*Article* **A Ponceau S Staining-Based Dot Blot Assay for Rapid Protein Quantification of Biological Samples**

**Dario Lucas Helbing 1,2,3, \* ,† , Leopold Böhm 1,2,3,† , Nova Oraha 1,2,3 , Leonie Karoline Stabenow 2,3 and Yan Cui 1,4,\***


**Abstract:** Despite the availability of a wide range of commercial kits, protein quantification is often unreliable, especially for tissue-derived samples, leading to uneven loading in subsequent experiments. Here we show that the widely used Bicinchoninic Acid (BCA) assay tends to underestimate protein concentrations of tissue samples. We present a Ponceau S staining-based dot-blot assay as an alternative for protein quantification. This method is simple, rapid, more reliable than the BCA assay, compatible with biological samples lysed in RIPA or 2x SDS gel-loading buffer, and also inexpensive.

**Keywords:** protein quantification; gel electrophoresis; Western blot; dot blot

Oraha, N.; Stabenow, L.K.; Cui, Y. A Ponceau S Staining-Based Dot Blot Assay for Rapid Protein Quantification of Biological Samples. *Gels* **2022**, *8*, 43. https://doi.org/

Academic Editor: Hiroyuki Takeno

**Citation:** Helbing, D.L.; Böhm, L.;

Received: 28 November 2021 Accepted: 1 January 2022 Published: 7 January 2022

10.3390/gels8010043

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

#### **1. Introduction**

Many methods exist for quantifying total protein contents in cell or tissue lysates. However, the conventional colorimetric methods, such as the Bicinchoninic Acid (BCA) [1], Lowry [2] and Bradford [3] assays, have several drawbacks: (1) their requirement for large volumes of samples ranging from 10 to 25 µL per replicate could be difficult to meet for samples with low protein contents, such as mouse nerve or lymph node lysates; (2) colorimetric assays tend to saturate if the protein concentration is too high; in this case, the samples must be diluted for repetition, consuming more materials and time; (3) the widely used SDS-based lysis buffers often contain bromophenol blue, a substance that interferes with most colorimetric assays, making colorimetric assays unsuitable for bromophenol blue-containing lysates [1,4]; and (4) some native components in biological samples, such as high concentrations of lipids in nerve or brain lysates, may interfere with the standard BCA assay [5]. Therefore, protein concentrations of biological samples determined by the common assays are often imprecise, resulting in unequal sample loading in subsequent experiments such as Western blots.

Here, we demonstrate a modified Ponceau S-based dot blot (PDB) assay that has distinct advantages over similar methods described elsewhere [6–8] concerning its ability to quantify proteins in cell or tissue lysates, its highly linear standard curve with a wide range from 0.25 to 12 µg/µL, and its overall improved workflow. We further show that protein loading based on the PDB assay was equal in Western blots, essentially eliminating the need for error-prone normalization of immunoblot band intensities to those of a loading control [9]. Therefore, the PDB assay is a reliable method for protein quantification, thereby facilitating subsequent experiments. Our work, in tandem with several other publications, demonstrates the versatility of dot blot in protein quantification and detection [10–12].

#### **2. Materials and Methods**

#### *2.1. Reagents*

Coomassie Blue solution: 0.05% (*w*/*v*) CBB-R250 (Carl Roth GmbH + Co. KG, Karlsruhe, Germany, #6862), 40% (*v*/*v*) ethanol, 10% (*v*/*v*) acetic acid in water.

Direct Blue solution: 0.008% (*w*/*v*) Direct Blue 71 (Merck KGaA, Darmstadt, Germany, #212407), 40% (*v*/*v*) ethanol and 10% (*v*/*v*) acetic acid in water.

Micro BCA Protein assay kit, Thermo Fisher Scientific Inc., Waltham, MA, USA, #23225. Pierce Bovine Serum Albumin Standard (BSA) ampules, Thermo Fisher Scientific Inc., Waltham, MA, USA, #23209.

Ponceau S solution: 0.1% (*w*/*v*) Ponceau S (Merck KGaA, Darmstadt, Germany, #P3504- 10G) in 5% (*v*/*v*) acetic acid.

RIPA buffer: Pierce RIPA buffer (Thermo Fisher Scientific Inc., Waltham, MA, USA) or self-made RIPA buffer (2.5% (*v*/*v*) 1 M Tris-HCl, 0.88% (*w*/*v*) NaCl, 1% (*w*/*v*) NP-40, 1% (*w*/*v*) sodium deoxycholate, 0.1% (*w*/*v*) SDS in water, adjusted to pH 7.6), 1 tablet/50 mL cOmplete protease inhibitor (Roche Diagnostics GmbH, Mannheim, Germany), 1 tablet/10 mL RIPA phosSTOP phosphatase inhibitor (Roche Diagnostics GmbH, Mannheim, Germany).

2x SDS Gel Loading buffer (2x SDS LB): 0.1 M Tris-HCl, 4% (*v*/*v*) SDS, 20% (*v*/*v*) Glycerol, 0.2% (*v*/*v*) Bromophenol blue (Carl Roth GmbH + Co. KG, Karlsruhe, Germany, #A512.1) in water, adjusted to pH 6.8.

#### *2.2. Ethics Statement*

The Thuringian State Office for Consumer Protection (TLV), Department of Health and Technical Consumer Protection has examined and prospectively approved the animal experiment application under the permit number 03-046/16. It is advised by the Thuringian Animal Protection Commission.

#### *2.3. Experimental Animals*

Mice (C57BL/6J) were housed under a 12 h dark/light cycle and had free access to food and water. All mice were handled in strict adherence to local governmental and institutional animal care regulations and euthanized by trained personnel using CO<sup>2</sup> inhalation. All efforts were made to minimize suffering.

