In situ observation of peptide bond formation at the water–air interface

Abstract

We report unambiguous spectroscopic evidence of peptide bond formation at the air–water interface, yielding a possible mechanism providing insight into the formation of modern ribosomal peptide bonds, and a means for the emergence of peptides on early Earth. Protein synthesis in aqueous environments, facilitated by sequential amino acid condensation forming peptides, is a ubiquitous process in modern biology, and a fundamental reaction necessary in prebiotic chemistry. Such reactions, however, are condensation reactions, requiring the elimination of a water molecule for every peptide bond formed, and are thus unfavorable in aqueous environments both from a thermodynamic and kinetic point of view. We use the hydrophobic environment of the air–water interface as a favorable venue for peptide bond synthesis, and demonstrate the occurrence of this chemistry with in situ techniques using Langmuir-trough methods and infrared reflection absorption spectroscopy. Leucine ethyl ester (a small amino acid ester) first partitions to the water surface, then coordinates with Cu²⁺ ions at the interface, and subsequently undergoes a condensation reaction selectively forming peptide bonds at the air–water interface.

Protein synthesis (condensation of amino acids through sequential peptide bond formation) is a fundamental and ubiquitous reaction in biology. Aqueous media are the required environments in which this chemistry takes place; however, protein synthesis is unfavorable in aqueous solution. In modern biology, the condensation reactions necessary in the formation of peptide bonds are facilitated catalytically by the large subunit of the ribosome. The mechanism of this catalyzed reaction originally proposed by Nissen et al. in 2000 (1) involved favorable orientation of peptide precursors, acid-base catalysis and transition-state stabilization, and the altered pK_a of the functional groups of the precursors caused by the reaction environment provided by the active site; such pK_a shifts had previously been seen in the active sites of other proteins (2). Since then, the original mechanism has been contested (3, 4). The acknowledged mechanistic function of the ribosome’s active site is its ability to bring the precursors in close proximity and orient them for reaction, with further mechanistic details remaining unresolved (4). Studies of peptide bond formation in the absence of modern biological machinery can give insight into the mechanism employed by the ribosome’s active site, as well as yield important information in the prebiotic route to the first peptides in the origin of life. The formation of a peptide bond (reaction R1 shown below) is a condensation reaction, eliminating a water molecule for each peptide bond formed, and thus faces both thermodynamic and kinetic constraints in bulk aqueous solution (5). R1

The equilibrium constant for peptide bond formation in water is extremely small and greatly favors the amino acid monomers (5, 6). Amino acid monomers have the added kinetic disadvantage of existing primarily as zwitterions at environmentally and physiologically relevant pH values in bulk aqueous solution (5, 7). Insightful experiments have been performed, yielding peptide bonds in anhydrous solvents with amino acid ester starting materials and copper(II) ion catalysis (8, 9). Transition metal ions are thought to have been components of the early ocean (10), with one source being the heavy meteoritic and cometary bombardment experienced by the early Earth, but the anhydrous solvents used in these studies are neither physiologically relevant nor likely to have been present on early Earth. The same mechanism was attempted in aqueous solution, but no peptide formation was detected (9). Polymer formation in aqueous environments would most likely have been necessary on early Earth because the liquid ocean would have been the reservoir of amino acid precursors needed for protein synthesis. In this work, the air–water interface is utilized as the auspicious environment for peptide bond formation, coupling the water surface with the bulk water reservoir of monomers. In situ surface-sensitive techniques are used here to observe the condensation reaction of a model system composed of a small, water-soluble amino acid ester (leucine ethyl ester) through Cu²⁺ coordination.

Air–water interfaces are found now, as on prebiotic Earth, at the surfaces of lakes, oceans, and atmospheric aerosols. The air–water interface (atmospheric aerosols in particular) has previously been proposed to be important in prebiotic chemistry (11–14) because it provides a unique environment for chemistry through its ability to concentrate and align biochemical precursors and to alter the state of ionization of surface species (15–19). Contemporary marine aerosols have been found to contain the amino acid precursors necessary for peptide bond chemistry (20), enabling the possibility for their use in such reactions. Further, the fluctuating conditions experienced by aerosols throughout their atmospheric lifetime, including evaporation of water, coagulation, and possibly reentry into the ocean, would naturally provide the compression of the surface layer shown in this work to be necessary for Cu²⁺ coordination leading to peptide bond chemistry (12).

