Investigation of the pH‐dependence of dye‐doped protein–protein interactions

Abstract

Proteins can dramatically change their conformation under environmental conditions such as temperature and pH. In this context, Glycoprotein's conformational determination is challenging. This is due to the variety of domains which contain rich chemical characters existing within this complex. Here we demonstrate a new, straightforward and efficient technique that uses the pH‐dependent properties of dyes‐doped Pig Gastric Mucin (PGM) for predicting and controlling protein–protein interaction and conformation. We utilize the PGM as natural host matrix which is capable of dynamically changing its conformational shape and adsorbing hydrophobic and hydrophilic dyes under different pH conditions and investigate and control the fluorescent properties of these composites in solution. It is shown at various pH conditions, a large variety of light emission from these complexes such as red, green and white is obtained. This phenomenon is explained by pH‐dependent protein folding and protein–protein interactions that induce different emission spectra which are mediated and controlled by means of dye–dye interactions and surrounding environment. This process is used to form the technologically challenging white light‐emitting liquid or solid coating for LED devices.

Abbreviations

B

Blue dye

BSM

bovine submaxillary mucin

DLS

dynamic light scattering

G

Green dye

OLED

organic light emitting diode

PGM

Pig Gastric Mucin

R

Red Dye.

Introduction

Mucins are of the family of glycoproteins which are present in many faunal niches from Invertebrates to Vertebrates. The primary role of these macromolecules is to protect various organs or cells from the outside environment by building low diffusivity mucus layer.1 In vertebrates, particularly in mammals, mucins have a vital role in protecting the gastric organs from the harming hydrochloric acid which is essential for digesting food but harmful to epithelial cells. The importance of mucin proteins can be viewed in evolution perspective since most mucins share conserved sequences.2, 3 Since that, a conformational study on particular type may be generalized to the entire mucin family.4 A member of this family, the Pig Gastric Mucin (PGM) is glycoprotein which belongs to the membrane and extracellular secreted proteins family with high molecular weight (2‐20 MDA). PGM structure is comprised of many repeating building blocks which consist of rich glycosylated oligosaccharide chains that make up approximately ∼80% of PGM Mw that is arranged in 5‐15 oligosaccharide's of fucose, galactose, N‐acetylgalactosamine, Nacetylglucosamine, and mannose.5, 6, 7 Those oligosaccharides are attached to the core protein by O‐glycosidic bonds in a “tree branches” manner. The core protein itself composed of a vast number of repeated sequences of threonine, serine, and proline. In addition, the presence of cysteine‐rich regions in hydrophobic pockets along the protein core has been reported.8, 9 Previous works10 ^, 11, 12 indicated that different pH conditions may induce crucial substantial PGM conformations, especially in acidic and neutral environments.11 Understanding this phenomenon is of particular importance since gastric‐mucins are subjected to a harsh acidic environment in vivo.

Characterizing and understanding the dynamic structure of mucin in these conditions is very challenging and not straightforward. Previous studies have only partly resolved the pH‐dependent conformation using several techniques such as dynamic light scattering (DLS),11 spectroscopic methods,12 rheology,13 Atomic Force14 and Transmission Electron Microscopy15 partly indicating on pH‐dependent folding and unfolding mechanism. We have previously suggested that incorporating light‐emitting dyes in this compound can lead to formation of controllable light‐emitting solid coatings which might be used for non‐biological applications such as a backlight coating for Liquid crystal displays and organic light‐emitting diodes (OLEDs).16, 17 One should note that obtaining white emission by means of appropriate phosphorous coating of blue LED is important since it will allow replacing the commonly used incandescent light bulbs and fluorescent tubes with much more efficient and cheap LEDs. Moreover, the ability to tune the exact color of the emission is needed for the LED market, since both warm (red‐rich) and cold (blue‐rich) white LEDs are used in different markets.18 Obtaining white light emission is a technologically challenging due to various self‐quenching and non‐radiative energy transfer mechanism that takes place in blended and closely separated dyes.19 We have previously suggested that proteins can help to solve this problem since they can host dyes while reducing significantly these unwanted effects.17

Here we demonstrate a simple and efficient method to study pH dependent protein–protein interactions in mucin protein family using a 2‐probe fluorescent method in which non‐radiative energy transfer processes between probes located at hydrophobic and hydrophilic sites are used to monitor protein‐protein interactions at various pH conditions. We also exploit this phenomenon to form a white‐light emitting coating. For this task, two types of dyes are used. These bind selectively to different domains in the PGM protein: hydrophilic oligosaccharides (Green dye, G, Fig. 1) and hydrophobic polypeptide pockets in the main core protein (Red dye, R, Fig. 1). Since PGM can complex hydrophobic molecules in aqueous solution,20 full solubilization of the two dyes is achieved.16, 17 This simple technique allows us to explore the folding mechanism by simple monitoring of the R and G fluorescence spectra and demonstrate a white light luminescence both in solution and on and in solid non‐woven matrix of gelatin‐PCL nanofibers placed on blue LED.