#### *2.4. Lysate Preparation*

Cell lysates were prepared by adding 2x SDS LB to cell culture plates after washing off cell debris and culture media and then collected with cell scrapers. Lysates were then transferred to Eppendorf tubes and vortexed for 10 s.

Sciatic nerves, brains and spleens were harvested from euthanized mice and lysed using ceramic beads in a Precellys 24 homogenizer (Bertin Instruments, Montigny-le-Bretonneux, France) in either RIPA buffer or 2x SDS LB.

#### *2.5. BCA Assay*

The microscale BCA assay was performed according to the manufacturer's instruction. Protein concentrations were calculated using the linear equation based on the trend line of the standard curve generated with Microsoft Excel.

#### *2.6. Dot Blot, Ponceau S Staining, Direct Blue Staining and Coomassie Blue Staining*

Lysates in 2x SDS LB were boiled for 8 min at 98 ◦C beforehand. Different volumes of lysates and BSA solution were applied dot-wise to dry nitrocellulose membranes. Membranes were then left to dry for 15 min. Membranes with samples lysed in 2x SDS LB were then washed three times (5 min each) in deionized (DI) water on a shaker before staining with Ponceau S solution for 1 min, and membranes with samples lysed in RIPA buffer were stained directly with Ponceau S solution for 1 min. Membranes were then briefly

washed with DI water to remove the Ponceau S solution. Membranes were then placed into a plastic film and scanned with an Epson Perfection V750 Pro scanner.

For re-staining of membranes with Direct Blue solution, membranes previously stained with Ponceau S were washed overnight with TBS-T, stained with Direct Blue staining solution for 5 min, washed with a solution of 40% (*v*/*v*) ethanol and 10% (*v*/*v*) acetic acid in DI water to remove non-specific background staining, and then imaged as described above. For re-staining of membranes with Coomassie Blue, membranes previously stained with Direct Blue were washed overnight with a solution of 50% (*v*/*v*) ethanol and 15% (*w*/*v*) 1 M sodium bicarbonate in water. Membranes were then stained with Coomassie Blue solution for 3 min and imaged as described above.

For experiments using diluted BSA solutions, a BSA solution was diluted 1:1 with either 0.5 M NaCl, PBS, TBS-T or a 0.1% (*v*/*v*) SDS solution, resulting in BSA solutions with a protein concentration of 1 µg/µL. BSA solutions were applied dot-wise onto a nitrocellulose membrane with different volumes to create a standard curve for each solution based on Ponceau S staining.

We included a movie demonstrating the experimental workflow and the analysis with Fiji in the Supplementary Materials (Video S1).

#### *2.7. Protein Quantification with Fiji and Microsoft Excel*

After transforming a source image into an 8-bit greyscale image and inverting it in ImageJ/Fiji, the rectangle tool and ROI manager were used to define a series of regions of interest (ROIs), covering each dot with a single ROI of the same size, and the integrated density of each ROI was measured. The integrated density values of technical replicates of the samples and the standards were averaged and, if necessary, corrected for the baseline signal by subtracting the trendlines' y-axis intercept. A standard curve was generated based on the mean integrated density values of the standards and the equation of a corresponding linear trendline used to calculate protein concentrations. The goodness of fit/coefficient of determination for each trendline was calculated in Microsoft Excel and displayed as R<sup>2</sup> in the corresponding figure.

#### *2.8. Cost Calculations*

We calculated the costs regarding the PDB assay and the micro BCA assay for measuring 12 biological samples and standards. Both the commercial BCA assay and a self-made variant were compared. The reagents for a self-made BCA assay were as follows:

Reagent A: 1% (*w*/*v*) sodium bicinchoninate, 2% (*w*/*v*) sodium carbonate, 0.16% sodium tartrate (*w*/*v*), 0.4% (*w*/*v*) NaOH, 0.95% (*w*/*v*) sodium bicarbonate in water, pH adjusted to 11.25.

Reagent B: 4% (*w*/*v*) cupric sulfate in water.

A range of total costs was calculated for each method to reflect the market price ranges of the required materials.

We calculated the costs of RIPA buffer and SDS Gel loading buffer per ml and compared the commercial and self-made RIPA buffers with the self-made SDS Gel loading buffer.

We determined the prices of reagents on the websites of established distributors and used the smallest available package size for our calculations. Basic laboratory chemicals such as Tris-HCl or NaCl were not included in our calculations. We included the protease and the phosphatase inhibitor prices in both cost calculations for the commercial and the self-made RIPA buffers.

#### *2.9. Immunoblotting*

Immunoblotting was performed as previously described [13]. Antibodies are listed in Table 1. Blots were developed with Pierce ECL Western Blotting Substrate (Thermo Fisher Scientific Inc., Waltham, MA, USA).


**Table 1.** List of antibodies.

## *2.10. Statistical Procedures and Figure Preparation*

The *p*-values were calculated by Student's *t*-test with Graphpad Prism 7.0. Statistical significance was accepted at *p* ≤ 0.05. All data are presented as mean +/− SEM (Standard error of the mean). Missing error bars are due to technical reasons when the error bar would be shorter than the height of a single datapoint. All figures were made with Graphpad Prism 7.0 and assembled in Adobe Photoshop CS6 or Microsoft PowerPoint.

#### **3. Results and Discussion**

To assess the reliability of the PDB assay for protein quantification, we spotted different volumes of a BSA solution (0.25–4 µg of BSA) on a nitrocellulose membrane. Ponceau S staining was performed after air-drying the membrane (Figure 1A). The sizes of the dots did not correlate with the total protein amounts (data not shown) and we therefore measured the integrated density, which is the sum of grey values in the region of interest. It should be noted that a gray value measured by ImageJ is opposite to the actual intensity, so images should be inverted before measurement. The integrated density and protein amount showed a linear trendline, indicated by a coefficient of determination *R <sup>2</sup>* near the optimal value of 1 (Figure 1B). For comparison, we performed a BCA assay, which also showed a trendline with R2~1 (Figure 1C), indicating that both assays perform well within their respective detection range. However, according to the manufacturer, beyond the maximum concentration of 40 µg/mL, the BCA assay—like any other colorimetric assay—is no longer reliable and further dilution of the sample is required. In contrast, the PDB assay yielded a linear standard curve up to 12 µg of the protein per dot (Figure S1).