In addition, the unfavorable equilibrium constant for peptide bond formation in bulk aqueous solution is shifted when the molecules experience a water-restricted reaction environment at the water surface. Although the exact surface pH of water is debated (21–23), the surface is known to alter the pK_a of surface-active molecules toward their neutral form (24, 25), which aids in the promotion of peptide bond chemistry at the interface by reducing zwitterion formation and alleviating the kinetic constraint on peptide bond synthesis. The air–water interface has been reported in the literature to have a catalytic role in peptide bond formation (26–29) using synthetic long-chain amino acid esters that are anchored to the surface by the polar groups attached to their 18-carbon-long hydrocarbon tails, thus forced to reside solely at the surface of the water. Reaction amongst the surface monomers was promoted in these studies (27–29) through surface compression, and supported by subsequent collection, drying, and analysis of the surface products. In the work presented here, infrared reflection absorption spectroscopy (IRRAS) and Langmuir trough methods were used to observe, in situ, complex formation with a metal cation followed by condensation chemistry leading to peptide bond formation occurring at the air–water interface. The observation of such condensation reactions in situ at the interface and with such a small activation group (an ethyl ester) on the starting amino acid precursor in an aqueous environment is unique.

Results and Discussion

Leucine ethyl ester is the short ester analog of leucine, one of the naturally occurring hydrophobic amino acids used in modern biochemistry. It is soluble in water, but exhibits a small amount of surface activity (Fig. S1), allowing it to be observed using IRRAS, a surface-sensitive IR spectroscopic technique that allows for vibrational characterization of molecules residing in the surface region. In this work, the “surface” or “interface” discussed is the reactive region of interest at or near the surface sampled by the IRRAS beam (30). The ester group of leucine ethyl ester serves to protect the carboxyl side of the molecule by eliminating the possibility of ionization to a carboxylate anion, and to activate the molecule as a whole for reaction. The short ester group gives a better leaving group for the reaction (ethanol rather than water; see reaction R2 below), but also serves to decrease the pK_a of the amine side of the molecule from 9.60 in the natural amino acid to 7.64 in the amino acid ester (5, 31), shifting the favorable form of the amine group toward neutral. R2

In this work, there is no disruption of the surface film during analysis, a problem faced in the collection procedures in previous studies of long-chain amino acid esters (27–29). We thereby demonstrate unambiguously that the chemistry occurred at the air–water interface. The long-chain amino acids used previously (27–29) were also highly synthetic molecules, which were unlikely to have been readily available on early Earth. Here, a two-carbon-long amino acid ester is used as a more plausible model for the initial abiotic peptide bond formation reactions because it contains a much smaller activation group and is also water-soluble. In addition, in these previous studies (27–29) the reactants were insoluble and their concentration was limited by the monolayer, which is deposited only at the surface. There is some debate in the literature as to whether the long-chain amino acid esters are capable of forming any peptides more than two monomers long (29) because of the confined nature of the surface film. Here, this problem is eliminated because there is a constant supply of monomers available for reaction adsorbing to the interface from the bulk solution, analogous to the reservoir of monomers provided by the ocean on early Earth. The chemistry observed in this work involves leucine ethyl ester at the water surface at the environmentally and physiologically relevant pH of 7.5, coordination of the leucine ester with Cu²⁺ through surface compression, and subsequent peptide bond formation at the air–water interface.

Copper–Leucine Ethyl Ester Complex Formation at the Air–Water Interface.

