We present an assay for easy quantification of metals introduced to samples prepared using immobilized metal affinity chromatography. The method uses hydroxynaphthol blue as the colorimetric metal indicator and a UV-Vis spectrophotometer as the detector.
Contamination of enzymes with metals leached from immobilized metal affinity chromatography (IMAC) columns poses a major concern for enzymologists, as many of the common di-and trivalent cations used in IMAC resins have an inhibitory effect on enzymes. However, the extent of metal leaching and the impact of various eluting and reducing reagents are poorly understood in large part due to the absence of simple and practical transition metal quantification protocols that use equipment typically available in biochemistry labs. To address this problem, we have developed a protocol to quickly quantify the amount of metal contamination in samples prepared using IMAC as a purification step. The method uses hydroxynaphthol blue (HNB) as a colorimetric indicator for metal cation content in a sample solution and UV-Vis spectroscopy as a means to quantify the amount of metal present, into the nanomolar range, based on the change in the HNB spectrum at 647 nm. While metal content in a solution has historically been determined using atomic absorption spectroscopy or inductively coupled plasma techniques, these methods require specialized equipment and training outside the scope of a typical biochemistry laboratory. The method proposed here provides a simple and fast way for biochemists to determine the metal content of samples using existing equipment and knowledge without sacrificing accuracy.
Since its inception by Porath and co-workers1, immobilized metal affinity chromatography (IMAC) has become a method of choice to quickly separate proteins based on their ability to bond with transition metal ions such as Zn2+, Ni2+, Cu2+, and Co2+. This is most commonly done via engineered poly-histidine tags and is now one of the most common chromatographic purification techniques for the isolation of recombinant proteins2. IMAC has also found applications beyond recombinant protein purification as a way to isolate quinolones, tetracyclines, aminoglycosides, macrolides, and β-lactams for food sample analysis3 and as a step in identifying blood-serum protein markers for liver and pancreatic cancers4,5. Not surprisingly, IMAC has also become a method of choice for the isolation of a number of native bioenergetics enzymes6,7,8,9,10. However, successful implementation of these purification methods for studies on enzymatically active bioenergetic proteins is dependent on the presence of negligible levels of metal cations leached from the column matrix into the eluate. The divalent metal cations commonly used in IMAC have known pathologic biological significance, even at low concentrations11,12. The physiological effect of these metals is most pronounced in bioenergetic systems, where they can prove lethal as inhibitors of cellular respiration or photosynthesis13,14,15. Similar issues are unavoidable for the majority of protein classes where residual contaminant metals can interfere with a protein's biological functions or characterization with biochemical and biophysical techniques.
While the levels of metal contamination under oxidizing conditions and using imidazole as an eluant are typically low16, protein isolations performed in the presence of cysteine reducing agents (DTT, β-mercaptoethanol, etc.) or with stronger chelators like histidine17,18 or ethylenediaminetetraacetic acid (EDTA) result in much higher levels of metal contamination19,20. Similarly, since metal ions in IMAC resins are frequently coordinated by carboxylic groups, protein elutions performed under acidic conditions are also likely to have much higher levels of metal contamination. Metal content in solutions can be assessed using atomic absorption spectroscopy (AAS) and inductively coupled plasma-mass spectrometry (ICP-MS) down to a limit of detection in the ppb-ppt range21,22,23,24. Unfortunately, AAS and ICP-MS are not realistic means for detection in a traditional biochemistry lab as those methods would require access to specialized equipment and training.
Previous work by Brittain25,26 investigated the use of hydroxynaphthol blue (HNB) as a way to identify the presence of transition metals in solution. However, there were several internal contradictions in the data20 and those works failed to offer an adequate protocol. Studies by Temel et al.27 and Ferreira et al.28 expanded on Brittain's work with HNB as a potential metal indicator. However, Temel developed a protocol that makes use of AAS for sample analysis, using HNB only as a chelating agent. Ferreira's study used the change in the HNB absorbance spectra at 563 nm, a region of the free-dye HNB spectra that overlaps heavily with the spectra of HNB-metal complexes at pH 5.7, making the assay sensitivity fairly low as well as resulting in relatively weak metal binding affinity20. To address issues in our own lab with Ni2+ leaching from IMAC, we have expanded the work done by Brittain25,26 and Ferreria28 to develop an easy assay capable of detecting nanomolar levels of several transition metals. We showed that HNB binds nickel and other common for IMAC metals with sub-nanomolar binding affinities and form 1:1 complex over a wide range of pH values20. The assay reported here is based on these findings and utilizes absorbance changes in the HNB spectrum at 647 nm for metal quantification. The assay can be performed in the physiological pH range using common buffers and instrumentation found in a typical biochemistry lab by using colorimetric detection and quantification of metal-dye complexes and the associated change in absorbance of the free-dye when it binds to metal.