**Figure 1.** Standard curves of the PDB and the BCA assays. (**A**) Ponceau S-stained dot blot. Variable volumes of a BSA solution (2 µg/µL) containing indicated BSA amounts were spotted in triplicate onto a nitrocellulose membrane. (**B**) Inverted standard curve based on Figure 1A. The experiment was repeated with three batches of BSA solutions. (**C**) Standard curve of three BCA assays performed with different BSA solutions. Error bars indicate SEM. R<sup>2</sup> = Coefficient of determination.

Visible in Figure S1, the bottom range of the PDB assay standard curve may lose fidelity if the total protein amount is below 1 µg and the respective BSA volumes are not spotted perfectly, which can occur if the applied volume per drop is small. Therefore, the BSA standards may be diluted if the samples are suspected of having low protein contents. To identify suitable diluents, we diluted the BSA standards with TBS-T, 0.1% SDS, PBS and 0.5 M NaCl, and compared the corresponding PDB assay standard curves (Figure 2).

As indicated by higher R<sup>2</sup> values and lower error deviations between technical replicates, TBS-T and 0.1% SDS in water outperformed PBS and 0.5 M NaCl (Figure 2B–E). Using diluted standards, samples with very low protein concentrations just above 0.125 µg/µL can be analyzed.

As described above, fixing the protein concentration while varying the volume yielded a linear standard curve in the PDB assay. Similarly, fixing the volume while varying the protein concentration also yielded a linear standard curve (Figure S2), suggesting that the integrated density (if not saturated) is determined by the protein amount and less affected by the dot size.

To test whether the PDB assay can be used to quantify proteins in lysates of biological tissues, spleens from four mice were lysed in 1 mL of RIPA buffer and a PDB assay was performed (Figure 3A–C). For comparison, the same lysates were also measured with a BCA assay (Figure 3C).

**Figure 2.** Comparison of diluents for the PDB assay. (**A**) Ponceau S-stained dot blot. BSA was diluted in different solutions to a final concentration of 1 µg/µL. Variable volumes of the solutions containing indicated BSA amounts were spotted in triplicate onto a nitrocellulose membrane. (**B**–**E**) Inverted standard curves based on Figure 2A. Error bars indicate SEM. R<sup>2</sup> = Coefficient of determination.

**Figure 3.** Underestimation of protein concentrations by the BCA assay. (**A**) Ponceau S-stained dot blot. Variable volumes of a BSA solution (2 µg/µL) containing indicated BSA amounts were spotted onto a nitrocellulose membrane. A cross indicates an incorrectly applied sample that was excluded for the analysis. (**B**) Ponceau S-stained dot blot. A quantity of 2 µL of spleen lysates from four mice was spotted in duplicates onto a nitrocellulose membrane. (**C**) Calculated protein concentrations of the spleen samples based on either the PDB or the BCA assay. (**D**) Retest of the protein concentrations of the spleen samples by different staining of the dot blot. Spleen samples containing 5 or 2.5 µg of total proteins (calculation based on either the PDB or the BCA assay), in addition to the BSA standards (2 µg/µL), were spotted in duplicate onto a nitrocellulose membrane. The blot was sequentially stained with Ponceau S, Direct Blue 71 and Coomassie Blue R250, with destaining between stainings. The integrated density values were then plotted: (**E**,**H**) for Ponceau S staining; (**F**,**I**) for Direct Blue 71 staining; (**G**,**J**) for Coomassie Blue R250 staining. Error bars indicate SEM. *p*-values were calculated using Student's *t*-test.

Both standard curves had to be extrapolated in order to calculate the protein concentrations of the spleen lysates because of high protein contents, which should generally be avoided. The BCA assay showed sample concentrations above the range of the respective standard curve. It would be necessary to repeat the measurement either with diluted samples or with another serial of BSA standards. In contrast, the calculation of the protein concentrations based on the PDB assay is less problematic because the standard curve for the PDB assay shows a linear correlation until 12 µg total protein amount per dot (Figure S1). It should be mentioned that, in routine experiments where the protein concentrations are estimated to be low, such high standards might not be necessary.

Strikingly, the protein concentrations of the spleen lysates based on the PDB assay were three to four times higher than those determined by the BCA assay (Figure 3C). To validate the results, we performed a PDB assay spotted with 5 and 2.5 µg of total proteins, which was based on the concentrations determined by either the BCA or the PDB assay (Figure 3D, left panel). Comparison of the integrated densities revealed a higher variability

of the dots based on the BCA assay (Figure 3E,H). Importantly, the integrated densities of the dots based on the BCA assay were generally much higher than those of the BSA standards. In contrast, the integrated densities of the dots based on the PDB assay were similar to those of the BSA standard (Figure 3E,H). This retest suggests that the PDB assay is more reliable than the BCA assay, whereas the latter may significantly underestimate the total protein contents of unknown samples

Protein inputs on blots can also be stained with other dyes. Different dyes may have different binding modes to proteins. Unequal binding to heterogenous proteins will lead to a bias in total protein quantification. Ponceau S binds to positively charged amino acids and nonpolar regions of proteins [14]; therefore, it stains heterogenous proteins relatively equally. To compare Ponceau S staining with two other staining methods, we destained the same blot and re-stained it with Direct Blue 71 [15] and Coomassie Blue R250 sequentially. The overall staining profiles and integrated densities were comparable between the three dyes, despite the fact that the BSA standards were less stained by Direct Blue 71 or Coomassie Blue R250 (Figure 3D–J). It is unclear whether this difference was due to each dye's binding preference or loss of BSA from the blot during extensive destaining. It is also apparent that Coomassie Blue R250 staining resulted in a high background. Due to its properties of having a simple recipe, fast procedure, relatively equal staining of heterogenous proteins, adequate sensitivity, and low background, Ponceau S staining is an excellent choice for protein quantification.