In Fig. 1A, the black trace shows the reflectance/absorbance spectrum of leucine ethyl ester at the air–water interface. The most prominent peak in the spectrum is the ester C═O stretch at 1,726 cm^-1, which remains prominent throughout the rest of the spectra shown in Fig. 1A. The remainder of Fig. 1A illustrates the evolution of leucine ethyl ester’s IRRAS spectrum through coordination with Cu²⁺ ions when the surface film is being held at a constant pressure of 15 mN/m using the mechanical barriers of the Langmuir trough. The coordination only occurs with the amine group on the molecule (proposed coordination complex shown in Fig. 2) as seen spectroscopically by the enhancements in the IRRAS bands below 1,200 cm^-1 in Fig. 1A. This is seen more clearly in the subtraction spectrum shown in Fig. 1B, in which the IRRAS spectrum of leucine ethyl ester in the absence of Cu²⁺ ions (shown in black) is subtracted from the Cu²⁺-coordinated IRRAS spectrum (shown in green). The subtraction spectrum that results shows only the changes caused by the coordination complex formed between leucine ethyl ester and the copper ions. From this spectrum, it is clear that there is no change at all above 1,200 cm^-1, but that there are four peaks that grow in below 1,200 cm^-1. These four peaks can all be assigned as associated with the amine side of the molecule through comparison with the literature infrared spectra of ethylenediamine complexes with Cu²⁺ ions (32) as follows: The two strongest peaks occurring at 987 and 1,037 cm^-1 are assigned to the CN stretching mode, and the two weaker peaks at 1,089 and 1,139 cm^-1 are assigned to the NH₂ twisting mode. This shows that through constant pressure, the copper ions form a coordination complex with the amine group of the leucine ethyl ester molecules at the surface. These results constitute an unambiguous spectroscopic characterization of the process of peptide bond formation at the water–air interface.

Fig. 1.

IRRAS spectra showing Cu²⁺ coordinated to the amine group of leucine ethyl ester over time, with compression. (A) Leucine ethyl ester: (black) in the absence of Cu²⁺ ions; (red) in the presence of Cu²⁺ immediately after beginning constant pressure of 15 mN/m; (blue) after 30 min of constant pressure; and (green) after 60 min of constant pressure. (B) Subtraction spectrum (gray) of uncoordinated leucine ethyl ester (black) from the Cu²⁺-coordinated leucine ethyl ester IRRAS spectrum (green) resulting in only the peaks enhanced through coordination.

Fig. 2.

Proposed Cu²⁺–leucine ethyl ester complex formed at the interface, as observed in the IRRAS spectra of Fig. 1.

The IRRAS spectrum of the copper-complex precursor also gives mechanistic insight into the peptide bond formation reaction at the surface. In anhydrous solutions, peptide bond formation is facilitated between two amino acid ester monomers via the formation of a chelate ring around a central Cu²⁺ ion, with coordination occurring only through the amine side of the molecule (8). It is well known that if in aqueous solutions the copper ion directly coordinates with the ester side of the molecule, hydrolysis of the ester bond is kinetically enhanced (33), resulting in the loss of the activation group utilized in the peptide bond formation reaction here. Because there is no enhancement or shift in the ester C═O stretch in Fig. 1, there is no copper ion coordination with the ester side of the molecule at the air–water interface (34, 35). This shows the importance of the environment provided by the water surface. The copper ions only coordinate with the amine group at the surface, thereby simultaneously avoiding hydrolysis of the ester and promoting peptide bond formation.

Compression-Promoted Peptide Bond Formation at the Air–Water Interface.

The complex between the amine and the Cu²⁺ ions only forms through compression of the surface layer, and was found to be a necessary precursor to the subsequent chemistry. No change was seen in the IRRAS spectrum in the absence of Cu²⁺ ions (Fig. S2) under the same experimental conditions presented here. It has been proposed in the literature that metal complexes with amines at the air–water interface induced through surface compression can result in an orientational change of the molecules residing at the surface (36). The orienting effect of the complex at the surface promotes subsequent condensation to form peptide products as observed in our studies. Fig. 3 shows a comparison between the IRRAS spectrum of the copper-complex precursor (Fig. 3A) and the condensation product formed at the surface (Fig. 3B). In the top spectrum (Fig. 3A), the most prominent peaks are the ester C═O band at 1,726 cm^-1 and the copper-complex amine peaks below 1,200 cm^-1, as described previously.

