This is the accepted version of the article: Carratalá, José Vicente; Cano Garrido, Olivia; Sánchez, Julieta María; [et al.]. «Aggregation-prone peptides modulate activity of bovine interferon gamma released from naturally occurring protein nanoparticles». New Biotechnolgy, Vol. 57 (July 2020), p. 11-19. DOI 10.1016/j.nbt.2020.02.001 This version is avaible at https://ddd.uab.cat/record/218927 under the terms of the license 1 Aggregation‐prone peptides modulate activity of bovine interferon gamma released from 2 naturally occurring protein nanoparticles 3 José Vicente Carrataláa,b, Olivia Cano‐Garridoa,b,c,1, Julieta Sáncheza,2, Cristina 4 Membradoa,b, Eudald Péreza,b, Oscar Conchillo‐Soléa, Xavier Dauraa,d, Alejandro Sánchez‐ 5 Chardie, Antonio Villaverdea,b,c, Anna Arísf, Elena Garcia‐Fruitósf, Neus Ferrer‐Mirallesa,b,c# 6 a 7 Institute for Biotechnology and Biomedicine, Autonomous University of Barcelona, 8 Bellaterra, Barcelona, Spain b 9 Department of Genetics and Microbiology, Autonomous University of Barcelona, 10 Bellaterra, Barcelona, Spain c 11 Bioengineering, Biomaterials and Nanomedicine Networking Biomedical Research Centre 12 (CIBER‐BBN), Bellaterra, Barcelona, Spain d 13 Catalan Institution for Research and Advanced Studies, Barcelona, Spain e 14 Microscopy Service, Autonomous University of Barcelona, Bellaterra, Barcelona, Spain. f 15 Department of Ruminant Production, Institute of Agrifood Research and Technology 16 (IRTA), Caldes de Montbui, Barcelona, Spain 1 17 Present address: Nanoligent SL. Eureka Building. Campus of the Autonomous University 18 of Barcelona. Bellaterra, Barcelona, Spain 1 2 19 Permanent address: National University of Cordoba, Faculty of Exact, Physical and Natural 20 Sciences, ICTA and Chemistry Department, CONICET Institute of Biological and 21 Technological Research (IIByT) Cordoba Argentina 22 #Correspondence: Neus Ferrer‐Miralles, neus.ferrer@uab.cat. 23 24 25 26 Abstract 27 Efficient protocols for the production of recombinant proteins are indispensable for the 28 development of the biopharmaceutical sector. Accumulation of recombinant proteins in 29 naturally‐occurring protein aggregates is detrimental to biopharmaceutical development. 30 In recent years, the view of protein aggregates has changed with the recognition that they 31 are a valuable source of functional recombinant proteins. In this study, bovine interferon‐ 32 gamma (rBoIFN‐) was engineered to enhance the formation of protein aggregates, also 33 known as protein nanoparticles (NPs), by the addition of aggregation‐prone peptides (APPs) 34 in the generally recognized as safe (GRAS) bacterial Lactococcus lactis expression system. 35 The L6K2, HALRU and CYOB peptides were selected to assess their intrinsic aggregation 36 capability to nucleate protein aggregation. These APPs enhanced the tendency of the 37 resulting protein to aggregate at the expense of total protein yield. However, fine physico‐ 38 chemical characterization of the resulting intracellular protein NPs, the protein released 2 39 from them and the protein purified from the soluble cell fraction indicated that the 40 compactability of protein conformations was directly related to the biological activity of 41 variants of IFN‐, used here as a model protein with therapeutic potential. APPs enhanced 42 the aggregation tendency of fused rBoIFN‐ while increasing compactability of protein 43 species. Biological activity of rBoIFN‐ was favored in more compacted conformations. 44 Naturally‐occurring protein aggregates can be produced in GRAS microorganisms as protein 45 depots of releasable active protein. The addition of APPs to enhance the aggregation 46 tendency has a positive impact in overall compactability and functionality of resulting 47 protein conformers. 48 49 Abbreviations 50 NPs, nanoparticles; APPs, aggregation‐prone peptides; GRAS, Generally Recognized as Safe; 51 rBoIFN‐ recombinant bovine IFN‐IBs, inclusion bodies; HSA, hot spot area; NHSA, 52 normalized hot spot area; a4vAHS, average aggregation‐propensity hot spot 53 54 Keywords 55 Interferon‐gamma, protein nanoparticles, protein aggregation, Lactococcus lactis, 56 Generally Recognized as Safe, conformational compactability 57 Introduction 3 58 The efficient production and purification of recombinant proteins in a wide range of 59 expression hosts has driven the launch of a large number of biopharmaceutical products. 60 One of the most‐studied and most‐used gene expression systems for biopharmaceutical 61 products is Escherichia coli [1,2]. Prokaryotic endotoxin‐free expression systems are being 62 explored to avoid the presence pro‐inflammatory contamination by lipopolysaccharide 63 (LPS) components of the outer leaflet of the outer membrane of E. coli, including E. coli LPS 64 mutant strains [3,4] and Generally Recognized As Safe (GRAS) microorganisms, such as 65 Lactococcus lactis [5‐7]. 66 During recombinant gene expression, the stress imposed on the protein quality control 67 machinery leads, in most cases, to the accumulation of the recombinant protein in 68 aggregates that form intracellular nanoparticles (NPs), known as inclusion bodies (IBs) [8‐ 69 10]. These are dynamic and complex nanostructures with a variable content of recombinant 70 protein [11‐13]. The trapped protein was formerly thought to be biologically inactive due to 71 aberrant protein conformations or inactive partially folded species incompatible with 72 biological activity. The recombinant protein can often be recovered, with low efficiency, 73 from the insoluble cell fraction by in vitro denaturing/refolding processes [14]. However, 74 this view of naturally occurring protein aggregates has changed radically since the detection 75 of biologically active protein embedded in these aggregates [15‐17]. The classic view of 76 protein aggregates as mere inactive folding intermediates has been transformed into one 77 of heterogeneous porous multimeric structures stabilized by a scaffold of cross beta‐sheet 78 structures containing conformers of the recombinant protein in which a spectrum of species 79 containing quasi‐native conformations are incorporated [9]. It has been reported that 4 80 biologically active protein species can be extracted from IBs, indicating the biologically 81 active nature of proteins forming these aggregates [18]. Hence, IBs are envisionaged as non‐ 82 toxic, biocompatible and mechanically stable materials from which biologically active 83 molecules of the recombinant protein can be released under mild solubilization and 84 physiological conditions [13,16,18‐21]. 