We have shown that the PDB assay is compatible with biological tissues lysed in RIPA buffer, but some researchers may prefer 2x SDS LB because some proteins are known to be insoluble in RIPA buffer, leading to an overall loss of 10% to 30% of all proteins [9]. We therefore tested the compatibility of the PDB assay with samples lysed in 2x SDS LB. Because an initial experiment with BSA diluted in 2x SDS LB did not yield a linear standard curve (data not shown), we decided to add a washing step to wash out the SDS from the dots, which we suspected interfered with the binding of Ponceau S to the proteins. Washing the membrane three times with DI water (5 min each) after drying it for 15 min allowed us to generate a standard curve similar to that based on undiluted BSA standards (Figure S3), allowing the PDB assay to measure samples lysed in 2x SDS LB. RIPA buffer emerged to be unsuitable to dilute BSA standards because it results in the formation of coffee rings on blots (Figure S4).

To compare the protein extraction efficiency between 2x SDS LB and RIPA buffer, we performed a PDB assay with tissues lysed in these two buffers and conducted a subsequent Western blot. Two sciatic nerves and cerebral hemispheres, each from one mouse (two mice in total), were pooled and lysed. Due to their high lipid content, both tissues were hard to lyse. We measured the total protein contents of the lysates by the PDB assay, and then loaded the lysates containing 50, 25, and 15 µg of total proteins for SDS-PAGE and Western blot (Figure 4C).

Ponceau S staining of the membrane after protein transfer revealed equal loading of the samples (Figure S5), indicating that the PDB assay performs similarly as lysates in RIPA buffer or 2x SDS LB. Interestingly, we observed striking differences in the band intensities between the nerve and the brain lysates. We hypothesize that this is due to a high abundance of albumin (the strong band below 70 kDa) and the IgG heavy (the strong band slightly above 55 kDa) and light chain (the strong band between 25 and 35 kDa) in the peripheral nerves, which are absent in the brains because of the blood–brain barrier [16,17]. This circumstance will lead to an overestimation of the actual protein content of the sciatic nerve and thus biased loading. This could be a major issue when comparing protein expression between the peripheral and the central nervous systems.

Our Western blot results showed that the cytoskeleton-associated protein merlin, as well as the nuclear protein Histone H1, the autophagic vesicle membrane proteins LC3A/B, and the myelin basic protein MBP, were better extracted by 2x SDS LB than by RIPA buffer (Figure 4C). In contrast, the cytoplasmic proteins MEK 1/2, ERK 1/2, GAPDH, and GAP-43 were indistinguishably extracted by the two buffers. These findings confirm the literature

reports that 2x SDS lyses biological tissues more effectively than RIPA buffer, which can be attributed to the insolubility of cytoskeleton-associated and extracellular matrix proteins in RIPA [12].

**Figure 4.** Western blots based on protein quantification with the PDB assay. (**A**,**B**) Ponceau S-stained dot blot. Undiluted BSA standards and BSA standards diluted 1:1 in 2x SDS lysis buffer were spotted in duplicate onto a membrane (fixed concentration, variable volumes). A quantity of 1 µL of sciatic nerve (SN) and brain samples lysed in 2x SDS LB or RIPA buffer was also applied onto the same membrane for quantification with the PDB assay. (**C**) The nerve and brain lysates containing 50, 25 or 15 µg of total proteins (based on the PDB assay) were loaded for SDS-PAGE and Western blot. After protein transfer to a nitrocellulose membrane, the membrane was also stained with Ponceau S.

To our surprise, p-ERK1/2 and p-MEK1/2 levels were higher in 2x SDS LB lysates than in RIPA lysates. We could not exclude the possibility that p-ERK1/2 and p-MEK1/2 in the respective nerves were indeed different. If this is not the case, the result suggests that RIPA buffer, which contains a designated phosphatase buffer, is inferior to 2x SDS LB in terms of phosphatase inhibition, which may be explained by the strong denaturing effect of SDS. This observation may be important for studying rapidly changing signaling processes, such as degeneration and regeneration in the nervous system [18,19], where 2x SDS LB may be a better choice for tissue lysis.

In the course of our experiments, we found that the PDB assay, especially when using 2x SDS LB instead of RIPA buffer to lyse biological tissue, can be much cheaper than the established workflow in our laboratory, which relies on the widely used BCA assay for protein quantification. Therefore, we calculated the costs for quantification of 12 samples with a commercial or a self-made BCA assay, or a PDB assay. In this study, we only used the commercial BCA assay, and the reagents can be self-made. The self-made BCA kit would cost between EUR 15.29 and 24.91, a commercial BCA kit was EUR 13.47, and the PDB assay cost only EUR 2.05 for quantification of 12 samples. From our experience, Ponceau S solution can be reused at least 20 times to stain transfer membranes or dot blots. A laboratory quantifying 1000 samples with the PDB instead of the BCA assay would save about EUR 950 (Figure S6). Because we have shown that direct lysis of biological tissues in 2x SDS LB is compatible with protein quantification by PDB assay and superior to RIPA buffer in terms of protein solubilization, we also checked the overall price difference between RIPA and 2x SDS LB, which amounted to a difference of between EUR 1380 and 2310 for 1000 lysates (Figure S7). One laboratory could save over EUR 2300 per 1000 lysates by switching from tissue lysis with RIPA buffer and protein quantification with the BCA assay to 2x SDS LB and our PDB assay.