Fig. 3.

IRRAS spectra of (A) leucine ethyl ester complexed with Cu²⁺ ions, and (B) the peptide product formed at the surface (red dotted line indicates new Amide I band of peptide bond formed in surface reaction).

After allowing the reaction to proceed at the surface overnight, the spectrum in Fig. 3B results. In this spectrum, the ester band is still prominent at 1,726 cm^-1, but there is a new prominent peak at 1,625 cm^-1, which is not seen in the leucine ethyl ester spectrum. This peak is assigned as the Amide I band of the peptide product formed by the condensation reaction of two or more leucine ethyl ester molecules. The Amide I band can yield secondary structure information about the peptide product because its position is very sensitive to its environment (35, 37, 38). The position of the Amide I band of the peptide in the IRRAS spectrum reveals that the peptide product is still coordinated to the Cu²⁺ ion that catalyzed its reaction, because the coordinated peptide should have an Amide I band at 1,625 cm^-1, whereas the uncoordinated peptide’s Amide I band would appear around 1,675 cm^-1 (35).

Concurrent with the peptide bond chemistry, floating crystals spontaneously formed at the surface and were subsequently manually collected and then analyzed using solid-state infrared spectroscopy. From the infrared spectra, which are shown in Fig. 4, it was determined that the collected material is a mixture of the peptide product formed at the surface and the leucine ethyl ester starting material from the bulk solution. The collected material (black) shows many of the same features as the leucine ethyl ester starting material (red), but exhibits a new peak at 1,625 cm^-1 and a new shoulder at 1,520 cm^-1, assigned to the Amide I and II bands, respectively. This Amide I band assignment (Fig. 4) corresponds directly with the Amide I band observed in the IRRAS surface spectrum (Fig. 3B), confirming the identity of the peptide product formed in the surface region.

Fig. 4.

Solid-state infrared spectra of collected solid formed at the surface (black), showing amide I (*) and amide II (**) bands of the peptide, and leucine ethyl ester commercial solid (red).

Conclusions

Here, we have unambiguously demonstrated peptide bond formation at the air–water interface using small, water-soluble amino acid esters. Condensation reactions that must eliminate water are thermodynamically unfavorable in aqueous bulk, and yet are ubiquitous and essential to life. In addition, peptide bond formation will not occur between two amino acids in their zwitterionic form, the predominate state in a bulk aqueous environment. Water–air interfaces, characteristic of the surface of oceans, lakes, and atmospheric aerosols, provide an auspicious environment for this condensation chemistry through their provision of a water-restricting environment, alteration of the ionization state of surface species, and ability to concentrate and align monomers. Through in situ spectroscopic measurements, we have identified that the peptide bond forms through the coordination of the amine group of leucine ethyl ester to Cu²⁺ ions at the surface, inducing an orientational change at the surface observed using IRRAS. Then, peptide bond formation occurs spontaneously at the surface of water, facilitated by the formation of the copper complex at the interface. This work gives insight into oligomeric peptide formation en route to the emergence of more complex biomolecules on early Earth, and reinforces the importance of orientation, alignment, and proximity in the functioning of modern ribosomal peptide bond synthesis.

Materials and Methods

Materials.

All reagents were purchased from Sigma-Aldrich Chemical Co. and used without further purification. A stock pH-8 buffer solution was prepared by dissolving 8.5·10^-4 mol of sodium phosphate monobasic monohydrate (NaH₂PO₄•H₂O, ACS grade) and 9.15·10^-3 mol of sodium phosphate dibasic heptahydrate (Na₂HPO₄•7H₂O, ACS grade) in 1 L deionized water. Then, copper(II) chloride (97%) was added to an aliquot of the buffer solution immediately before the experiments were performed to yield a final concentration of 1 mM Cu²⁺ and a final pH of 7.5. Leucine ethyl ester hydrochloride (99 + %) was then added to the solution to a concentration of 0.06 M of leucine ethyl ester. This final solution was then sonicated until a transparent solution resulted.