85 Interest in the possibility of controlling the aggregation of recombinant proteins in these 86 types of nanostructures is increasing, and several aggregation‐prone peptides (APPs) have 87 been identified for fusion with recombinant proteins to enhance the aggregation process in 88 the producing cell [22]. In this study, interferon (IFN)‐ was selected as a model protein in 89 order to study the effect of the addition of APPs in naturally occurring protein aggregates 90 due to interest in this activity in biomedicine and its potential use in animal health. IFN‐ is 91 the sole type II IFN. IFN‐ secretion by natural killer (NK) cells and antigen‐presenting cells 92 enhances the innate immune response, while T‐lymphocytes are involved in the secretion 93 of IFN‐ in the adaptive immune response [23,24]. The activity of IFN‐ depends on its 94 interaction, as a dimer, with the IFN‐ receptor (IFNGR). Approved recombinant human IFN‐ 95  can be obtained from the E. coli expression system, but novel protein formulations need 96 to be developed in GRAS expression systems due to safety concerns. In most reported 97 studies of the expression and purification of IFN‐, the recombinant protein is recovered 98 from the purified IBs through extensive denaturation‐refolding processes [25‐28]. 5 99 In this work, the mature form of bovine IFN‐ (rBoIFN‐ protein (UniProtKB P07353, 100 residues 24 to 166) was produced in GRAS lactic acid bacteria (L. lactis) in the form of protein 101 NPs. 102 The ability of APPs fused to rBoIFN‐ to enhance the aggregation propensity of the 103 recombinant cytokine was analyzed and the link assessed between the biological activities 104 contained in protein NPs of IFN‐ variants and their physicochemical characteristics. It was 105 found that the activity of the IBs is related to the specific biological activity of the 106 recombinant protein they contain, whereas the proportion of released protein is not the 107 main factor. The data presented illustrate the potential of endotoxin‐free protein NPs as 108 active biomaterials to formulate, at the nanoscale level, releasable proteins of biomedical 109 interest. 110 111 112 113 Materials and methods 114 Bacterial strains and plasmids 115 E. coli MC4100 (StrepR) [29] was used for cloning genes for protein production in L. lactis. 116 E. coli DH5 was used for cloning genes in E. coli. L. lactis cremoris NZ9000 (Boca Scientific, 117 MA, USA), and ClearColi® BL21(DE3) (Lucigen, WI, USA) were used in experiments for each 118 expression system. Gene sequences were codon optimized for the L. lactis expression host 6 119 (Geneart, MA; USA, Suppl. Figure S1). For L. lactis expression vectors, IFN‐ of bovine origin 120 (Bos Taurus; NM_ 174086.1 in Suppl. Figure S1) was cloned into the CmR pNZ8148 plasmid 121 (MoBiTech, Goettingen, Germany) as described in Supplementary Materials and Methods 122 and [6]. In addition, fusions of rBoIFN‐ with APPs were constructed (rBoIFN‐_L6K2, 123 rBoIFN‐_HALRU and rBoIFN‐_CYOB; Figure 1). L6K2 is a surfactant‐like peptide with 124 aggregating properties [30]. HALRU and CYOB are aggregating‐prone peptides from 125 Cytochrome bo3 ubiquinol oxidase subunit 1 from E. coli (UniProtKB P0ABI8), HALRU: 126 Aragonite protein AP7 (UniProtKB Q9BP37) selected with AGGRESCAN [31] (see Table 1). 127 For Clearcoli®, the L. lactis codon‐optimized bovine IFN‐ gene was cloned into pETDuet‐1 128 (Novagen, WI, USA) (Suppl. Figure S2 and Supplementary Materials and Methods). The 129 recombinant proteins were produced as the mature form of the IFN‐ (from Gln24 to 130 Thr166; NP_776511.1) (Figure S2). All genes were C‐terminally fused to a His‐tag for 131 detection and quantification by western blot analysis and a linker with a predicted random 132 coil conformation was positioned between the IFN‐ and APP as previously described [30]. 133 134 Selection of APPs 135 APPs were selected by scanning the Disprot v6.02 database [32] with AGGRESCAN 136 software [31]. The selection was based on the assumption that APPs in solvent‐exposed 137 regions were the best candidates for the purposes of this study. Two unstructured regions 138 were selected from two different proteins namely CYOB: Cytochrome bo3 ubiquinol oxidase 139 subunit 1 from E. coli (UniProtKB P0ABI8) and HALRU: Aragonite protein AP7 (UniProtKB 7 140 Q9BP37). CYOB was selected as the peptide displaying the highest hot spot area (HSA). 141 HALRU showed a high normalized hot spot area (NHSA) and average aggregation‐propensity 142 hot spot (a4vAHS) while maintaining a significantly high HSA value relative to the other 143 identified peptides. L6K2 was selected based on previous experimental results [30] after 144 analysis with AGGRESCAN showed that this peptide had a high normalized HSA (NHSA) and 145 high average aggregation‐propensity hot spot (a4vAHS) despite having shorter sequence 146 (Table 1 and Figure 1a). 147 148 Production and purification of rBoIFN‐ protein from the soluble cell fraction 149 Cultures of ClearColi® BL21 (DE3) cells transformed with the plasmid pETDuet‐rBoIFN‐ 150 (Supplementary Materials and Methods) were incubated in a shake flask at 37 °C and 151 250 rpm in LB medium supplemented with 100 μg/ml ampicillin. Protein expression was 152 induced by adding 1 mM isopropyl‐β‐D‐thiogalactopyranoside (IPTG). The cultures were 153 then incubated at 20 °C and 250 rpm overnight for protein production. Cells were collected 154 by centrifugation (15 min, 6,000 x g, 4 °C), and soluble rBoIFN‐ protein was purified as 155 described in Supplementary Materials and Methods. Protein expression of L. lactis cells 156 transformed with plasmid containing the rBoIFN‐ gene was induced and purified as 157 described in Supplementary Materials and Methods and [6]. The control protein rBoIFN‐ 158 _Std, produced in E. coli was obtained from R&D Systems (2300‐BG‐025, R&D Systems, 159 MN, USA). 160 8 161 Production and purification of rBoIFN‐ protein nanoparticles. 162 L. lactis cells transformed with expression plasmids (pNZ8148‐rBoIFN‐, pNZ8148‐ 163 rBoIFN‐L6K2pNZ8148‐rBoIFN‐HALRU and pNZ8148‐rBoIFN‐_CYOB) were grown as 164 above. NP production was induced by adding 12.5 ng/ml nisin (Sigma‐Aldrich, MO, USA) to 165 L. lactis cultures. After induction, the cultures were grown for 5 h. 166 The protein NPs were purified using the protocol described previously (Supplementary 167 Materials and Methods and [6]). 