Introduction

Hexa-histidine tag (His-tag) is widely used to detect and purify recombinant proteins expressed in Escherichia coli (E. coli) (Terpe 2003). The His-tag allows for easy and fast purification owing to the strong affinity of histidine sequences to nickel-chelated nitrilotriacetic acid (Soh 2008). Standard techniques to identify recombinant proteins during and after purification are western blotting and ELISA using anti-His-tag antibodies, both of which are laborious and time-consuming processes. Therefore, alternative sensors have been developed for specific detection of His-tagged proteins. They include gold nanorod sensors using surface plasmon resonance (Wasowicz et al. 2008), impedance spectroscopy (Wasowicz et al. 2010), and field effect transistors (Liu et al. 2010). These detection systems are sufficiently sensitive (detection limits, ~5–60 μg/ml), and fast (10–30 min), but these are based on non-traditional equipment for a molecular biology laboratory. Immunochromatography (IC) is a rapid, sensitive, and economic screening method which has been developed for wide ranging applications including the immediate detection of disease markers (Li et al. 2014), virus infections (Peirui et al. 2014), and toxins in food (Lyubavina et al. 2014). At present, gold nanoparticles or quantum dots are used as probes to detect conjugation of target molecules to their specific antibodies spotted on membrane strips for IC (Hu et al. 2014; Berlina et al. 2013). However, metal nanoparticles tend to aggregate easily in salt-containing buffer solutions optimized for protein solvation due to their electrodynamic properties (Wang et al. 2012). This tendency has led to failures in applying the IC assay to particular target proteins. In this study, we introduced colorimetric dye-labeled cellulose nanobeads (CNBs) as detection probes for the quantification of His-tagged proteins in IC assay. The multiple hydroxyl groups on the glucose units of CNBs are expected to form hydrogen bonds with water, resulting in mono-dispersed CNBs in the buffer solution. A recombinant ATG8, which is known as an ubiquitin-like protein required for the formation of autophagosomal membranes (Shpilka et al. 2011; Klionsky and Schulman 2014), was detected by a competitive IC method, and the application of this IC assay to E. coli extracts was also examined. Though it was optimized for recombinant ATG8 detection, the potential to detect His-tag residues in recombinant proteins without protein-specific antibodies was assessed.

Materials and methods

Preparation of detection probes for competitive immunochromatographic (IC) assay

Detection probes for competitive IC assays were prepared by conjugation of hexa-histidine-tagged (His-tag) ATG8 (Sino Biological Inc., Beijing, China) on the surface of CNBs (CNBs; NanoAct, DCN Diagnostics, Carlsbad, CA) by hydrophobic interaction. ATG8 was incubated at 0.5 μg/ml with 0.01 % (w/v) CNBs dispersed in phosphate-buffered saline (PBS) for 30 min at room temperature. The unbound proteins were discarded by centrifugation at ~5,000×g and washed with PBS twice. Bovine serum albumin (BSA; 10 %, w/v) dissolved in PBS was added to the ATG8-conjugated CNB particles (A-CNBs) to give 1 % (w/v). The unbound BSA was removed by centrifugation at 500×g and washed twice with PBS. The particles were resuspended in the starting volume of PBS with 0.05 % (v/v) Tween 20.

Characterization of the detection probes

The morphology and size of bare CNBs and A-CNBs were analyzed by field emission scanning electron microscope (FE-SEM; S-4800, Hitachi). The hydrodynamic size and zeta potential of A-CNBs and bare CNBs were measured by dynamic light scattering (DLS; ZetaSizer NanoZS, Malvern Instruments).

Preparation of IC assay strips

The IC assay strips consisted of an absorption pad and nitrocellulose membrane (Fig. 1). Using a micropipette for the control and test dots, 0.3 μl of 0.1 μg/ml anti-BSA antibody (upper) and anti-His-tag antibody (below), respectively, were spotted onto the nitrocellulose membrane, and dried in a strictly controlled dessicator with a humidity of 50–60 %. The diameter of the spotted dot was ca. 3 mm.