1. Assay component preparation
2. Sample preparation and measurement
3. Metal quantification
The spectrum of free HNB at neutral pH (black line) and representative spectra of fractions assayed for Ni2+ from the isolation of MSP1E3D129 are shown in Figure 2. A successful assay series should demonstrate a decreased absorbance at 647 nm compared to the HNB control, which corresponds to the formation of HNB complexes in the presence of a transition metal. A failed assay would be indicated by an increase in absorbance at 647 nm. Alternatively, more than 90% decrease from initial absorbance at 647 nm would indicate too high metal content and a need to assay more diluted fractions. An assay with no absorbance changes from the free HNB control does not necessarily indicate a failure. It is possible that samples contain essentially no leached metals. However, this is unlikely and any sample showing no absorbance change should be prepared and measured again, preferably with less dilution, to confirm the result. In total, most failures to observe the expected absorbance change can be attributed to improper pipetting during sample preparation, an inadequate incubation time prior to measurement, or pH values outside the recommended 7-12 range.
To demonstrate the application of this assay, we analyzed 2 His-tag membrane scaffold proteins MSP1E3D1 (isolated as in Denisov, I. G. et al.29), MSP2N2 (isolated as in Grinkova, Y. V.30), and a novel 3-heme c-type cytochrome GSU0105 from Geobacter sulfurreducens, which was recombinantly expressed in E.coli and eluted with 500 mM imidazole. The elution profiles of Ni2+ from a Ni-NTA resin column (see Table of Materials) and the associated protein elution profiles for these 3 proteins are shown in Figure 3. Any protein will have a unique nickel elution that may or may not align with the protein elution profile as measured at 280 nm. For example, Figure 3C shows that the protein and Ni2+ content of each fraction for GSU0105 are significantly shifted from one another while the fractions for MSP1E3D1 and MSP2N2 (Figure 3A,B) that contain the most protein also have the highest Ni2+ content. Figure 3A,B also illustrate that metal content may not be evenly distributed among fractions collected using IMAC. Depending on column packing, the composition of the elution buffer, pumping equipment and conditions, it is possible to have metal elute in consecutive fractions at very different concentrations independent of the protein content of those fractions.
Figure 1: Structure of hydroxynaphthol blue (HNB). In the functional pH range of the assay, all sulfonate groups and one of the hydroxyl groups are ionized. Please click here to view a larger version of this figure.
Figure 2: Absorbance spectra for selected MSP1E3D1 fractions and HNB. The relative absorbance of three fractions of MSP1E3D1 (colors) compared to an HNB control (black thick line) are shown. Samples were prepared in 20 mM Tris, pH 7.5. Please click here to view a larger version of this figure.
Figure 3: Ni2+ quantification for 3 representative His-tag proteins. Protein and Ni2+ elution profiles for (A) MSP1E3D1, (B) MSP2N2, and (C) GSU0105. Protein elution was performed using 300 mM, 300 mM, and 500 mM imidazole, respectively. Ni2+ quantification was performed in 20 mM Tris, pH 7.5. Please click here to view a larger version of this figure.
Colorimetric detection of metals using HNB provides a simple way to quantify the degree of protein contamination by transition metal ions from IMAC resins. As we established in Ref. 20, Ni2+ binds to HNB with 1:1 stoichiometry and the dissociation constant for the Ni-HNB complex changes with pH. However, the complex Kd is in the sub-nM range for all recommended (7-12) pH values. In practical terms, it means that all Ni2+ in any tested fractions will bind to HNB as long as no other strong chelating reagents, like EDTA, are present. All these properties together result in linear Ni2+ titration curves, which we experimentally observed. In that report20, we also established that the spectral changes due to metal-dye complex formation will be the same over the entire 7-12 pH range. The detection is limited by the minimal reliable absorbance change measurements (10-4 – 10-3 OD depending on the spectrophotometer used) corresponding to 2.7-27 nM Ni2+. The upper range is limited by the amount of HNB present. In our work, we use ~15 µM, corresponding to ~0.6 OD at 647 nm. However, this can be increased up to 50-80 µM HNB, if needed. Practically, we observed Ni2+ contamination levels in chromatographic fractions comparable or higher than the upper limit forcing us to make 10- to 50-fold dilutions of assayed fractions. However, this additional dilution step can increase relative errors while determining the nickel concentration in a fraction.