168 Quantitative protein analysis 169 Recombinant proteins were quantified by denaturing SDS‐PAGE as described previously 170 (Supplementary Materials and Methods and [33]). In addition, the yields of purified proteins 171 in each of the formats are shown in Table S1. 172 173 Ultrastructural characterization 174 To characterize the morphometry of the NPs, microdrops of protein aggregate 175 suspensions were deposited for 2 min on silicon wafers (Ted Pella Inc.), air‐dried and 176 observed in a near‐native state under a field emission scanning electron microscope 177 (FESEM) Zeiss Merlin (Zeiss, Obercochen, Germany) operating at 1 kV. Micrographs were 178 acquired with a high‐resolution in‐lens secondary electron (SE) detector. Images were taken 179 at magnifications ranging from 20,000x to 80,000x. 180 181 Z potential analysis 9 182 Z potential (ZP) characterization of each kind of protein NP was carried out by Dynamic 183 Light Scattering (DLS) (Zetasizer Nano ZS, Malvern Instruments Ltd, Malvern, UK). To 184 prevent the electrodes from burning, the samples were prepared in deionized (MilliQ) 185 water. Each sample was analyzed in triplicate. 186 187 Determination of rBoIFN‐γ biological activity in bovine cells 188 The different rBoIFN‐ formulations described here were analyzed by a modified 189 kynurenine bioassay (Supplementary Materials and Methods and [34]). The 190 antiproliferative activity of IFN‐ in this assay is related to the induction of the expression of 191 the indoleamine 2,3‐dioxygenase 1 (IDO1) gene, which is the first and rate‐limiting enzyme 192 in tryptophan catabolism. IDO1 catalyzes oxidative cleavage of tryptophan to N‐ 193 formylkynurenine. Following a hydrolysis step, the latter is transformed into L‐kynurenine 194 by Ehrlich’s reagent, giving a yellow‐colored compound absorbing at 490 nm [35]. The 195 absorbance vs IFN‐ concentration (nmol/L) curves were adjusted to Eq. 1 [20]. Abs490 is 196 the absorbance at 490 nm, which represents an indirect measurement of IFN‐ binding to 197 the receptor, Absmax is the maximal binding of IFN‐ to the receptor, and KD is the 198 equilibrium dissociation constant. A low value of KD indicates high IFN‐ affinity to the 199 receptor. 200 Abs𝑚𝑎𝑥 x IFN 201 Abs490 (1) IFN KD 202 203 Assay of protein solubilization from protein nanoparticles 10 204 The rBoIFN‐ protein NPs (rBoIFN‐_L, rBoIFN‐_L6K2, rBoIFN‐_CYOB and rBoIFN‐ 205 _HALRU) were solubilized in PBS. In all cases, the concentration was adjusted to 1 µmol/L. 206 After manual agitation, each sample was incubated at 37 °C for 96 h to reproduce the 207 conditions used during the biological activity analysis. Protein concentration was quantified 208 and the biological activity determined at a single concentration (0.72 nmol/L) as described 209 in previous section. 210 211 Interferon size determination 212 The volume size distribution of IFN‐ was determined by DLS. A 60‐µl aliquot (stored 213 at ‐80 °C) was thawed, and the volume size distribution of each protein format was 214 immediately determined at 633 nm (Zetasizer Nano ZS, Malvern Instruments Ltd, Malvern, 215 UK). 216 217 Analysis of protein conformation by intrinsic tryptophan fluorescence 218 Fluorescence spectra were recorded on a Cary Eclipse spectrofluorometer (Agilent 219 Technologies, CA, USA). A quartz cell with a 10‐mm path length and a thermostatic holder 220 was used. The excitation and emission slits were set at 5 nm. The excitation wavelength 221 (ex) was set at 295 nm. Emission spectra were acquired within a range from 310 to 550 nm. 222 The protein concentration was 14 mol/L in PBS DEFINE. To evaluate conformational 223 differences between the proteins, the center of spectral mass (CSM), was applied, the 224 weighted average of the fluorescence spectrum peak. The CSM was calculated for each of 11 225 the fluorescence emission spectra [36] according to Eq.2, where Ii is the fluorescence 226 intensity measured at wavelength i. 227 228    .I i i (2) I i 229 CSM values were analyzed at room temperature and under thermal heating at 5 °C/min 230 rate. 231 232 Statistical analysis 233 Prior to the use of parametric tests, normality and homogeneity of variances were tested 234 using the Shapiro‐Wilk test for all quantitative data or the Levene test for raw or 235 transformed data. Divergences between groups were tested with one‐way ANOVA, and 236 pairwise comparisons were made with Student’s t tests. The results were expressed as the 237 arithmetic mean for non‐transformed data ± the standard error of the mean (x̅ ± SEM), 238 except otherwise stated. 239 The least squares method was applied to fit functions through a regression analysis to 240 determine the KD values according to Eq. 1. Significance was accepted at p < 0.05, and 241 Bonferroni correction was applied for sequential comparisons. All statistical analyses were 242 performed with SPSS v. 18 for Windows. 243 244 Results and Discussion 245 Production of rBoIFN‐ in L. lactis 12 246 In L. lactis, most of the rBoIFN‐ protein was detected in the soluble cell fraction in the 247 absence of any APP (Figure 1b, upper panel). This observation is in agreement with previous 248 results for the expression of the natural DNA sequence of the bovine IFN‐ gene in E. coli 249 [37]. The presence of the APPs in the recombinant protein caused a noticeable shift of the 250 final products toward the insoluble cell fraction, as expected (Figure 1b, lower panel). The 251 purity of the protein aggregates ranged between 50‐60 % in all constructs (Suppl. Figure 252 S3). The APP resulting in the highest aggregation tendency was the L6K2 peptide. In 253 addition, the presence of an APP tag also had a negative effect on the total recombinant 254 protein produced in the cell (Figure 1a, upper panel). The best APP in terms of aggregation 255 propensity and protein yield in the insoluble cell fraction, corresponded to the IFN‐ L6K2 256 formulation. The performance of this surfactant‐like peptide exceeded the predicted 257 aggregation‐prone capabilities of CYOB and HALRU peptides (Table 1). 