation ments were inserted into the MCS1 after digestion with propensity of recombinant proteins in biological formu- NcoI and HindIII. All L6K2-containing protein versions lations based on IBs and open up the possible exploration included a linker (PT) between GFP and the L6K2 pep- of synergic activities of hybrid protein aggregates, pro- tide or derivatives as previously described [30]. duced in bacterial cell factories, for biomedical and nano- biotechnological purposes. Expression of recombinant proteins in ClearColi cells ClearColi BL21 (DE3) was selected as expression host Methods for the different versions of the fluorescent proteins. The All protein designs were cloned in pETDuet™-1 plasmid Molecular cloning same conditions were applied in all cases. Briefly, after transformation with the corresponding expression vector, (Novagen), except for H6iRFP, GFPH6 an GWH1GFPH6, bacterial cells were allowed to growth in lysogenic broth mid pET22b (Novagen). For all pETDuet™-1 derived which were cloned into NdeI and HindIII sites of plas- (LB) medium supplemented with 100 µg/ml ampicillin in a shake flask (250 rpm) at 37 ºC. When cultures reached were inserted in either, MCS1 or MCS2 of pETDuet™-1 expression vectors, protein-coding DNA fragments an optical density of approximately 0.5–0.6, protein expression was induced by adding 1 mmol/L isopropyl- plasmid. In the case of H6GFPL6K2, H6GFP(L6K2) β-D-thiogalactopyranoside (IPTG). Protein samples X2, H6GFP(L6K2)X3, H6GFPL12K4, H6GFPL18K6 were taken at the indicated times (h) postinduction. In and VP1GFPH6, digestion was performed with NdeI all cases, bacterial OD was measured and adjusted to 1, and XhoI and insertion into the MCS2. On the other subsequently cells were collected by centrifugation (5 hand, H6iRFPL6K2 and VP1EBFP2H6 were digested minutes, 1,200 g). Resuspended cells were processed for Carratalá et al. Microb Cell Fact (2021) 20:30 Page 8 of 10 confocal microscopy visualization or to evaluate the rela- with 0.1 mmol/L isopropyl-β-D-thiogalactopyranoside tive protein distribution between the insoluble and solu- (IPTG) at 20 ºC, overnight. The cell pellet was collected ble cell fractions, in those cases, the expression time was (6000 g, 4 ºC, 15 minutes) and resuspended in wash set at 3 h. buffer (20 mmol/L Tris-HCl, pH 8.0, 500 mmol/L NaCl, 10 mmol/L imidazole) with ethylenediamine tetra-ace- Evaluation of protein aggregation propensity tic acid-free protease-inhibitor (complete EDTA-Free, Bacterial pellets harboring the expressed proteins were Roche). Cells were then disrupted by sonication (1 round resuspended in 1 mL of PBS until a homogeneous sus- of 2 min at 10 % amplitude and 10 rounds of 2 min at pension was achieved. Bacterial cell disruption was car- 15 % of amplitude) and cell debris was separated from ried out by sonication (1 round of 1 min at 10 % amplitude soluble fraction by centrifugation (15,000 g at 4 ºC, 45 and 1 round of 1 min at 15 % amplitude). Then, the solu- minutes). After filtration (0.22 µm), the His-tagged pro- ble and insoluble cell fractions were separated by cen- teins were purified from the soluble fraction by His tag trifugation (15 min, 15,000 g at 4 °C). The insoluble cell affinity chromatography using HiTrap Chelatin HP 1 ml fraction, containing the cell debris, was resuspended in column (GE Healthcare) in an ÄKTA purifier FPLC (GE 1 mL of PBS, after that, a small aliquot of both fractions, Healthcare). The purified fraction was obtained after elu- soluble and insoluble, was mixed (1:1) with Laemmli tion with a linear gradient of 20 mmol/L Tris-HCl pH 8.0, buffer. Soluble samples were boiled at 90 ºC for 10 min, 500 mmol/L NaCl, 500 mmol/L imidazole. The purity of while the insoluble samples were boiled for 40 min. The the different samples was analyzed by TGX gel chemistry processed samples were charged on SDS-PAGE gels and and Western Blotting. The selected fractions were mixed analyzed by Western Blotting with an anti-His mono- and dialyzed against sodium bicarbonate buffer with salt script). Images were acquired with the ChemiDoc™ clonal antibody (His Tag Antibody, mAb, Mouse, Gen- (166 