The time needed for a PDB assay or a BCA assay is comparable: both assays will take around 45 min from sample preparation to measurement with a plate reader (BCA assay) or to scanning (PDB assay); the subsequent analysis including standard curve preparation, linear regression, and calculation of protein concentrations is almost the same.

Previously, Morcol et al. described a similar method to quantify purified proteins [8]. Here, we demonstrate an improved PDB assay that can be used to quantify tissue lysates and is compatible with 2x SDS LB. The PDB assay has the following advantages: (1) it can be easily adjusted to quantify samples with very high or low protein contents; (2) it requires far fewer materials than the BCA assay, saving valuable samples; (3) it only needs a simple laboratory equipment; and (4) it is significantly cheaper and more reliable than the BCA assay, thereby facilitating downstream experiments relying on correct sample inputs.

#### **4. Conclusions**

The PDB assay is a cheap and reliable method for quantifying proteins in tissue lysates.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/ 10.3390/gels8010043/s1. Figure S1: Extended range for the PDB assay; Figure S2: The integrated density is determined by the protein amount instead of the dot size; Figure S3: Standard curves related to Figure 4A,B; Figure S4: BSA standards diluted with RIPA buffer forms "coffee rings" on dot blots; Figure S5: Ponceau S staining of the membrane after SDS-PAGE and protein transfer, uncropped, related to Figure 4C; Figure S6: Cost comparison between the PDB and the BCA assays; Figure S7: Cost comparison between the PDB assay with samples lysed in 2x SDS LB and the BCA assay with samples lysed in RIPA buffer; Video S1: Workflow of the PDB assay.

**Author Contributions:** Conceptualization, D.L.H., L.B. and Y.C.; methodology, D.L.H., L.B., Y.C. and N.O.; software D.L.H. and L.B.; validation D.L.H. and L.B.; formal analysis D.L.H. and L.B.; investigation, D.L.H. and L.B.; resources, D.L.H., L.B. and Y.C.; data curation, D.L.H. and L.B.; writing—original draft preparation, D.L.H., L.B., Y.C. and L.K.S.; writing—review and editing, Y.C., D.L.H., L.B. and L.K.S.; visualization, D.L.H., L.B. and N.O.; supervision, D.L.H. and Y.C.; project administration, D.L.H., L.B. and Y.C.; funding acquisition, Y.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** FLI is a member of the Leibniz Association and is financially supported by the Federal Government of Germany and the State of Thuringia. This work was supported in part by funding from the Deutsche Forschungsgemeinschaft (DFG) granted to YC (MO 1421/5-1).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** All data underlying the results are available as part of the article and no additional source data are required.

**Acknowledgments:** The authors would like to thank Helen Morrison for providing laboratory space and support, and Debra Weih for critical reading and editing of the manuscript.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


**Ajamaluddin Malik 1, \*, Javed Masood Khan 2 , Abdullah S. Alhomida <sup>1</sup> and Mohammad Shamsul Ola 1**


**Abstract:** Alpha-crystallin protein performs structural and chaperone functions in the lens and comprises alphaA and alphaB subunits at a molar ratio of 3:1. The highly complex alpha-crystallin structure challenges structural biologists because of its large dynamic quaternary structure (300–1000 kDa). Camel lens alpha-crystallin is a poorly characterized molecular chaperone, and the alphaB subunit possesses a novel extension at the N-terminal domain. We purified camel lens alpha-crystallin using size exclusion chromatography, and the purity was analyzed by gradient (4–12%) sodium dodecyl sulfate–polyacrylamide gel electrophoresis. Alpha-crystallin was equilibrated in the pH range of 1.0 to 7.5. Subsequently, thermal stress (20–94 ◦C) was applied to the alpha-crystallin samples, and changes in the conformation and stability were recorded by dynamic multimode spectroscopy and intrinsic and extrinsic fluorescence spectroscopic methods. Camel lens alpha-crystallin formed a random coil-like structure without losing its native-like beta-sheeted structure under two conditions: >50 ◦C at pH 7.5 and all temperatures at pH 2.0. The calculated enthalpy of denaturation, as determined by dynamic multimode spectroscopy at pH 7.5, 4.0, 2.0, and 1.0 revealed that alpha-crystallin never completely denatures under acidic conditions or thermal denaturation. Alpha-crystallin undergoes a single, reversible thermal transition at pH 7.5. The thermodynamic data (unfolding enthalpy and heat capacity change) and chaperone activities indicated that alpha-crystallin does not completely unfold above the thermal transition. Camels adapted to live in hot desert climates naturally exhibit the abovementioned unique features.

**Keywords:** alpha-crystallin; dynamic multimode spectroscopy; circular dichroism; fluorescence; thermal stability

#### **1. Introduction**

Alpha-crystallin belongs to the small heat shock protein (sHsp) superfamily, is highly expressed in the eye lens, and has at least two known functions. First, alpha-crystallin is a structural protein that maintains an appropriate refractive index (ability to focus light on the retina). Second, as a molecular chaperone, it maintains lens clarity throughout the lifespan of an organism [1]. Eye lens proteins are frequently exposed to environmental stress, including UV-radiation and high temperatures. The mature lens fibers lack a protein folding machinery and all organelles to minimize light scattering. Consequently, no new protein can be synthesized, and damaged proteins cannot be replaced. Therefore, the eye lens in all organisms must maintain damaged proteins such as alpha-crystallin in a soluble state throughout life. The camel has adapted to thrive in extreme desert climates, which includes high temperatures, solar radiation, dryness, and low nutrition. High ambient temperature and UV–Vis radiation may increase the lens temperature and induce protein misfolding and aggregation [2]. Epidemiological studies have shown a positive association between early-onset and a high grade of cataracts and prolonged sunlight exposure [3,4].