258 259 Nanoarchitectonic characterization of protein nanoparticles 260 The morphometry of purified protein NPs of the rBoIFN‐ variants was examined by 261 FESEM (Figure 2a). The images revealed the presence of multimeric complexes comprising 262 discrete NPs in addition to isolated protein NPs (inset Figure 2a). The NPs were similar to 263 rBoIFN‐ protein NPs obtained previously in this expression system [6]. ZP measurements 264 showed that all of the NPs presented negatively charged surfaces with negative values 265 ranging from ‐38 to ‐28 mV (Figure 2b), indicating the stability of the NP suspension. The 266 higher values of ZP obtained for the IFN‐ variants provide information about particle 13 267 stability, as NPs displaying higher ZP values (>+30 mV or <‐30 mV) exhibit increased stability 268 due to greater electrostatic repulsion between particles [38]. 269 270 Biological activity of soluble IFN‐ and NPs of IFN‐ 271 The activity of IFN‐ is usually determined by an antiviral assay [39]. However, alternative 272 assays have been developed to simplify the procedure. One approach to evaluate IFN‐ 273 activity mediated by IFN‐ receptor binding is the detection of L‐kynurenine. The activity of 274 IFN‐is highly species‐specific and, a specific assay for the bovine IFN‐ was developed and 275 validated in this study [34]. For validation, the activity of three soluble rBoIFN‐ proteins 276 was tested (Figure 3a). rBoIFN‐_Std exhibited the lowest dissociation constant (KD) among 277 the proteins purified from the soluble cell fraction (Figure 3a) with a similar value 278 determined for the human IFN‐ [40]. The difference in this parameter with in‐house IFN‐ 279 produced in Clearcoli (rBoIFN‐_E. coli) may be related to the absence of C‐terminal variants 280 in this sample, the effect of the fused His‐tag at the C‐terminus, or other variables [41]. The 281 protein obtained from the L. lactis expression system displayed less activity, which may be 282 due to differences in the folding efficiency during the production process among prokaryotic 283 expression systems [42,43]. Once the activity assay was validated, the biological activity 284 contained in the IFN‐ protein NPs produced in L. lactis was determined. The results showed 285 that all cells were able to elicit responses to the presence of the protein NPs, and the IFN‐ 286 _L6K2 formulation displayed the highest initial rate and kynurenine production (Figure 3b). 287 The addition of HALRU and CYOB APP to IFN‐ had a moderate effect on the cell response. 14 288 It is of interest to know why the sample corresponding to protein NPs of IFN‐_L6K2 had the 289 highest activity and initial rate, even compared with commercial IFN‐. Consistent with this 290 observation, a previous analysis of the activity of recombinant ‐galactosidase produced in 291 E. coli in the form of protein NPs revealed higher specific activity than the corresponding 292 soluble version of the protein [15]. However, protein NPs obtained from E. coli have not 293 been characterized in detail. The activity displayed by E. coli IBs has been attributed to the 294 release of a spectrum of conformers of the recombinant protein, which leaves a scaffold 295 that is resistant to proteolysis and has an extensive cross‐beta‐pleated sheet conformation 296 [44,45]. For protein NPs of rBoIFN‐ produced in L. lactis, 30‐40 % of the material is resistant 297 to proteolysis, indicating that the protein NPs obtained in this expression system follow 298 similar principles to the E. coli system [6]. Thus, the activities displayed by the protein NPs 299 are probably due to the partial release of the IFN‐ that forms part of the macromolecular 300 complex [46]. To evaluate better the ability of the protein NPs to release protein, they were 301 incubated in PBS for 96 h to emulate the protein release conditions established during the 302 biological activity assay (see the experimental design used to obtain the different protein 303 samples in Figure 3c). Release of 52.67 %, 5.30 %, 0.42 % and 0.46 % was observed for IFN‐ 304 , IFN‐_L6K2, IFN‐_HALRU and IFN‐_CYOB NPs, respectively. In order to analyze the 305 specific activity of the proteins released from the protein NPs, an activity assay was 306 performed and the results compared with proteins obtained directly from the soluble cell 307 fraction (Figure 3d). The results showed that the maximal specific activity corresponded to 308 the IFN‐_L6K2 protein released from NPs. In addition, the comparison of the specific 309 activity of the rBoIFN‐ protein produced in L. lactis and purified from the soluble cell 15 310 fraction with that of the corresponding protein released from NPs suggested that the 311 released protein elicited better conformational performance (compare the second and last 312 bars in Figure 3d). 313 The addition of APPs to the rBoIFN‐protein improved the aggregation profile of the 314 produced protein (Figure 1b). However, the presence of this type of peptide had a negative 315 effect on the overall production of the protein and, in the case of HALRU and CYOB, a major 316 impact on biological activity (Figure 3). From this, AGGRESCAN software is able to predict 317 the propensity of the resulting APP‐containing recombinant IFN‐ to aggregate and is a 318 reliable tool for analyzing solubility performance in the design of recombinant genes [31]. 319 320 Physicochemical characterization of soluble IFN‐ and nanoparticles of IFN‐ 321 The precise physicochemical analysis of recombinant proteins is important for safety 322 concerns [47,48]. To further analyze the protein in different formats, DLS measurements 323 were performed (Figure 4a‐4d). The rBoIFN‐_Std exhibited a peak with a maximum at 324 7.6 nm, similar to the peak at 6.13 nm for the IFN‐ produced in L. lactis. This configuration 325 (6‐8 nm) might correspond to the dimeric form of the cytokine. However, the IFN‐ obtained 326 from E. coli showed a tendency towards a larger size. Therefore, the specific activity of the 327 different rBoIFN‐formats is not simply linked to the dimeric configuration, which is the 328 functional conformation when binding to the cell receptor, and some other variables might 329 be involved. When analyzing the size of the purified NPs, a peak above 1,000 nm was 330 detected, exceeding the upper sensitivity limit of the equipment (Figure 4b). The NPs were 16 331 clustered in higher‐order complexes from monomeric versions of 200 nm (Figure 2a). All 332 samples exhibited the same profile. After solubilization of the protein embedded in the NPs, 333 the size of the remaining material remained above 1,000 nm since the scaffold of the NPs 334 retained the overall structure after the protein was released (Figure 4d). The released 335 protein