mmol/L NaHCO3, pH 8.0, 333 mmol/L NaCl) and protein amounts were quantified by Bradford assay. Touch Imaging System (Bio-Rad) and further processing was performed with Image Lab Software. Percentage of Antimicrobial activity of APP‑containing recombinant aggregation was calculated based on the numerical band proteins intensity value obtained from blotting membrane images. The antimicrobial activity of H6GFPL6K2, H6GFP(L6K2) For each expression time, the total amount of protein X2 and GWH1GFPH6 was evaluated against three bac- (100 %) was considered as the sum of the band intensi- terial species, E. coli, S. aureus, and M. luteus, using the ties in both, soluble and insoluble cell fractions. There- broth micro-dilution method. Different two-fold dilu- fore, percentage of aggregation can be estimated from the tions of the proteins, ranging from 0.06 to 8 µmol/L, were band intensity value in the insoluble cell fraction. seeded in 96-well plates for each bacterial species. After that, ­106 CFU/mL of the corresponding bacteria were Visualization of recombinant proteins in ClearColi cells inoculated in each well. Maximal growth was achieved Bacterial pellets harboring the expressed proteins were in control wells with no protein. Each concentration was resuspended in 500 µL of PBS containing 4 % formal- evaluated in technical duplicates. Wells with 100 µL of dehyde. Then, resuspended samples were incubated 10 Mueller Hinton Broth Cation-adjusted medium (MHB- slide, a small drop of ProLong™ Gold Antifade Mountant minutes at RT and washed twice with PBS. In a glass 2, Sigma-Aldrich) were considered as blank solution. Growth conditions were stablished in 18 h at 37 ºC. The available BacTiter-Glo™ Microbial Cell Viability Assay (Thermo) was mixed with 5 µL of the bacterial suspen- bacterial viability was evaluated using the commercially sion. The resultant solution was covered with a cover- slip and fixed to avoid dehydration. The observation of (Promega) following the manufacturer’s instructions. the fluorescent proteins inside bacteria was recorded Luminescence was measured using the Multilabel Plater by TCS-SP5 confocal laser scanning microscopy (Leica Reader VICTOR3 (PerkinElmer). Microsystems). Images were processed using the ImageJ software. Colocalization analysis of fluorescent pro- Supplementary Information teins in ClearColi cells were performed by measuring The online version contains supplementary material available at https​://doi. the overlap coefficients of 10 regions of interest (ROIs) org/10.1186/s1293​4-021-01524​-3. which were compared by one-way analysis of variance (ANOVA). Additional file 1: Figure S1. Extended confocal microscopy visualization of E. coli ClearColi cells producing the recombinant fluorescent proteins. a Visual field of cells producing L6K2-containing constructs and VP1GFPH6 Purification of soluble recombinant proteins fused to APP (aggregation control) analyzed in Fig. 2. b Visual fields of cells producing For purification of H6GFPL6K2, H6GFP(L6K2)X2 simultaneously GFP and EBFP2 constructs analyzed in Fig 5. Scale bar indicates 4 μm. and GWH1GFPH6, protein expression was induced Carratalá et al. Microb Cell Fact (2021) 20:30 Page 9 of 10 Barcelona, Spain. 3 Bioengineering, Biomaterials and Nanomedicine Network‑ Additional file 2: Figure S2. PEP-FOLD server-generated models of ing Biomedical Research Centre (CIBER-BBN), 08193 Bellaterra, Barcelona, aggregation prone peptides fused to GFP scaffold protein. a Regular poly‑ Spain. peptide helices in a right-handed alpha-helical conformation are shown. All structures are reproduced at the same scale. b Helical conformation of Received: 3 August 2020 Accepted: 21 January 2021 two L6K2-containing peptides (blue) in the presence of PT linker (grey). Additional file 3: Figure S3. DNA sequences of recombinant genes H6iRFPL6K2 and H6GFPL6K2 used in the study. Translate tool from Expasy was used to obtain corresponding amino acid sequences. Clustalw was run to align amino acid sequences and Swiss Model to display 3D struc‑ tures of the recombinant proteins. References Additional file 4: Figure S4. DNA sequence of recombinant gene VP1EB‑ 1. Schramm FD, Schroeder K, Jonas K. Protein aggregation in bacteria. FEMS FP2H6 used in the study. Translate tool from Expasy was used to obtain Microbiol Rev. 2020;44:54–72. corresponding amino acid sequence. Clustalw was run to align amino acid 2. Wallace EWJ, Kear-Scott JL, Pilipenko EV, Schwartz MH, Laskowski PR, sequence of GFP and EBFP2. Swiss Model was used to display 3D structure Rojek AE, et al. Reversible, specific, active aggregates of endogenous of EBFP2 protein. proteins assemble upon heat stress. Cell. 2015;162:1286–98. Additional file 5: Figure S5. Relative solubility (%) of recombinant pro‑ 3. Chiti F, Dobson CM. Protein misfolding, amyloid formation, and human teins. Detection of H6GFPL6K2, VP1GFPH6 and VP1EBFP2H6 in the soluble disease: a summary of progress over the last decade. Annu Rev Biochem. and insoluble cell fractions of ClearColi analyzed by Western Blotting. 2017;86:27–68. Equivalent number of transformed ClearColi cells were lysed and soluble 4. Dannies PS. Prolactin and growth hormone aggregates in secretory and insoluble cells fractions were separated. All proteins were detected granules: the need to understand the structure of the aggregate. Endocr with anti-his antibody. Data are presented as mean ± STD of biological Rev. 2012;33:254–70. triplicate measurements. 5. Corchero JL. Eukaryotic aggresomes: from a model of conformational dis‑ eases to an emerging type of immobilized biocatalyzers. Appl Microbiol Biotechnol. 2016;100:559–69. Abbreviations 6. Singh A, Upadhyay V, Upadhyay AK, Singh SM, Panda AK. Protein recovery IBs: Inclusion bodies; APP: Aggregation-prone peptide; iRFP: Near-infrared from inclusion bodies of Escherichia coli using mild solubilization process. fluorescent protein; GFP: Green fluorescence protein; HS: Hot spot; HAS: Hot Microb Cell Fact. 2015;14:41. spot area; NHSA: Normalized HAS; a4vAHS: Average aggregation-propensity in 7. Kuriata A, Iglesias V, Pujols J, Kurcinski M, Kmiecik S, Ventura S. Aggres‑ each HS; AMP: Antimicrobial peptide. can3D (A3D) 2.0: prediction and engineering of protein solubility. Nucleic Acids Res. 2019;47:W300–7. 8. Conchillo-Solé O, de Groot NS, Avilés FX, Vendrell J, Daura X, Ventura S. Acknowledgements AGGRESCAN: a server for the prediction and evaluation of “hot spots” of The authors acknowledge ICTS ‘‘NANBIOSIS”, more specifically the Protein aggregation in polypeptides. BMC Bioinformatics. 2007;8:65. Production Platform of CIBER-BBN/IBB, at the UAB sePBioEs scientific-technical 9. Tsolis AC, Papandreou NC, Iconomidou VA, Hamodrakas SJ. A consensus service (http://www.nanbi​osis.es/unit/u1-prote​in-produ​ction​-platf​orm-ppp/) method for the prediction of “aggregation-prone” peptides in globular and the UAB scientific-technical services SM (https​://sct.uab.cat/micro​scopi​a/) proteins. PLoS One. 2013;8:e54175. and SCAC (https​://sct.uab.cat/scac/). 10. Singh SK. Impact of product-related factors on immunogenicity of bio‑ therapeutics. J Pharm Sci. 2011;100:354–87. Authors’ contributions 11. Schlesinger O, Chemla Y, Heltberg M, Ozer E, Marshall R, Noireaux V, JVC and NFM designed the experiments. JVC, AC and EH performed the et al. Tuning of recombinant protein expression in Escherichia coli by experiments. JVC and NFM analyzed and interpreted the data. NFM wrote the manipulating transcription, translation initiation rates, and incorporation manuscript. AVC, EH and JVC revised the manuscript. All authors read and of noncanonical amino acids. ACS Synth Biol. 2017;6:1076–85. approved the final manuscript. 12. Rosano GL, Ceccarelli EA. Recombinant protein expression in Escherichia coli: advances and challenges. Front Microbiol. 2014;5:172. Funding 13. Calloni G, Chen T, Schermann SM, Chang H-C, Genevaux P, Agostini F, This work was supported by INIA (MINECO,Spain) to NFM (grant RTA2015- et al. DnaK functions as a central hub in the E. coli chaperone network. 