Advantages and Disadvantages of His-Tagged Beta-Galactosidase

23 β -Galactosidase is one of the most important biotechnological enzyme used in the dairy industry, 24 pharmacology and in molecular biology. In our laboratory we have overexpressed a recombinant β - 25 galactosidase in Escherichia coli ( E. coli ). This enzyme differs from its native version ( β -Gal WT ) in that 26 6 histidine residues have been added to the carboxyl terminus in the primary sequence ( β -Gal His ), which 27 allows its purification by immobilized metal affinity chromatography (IMAC). In this work we 28 compared the functionality and structure of both proteins and evaluated their catalytic behavior on the 29 kinetics of lactose hydrolysis. We observed a significant reduction in the enzymatic activity of β -Gal His 30 with respect to β -Gal WT . Although, both enzymes showed a similar catalytic profile as a function of 31 temperature, β -Gal His presented a higher resistance to the thermal inactivation and evidenced greater 32 half- life time compared to β -Gal WT . At room temperature, β -Gal His showed a fluorescence spectrum 33 compatible with a partially unstructured protein however, it exhibited a lower tendency to the thermal- 34 induced unfolding with respect to β -Gal WT . Analytical ultracentrifugation experiments demonstrated that 35 the population of β -Gal His molecules exhibited a higher proportion of monomers and a lower proportion 36 of tetrameric species with respect to the His-tag free protein. The impairment of tetramerization may 37 would explain the negative effect of the presence of His-tag on the enzymatic activity. In addition, the 38 present results, analyzed in the context of the available literature, suggest that the effect of the His-tag 39 is protein-specific. 40 41 42

4°C and the supernatant was then incubated with 1 mL Ni-NTA beads (ProBond TM resin Invitrogen) for 103 3h at 4ºC by tumbling. Then the beads were poured in cartridges and washed with imidazole solution of 104 0, 60, 100, 200 and 400 mM. Each elution sample was dialyzed against 0.05 M phosphate buffer (pH 105 6.8) at 4ºC for 20 h with 3-4 changes of buffer. 106 The sample eluted with the 200 mM imidazole contained a considerable amount of highly pure β-Gal 107 tested by SDS-PAGE and finally the pure enzyme was stored at 4ºC in dialysis buffer (at 3 mg/mL). 108 109 β-Gal SDS-PAGE 110 The β-Gal purity was evaluated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-111 PAGE) (10%). Samples were heated at 100 °C during 5 min. in loading buffer 23  The reaction was carried out in phosphate buffer 0.05 M, pH 6.8 using monohydrated Lactose (Anedra) 121 as substrate. It was initiated by the addition of 0.01 mL of the enzyme preparation containing (0.001 mg 122 to 0.01 mg of β-Gal) and incubated for 20 min at 37°C. Then, the reaction was quenched by boiling 5 123 min. β-Gal activity was measured by quantitative analysis of the glucose released, determined by the 124 glucose oxidase method 25 . One unit of enzyme activity was defined as the amount of enzyme required 125 to produce 1 μmol of glucose/min in the described conditions. 126 The values of KM and Vmax were determined by fitting the Michaelis-Menten equation to the Vo versus The thermal activity profiles of β-GalWT and β-GalHis were evaluated from 25°C to 65°C at pH 6.8. The 131 other reaction conditions were described above. 132 133 2.2.3.3. Thermal stability of β-Gals. 134 The thermal stability of β-GalWT and β-GalHis was studied by measuring the residual catalytic activity 135 determined as described above after a pre-incubation period of 20 min. at different temperatures (20 -136 60ºC Residual activity was measured as described above, a closed cap tube containing 3 µl of β-Gal (0.01 140 g/L), preincubated at 50ºC during a time ranging from 2 to 20 min. After this time, 300 µl of lactose 141 (200 mM) was added and the specific enzyme activity was measured as described above at 37º. The 142 profile of residual activity was plotted vs time. 143 A single exponential decay mathematical approach was applied to fit the curve of specific activity (SA) 144 vs time, in order to determine the half-life time (τ1/2). 145 Fluorescence spectra were recorded in a Fluoromax Spex-3 JovinYvon (Horiba, New Jersey, USA) 160 spectrofluorimeter. A quartz cell with 10 mm path length and a thermostated holder was used. The slits 161 and λex were set at 2 nm and 290 nm, respectively. Emission spectra were acquired within the 300-400 162 nm range. Protein concentration used was 0.2 mg/ml. Raman scattering contribution from water was 163 subtracted in all spectra. To facilitate comparisons, max was determined but also the center of spectral 164 mass (CSM) was calculated for each fluorescence emission spectra 26    where vt is the sedimentation rate, r is the distance of the particle from the axis of rotation, M the molar 188 mass of the particle and ω is the angular velocity of the rotor equation (6). S serves to normalize the vt 189 of a particle by the acceleration applied to it (r. ω 2 ). The resulting value is independent on the 190 acceleration but depends on the properties of the particle (the mass m and the hydrodynamic radius r0) 191 and the viscosity of the medium (η) where it is suspended as shown by equation (7). Experimentally, 192 the absorbance profile obtained by analytical ultracentrifugation is described by the Lamm equation

