Metals are involved in many physiological processes, with many proteins requiring a metal cofactor to perform their functions. [[1], [2], [3], [4], [5]] The formation of a metal binding site is a result of the overall three-dimensional structure of the protein arranging in such a way to bring the required amino acids to the correct spatial orientation. [6] The metal cofactor can assist in the protein folding process [[7], [8], [9]] or can be required for formation of the tertiary structure. [10,11] Metallothioneins (MTs) are specialist proteins that bind multiple metal ions in binding sites that only form upon the metallation.

MTs were initially discovered as an equine Cd(II)-binding protein [12] and isolated in humans as a Zn(II), Cd(II), and Cu(I) protein. [13] Since then, MTs have been isolated from numerous organisms and are now considered ubiquitous in life. MTs are cysteine-rich proteins thought to be involved in metal homeostasis and heavy metal detoxification [14,15], with upregulation occurring via the MTF-1 pathway in some vertebrate isoforms. [16] MTs are intrinsically disordered in the absence of metals (apo-MTs) but will form more ordered structures in the presence of metals. Thus, the binding of metals is sometimes referred to metal-induced folding. [17] These structures formed depend on factors such as the species of MT, the metal identity, and the metal stoichiometry.

The most well-known structure comes from the single available mammalian MT crystal structure of Cd5Zn2MT, which exhibited a two-domain structure. [[18], [19], [20]] Mammalian MTs contain 20 conserved cysteine residues. [21] The N-terminal β-domain binds three divalent metals with an M3S9 stoichiometry and the C-terminal α-domain binds four divalent metals with an M4S11 stoichiometry. These cluster structures involve a network of both bridging (one cysteine bound to two metals) and terminal (one cysteine bound to only one metal) cysteines. [[22], [23], [24]]

Although mammalian MTs fully metallated with Zn(II) and Cd(II) have been studied extensively, very much less is known about the relevant apo-MT and partially metallated species that likely form in vivo. [[25], [26], [27]] However, due to their fluxional nature, these structures are more difficult to characterize than those of fully metallated MTs. ESI-MS is a powerful technique to study partially metallated states of MT due to the ability to observe individual species and their relative abundances rather than an average in solution.

Through electrospray ionization mass spectrometry (ESI-MS) and circular dichroism (CD) spectroscopy methods, it has been shown that mammalian MTs can bind Zn(II) or Cd(II) in one of two different pathways. [[28], [29], [30], [31], [32]] The “cluster” pathway describes a cooperative formation of M4S11 in the α-domain, followed by M3S9 in the β-domain without the formation of any stable intermediates. [28,29] The non-cooperative pathway involves the sequential binding of each metal initially to terminal cysteines, where all intermediates are observed. [28,[30], [31], [32]] It has been proposed that up to five metals each coordinate to four separate terminally bound cysteines, forming Zn5S20 or Cd5S20 structures, followed by rearrangement into bridging metal-thiolate clusters upon further binding. [28,30]

These two pathways are highly pH dependent. [28] Cd(II) binds in the cooperative pathway below pH 6, and in the non-cooperative pathway above pH 8. Between these pHs, a sigmoidal transition between these pathways is observed. A mixed pathway is observed in this range. Zn(II) binds in the cooperative pathway below pH 4, and in the non-cooperative pathway above pH 6. Again, a mixed pathway is observed between these pHs. At physiological pH (7.4), Cd(II) binds in a mixed cooperative and non-cooperative pathway, while Zn(II) binds entirely non-cooperatively.

Tandem MS studies have been performed in an attempt to elucidate the partially metallated binding sites of Zn(II) and Cd(II). [[33], [34], [35]] These studies suggest that Zn(II) and Cd(II) have different binding pathways, with Cd4MT localized in the α-domain and Zn4MT spanning both domains. [[33], [34], [35]] Conversely, similar studies on Zn4MT and Cd4MT structures suggest higher affinities of these metals to the middle of the protein, between Asn18 to Cys38 in MT2. [36]

The change in metallation kinetics under various conditions can also provide insight into the binding reactions. The pH-dependent binding kinetics of both Zn(II) and Cd(II) to apo-MT, which occur on the millisecond timescale, have been observed previously using stopped flow methods at 25 °C. [17] That study was concerned with describing the folding of MT in terms of the metal binding pathway, contrasting the formation of secondary structural elements and a hydrophobic core in typical proteins. At neutral pH, much of the binding reaction was completed within the dead time of the instrument used and thus no kinetic data at physiological pH (7.4) for apo-βαMT were obtained. [17]

The knowledge of pH-dependent binding pathways as well as pH-dependent binding kinetics prompted our group to study the kinetics of non-cooperative binding at pH 8 and cooperative cluster formation at pH 5. Stopped-flow methods at low temperatures overcame the previous limitation preventing kinetic observation at high pHs. [5] These results showed the non-cooperative Cd(II) binding at pH 8 was faster than the cooperative Cd4MT cluster formation at pH 5. It was proposed that this difference in rate was due to the concerted cluster formation involving 4 Cd(II) requiring more time than the binding of individual Cd(II) non-cooperatively. However, the likely role of thiol protonation and solution H+ concentration driving this process was not thoroughly considered.

As well, Cd(II) metallation in the presence of a denaturant was shown slowed the binding, implying that the structure of the initial apo-MT is optimal for its metal-binding function. [5,37] This is significant, because apo-MT appears to take on a compact structure which can be unfolded by a denaturant. [[38], [39], [40], [41], [42], [43], [44]] The mammalian MT isoforms MT1a and MT3 have recently been shown to have differences in their apo structures, with the cysteines of MT3 existing in a more collapsed conformation with the cysteines buried in the center of the protein. [43] Stopped-flow kinetic studies have shown that apo-MT3 binds Cd(II) with a rate constant approximately 1.7 times larger than apo-MT1a. [43] This also gives an indication that the apo-MT structure impacts the rate of metal binding.

Although changes in kinetics can indicate both structural changes in the protein and changes in the pathway of metal binding, the information gained using stopped-flow kinetics monitored by absorption spectroscopy alone is limited. This is because only a single kinetic trace is obtained, which corresponds to seven overlapping bimolecular reactions. Only an average of M1-7MT species is observed and only a single rate constant can be reported. For this reason, ESI-MS is an excellent technique to pair with stopped-flow methods. Stopped-flow provides quantitative data measured in the solution phase which can be combined with the level of detail seen using ESI-MS.

In this paper, stopped-flow kinetics and ESI-MS are combined to probe the pH and structural dependence of apo-MT on the rates of Zn(II) and Cd(II) binding to apo-MT. The effect of different pH conditions from pH 5 to pH 7.4 and implications for the rate of different binding pathways are reported and discussed. The rates of Zn(II) and Cd(II) metallation as a function of GdmCl concentration are reported. The different structures formed during a titration of Zn(II) and Cd(II) into native and denatured apo-MT are monitored using CD spectroscopy. ESI-MS studies reveal the metal-thiolate stoichiometries in the first few metallation steps of Zn(II) and Cd(II) to native and denatured apo-MT.

To date, the dependence of Zn(II) and Cd(II) metallation properties on the initial apo-MT structure has not been studied using both stopped-flow kinetics to analyze the difference in rate constants and ESI-MS to visualize the individual MxSy formed. Additionally, the pH-dependence of both Zn(II) and Cd(II) binding to apo-MT has not been fully described.