00064-C02), AGAUR to AV (2017 SGR-229), from Bioengineering, Biomaterials Cell Rep. 2012;1:251–64. and Nanomedicine Networking Biomedical Research Centre (CIBER-BBN), 14. Carratalá JV, Cano-Garrido O, Sánchez J, Membrado C, Pérez E, financed by the Carlos III Health Institute, Spain. The authors are also indebted Conchillo-Solé O, et al. Aggregation-prone peptides modulate activity to CERCA Programme (Generalitat de Catalunya). JVC. received a pre-doctoral of bovine interferon gamma released from naturally occurring protein fellowship and a short-term research grant from UAB. nanoparticles. N Biotechnol. 2020;57:11–9. https​://doi.org/10.1016/j. nbt.2020.02.001. Availability of data and materials 15. Peternel S, Grdadolnik J, Gaberc-Porekar V, Komel R. Engineering inclusion The datasets used and/or analyzed during the current study are available from bodies for non denaturing extraction of functional proteins. Microb Cell the corresponding author on reasonable request. Fact. 2008;7:34. 16. Pesarrodona M, Fernández Y, Foradada L, Sánchez-Chardi A, Conchillo- Ethics approval and consent to participate Solé O, Unzueta U, et al. Conformational and functional variants of CD44- Not applicable. targeted protein nanoparticles bio-produced in bacteria. Biofabrication. 2016;8:025001. Consent for publication 17. Céspedes MV, Cano-Garrido O, Álamo P, Sala R, Gallardo A, Serna N, et al. Not applicable. Engineering secretory amyloids for remote and highly selective destruc‑ tion of metastatic foci. Adv Mater. 2020;32:e1907348. Competing interests 18. Unzueta U, Seras-Franzoso J, Céspedes MV, Saccardo P, Cortés F, Rueda F, The authors declare that they have no competing interests. et al. Engineering tumor cell targeting in nanoscale amyloidal materials. Nanotechnology. 2017;28:015102. Author details 19. Unzueta U, Cespedes MV, Sala R, Alamo P, Sánchez-Chardi A, Pesarro‑ 1 Institute for Biotechnology and Biomedicine, Autonomous University dona M, et al. Release of targeted protein nanoparticles from func‑ of Barcelona, 08193 Bellaterra, Barcelona, Spain. 2 Department of Genetics tional bacterial amyloids: A death star-like approach. J Control Release. and Microbiology, Autonomous University of Barcelona, 08193 Bellaterra, 2018;279:29–39. Carratalá et al. Microb Cell Fact (2021) 20:30 Page 10 of 10 20. Vázquez E, Corchero JL, Burgueño JF, Seras-Franzoso J, Kosoy A, Bosser R, 37. Zhao Q, Xu W, Xing L, Lin Z. Recombinant production of medium- to et al. Functional inclusion bodies produced in bacteria as naturally occur‑ large-sized peptides in Escherichia coli using a cleavable self-aggregating ring nanopills for advanced cell therapies. Adv Mater. 2012;24:1742–7. tag. Microb Cell Fact. 2016;15:136. 21. Pesarrodona M, Jauset T, Díaz-Riascos ZV, Sánchez-Chardi A, Beaulieu 38. Chou HT, Kuo TY, Chiang JC, Pei MJ, Yang WT, Yu HC, et al. Design and M-E, Seras-Franzoso J, et al. Targeting antitumoral proteins to breast synthesis of cationic antimicrobial peptides with improved activity and cancer by local administration of functional inclusion bodies. Adv Sci. selectivity against Vibrio spp. Int J Antimicrob Agents. 2008;32:130–8. 2019;6:1900849. 39. Serna N, Sánchez-García L, Sánchez-Chardi A, Unzueta U, Roldán M, Man‑ 22. Slouka C, Kopp J, Spadiut O, Herwig C. Perspectives of inclusion bodies gues R, et al. Protein-only, antimicrobial peptide-containing recombinant for bio-based products: curse or blessing? Appl Microbiol Biotechnol. nanoparticles with inherent built-in antibacterial activity. Acta Biomater. 2019;103:1143–53. 2017;60:256–63. 23. Kopp J, Kolkmann A-M, Veleenturf PG, Spadiut O, Herwig C, Slouka C. 40. Yu H, Li H, Gao D, Gao C, Qi Q. Secretory production of antimicrobial pep‑ Boosting Recombinant Inclusion Body Production-From Classical Fed- tides in Escherichia coli using the catalytic domain of a cellulase as fusion Batch Approach to Continuous Cultivation. Front Bioeng Biotechnol. partner. J Biotechnol. 2015;214:77–82. 2019;7:297. 41. Gordon YJ, Romanowski EG, McDermott AM. A review of antimicrobial 24. Kopp J, Slouka C, Strohmer D, Kager J, Spadiut O, Herwig C. Inclusion peptides and their therapeutic potential as anti-infective drugs. Curr Eye body bead size in E. coli controlled by physiological feeding. Microorgan‑ Res. 2005;30:505–15. isms. 2018;6:116. 