showed a narrow dispersion ranging from the dimeric size of the protein identified 336 in the samples obtained from the commercial IFN‐ or the soluble version purified from L. 337 lactis detected in the upper panel of Figure 4a (Figure 4c). In addition, the polydispersity 338 index (PI) of these samples was higher than that of the soluble IFN‐ versions. The PI 339 corresponds to an estimate of the width of the distribution, and higher values of PI are 340 consistent with the data showing a pool of conformers in the folding of recombinant 341 proteins when the proteins are produced in the cell [49]. In contrast, in the protein versions 342 purified from the soluble cell fraction, the downstream processing based on affinity 343 chromatography selects only a narrow collection of conformers (only those that are able to 344 bind to the Ni2+ in the resin). This indicates that the protein obtained during solubilization 345 assays from protein NPs is more representative of the diversity in conformations of a single 346 protein that are produced in the expression system. 347 To further analyze the link between the physicochemical properties and the specific 348 activity of the proteins, the fluorescence emission of Trp was recorded. Each fluorescence 349 emission spectrum was transformed into a CSM value. This parameter is related to the 350 relative exposure of the Trp to the protein environment. The maximum red‐shift in the CSM 351 of the Trp spectrum is compatible with large solvent accessibility [50‐52], whereas the blue 352 shift in the CSM corresponds to a Trp hidden in a more hydrophobic milieu [53]. The mature 17 353 form of BoIFN‐ has a unique Trp at position 36, which is partially buried in the 3D structure 354 of the protein (PDB 1D9C) [54] and is not involved in either monomer or in cytokine‐ 355 receptor interactions, as shown in the 3D structure of the human tetrameric complex of the 356 cytokine dimer with the receptor (PDB 1FG9) [55]. A remarkable aspect of the intrinsic 357 fluorescence analysis is that all the rBoIFN‐ variants within the NPs or after solubilization 358 from the protein NPs exhibited lower CSM values than the samples obtained from the 359 soluble fraction (Table 2). These results suggest that the protein forming part of the NPs 360 and the protein solubilized from the aggregates have a more compact conformation than 361 the soluble version. The most active IFN‐ soluble version corresponded to the commercial 362 IFN‐, which had the lowest CMS due to its highly compacted structure. The proteins 363 obtained from the soluble fraction of E. coli and L. lactis exhibited higher CMS values than 364 the commercial protein. These differences might be related to the distinct sizes detected 365 (Figure 4a). The rBoIFN‐_E. coli was approximately three times larger than the same 366 protein produced in L lactis, indicating that the Trp residue was located in a more polar 367 environment compared with the L lactis form (Table 2). For the protein originating from the 368 particulate form, a blue shift was observed compared with the soluble versions, and the 369 CSM increased as it was resolubilized (lines 4 and 6 of Table 2). The CSM value of the 370 solubilized rBoIFN‐_L. lactis protein sample did not reach that of the soluble counterpart 371 (lines 3 and 4 of Table 2). When the APPs were incorporated in the engineered protein 372 constructs, the solubilized proteins showed a decrease in the CSM values compared with 373 the protein NPs samples (lines 5 and 7). This behavior suggests a possible self‐arrangement 374 of the tag within the protein that could replace water molecules and increase the 18 375 hydrophobicity of the Trp environment. The CYOB construct (line 6 of Table 2) required a 376 specific analysis as this tag contributes five additional Trp residues to the whole protein 377 structure. In this case, the solubilized protein spectrum exhibited a modest red shift (higher 378 CMS value) compared with the particulate form, indicating that the solubilization process 379 exposed some of the Trp residues to a hydrophilic environment. The CSM values of the 380 CYOB and HALRU protein NPs remained unaltered after solubilization (Table 2, lines 6 and 381 7). These data are in accordance with the higher stability of the particulate forms, which 382 exhibited low levels of protein release. 383 The NP form of IFN‐ also favored the specific activity (insets, Figure 3, L. lactis and non‐ 384 tagged rBoIFN). This phenomenon is not only due to more active conformation of the 385 protein (Figure 4 and Table 2, line 3 vs line 4) [56] but also to the heterogeneous distribution 386 of the protein and the ability of the protein NPs to increase the effective concentration of 387 protein in the proximity of the receptor. Moreover, the formulation containing L6K2 was 388 the most efficient, even compared with the commercial protein. Solubilization clearly 389 conferred the most active and altered conformation of the protein without the tag. 390 Although a low percentage of protein was released from the NPs containing L6K2, at least 391 in PBS, this released protein seems to be sufficient to surpass the activity of the released 392 protein without a tag (Figure 3). Furthermore, the CSM thermal profile of the released 393 proteins demonstrated that L6k2 tag not only confers the highly compact structure (low 394 CSM value, Table 2) but also contribute to a unique and complete thermal unfolding profile 395 (Figure S4). 19 396 Another interesting aspect is the effect of the size of the tag on the structure‐function of 397 the protein. The incorporation of a tag larger than 17 amino acids beyond the linker (Figure 398 1a) could generate steric problems preventing the interaction of tagged IFN‐ with the 399 receptor. As shown in Figure 1, L6K2 is only 8 amino acids, compared with 17 amino acids 400 for HALRU and 38 amino acids for CYOB. The short size of the L6K2 tag might reduce the 401 difficulty of the interaction between L6K2‐IFN‐ and the receptor compared with the longer 402 IFN‐ tags since the C‐terminal end of the protein, where the APPs are fused, is located in 403 close proximity to the receptor in the 3D structure in PDB 1FG9. 