Though we have not investigated the details, it appeared that binding of other metals used in IMAC resins (Co, Zn, Fe) also has sub-µM dissociation constants and virtually no overlap between dye and dye-metal absorbance spectra at 647 nm, the peak wavelength of free HNB. Therefore, complete metal binding to the dye and the associated spectral changes of the dye can be used for absolute metal determination over the entire recommended pH range.
Execution of the protocol is straightforward and depends most heavily on proper laboratory technique. Modern spectrophotometers have highly linear responses and dynamic ranges of 3-4 orders of magnitude. Consequently, the most likely place for the introduction of error in the method is through the pipetting steps for sample preparation. As described in this text, the method is based on the quantification of metal content based on the difference in the HNB absorbance peak at 647 nm from a free HNB control and samples with HNB complexed with a metal. If care is not taken to accurately pipette the HNB aliquots or the buffer volumes, comparison of the control and sample absorbances at 647 nm becomes a point of error. Similarly, poor pipetting of protein fractions for sample preparation can skew the perceived concentration of metal in a fraction. It is recommended that, due to the sensitivity of the assay, any pipettes being used for analyses where precise quantification is required be calibrated prior to use.
The primary limitations of the method come with the functional pH range of the assay and the presence of strong chelating agents. The assay is best utilized in a pH range from 7-12. Below pH 7, the spectrum of the free HNB dye changes, losing the peak at 647 nm used for quantification20. Above pH 12, many metal hydroxides begin to precipitate, including those of metals commonly found in IMAC resins, making quantification slower and less reproducible. While the alkaline maximum does not pose a significant problem as purification procedures rarely call for such a high pH, the acidic minimum is more likely to be a limiting factor. Since the detection limits for Ni2+ and other transition metals are approximately 1000-fold lower than metal contamination levels demonstrated above (Figure 3), the low pH limit for the assay can be circumvented by dilution of the assayed acidic protein fractions in buffers with neutral pH values and sufficiently high buffering capacity. Alternatively, the pH of analyzed fractions can be adjusted or the HNB stock solution can be more strongly buffered to maintain the desired pH after mixing.
If the isolation procedure for the protein being purified requires the use of reagents with known or suspected transition metal chelating properties, a modification of the method would be necessary to allow for proper quantification of leached metals. A standard curve would need to be prepared using the chelating agent used for protein elution and known concentrations of metal standards to accurately quantify the concentration of leached metal in the presence of the chelator. An example of metal quantification in the presence of histidine is available in Kokhan & Marzolf20.
Accurate quantification of metals in biological samples is still largely dependent on the use of analytical techniques and instrumentation, such as AAS and ICP-MS, that remain outside the realm of the typical biochemist31,32. Bonta et al. have described the simple preparation of biologic samples on common filter paper for analysis by ICP-MS, however, their method still relies on non-standard instrumentation for a biochemist31. The method we describe allows for the measurement of metal content in a sample to be taken without additional training on new instrumentation or outsourcing to others. Similar colorimetric protocols have been developed for metal analysis in biological samples33. However, the method described by Shaymal et al.33 relies on a fluorescence assay using a commercially unavailable fluorescent probe that gives a higher limit of detection than that in this paper. Considering the relative ease by which the described method can be performed and the recent interest in the development of portable metal detecting protocols for aqueous samples34,35, it could be easily adapted for field testing of water samples. As a portable test, our method could be modified for use with a portable spectrophotometer for quantification or as a qualitative measure to identify samples for further analysis at a fixed testing location.
The authors have nothing to disclose.
This material is based upon work supported by the National Science Foundation under Grant MCB-1817448 and by an award from the Thomas F. and Kate Miller Jeffress Memorial Trust, Bank of America, Trustee and specified donor Hazel Thorpe Carman and George Gay Carman Trust.
2xYT broth | Fisher Scientific | BP9743-500 | media for E.coli growth |
HEPES, free acid | BioBasic | HB0264 | alternative buffer |
HisPur Ni-NTA resin | Thermo Scientific | 88222 | |
Hydroxynaphthol blue disoidum salt | Sigma-Aldrich | 219916-5g | |
Imidazole | Fisher Scientific | O3196-500 | |
Imidazole | BioBasic | IB0277 | |
MOPS, free acid | BioBasic | MB0360 | alternative buffer |
Sodium chloride | Fisher Scientific | S271-500 | |
Sodium phosphate | Fisher Scientific | S369-500 | alternative buffer |
Tricine | Gold Bio | T870-100 | |
Tris base | Fisher Scientific | BP152-500 | |
Triton X-100 | Sigma-Aldrich | T9284-500 |