β-Gals functionality. 206
In most cases, the incorporation of His-tag has resulted in great increases in the efficiency of the protein 207 purification process however, sometimes, it hampers the protein functionality. Then, several 208 experiments were conducted to remove the additional histidine residues but drawbacks remain if non-209 wild type amino acids remain in the protein 18 . In our experiments we evaluate the enzymatic activity of 210 the wild type β-GalWT, the commercial enzyme from Sigma, and a recombinant β-GalHis that we β-GalHis exhibits lower specific activity with respect to β-GalWT (Fig.1, Table 1); a Vmax decrease of 213 around 60 % is observed for β-GalHis with respect to β-GalWT. On the other hand, the presence of histidine 214 residues also seems to moderately favor the affinity of the active sites for lactose (β-GalHisKM < -215 GalWTKM), although, the difference observed in the later parameter was not statistically significant. We 216 ruled out the possibility that the difference in the kinetic parameters were due to impurities in the protein 217 sample. SDS-PAGE (Fig 1.b) showed that both samples only present the -gal monomer (116 kD) band.   denaturation. Results are shown in Fig. 2 and Table 1. Both enzymes exhibit the same catalytic profile 249 vs reaction temperature with an optimal activity at around 45°C (Fig. 2a). Moreover, the negative effect 250 of high temperature was much more important for the wild type than for the His-tagged enzyme: when 251 the enzymatic activity is measured at 50°C, β-GalHis maintains approximately 50% while β-GalWT only resistance to temperature inactivation compared to β-GalWT as shown in Fig. 2b. In order to evaluate this 254 difference, the half-life time of each enzyme (Fig. 2c) was studied. The preincubation of the enzymes at 255 50ºC β-GalHis reflects this tendency and β-GalHis seems to be more stable with respect to β-GalWT (Table  256 1). The effect of His-tag on proteins thermal stability is different depending on the protein but in general, 257 a negative effect is observed. Chen and his collaborators 36  On the other hand, Booth and coworkers 18 using differential scanning fluorimetry studied the 262 thermostability for a set of ten different N-Terminal His-tagged proteins and found that in almost all 263 cases, the His-tagged version was less stable than the wild type protein. This scenario encourages us to explore the structural features underlying the differential behavior of β-265 GalWT and β-GalHis.  Table 1).