404 In the recombinant protein production platform, the general consensus for improving 405 protein yield is to improve the solubility of the protein. However, solubility and 406 conformational quality are not necessarily coincident parameters [57]. The functionalities 407 of the protein obtained from the soluble cell fraction or the protein NPs of rBoIFN‐_L. lactis 408 in the present work supported these previous findings, as the protein obtained from the 409 soluble cell fraction was less active than that recovered from the protein NPs. The 410 compactabilities of the conformations of these proteins were in agreement with their 411 dissimilar biological activity. Therefore, results obtained in this study may indicate that the 412 compactability of protein conformations is a significant parameter related to stability and 413 function [58,59]. 414 415 Conclusions 20 416 In this study, it was demonstrated that the addition of aggregation‐prone peptides 417 (APPs) promoted the production of naturally occurring protein nanoparticles (NPs) of 418 interferon gamma (IFN‐) in the generally recognized as safe (GRAS) Lactococcus lactis 419 expression system. The fine physico‐chemical characterization of the resulting proteins 420 revealed that conformational compactability was directly related to the biological 421 performance of the recombinant IFN‐. 422 423 Acknowledgments 424 This work was supported by grants from INIA, MINECO, Spain to N.F.M. and E.G.F. 425 (RTA2015‐00064‐C02‐01 and RTA2015‐00064‐C02‐02). The authors acknowledge financial 426 support granted to A.V. from AGAUR (2017 SGR‐229) and from Bioengineering, Biomaterials 427 and Nanomedicine Networking Biomedical Research Centre (CIBER‐BBN), financed by the 428 Carlos III Health Institute, Spain, with assistance from the European Regional Development. 429 We are also indebted to the CERCA Programme (Generalitat de Catalunya) and European 430 Social Fund for supporting our research. J.V.C. received a pre‐doctoral fellowship from UAB, 431 O.C.G. received a PhD fellowship from MECD (FPU), and E.G.F. received a post‐doctoral 432 fellowship from INIA (DOC‐INIA). AV has been distinguished with an ICREA ACADEMIA 433 Award. The authors also acknowledge ICTS ‘‘NANBIOSIS”, more specifically the Protein 434 Production Platform of CIBER‐ BBN/IBB, at the UAB sePBioEs scientific‐technical service 435 (http://www.nanbiosis.es/unit/u1‐protein‐production‐platform‐ppp/) and the UAB 436 scientific‐technical services LLEB, SM and SCAC 21 437 (https://www.uab.cat/web/research/scientific‐technical‐services/all‐scientific‐technical‐ 438 services‐‐1345667278676.html). The authors would like to thank Milena Tileva for her 439 helpful advice on technical issues related to the experimental adjustment of the IFN‐ 440 activity bioassay. Special thanks to Sandra Párraga‐Ferrer for the design of Figure 3c. 441 Author contributions 442 E. Garcia‐Fruitós and N. Ferrer‐Miralles designed and supervised the experiments. J.V. 443 Carratalà, O. Cano‐Garrido, J. Sánchez, C. Membrado, E. Pérez, O. Conchillo‐Solé and A. 444 Sánchez‐Chardi performed the experiments. J. V. Carratalà, O. Cano‐Garrido, J. Sánchez and 445 N. Ferrer‐Miralles analyzed the data. All authors edited the manuscript. N. Ferrer‐Miralles 446 wrote the paper. 447 448 References 449 [1] Baeshen MN, Al‐Hejin AM, Bora RS, Ahmed MM, Ramadan HA, Saini KS et al. 450 Production of Biopharmaceuticals in E. coli: Current Scenario and Future Perspectives. 451 J Microbiol Biotechnol 2015;25:953‐62. https://doi.org/10.4014/jmb.1412.12079 452 [2] Sanchez‐Garcia L, Martin L, Mangues R, Ferrer‐Miralles N, Vazquez E, Villaverde A. 453 Recombinant pharmaceuticals from microbial cells: a 2015 update. Microb Cell Fact 454 2016;15:33. https://doi.org/10.1186/s12934‐016‐0437‐3 455 [3] Rueda F, Cespedes MV, Sanchez‐Chardi A, Seras‐Franzoso J, Pesarrodona M, Ferrer‐ 456 Miralles N et al. Structural and functional features of self‐assembling protein 457 nanoparticles produced in endotoxin‐free Escherichia coli. Microb Cell Fact 458 2016;15:59. 10.1186/s12934‐016‐0457‐z [doi] 459 [4] Mamat U, Wilke K, Bramhill D, Schromm AB, Lindner B, Kohl TA et al. Detoxifying 460 Escherichia coli for endotoxin‐free production of recombinant proteins. Microb Cell 461 Fact 2015;14:57. 10.1186/s12934‐015‐0241‐5 [doi] 22 462 [5] Garcia‐Fruitos E. Lactic Acid Bacteria: a promising alternative for recombinant protein 463 production. Microb Cell Fact 2012;11:157. 10.1186/1475‐2859‐11‐157 [doi] 464 [6] Cano‐Garrido O, Sanchez‐Chardi A, Pares S, Giro I, Tatkiewicz WI, Ferrer‐Miralles N et 465 al. Functional protein‐based nanomaterial produced in microorganisms recognized as 466 safe: A new platform for biotechnology. Acta Biomater 2016;43:230‐9. 467 https://doi.org/10.1016/j.actbio.2016.07.038 468 [7] Ferrer‐Miralles N, Villaverde A. Bacterial cell factories for recombinant protein 469 production; expanding the catalogue. Microb Cell Fact 2013;12:113. 470 https://doi.org/10.1186/1475‐2859‐12‐113 471 [8] Kopito RR. Aggresomes, inclusion bodies and protein aggregation. Trends Cell Biol 472 2000;10:524‐30. 473 [9] Rinas U, Garcia‐Fruitos E, Corchero JL, Vazquez E, Seras‐Franzoso J, Villaverde A. 474 Bacterial Inclusion Bodies: Discovering Their Better Half. Trends Biochem Sci 475 2017;42:726‐37. https://doi.org/10.1016/j.tibs.2017.01.005 476 [10] Rodriguez‐Carmona E, Mendoza R, Ruiz‐Canovas E, Ferrer‐Miralles N, Abasolo I, 477 Schwartz S Jr et al. A novel bio‐functional material based on mammalian cell 478 aggresomes. Appl Microbiol Biotechnol 2015;99:7079‐88. 479 https://doi.org/10.1007/s00253‐015‐6684‐0 480 [11] Garcia‐Fruitos E. Inclusion bodies: a new concept. Microb Cell Fact 2010;9:80. 481 https://doi.org/10.1186/1475‐2859‐9‐80 482 [12] Jurgen B, Breitenstein A, Urlacher V, Buttner K, Lin H, Hecker M et al. Quality 483 control of inclusion bodies in Escherichia coli. Microb Cell Fact 2010;9:41. 484 https://doi.org/10.1186/1475‐2859‐9‐41 485 [13] Singh A, Upadhyay V, Upadhyay AK, Singh SM, Panda AK. Protein recovery from 486 inclusion bodies of Escherichia coli using mild solubilization process. Microb Cell Fact 487 2015;14:41. https://doi.org/10.1186/s12934‐015‐0222‐8 488 [14] Yamaguchi H, Miyazaki M. Refolding techniques for recovering biologically active 489 recombinant proteins from inclusion bodies. Biomolecules 2014;4:235‐51. 