**Citation:** Malik, A.; Khan, J.M.; Alhomida, A.S.; Ola, M.S. Modulation of the Structure and Stability of Novel Camel Lens Alpha-Crystallin by pH and Thermal Stress. *Gels* **2022**, *8*, 273. https:// doi.org/10.3390/gels8050273

Academic Editors: Hiroyuki Takeno and Vijay Kumar Thakur

Received: 14 March 2022 Accepted: 22 April 2022 Published: 27 April 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Therefore, the role of the camel lens in maintaining lenticular alpha-crystallin in a soluble state throughout its entire life presents challenges. Unfolding and aggregation of lenticular proteins cause lens opacity, resulting in cataract formation.

Lens alpha-crystallin is a large, heterogeneous multimeric protein comprising two subunits (alphaA and alphaB chains), each approximately 175 amino acids and exhibiting 60% homology. In the human eye lens, alpha-crystallin comprises 15–50 subunits of two homologous forms, alphaA and alphaB, each approximately 20 kDa [5,6] and at a 3:1 molar ratio [5]. Camels have evolved uniquely (anatomically, physiologically, and biochemically) to adapt to the scorching climate where most other mammals cannot survive. To our best knowledge, the camel eye lens has at least two novel features: the recruitment of high levels of taxon-specific zeta-crystallin and the presence of an extended N-terminal domain in the alphaB-crystallin protein. Camel alphaA-crystallin (XP\_010998042.1) comprises 173 amino acid residues, identical to human alphaA-crystallin. However, camel alphaBcrystallin (XP\_010984284.1) contains an additional 44 residues compared with human alphaB-crystallin (219 residues vs. 175 residues, respectively) [7].

The expression of alphaA-crystallin is primarily lens specific; in other tissues, it is expressed in trace amounts. By contrast, alphaB-crystallin expression is stress-inducible and widespread throughout the body, particularly in the heart, muscle, and brain [8,9]. AlphaB-crystallin overexpression is linked to several protein misfolding and neurodegenerative diseases, including myopathies [10,11], Parkinson's disease [12,13], Alzheimer's disease [14,15], Creutzfeldt–Jakob disease [16,17], multiple sclerosis [18,19], and cancer [20,21].

Alpha-crystallin acts as a "holdase" in an ATP-independent manner [1,22]. The size of the hetero-oligomeric quaternary structure of alpha-crystallin is diverse with an average molecular weight of 700 kDa, and its size ranges from 300 to 1000 kDa [5,23]. The size variation is caused by several factors (e.g., pH, ionic strength, temperature, and metal ions). Temperature is a critical factor for alpha-crystallin oligomerization [23,24]. Recently, we reported that the chaperone activity of camel alpha-crystallin is activated in a stepwise manner during heat stress [7]. Moreover, camel alpha-crystallin retains a native beta-sheeted dominant secondary structure up to 50 ◦C. High thermal stress (above 50 ◦C) leads to a structural transition in alpha-crystallin with a gain of a randomcoiled-like structure without losing beta-sheeted content [7]. In previous studies of alphacrystallin, temperature was associated with the single thermal transition and activation of its chaperone activity [25,26].

Thermal transition was reported to occur above 50 ◦C with a transition mid-point (*T*m) of approximately 61–63 ◦C [7,27–29]. Interestingly, alpha-crystallin efficiently retained its function during and above thermal transition [7,30,31]. Surprisingly, the calculated enthalpy of denaturation (∆H) for alpha-crystallin at pH 7.5 using different techniques [Differential scanning calorimetry (DSC), dynamic multimode spectroscopy (DMS), and Fourier transform infrared spectroscopy (FTIR)] was significantly lower than the theoretically estimated enthalpy of denaturation [7,32]. The experimentally calculated heat capacity change of alpha-crystallin denaturation (∆Cp) was also less than half of the theoretical (∆Cp) value [26,32,33]. Several reports have shown the structural integrity of alpha-crystallin under thermal denaturation at pH 7.5. However, ambiguities exist regarding the folding species of alpha-crystallin above the transition. Whether alpha-crystallin is fully unfolded, partially unfolded, or retains a native-like structure remains unclear [28,29,34,35]. Alphacrystallin protein presents challenges for structural determination, and its crystal structure is unavailable. In particular, camel lens alpha-crystallin comprises an extended N-terminal domain and is poorly characterized. In the present study, we used multi-spectroscopic techniques to characterize the folding and thermodynamic characteristics of alpha-crystallin in the pH range of 1.0–7.5 and temperature range of 20–94 ◦C. Many proteins are unfolded at low pH due to loos of electrostatic interactions. Several types of forces such as ionic, hydrophobic, H-bond, and covalent interactions are responsible to maintain the structure–function relationship of the proteins. The change in medium may perturb these

interactions and result in protein unfolding. In this study, we have evaluated the role of pH and temperature on the stability of alpha-crystallin.

#### **2. Materials and Methods**

Superdex 200 and Superdex 75 prepacked columns were obtained from GE Healthcare Life Sciences, Chicago, USA. The 4–12% gradient SDS-PAGE gels were purchased from Invitrogen and Bradford's reagent was obtained from Pierce. All other chemicals were of analytical grade.

#### *2.1. Extraction and Purification of Alpha-Crystallin from Camel Lens*

Fresh camel eye lenses were obtained from a local slaughterhouse and transported under chilled conditions. Two lenses were gently stirred in 50 mL of extraction buffer [20 mM sodium phosphate buffer (pH 7.8) containing 0.2 mM EDTA] for 30 min to extract the soluble lens protein. The supernatant was collected after centrifugation at 13,000 rpm for 15 min. Alpha-crystallin was purified using two different size exclusion chromatography columns (Superdex 200 and Superdex 75 gel filtration columns). The Superdex 200 and Superdex 75 gel filtration columns were equilibrated with 20 mM sodium phosphate buffer (pH 7.8) containing 0.2 mM EDTA. The clear soluble lens extract was passed through a Superdex 200 column, and the purity of the eluted fractions was evaluated by 4–12% gradient sodium dodecyl sulfate–polyacrylamide gel electrophoresis. Subsequently, fractions containing relatively pure alpha-crystallin were pooled and passed through a Superdex 75 gel filtration column. The fraction purity was re-analyzed by 4–12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis [7]. The pure fractions were concentrated to 8 mg/mL and stored at −20 ◦C. The protein was quantified using the Bradford assay.