Figure 1.

Schematic presentation of the complexation process of red and green dyes inside PGM protein.

Results and Discussion

Figure 2 shows fluorescence spectra taken at different pH values measured for RG dyes dissolved in representative buffer solutions [Fig. 2(A)] and uncompleted in PGM [Fig. 2(B)]. At pH = 2 for the complexed solution, a distinct red and green emission is present. For the complexed solution at this pH, broad emission spectra at the red band, accompanied by a decrement of the green band are observed. For un‐complexed solution at pH = 6 green emission and red emission quenching is measured, while for the complexed structure, distinct yellow‐; green‐ and red‐bands appear. At pH = 10, the spectral emission of un‐complexed solution is similar to pH = 6, while for the complexed structure, green, yellow and smeared red bands are present.

Figure 2.

a) Fluorescence spectra of red and green dyes under different pH conditions. Blue, pH = 2 (λ _max = 677 nm, λ _max =517 nm), Orange, pH = 6 (λ _max =558 nm, λ _max =479 nm, intensity multplied by 10 for clarity) and pH = 10 (Grey, λ _max =556 nm and λ _max =474 nm). b) Fluorescence spectra of red and green dye complexed in PGM protein under the same pH conditions pH = 2 (broad peak λ _max =687 nm), pH = 6 (λ _max =707 nm, λ _max =603 nm and λ _max =508 nm) and pH = 10 (broad peak λ _max =704 nm, λ _max =620 nm and λ _max =521 nm).

The results can be interpreted as follows: it is evident that the water soluble PGM successfully incorporates the hydrophobic R dye that is insoluble in aqueous solutions as shown by previous studies.16, 17, 20 Thus, the R dye aggregates strongly in aqueous solutions which is manifested by its quenching [Fig. 2(A)], while no aggregates are observed in the complexed PGM solution [Fig. 2(B)].21 One should also note that the hydrophilic G dye emits more light in an acidic environment than basic one due to different degree of ionization of the 8‐hydroxyl group of the G dye22 (For a detailed study of pH dependent Fluorescence of R and G dyes see Figure S1.)

Notably, the PGM‐dyes complex at pH = 6 results in a continuous spectrum which gives rise to white emission which can be utilized as a coating material for white LED (Fig. 3). Moreover, the presented technique allows us to tune the emission by changing the pH of the solution from acidic to basic to obtain various emitted spectra; red, green and cold white.

Figure 3.

White light emission: solution and solids state complexes. (a). White light emission from solution (excitation −404 nm). (b,c) Preparation of a solid coating for white LED. (c) PCL (top) was used as basic material for formation of solid mat using electrospinning (d, scanning electron microscopy image), G‐ and R‐ complexes were then introduced to the mat. (c) blue LED was used to excite the mat giving rise to RGB emission (white light). (d) Optical image showing white emission from the structure.

Next, a mechanism for the pH‐dependent emission of the complexed dyes is suggested (Fig. 4): at basic‐neutral environment, high green emission signal is obtained while the intensity of the red signal depends on the ability of the PGM to separate the red dye molecules from each other. This process yields white light emission which is formed by optical excitation by the blue wavelength of the red and the green fluorescence at neutral conditions giving rise to R, G and B emission (white). At acidic pH, the intensity of the green band is small; hence, the main contribution to the emission spectrum will originate from the red dye. Previous reports11 indicated that PGM exhibits substantial conformational pH‐induced mechanism. This mechanism includes unfolding of the protein skeleton's hydrophobic pockets of PGM at low pH values and reversible folding of these pockets at high ones. Additional studies which used AFM and TEM techniques confirmed that under acidic pH, PGM shows dense cluster form, while under basic pH it has stretched to semi‐fibrous one.14, 15 Sotres et al.23 also tested hydrophobic dependency of another member in mucin family of proteins, the bovine submaxillary mucin (BSM) and concluded that hydrophobic interactions are strongly present in BSM under acidic environment. Single‐probe intensity analysis demonstrated by Cao and co‐workers12 supported the folding‐unfolding mechanism. It was suggested that this process plays a major role in mucin's hydrophobic protein–protein interactions. To verify the proposed mechanism, the nature of the doping sites (polypeptide chain or polysaccharide) was investigated using deglycosylation of the proteins24 (Fig. 4).

Figure 4.

(a) Mechanism for pH conformational changes under different pH regimes. Circle represents hydrophobic inter‐protein interaction sites. See text for details. (b) Quantification of PGM polysaccharids performed by Schiff assay, (c) Comparison between fluorescence spectra of untreated and deglycosylated doped PGM.