Fig. 1
figure 1

A schematic representation of the immunochromatography (IC) assay with cellulose nanobeads (CNBs)

Full size image

Preparation of the standard curve

ATG8 proteins were diluted in PBS from 0.12 to 30 μg/ml. A 5 μl sample containing ATG8 was mixed with 45 μl A-CNB suspension (0.01 %, w/v) in 96-well plate. The end of assay strip was then immersed in the suspension mixed with the sample and A-CNBs were run along the strip. After 5 min, the strip was removed from the well and the spot image was captured by an image scanner (ImageQuant 350, GE Healthcare). The intensity of the spot was quantified by image analyzer (UN-SCAN-IT gel, ver. 6.1, Silk Scientific Co., Orem, UT). The signal value for each sample concentration was determined as the ratio of the intensity of the test to the control spot. A standard curve was prepared by plotting protein concentration versus signal value as calculated by image analysis.

Analysis of His-tagged protein concentration in complex mixtures

ATG8 proteins in complex mixtures were prepared by spiking ATG8 into 1 % (v/v) E. coli extract (Sigma-Aldrich Co.) in PBS. The signal value for each sample concentration was determined as the ratio of the intensity of the test to the control spot. The concentration of ATG8 in the E. coli extract was deduced from the standard curve of pure ATG8.

Cross-reactivity with other His-tagged proteins

To analyze the ability of this IC assay method to identify other His-tagged proteins, recombinant carbonic anhydrase I (CA-I; Cusabio, Hubei, China) and fructose bisphosphate aldolase (FBA; Cusabio) were prepared from 0.12 to 30 μg/ml. These samples were mixed with A-CNBs and run through the IC strip. The signal values were determined by measuring spot intensity. The concentration estimated by the IC assay was deduced from the ATG8 standard curve, and compared with the nominal concentration of the samples.

Results and discussion

Characterization of cellulose nanobeads (CNBs) conjugated by hexa-histidine-tagged (His-tag) proteins

The morphology and size distribution of the CNBs and (A-CNBs) were analyzed by scanning electron microscopy (SEM; Fig. 2a, b) and dynamic light scattering (DLS; Fig. 2c, d). The hydrodynamic size of CNBs dispersed in PBS was similar to the monomer size observed in SEM images, which indicated that the particles were nearly mono-dispersed in PBS. The larger radius of A-CNBs compared to that of bare CNBs could be explained by differences in protein conjugation on the CNB surfaces. The decreased negativity of A-CNBs measured by zeta potential was also attributed to the coverage of proteins on the electronegative hydroxyl residues of cellulose on the CNB surfaces.

Fig. 2
figure 2

Scanning electron microscopy (SEM) and dynamic light scattering (DLS) analyses of bare cellulose nanobeads, CNBs (a, c) and ATG8-conjugated CNBs, A-CNBs (b, d)

Full size image

Preparation of standard curves

The original ATG8 in sample solution was expected to compete with A-CNBs for anti-His-tag antibodies spotted on the strip, resulting in an inverse relationship between ATG8 concentration and colorimetric signal. When IC assays were performed with 0.12–30 μg/ml (7.5 nM to 19 μM) ATG8 in PBS, competitive binding of A-CNBs with the original ATG8 in the sample was occurred (Fig. 3a).

Fig. 3
figure 3

a The representative image of the immunochromatographic (IC) assay for pure ATG8; b A standard curve of pure ATG8 based on the signal value calculated from the relative intensity of the test to control spots (n = 3)

Full size image

The intensity of control spots dotted with anti-BSA antibody was increased as the ATG8 concentration increased. This could be attributed to the decreased amount of available A-CNBs to bind to the control spot after A-CNBs were bound to the test spot. Because the control spot intensity also varied in response to the sample concentration, the signal value from the IC assay was calculated by the relative intensity of the test to control spots.

A standard curve obtained using the signal values is shown in Fig. 3b. An inverse relationship between ATG8 concentration and the signal value was obtained and fitted by a negative logarithmic model (y = −0.193 ln(x) + 0.78) with a high correlation coefficient (R 2 = 0.9982).