490 https://doi.org/10.3390/biom4010235 491 [15] Garcia‐Fruitos E, Gonzalez‐Montalban N, Morell M, Vera A, Ferraz RM, Aris A et al. 492 Aggregation as bacterial inclusion bodies does not imply inactivation of enzymes and 493 fluorescent proteins. Microb Cell Fact 2005;4:27. https://doi.org/10.1186/1475‐2859‐ 494 4‐27 23 495 [16] Jevsevar S, Gaberc‐Porekar V, Fonda I, Podobnik B, Grdadolnik J, Menart V. 496 Production of nonclassical inclusion bodies from which correctly folded protein can be 497 extracted. Biotechnol Prog 2005;21:632‐9. https://doi.org/10.1021/bp0497839 498 [17] Nahalka J, Nidetzky B. Fusion to a pull‐down domain: a novel approach of 499 producing Trigonopsis variabilisD‐amino acid oxidase as insoluble enzyme aggregates. 500 Biotechnol Bioeng 2007;97:454‐61. https://doi.org/10.1002/bit.21244 501 [18] Gifre‐Renom L, Cano‐Garrido O, Fabregas F, Roca‐Pinilla R, Seras‐Franzoso J, Ferrer‐ 502 Miralles N et al. A new approach to obtain pure and active proteins from Lactococcus 503 lactis protein aggregates. Sci Rep 2018;8:13917. 10.1038/s41598‐018‐32213‐8 504 [doi];10.1038/s41598‐018‐32213‐8 [pii] 505 [19] Cespedes MV, Fernandez Y, Unzueta U, Mendoza R, Seras‐Franzoso J, Sanchez‐ 506 Chardi A et al. Bacterial mimetics of endocrine secretory granules as immobilized in 507 vivo depots for functional protein drugs. Sci Rep 2016;6:35765. 508 https://doi.org/10.1038/srep35765 509 [20] de Jong LA, Uges DR, Franke JP, Bischoff R. Receptor‐ligand binding assays: 510 technologies and applications. J Chromatogr B Analyt Technol Biomed Life Sci 511 2005;829:1‐25. 10.1016/j.jchromb.2005.10.002 [doi] 512 [21] Peternel S, Komel R. Active protein aggregates produced in Escherichia coli. Int J 513 Mol Sci 2011;12:8275‐87. https://doi.org/10.3390/ijms12118275 514 [22] Krauss U, Jager VD, Diener M, Pohl M, Jaeger KE. Catalytically‐active inclusion 515 bodies‐Carrier‐free protein immobilizates for application in biotechnology and 516 biomedicine. J Biotechnol 2017;258:136‐47. 517 https://doi.org/10.1016/j.jbiotec.2017.04.033 518 [23] Le PC, Genin P, Baines MG, Hiscott J. Interferon activation and innate immunity. 519 Rev Immunogenet 2000;2:374‐86. ncbi.nlm.nih.gov/pubmed/11256746 520 [24] Schoenborn JR, Wilson CB. Regulation of interferon‐gamma during innate and 521 adaptive immune responses. Adv Immunol 2007;96:41‐101. 522 https://doi.org/10.1016/S0065‐2776(07)96002‐2 523 [25] Khalilzadeh R, Shojaosadati SA, Maghsoudi N, Mohammadian‐Mosaabadi J, 524 Mohammadi MR, Bahrami A et al. Process development for production of 525 recombinant human interferon‐gamma expressed in Escherichia coli. J Ind Microbiol 526 Biotechnol 2004;31:63‐9. https://doi.org/10.1007/s10295‐004‐0117‐x 527 [26] Perez L, Vega J, Chuay C, Menendez A, Ubieta R, Montero M et al. Production and 528 characterization of human gamma interferon from Escherichia coli. Appl Microbiol 529 Biotechnol 1990;33:429‐34. 24 530 [27] Vaiphei ST, Pandey G, Mukherjee KJ. Kinetic studies of recombinant human 531 interferon‐gamma expression in continuous cultures of E. coli. J Ind Microbiol 532 Biotechnol 2009;36:1453‐8. https://doi.org/10.1007/s10295‐009‐0632‐x 533 [28] Kumar M, Singh M, Singh SB. Optimization of conditions for expression of 534 recombinant interferon‐gamma in E.coli. Mol Biol Rep 2014;41:6537‐43. 535 https://doi.org/10.1007/s11033‐014‐3537‐3 536 [29] Thomas JG, Baneyx F. Roles of the Escherichia coli small heat shock proteins IbpA 537 and IbpB in thermal stress management: comparison with ClpA, ClpB, and HtpG In 538 vivo. J Bacteriol 1998;180:5165‐72. 539 [30] Zhou B, Xing L, Wu W, Zhang XE, Lin Z. Small surfactant‐like peptides can drive 540 soluble proteins into active aggregates. Microb Cell Fact 2012;11:10. 541 https://doi.org/10.1186/1475‐2859‐11‐10 542 [31] Conchillo‐Sole O, de Groot NS, Aviles FX, Vendrell J, Daura X, Ventura S. 543 AGGRESCAN: a server for the prediction and evaluation of "hot spots" of aggregation 544 in polypeptides. BMC Bioinformatics 2007;8:65. https://doi.org/10.1186/1471‐2105‐8‐ 545 65 546 [32] Piovesan D, Tabaro F, Micetic I, Necci M, Quaglia F, Oldfield CJ et al. DisProt 7.0: a 547 major update of the database of disordered proteins. Nucleic Acids Res 2017;45:D219‐ 548 D227. https://doi.org/10.1093/nar/gkw1056 549 [33] Cano‐Garrido O, Rueda FL, Sanchez‐Garcia L, Ruiz‐Avila L, Bosser R, Villaverde A et 550 al. Expanding the recombinant protein quality in Lactococcus lactis. Microb Cell Fact 551 2014;13:167. https://doi.org/10.1186/s12934‐014‐0167‐3 552 [34] Spekker K, Czesla M, Ince V, Heseler K, Schmidt SK, Schares G et al. Indoleamine 553 2,3‐dioxygenase is involved in defense against Neospora caninum in human and bovine 554 cells. Infect Immun 2009;77:4496‐501. https://doi.org/10.1128/IAI.00310‐09 555 [35] Boyanova M, Tsanev R, Ivanov I. A modified kynurenine bioassay for quantitative 556 determination of human interferon‐gamma. Anal Biochem 2002;308:178‐81. 557 [36] Lakowicz JR, Kusba J, Wiczk W, Gryczynski I, Szmacinski H, Johnson ML. Resolution 558 of the conformational distribution and dynamics of a flexible molecule using 559 frequency‐domain fluorometry. Biophys Chem 1991;39:79‐84. 560 [37] Li GY, Xiao ZZ, Lu HP, Li YY, Zhou XH, Tan X et al. A simple method for recombinant 561 protein purification using self‐assembling peptide‐tagged tobacco etch virus protease. 562 Protein Expr Purif 2016;128:86‐92. https://doi.org/10.1016/j.pep.2016.08.013 563 [38] Bhattacharjee S. DLS and zeta potential ‐ What they are and what they are not? J 564 Control Release 2016;235:337‐51. https://doi.org/10.1016/j.jconrel.2016.06.017 25 565 [39] Meager A. Biological assays for interferons. J Immunol Methods 2002;261:21‐36. 566 [40] Bomsztyk K, Stanton TH, Smith LL, Rachie NA, Dower SK. Properties of interleukin‐1 567 and interferon‐gamma receptors in B lymphoid cell line. J Biol Chem 1989;264:6052‐7. 568 [41] Hess AK, Saffert P, Liebeton K, Ignatova Z. Optimization of translation profiles 569 enhances protein expression and solubility. PLoS One 2015;10:e0127039. 570 https://doi.org/10.1371/journal.pone.0127039 571 [42] Boumaiza M, Colarusso A, Parrilli E, Garcia‐Fruitos E, Casillo A, Aris A et al. Getting 572 value from the waste: recombinant production of a sweet protein by Lactococcus lactis 573 grown on cheese whey. Microb Cell Fact 2018;17:126. 574 https://doi.org/10.1186/s12934‐018‐0974‐z 575 [43] Marini G, Luchese MD, Argondizzo AP, de Goes AC, Galler R, Alves TL et al. 576 Experimental design approach in recombinant protein expression: determining 577 medium composition and induction conditions for expression of pneumolysin from 578 Streptococcus pneumoniae in Escherichia coli and preliminary purification process. 