#### *2.2. Equilibration of Alpha-Crystallin at Different pH Values*

Camel lens alpha-crystallin (0.3 mg mL−<sup>1</sup> ) was equilibrated overnight with a 20 mM buffer (pH 1.0–7.5) at room temperature. To obtain the desired pH, the following buffers were used: KCl-HCl (pH 1.0 and 1.5), Gly-HCl (pH 2.0–3.0), acetate buffer (pH 4.0–5.0), and phosphate buffer (pH 6.0–7.5).

#### *2.3. Far-UV CD Spectroscopy of Alpha-Crystallin at Different pH Values*

Far-UV CD measurements of alpha-crystallin equilibrated at different pH values were obtained using a ChirascanPlus spectropolarimeter (Applied Photophysics Ltd., London, UK) and coupled with a Peltier temperature controller. The far-UV CD spectra of alphacrystallin were measured at a concentration of 0.3 mg/mL in a 1-mm-pathlength cuvette at 22 ◦C. Three spectra for each sample were scanned from 200 to 250 nm with a 1-nm bandwidth, and the data were collected at 0.5 s per point. The air baseline and buffer background were subtracted from each spectrum of alpha-crystallin.

#### *2.4. Intrinsic Fluorescence Spectroscopy of Alpha-Crystallin at Different pH Values*

The tryptophan fluorescence emission spectra of alpha-crystallin at different pH values were measured at room temperature using a Cary Eclipse Fluorescence Spectrophotometer (Agilent Technologies, Santa Clara, CA, USA) coupled with a Peltier temperature controller [36]. Alpha-crystallin (0.1 mg/mL) at different pH values (1.0 to 7.5) in a 10-mmpathlength cuvette was excited at 295 nm (bandwidth, 5 nm each) to record tryptophan fluorescence emission spectra.

#### *2.5. Dynamic Multimode Spectroscopy of Alpha-Crystallin at Different pH Values*

DMS was performed using a ChirascanPlus spectrophotometer [37]. Based on the observation of major secondary and tertiary structural transitions in camel lens alphacrystallin with respect to pH values, four different pH values (1.0, 2.0, 4.0, and 7.5) were selected for detailed spectroscopic and thermodynamic studies. Camel lens alpha-crystallin (0.2 mg/mL) was dissolved in 20 mM buffer at pH 1.0, 2.0, 4.0, and 7.5, and temperaturedependent conformational changes were measured in 1-mm-pathlength cuvettes using internal temperature probes. Alpha-crystallin was gradually heated from 20 ◦C to 94 ◦C at a rate of 1 ◦C/min, and far-UV CD spectra were recorded between 200 and 250 nm. The thermal transition data were processed using Chirascan Global 3 software provided by the manufacturer.

#### *2.6. ANS (8-Anilino-1-naphthalene sulfonate) Fluorescence Measurements of Alpha-Crystallin at Different Temperatures and pH Values*

ANS fluorescence of alpha-crystallin (0.2 mg mL−<sup>1</sup> ) at pH 1.0, 2.0, 4.0, and 7.5, respectively, were recorded at different temperatures (5 ◦C increments at each step) from 20 ◦C to 90 ◦C using a Peltier-controlled Cary Eclipse Fluorescence Spectrophotometer. ANS (50 µM) was added to the alpha-crystallin samples at pH 1.0, 2.0, 4.0, and 7.5, respectively, and the samples were equilibrated for 1 h at room temperature. The solution temperature was monitored using an internal temperature probe. ANS-treated alpha-crystallin was gradually heated and allowed to equilibrate for 5 min at each temperature step. The ANS fluorescence emission spectra were recorded between 400 and 600 nm (5.0-nm slit) by exciting the samples at 375 nm (2.5-nm slit).

#### **3. Results**

#### *3.1. Effect of pH on the Secondary Structure of Camel Lens Alpha-Crystallin*

Pure alpha-crystallin was obtained as previously described [7]. Far-UV CD (200–250 nm) was used to characterize the effect of acidic pH on the secondary structure of alphacrystallin (0.3 mg mL−<sup>1</sup> ) (Figure 1). The far-UV CD spectra of alpha-crystallin at pH 7.5 revealed a single negative minimum at 217, which is a characteristic feature of beta-sheeted proteins. Changes in the negative minima of alpha-crystallin were insignificant, between pH 4.0 and 7.5 (Figure 1, inset), but the ellipticity at 217 nm gradually decreased as the pH was reduced from 7.5 to 5.0, indicating a loss of secondary structure (beta-sheeted structure) (Supplementary Figure S1). The maximum loss of secondary structure was observed at pH 5.0. The alpha-crystallin quickly regained a beta-sheeted structure below pH 5.0, particularly at pH 4.5, 4.0, and 3.0, respectively, which was close to the native secondary structure. Interestingly, at pH 3.0, 2.5, and 2.0, respectively, the alpha-crystallin secondary structure transformed into a random coil structure without affecting its betasheeted core structure (Figure 1). When alpha-crystallin was further incubated at pH 1.0 and 1.5, negative ellipticity was regained to that of native alpha-crystallin (Figure 1, inset). The far-UV CD data indicated that the secondary structure of alpha-crystallin was unstable with respect to pH changes. Figure 1 shows the different folding states of alpha-crystallin in which the far-UV CD minima varied with pH: native state (pH 7.5), beginning of the random-coil-like structure (pH 4.0), random-coiled structure (pH 2.0), and gain of the native-like secondary structure (pH 1.0).