42. Raheem N, Straus SK. Mechanisms of Action for Antimicrobial Pep‑ 25. Wang X, Zhou B, Hu W, Zhao Q, Lin Z. Formation of active inclusion bod‑ tides With Antibacterial and Antibiofilm Functions. Front Microbiol. ies induced by hydrophobic self-assembling peptide GFIL8. Microb Cell 2019;10:2866. Fact. 2015;14:88. 43. Kumar P, Kizhakkedathu JN, Straus SK. Antimicrobial peptides: diversity, 26. Morell M, Bravo R, Espargaró A, Sisquella X, Avilés FX, Fernàndez-Busquets mechanism of action and strategies to improve the activity and biocom‑ X, et al. Inclusion bodies: specificity in their aggregation process and patibility in vivo. Biomolecules. 2018;8:4. amyloid-like structure. Biochim Biophys Acta. 2008;1783:1815–25. 44. Wimley WC. Describing the mechanism of antimicrobial peptide action 27. García-Fruitós E, González-Montalbán N, Morell M, Vera A, Ferraz RM, Arís with the interfacial activity model. ACS Chem Biol. 2010;5:905–17. A, et al. Aggregation as bacterial inclusion bodies does not imply inacti‑ 45. Torcato IM, Huang Y-H, Franquelim HG, Gaspar D, Craik DJ, Castanho vation of enzymes and fluorescent proteins. Microb Cell Fact. 2005;4:27. MARB, et al. Design and characterization of novel antimicrobial peptides, 28. Arié J-P, Miot M, Sassoon N, Betton J-M. Formation of active inclusion R-BP100 and RW-BP100, with activity against Gram-negative and Gram- bodies in the periplasm of Escherichia coli. Mol Microbiol. 2006;62:427–37. positive bacteria. Biochim Biophys Acta Netherlands. 2013;1828:944–55. 29. Nahalka J, Nidetzky B. Fusion to a pull-down domain: a novel approach 46. Elliott AG, Huang JX, Neve S, Zuegg J, Edwards IA, Cain AK, et al. An of producing Trigonopsis variabilis D-amino acid oxidase as insoluble amphipathic peptide with antibiotic activity against multidrug-resistant enzyme aggregates. Biotechnol Bioeng. 2007;97:454–61. Gram-negative bacteria. Nat Commun. 2020;11:3184. 30. Zhou B, Xing L, Wu W, Zhang X-E, Lin Z. Small surfactant-like peptides can 47. Martínez-Alonso M, Toledo-Rubio V, Noad R, Unzueta U, Ferrer-Miralles drive soluble proteins into active aggregates. Microb Cell Fact. 2012;11:10. N, Roy P, et al. Rehosting of bacterial chaperones for high-quality protein 31. Lamiable A, Thévenet P, Rey J, Vavrusa M, Derreumaux P, Tufféry P. PEP- production. Appl Environ Microbiol. 2009;75:7850–4. FOLD3: faster de novo structure prediction for linear peptides in solution 48. Rueda F, Gasser B, Sánchez-Chardi A, Roldán M, Villegas S, Puxbaum V, and in complex. Nucleic Acids Res. 2016;44:W449-54. et al. Functional inclusion bodies produced in the yeast Pichia pastoris. 32. Shen Y, Maupetit J, Derreumaux P, Tufféry P. Improved PEP-FOLD Microb Cell Fact. 2016;15:1–2. approach for peptide and miniprotein structure prediction. J Chem 49. Rodríguez-Carmona E, Mendoza R, Ruiz-Cánovas E, Ferrer-Miralles Theory Comput. 2014;10:4745–58. N, Abasolo I, Schwartz S, et al. A novel bio-functional material 33. Thévenet P, Shen Y, Maupetit J, Guyon F, Derreumaux P, Tufféry P. PEP- based on mammalian cell aggresomes. Appl Microbiol Biotechnol. FOLD: an updated de novo structure prediction server for both linear and 2015;99:7079–88. disulfide bonded cyclic peptides. Nucleic Acids Res. 2012;40:W288-93. 50. Jäger VD, Kloss R, Grünberger A, Seide S, Hahn D, Karmainski T, et al. Tailor‑ 34. Poon DKY, Withers SG, McIntosh LP. Direct demonstration of the flexibility ing the properties of (catalytically)-active inclusion bodies. Microb Cell of the glycosylated proline-threonine linker in the Cellulomonas fimi Fact. 2019;18:33. Xylanase Cex through NMR spectroscopic analysis. J Biol Chem United States. 2007;282:2091–100. 35. Huang Y, Huang J, Chen Y. Alpha-helical cationic antimicrobial peptides: Publisher’s note relationships of structure and function. Protein Cell. 2010;1:143–52. Springer Nature remains neutral with regard to jurisdictional claims in pub‑ 36. Gaglione R, Pane K, Dell’Olmo E, Cafaro V, Pizzo E, Olivieri G, et al. Cost- lished maps and institutional affiliations. effective production of recombinant peptides in Escherichia coli. N Biotechnol. 2019;51:39–48.