579 BMC Biotechnol 2014;14:1. https://doi.org/10.1186/1472‐6750‐14‐1 580 [44] Garcia‐Fruitos E, Aris A, Villaverde A. Localization of functional polypeptides in 581 bacterial inclusion bodies. Appl Environ Microbiol 2007;73:289‐94. 582 https://doi.org/10.1128/AEM.01952‐06 583 [45] Upadhyay AK, Murmu A, Singh A, Panda AK. Kinetics of inclusion body formation 584 and its correlation with the characteristics of protein aggregates in Escherichia coli. 585 PLoS One 2012;7:e33951. 10.1371/journal.pone.0033951 [doi] 586 [46] Unzueta U, Cespedes MV, Sala R, Alamo P, Sanchez‐Chardi A, Pesarrodona M et al. 587 Release of targeted protein nanoparticles from functional bacterial amyloids: A death 588 star‐like approach. J Control Release 2018;279:29‐39. 589 https://doi.org/10.1016/j.jconrel.2018.04.004 590 [47] Hunig T. The storm has cleared: lessons from the CD28 superagonist TGN1412 trial. 591 Nat Rev Immunol 2012;12:317‐8. https://doi.org/10.1038/nri3192 592 [48] Suntharalingam G, Perry MR, Ward S, Brett SJ, Castello‐Cortes A, Brunner MD et al. 593 Cytokine storm in a phase 1 trial of the anti‐CD28 monoclonal antibody TGN1412. N 594 Engl J Med 2006;355:1018‐28. https://doi.org/10.1056/NEJMoa063842 595 [49] Yu Z, Reid JC, Yang YP. Utilizing dynamic light scattering as a process analytical 596 technology for protein folding and aggregation monitoring in vaccine manufacturing. J 597 Pharm Sci 2013;102:4284‐90. 10.1002/jps.23746 [doi];S0022‐3549(15)30802‐9 [pii] 598 [50] Li TM, Hook JW, III, Drickamer HG, Weber G. Plurality of pressure‐denatured forms 599 in chymotrypsinogen and lysozyme. Biochemistry 1976;15:5571‐80. 26 600 [51] Mohana‐Borges R, Silva JL, Ruiz‐Sanz J, de Prat‐Gay G. Folding of a pressure‐ 601 denatured model protein. Proc Natl Acad Sci U S A 1999;96:7888‐93. 602 [52] Ruan K, Weber G. Hysteresis and conformational drift of pressure‐dissociated 603 glyceraldehydephosphate dehydrogenase. Biochemistry 1989;28:2144‐53. 604 [53] Sanchez JM, Nolan V, Perillo MA. beta‐galactosidase at the membrane‐water 605 interface: a case of an active enzyme with non‐native conformation. Colloids Surf B 606 Biointerfaces 2013;108:1‐7. https://doi.org/10.1016/j.colsurfb.2013.02.019 607 [54] Randal M, Kossiakoff AA. The 2.0 A structure of bovine interferon‐gamma; 608 assessment of the structural differences between species. Acta Crystallogr D Biol 609 Crystallogr 2000;56:14‐24. 610 [55] Thiel DJ, le Du MH, Walter RL, D'Arcy A, Chene C, Fountoulakis M et al. Observation 611 of an unexpected third receptor molecule in the crystal structure of human interferon‐ 612 gamma receptor complex. Structure 2000;8:927‐36. 613 [56] Flores SS, Nolan V, Perillo MA, Sanchez JM. Superactive beta‐galactosidase 614 inclusion bodies. Colloids Surf B Biointerfaces 2019;173:769‐75. 615 https://doi.org/10.1016/j.colsurfb.2018.10.049 616 [57] Martinez‐Alonso M, Garcia‐Fruitos E, Villaverde A. Yield, solubility and 617 conformational quality of soluble proteins are not simultaneously favored in 618 recombinant Escherichia coli. Biotechnol Bioeng 2008;101:1353‐8. 619 https://doi.org/10.1002/bit.21996 620 [58] Ghobadi S, Ashrafi‐Kooshk MR, Mahdiuni H, Khodarahmi R. Enhancement of 621 intrinsic fluorescence of human carbonic anhydrase II upon topiramate binding: Some 622 evidence for drug‐induced molecular contraction of the protein. Int J Biol Macromol 623 2018;108:240‐9. https://doi.org/10.1016/j.ijbiomac.2017.12.011 624 [59] Wang L, Dong Q, Zhu Q, Tang N, Jia S, Xi C et al. Conformational Characteristics of 625 Rice Hexokinase OsHXK7 as a Moonlighting Protein involved in Sugar Signalling and 626 Metabolism. Protein J 2017;36:249‐56. https://doi.org/10.1007/s10930‐017‐9718‐x 627 628 629 FIGURE AND TABLE LEGENDS (DO NOT SUBMIT AS A SEPARATE FILE) 630 FIGURE 1 IFN‐constructs produced in E. coli and L. lactis (a) The general organization of 631 IFN‐constructs is configured (IFN‐)‐Linker‐APP‐H6. The amino acid sequences of the linker 27 632 and the APPS are shown below the schematic representation of the protein designs. (b) 633 Quantification of the production of IFN‐ in IB‐like nanoparticles in L. lactis (top) and 634 solubility (bottom) of IFN‐ in L. lactis. Significant results are shown as * p ≤ 0.05 and ** 635 p ≤ 0.005. 636 637 FIG 2 (a) Ultrastructural characterization by FESEM of protein aggregates and purified 638 protein nanoparticles of rBoIFN_, rBoIFN‐_CYOB, rBoIFN‐_HARLU and rBoIFN‐_L6K2. 639 Scale bars correspond to 200 nm. (b) ZP of purified protein nanoparticles. 640 641 FIG 3 Kynurenine levels measured by absorbance at 490 nm after treatment of EBTr cells 642 for 96 h with increasing amounts of rBoIFN‐ from different origins. (a) Soluble rBoIFN‐ 643 produced in the indicated expression system. (b) Protein nanoparticles of rBoIFN‐ 644 produced in L. lactis. The KD values are indicated in the plot. (c) Schematic representation 645 of the protein samples used in the activity assays: soluble protein obtained from the soluble 646 cell fraction, protein NPs purified from the insoluble cell fraction, soluble protein obtained 647 from the protein NPs, and the NP core after a resolubilization procedure. (d) Comparison of 648 the activity between rBoIFN‐ protein obtained from solubilization of protein NPs and 649 purified rBoIFN‐ from the soluble cell fraction as indicated at 0.72 nmol/L. Different letters 650 depict differences between proteins (p < 0.001) except rBoIFN‐ from protein NPs and 651 rBoIFN‐_E (p = 0.024). 28 652 653 FIG 4 Recombinant IFN‐ sizes in different supramolecular arrangements (purified soluble 654 IFN‐ and INF‐ IBs). (a) Soluble rBoIFN‐ from different origins: commercial rBoIFN‐_Std, 655 in‐house rBoIFN‐ from E. coli and L. lactis. (b) rBoIFN‐ IBs produced in L. lactis. (c) PBS 656 solubilized rBoIFN‐ from IBs after interferon release. (d) Scaffold of rBoIFN‐ IBs incubated 657 for 96 hours at 37 °C. The mean size and polydispersity index are indicated in brackets. The 658 average size data of the soluble proteins were analyzed by one‐way ANOVA (t corresponds 659 to P  0.07). 660 Table 1 Selection of APPs from predictions of “hot spots (HS)” of aggregation in 661 polypeptides by AGGRESCAN [31]. CYOB: Cytochrome bo3 ubiquinol oxidase subunit 1 662 from E. coli, HALRU: Aragonite protein AP7. NA: Not applicable. HS: hot spot. HSA: hot 663 spot area. NHSA: normalized HSA. a4vAHS: average aggregation‐propensity in each HS. 664 Table 2 Center of spectral mass (CSM) of IFN‐ protein preparations in soluble formats 665 or in protein NPs analyzed before and after the resolubilization protocol. 666 667 29