The degree of deglycosylation was assessed using periodate/Schiff reagent test.25 We have found that the amount of the remaining polysaccharides was ∼10% [Fig. 4B,C]. Next, the R dye was incorporated into deglycosylated PGM, and emission spectrum was recorded [Fig. 5(B)]. A decrement in the emission intensity was detected for the deglycosylated PGM‐dye complex. This observation can be explained by the conformational change that occurs upon deglycosylation.14 It can be attributed to the folding of the protein into a globular form, which can block some of the pockets, so the R dye wouldn't be able to enter. Since the overall emission is more than 10% of the pure PGM‐dye complex emission, it can be concluded that the R dye is incorporated by the hydrophobic pockets and not by the polysaccharide chains.

Conclusions

Our observations clearly support the current model12 (Fig. 4): In an acidic environment, the unfolding of the pockets drastically shortens the distances between the red dyes which lead to dye aggregation and thus decreases the red band intensity [see Fig. 2, Fig. S2(B)]. The quenching of the red dye fluorescence signal follows the aggregation‐induced fluorescence quenching mechanism (Fig. 4). At higher pH, reversible separation of the protein–protein matrix takes place, which leads to stabilization of hydrophobic pockets of each protein. The intensity of the red dye emission increases due to the folding of the proteins back to their native form in which the distances between the dyes increase. This leads to a larger contribution of unquenched R and G to the emission band and formation of white emission spectra.

To summarize, in this work we demonstrate simple tagging method that can be used for better understanding the complexity of protein‐protein interactions. For PGM and other mucins, it can be concluded that the hydrophobic pockets play a crucial role in these interactions. A white emitting coating for LEDs was demonstrated, the emission can be tuned by varying the ratio between the dye‐incorporated proteins.

Materials and Methods

Complex preparation

Stock solutions of 30 mM of each dye (Green, G,—8‐Hydroxypyrene‐1,3,6‐trisulfonic acid trisodium salt (Alfa‐Aesar) and Red, R,—4‐(Dicyanomethylene)‐2‐methyl‐6‐(4dimethylaminostyryl)−4H‐pyran(Sigma‐Aldrich)) were dissolved in mL DMSO(Sigma‐Aldrich). PGM (Sigma‐Aldrich) solutions (2 mg/mL) were separately mixed with R and G (150 µM) dyes and stirred overnight in dark. Doped PGM solutions of the two different dyes were mixed in ratio of 1:1 and adjusted to desired pH value by addition of glycine buffer giving pH values of 2‐10 (Fig. 1).

Complex cleaning

To separate between the doped PGM and non‐complexed dyes, a passive membrane dialysis method was applied. Shortly, membranes (Sigma‐Aldrich) with retention limit of particles heavier than 12 kDa were prepared according to standard preparation procedure. Next, doped PGM solution was loaded inside the freshly prepared membranes and put into glycine buffer solution of 20 mM with a pH value identical to PGM solution pH. Membranes were gently stirred overnight at the dark and RT. Finally, doped PGM solutions were extracted from the membranes and stored in 8–10°C.

Spectroscopic measurements

All fluorescence measurements were taken in fluorolog‐Horiba modular spectrofluorometric at excitation wavelength 456 nm and slit width of 5 nm. All absorbance measurements were taken in UV–VIS carry 5000 in RT with scanning range of 200–800 nm. Fluorescence of protein–dyes complexes of 20 mg/10 mL protein with 150 µM dyes (per dye) was measured with appropriate pH value calibrated by glycine buffer. All the measurements were taken in room temperature with co‐equilibration time of 24 h.

White light emitting coating preparation

A white light emitting coating made from PGM + DYES complex was prepared in two phases; the first phase included preparation of holding matrix made from hydrophilic nanofibers. The second phase included preparation of PGM + DYES complex. Hydrophilic nanofibers were prepared according to procedure reported elsewhere.26 Shortly, gelatin type (Sigma‐Aldrich) and Polycaprolactone (PCL, molecular weight 80,000, Sigma‐Aldrich) were dissolved separately in acetic acid at a concentration of 10% (w/v) with sufficient stirring at room temperature for 24 h. Next, the two solutions were mixed in 50:50 volume ratios and electrospun. Next, a blend of G‐ and R‐PBM was drop cast on the electrospun mat and dried. Finally, the coating was placed on a blue LED (emission wavelength 404 nm) giving rise to a white‐LED formation (Fig. 3).

Deglycosylation

10 mg/mL PGM solution was deacetylated for 30 minutes in 0.1M NaOH solution followed by neutralization with HCl. Acetic acid was added to 0.1M concentration and was titrated with NaOH to pH 4.5. Next, NaIO4 was added to final concentration of 0.1M, overnight at 4˚C. The reactive aldehydes were then neutralized, by stirring with 2% glycine solution for 30 min, followed by β‐elimination with 0.1M NaOH at room temperature, for 30 min. The deglycosylated proteins were dialyzed vs. H2O overnight, separated from the solution by centrifugation and lyophilized. Quantification of the amount of the remaining polysaccharides was done by the sodium periodate/Schiff assay.25

Supporting information

Supporting Information