Analysis of His-tagged protein concentration in E. coli crude extract

It is important to evaluate the concentration of His-tagged proteins in complex mixture such as E. coli extract. To avoid problems caused by high protein contents during chromatography, the total proteins in the sample solution should be optimized. In this study, the E. coli extract was diluted to 1 % (v/v) to prevent particle aggregation and nonspecific adsorption of proteins to the antibody spot or detection probes.

Figure 4a shows the IC assay results for ATG8 in 1 % E. coli extract. The signal values were determined by the relative intensity of the test to control spots, and the result was very similar to that of pure ATG8 as shown in Fig. 4b, with a high correlation coefficient (R 2 = 0.9906). To evaluate the effect of E. coli extract on the IC assay results, the sample concentration deduced by standard curve in Fig. 3b was plotted for the nominal concentration of spiked ATG8 (Fig. 4c). A linear relationship was observed, expressed by y = 1.1418x + 1.0449. The slope of this equation (a = 1.1418) was nearly one, which indicated that the signal value of His-tagged proteins in E. coli extract was nearly proportional to the standard curve with pure His-tagged proteins. This result implied that the detection probes for ATG8 quantification could be applied with high accuracy to complex mixtures such as E. coli extract.

Fig. 4
figure 4

a The representative image of immunochromatography (IC) assay strip and b a plot showing the concentration-dependent signal value for ATG8 in 1 % E. coli extract (n = 3); c The relationship between the nominal and calculated concentrations deduced by the ATG8 standard curve (n = 3)

Full size image

In the conventional purification process of recombinant proteins, the His-tagged protein concentration in E. coli extract is 4–5 mg/ml (Palmer and Wingfield 2004). Therefore, the detection range should be under ~40–50 μg/ml of target proteins considering the dilution ratio of the E. coli extract used in this study. The lowest concentration detected by the IC assay (0.12 μg/ml) was less than one-hundredth the required concentration under the above assumption.

As a result, the IC assay could be performed rapidly (within 2 min) to detect His-tagged proteins during purification from E. coli extract, without the laborious and time-consuming work involved in antibody binding and washing of non-specific substrates that are required for western blotting or ELISA.

Application to other His-tagged proteins

Because the anti-His-tag antibody was used for detection probe on the test spot, this IC assay could be applied to other His-tagged proteins, in theory. This was confirmed by the successful application of this system to carbonic anhydrase-I (CA-I) and FBA.

Table 1 shows the calculated concentrations of CA-I and FBA deduced by the IC standard curve using A-CNBs. Both CA-I and FBA were estimated to have a positive relationship between the input and calculated concentration deduced by the standard curve in Fig. 3b. However, the recovery differed between CA-I and FBA. These results could be explained by the difference in the molecular weight (MW) among ATG8 (15.9 kDa), CA-I (28.3 kDa), and FBA (39.4 kDa). Because the standard curve was prepared with ATG8, which was the protein with the smallest MW, the mole number of His-tag moieties for ATG8 would be largest compared to the same mass of CA-I and FBA. This interpretation was supported by the observation that the lowest recovery was observed for FBA, which had the largest MW.

Table 1 Concentration and its recovery for carbonic anhydrase-I (CA-I) and fructose bisphosphate aldolase (FBA) using ATG8-conjugated cellulose nanobeads (A-CNBs) by immunochromatograpic (IC) assay
Full size table

Although the IC assay system was prepared by optimization for ATG8 by controlling the conditions for conjugation, blocking agent, and antibody concentration, to achieve lowest detection limit and widest concentration range with linearity, other His-tagged proteins were also detectable by the same procedure. These results indicated that IC assay system prepared in this study could have the cross reactivity to various His-tagged proteins without protein-specific antibodies. However, the sensitivity was decreased for these alternative proteins.

Conclusions

A quantitative IC assay using dye-labeled polymers was developed to measure the concentration of recombinant hexa-histidine-tagged (His-tag) proteins with high regression coefficients. It is a simple and fast method and suitable for the monitoring of target molecules in E. coli cell lysates as a pre-test for western blotting and ELISA. To our knowledge, the assay presented in this report is the first IC assay developed for the detection of recombinant His-tagged proteins.

This assay system has the potential to reduce the workload associated with recombinant protein expression considerably. In addition, this study is an example of the successful application of dye-labeled polymer beads to the probes for an IC assay.