Fluorescence Spectroscopy. 278
The fluorescence emission of Trp is sensitive to the polarity of the environment and can be used to detect 279 conformational changes in proteins 26 . β-Gal is a tetrameric protein with 39 Trps residues per protomer 280 30 . β-GalWT fluorescence spectrum shows a λmax at 345 nm and CSM=356 nm (Fig.3a, Table 2) which is 281 compatible with Trp residues localized in highly polar environment if compared with the spectrum of 282 Trp localized in a buried or non-polar medium (λmax~325 nm). It is noteworthy that we tested the Mg 2+ 283 free protein. This condition could contribute to the higher value of λmax in β-GalWT spectrum compared 284 to those previously reported for the same protein 37 . For β-GalHis, the spectrum shows an important 285 bathocromic shift, with a λmax at 347 nm and CSM=360 nm (Fig. 3a, Table1). These results indicate that 286 Trp residues became, on average, more accessible to the solvent and that the His-tag could induce at 287 least a partial unfolding of β-Gal or, on average, a highly hydrated structure. 288 However, when λmax was evaluated along heating (Fig. 3b) we observe that the His-tagged protein shows 289 a greater thermal stability with respect to the wild type protein (compare the magnitude of the thermal 290 λmax increase up to 50ºC in both samples). In our laboratory we have also described an increase in the 291 thermal stability of -GalWT when interacting with model membranes or in molecular crowded 292 conditions 37,38 . The present results indicate that the covalent modification on -GalWT leading to -293 GalHis also accounts for the increase resistance to inactivation that displays the latter compared with the 294 former (Fig. 2b). Moreover, at 50ºC and above we observe the major differences in the max value 295 between both proteins. It is noteworthy that at 50ºC β-GalHis shows lower values of λmax with respect to 296 β-GalWT. and as consequence, an opposite structural behavior respect to what was observed at 25ºC. hand, Wu and Filutowics 43 proposed that His tagged proteins may differ from their wild type 308 counterparts in dimerization/oligomerization and afterwards, many researchers lead to similar 309 conclusion. Also, His-tag promoted the dimerization of HSC70 C30ΔL-His but had no effect on the elution 310 profile of HSC70 C30WT-His(+), compared to their respective untagged forms 44 . Moreover, Kenig and 311 her colleagues 45 demonstrated that the His-tagged proteins were located intracellularly as soluble 312 proteins and also in an aggregated form as inclusion bodies. In contrast, the non-tagged proteins were 313 found only in the soluble form and this fraction was used for further purification studies. With those 314 concepts in mind we decided to evaluate the oligomeric state of our β-Gals. 315 In order to evaluate the effect of His-tag on the supramolecular organization of the enzyme we developed 318 AUC. Both enzymes presented a heterogeneous profile, but the presence of histidine seems to increase 319 the population of monomers (Table 2, Fig.4). Besides, it was remarkable the lower proportion of the 320 tetramers' population found in GalHis if compared with β-GalWT. These phenomena contribute to explain 321 the negative effect of the presence of a His-tag on the enzyme activity.  Another important finding was the lower size (or hydrodynamic radius) of the β-GalHis tetramer if 331 compared with that of β-GalWT which may also be associated with an inactive conformational structure 332 of the protein. All these results could explain the decreased in His-tagged protein functionality. This 333 statement is supported by the fact that the active site is made up of elements from two subunits of the 334 tetramer, and disassociation of the tetramer into dimers removes critical elements of the active site 30 . 335 336

Conclusions 337
In the present work, through kinetic, fluorometric and ultracentrifugation analysis, we demonstrated that 338 the addition of a His-tag to E. coli β-Gal reduces the catalytic activity possibly through an impairment 339 of the proper acquisition of the typically active quaternary structure (tetramers). However, the tendency 340 to form higher oligomeric structures may explain the improvement in the thermal structural and 341 functional stability. It is important to consider that, in view of the present study and the available 342 literature, the effect of His-tag seems to be protein-specific. 343 344 Declarations 345

Ethics approval and consent to participate 346
Not applicable 347

Consent for publication 348
Not applicable 349

Availability of data and materials 350
The datasets during and/or analysed during the current study available from the corresponding author 351 on reasonable request.   Thermal and temporal inactivation pro les of -Gals.a) Speci c activity measured at the indicated temperature within the range 22ºC -65ºC. b) β-Gal speci c activity measured at 37ºC after preincubation for 20 min. at different temperatures (20-60ºC). In each data set normalization was done with respect to the corresponding speci c activity preincubated at 30ºC, which was taken as the unity. c) Catalytic activity was evaluated at 37ºC, pH 6.8 and lactose 200 mM after preheating the proteins (0.01 g/L) at 50ºC during a x time period (20-60 min). Exponential decay curves (dashed lines) are the tness to the experimental points (black and white circles) and allowed estimating the half-life times (shown in Table   1).

Figure 3
Structural analysis of -Gals a) Intrinsic uorescence spectra of of -GalHis (dotted line) and -GalWT (full line) proteins at 0.2 g/L at 25°C. b) Effect of temperature on the max of -GalHis () and -GalWT ().