Instrumentation.

Langmuir trough.

The Langmuir trough used in these studies was purchased from NIMA (KSV NIMA) and consisted of a PTFE trough (145 by 70 by 0.5 mm) coupled to two computer-controlled PTFE barriers. The barriers were limited at their open position to a surface area of 70 cm². The Langmuir trough was equipped with a Wilhelmy balance, which allows for simultaneous measurement of surface pressure during surface area changes induced by the mechanical PTFE barriers. The result is a surface pressure–area (π–A) isotherm, yielding interfacial thermodynamic information.

IRRAS.

The Langmuir trough described above was coupled with a custom-built IRRAS setup. The IRRAS spectra were obtained using the external port of a commercial Bruker Tensor 27 FTIR spectrometer, with all external optics and equipment (including the Langmuir trough) being constantly purged with dry house air. The IR beam exited the spectrometer and was passed through a CaF₂ lens after which it was reflected off of two 2-inch gold mirrors directing it onto the air–aqueous interface and then directing the reflected light to a liquid nitrogen–cooled MCT detector. The two gold mirrors were positioned over the Langmuir trough so that the IR beam was incident on the air–aqueous interface at an angle of 22° relative to the surface normal. This angle is within the optimum angles found to be ideal for unpolarized light incident on an air–aqueous interface (39). Spectra were then collected and averaged over 200 scans with 1-cm^-1 resolution. A single-channel spectrum of the surface of the buffer solution (in the absence of leucine ethyl ester) was used as the background spectrum. The IRRAS spectra here are shown as reflectance absorbance (RA) spectra, where RA = - log(R/R_o), with R being the IR reflectivity of the surface of interest, and R_o being the IR reflectivity of the background surface. The details of such reflectivities can be described by the Fresnel equations, which are presented elsewhere (40). With an angle of incidence of 22° relative to the surface normal and unpolarized light on an air–aqueous interface, the absorption bands will be negative.

It is important to note that the IRRAS technique does not only sample the top layer of molecules at the air–water interface. Rather, it samples the surface region, which, for soluble molecules adsorbing to the surface (such as leucine ethyl ester), can be as deep as 1–2 μm (30). Other surface techniques, such as vibrational sum-frequency generation, arguably have smaller probe depths, but for the work presented here, the IRRAS technique samples the reactive region of interest.

Solid-State FTIR.

Solid-state infrared spectra were obtained using the internal chamber of the Bruker Tensor 27 FTIR spectrometer purged with dry house air. Infrared spectra were taken with the samples pressed between two CaF₂ windows, averaging over 50 scans with a 1-cm^-1 resolution.

Methods.

The prepared solutions were spread on the Langmuir trough and the system was allowed to purge for 30 min. Then, the mechanical barriers of the Langmuir trough were compressed to hold the surface at a constant surface pressure of 15 mN/m for at least 1 h, during which IRRAS spectra were taken every 10 min. The surface pressure of 15 mN/m was chosen because it was found to put some strain on the surface molecules, and subsequently to force them to orient to allow for complex formation with Cu²⁺ and subsequent condensation reaction, without excessive loss of surface molecules back into the bulk solution. Without compression, the adsorbed molecules were not observed to form complexes with the copper ions or react in any way. Afterward, the barriers were returned to their fully open positions (at a surface area of 70 cm²), and the surface molecules were allowed to react overnight. Then, after the addition of some buffer solution underneath of the surface film to compensate for the decrease in the surface level caused by evaporation overnight, another IRRAS spectrum was taken of the surface. Surface crystals that spontaneously formed overnight were then collected, dried out, and analyzed using solid-state FTIR. IRRAS spectra of leucine ethyl ester in the absence of copper(II) ions were also taken for comparison following the procedures outlined above (spectra shown in Fig. S2).