‐ − ‐ ‐ ‐ ‐ **Figure 1.** Far-UV circular dichroism (CD) spectra of 0.3 mg mL−<sup>1</sup> of alpha-crystallin at pH 1.0 to 7.5. Changes in the ellipticity minima at different pH values are plotted in the inset figure. Below pH 3.5, alpha-crystallin gained a random coil-like structure but was restored to a native-like structure at pH 1.0.

#### *3.2. Effect of pH on the Tertiary Structure of Alpha-Crystallin*

*‐* ‐ ‐ ‐ ‐ ‐ ‐ ‐ λ ‐ λ ‐ ‐ ‐ ‐ ‐ ‐ ‐ λ ‐ ‐ ‐ ‐ λ ‐ ‐ λ Intrinsic fluorescence spectroscopy was used to investigate changes in the alphacrystallin tertiary structure with respect to pH changes. Measurements of intrinsic fluorescence are useful readouts of the microenvironment surrounding aromatic residues and provide information regarding even subtle changes in the tertiary structure of proteins [38–40]. Figure 2 shows the tryptophan fluorescence spectra of alpha-crystallin at pH 1.0 to 7.5, revealing that alpha-crystallin at pH 7.5 exhibited a maximum fluorescence intensity at 336 nm. This finding confirmed that alpha-crystallin existed in a well-folded form. As the pH was reduced, the fluorescence emission maximum (λmax) of alpha-crystallin was unchanged up to pH 5.5. However, below pH 5.5 down to pH 2.0, a gradual redshift in the λmax was observed, indicating exposure of tryptophan residues to the polar environment (Figure 2, inset). The redshift in the wavelength maximum occurs when the microenvironment surrounding tryptophan residues becomes polar (aqueous), indicating protein unfolding or a loss of protein tertiary structure. The maximum redshift of alpha-crystallin was found at pH 2.0, indicating that the alpha-crystallin tertiary structure was maximally lost. Interestingly, at pH values below 2.0, the λmax of the alpha-crystallin returned to 336 nm (i.e., the native-like structure), indicating that alpha-crystallin was again refolded at pH 2.0. The λmax of alpha-crystallin was 352 nm; in the completely unfolded state (in 6 M GdnHCl), alpha-crystallin showed a λmax of 363 nm, indicating a partially unfolded state of alpha-crystallin at pH 2.0.

‐

‐ ‐ ‐ − ‐ λ **Figure 2.** Tryptophan fluorescence spectra of camel lens alpha-crystallin at pH 1.0–7.5. Alphacrystallin (0.1 mg mL−<sup>1</sup> ) equilibrated at different pH values was excited at 295 nm. Emission spectra were collected from 300 to 400 nm at room temperature (bandwidth 5 nm each). Each spectrum recorded at a different pH is color-coded. The inset figure shows the λmax plot with respect to pH.

#### *3.3. Changes in Surface Hydrophobicity at Selected pH Values*

‐ ‐ ‐ ‐ ‐ ‐ − ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ The partially unfolded states or molten globule states of proteins are frequently characterized by measuring changes in ANS fluorescence. ANS has a significantly lower binding affinity with native and fully denatured proteins because the appropriate hydrophobic patches are unavailable for ANS binding. However, the partially unfolded or molten globule state of the protein exposes hydrophobic patches and provides a suitable environment for ANS binding and producing high ANS fluorescence intensity [41,42]. The exposure of hydrophobicity of alpha-crystallin (0.2 mg mL−<sup>1</sup> ) at four different pH values (1.0, 2.0, 4.0, and 7.5, respectively) was measured at room temperature (Figure 3). Alpha-crystallin in the native state (pH 7.5) exhibited a poor ANS binding signal, indicating that alpha-crystallin in the native state has low surface hydrophobicity (Figure 3). This finding indicates that alpha-crystallin is well-folded. However, at pH 4.0, 2.0, and 1.0, respectively, the fluorescence intensity of ANS increased in response to a change in pH, confirming that the surface hydrophobicity of alpha-crystallin was increased with respect to the change in pH. The increased surface hydrophobicity with respect to pH resulted from the formation of the molten, globule-like state of alpha-crystallin.

‐ ‐ ‐ − μ ‐ **Figure 3.** Surface hydrophobicity measurements of alpha-crystallin at four selected pH values. Alphacrystallins (0.2 mg mL−<sup>1</sup> ) equilibrated at pH 1.0, 2.0, 4.0, and 7.5, respectively, were treated with 50 µM ANS. The samples were excited at 375 nm (2.5 nm slit) and the emission spectra were recorded between 400 and 600 nm (5.0 nm slit).

#### *‐ 3.4. Thermodynamic and Spectroscopic Properties of Alpha-Crystallin at Different pH Values as Determined by Dynamic Multimode Spectroscopy*

‐ ‐ ‐ ‐ ‐ ‐ − ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ The spectroscopic and thermodynamic properties of alpha-crystallin at four different pH values (1.0, 2.0, 4.0, and 7.5) were examined by DMS [43]. Alpha-crystallin revealed distinct secondary and tertiary structures at pH 1.0, 2.0, 4.0, and 7.5, respectively. Therefore, we selected these pH values to evaluate the thermodynamic and folding characteristics of alpha-crystallin. Alpha-crystallin samples (0.2 mg mL−<sup>1</sup> ) at pH 1.0, 2.0, 4.0, and 7.5, respectively, were heat-stressed from 20 ◦C to 94 ◦C at a rate of 1 ◦C/min under identical conditions. The far-UV CD spectra (200 to 250 nm) were recorded as a function of temperature. Figure 4A–D shows the changes in the secondary structure conformation of alpha-crystallin at different temperatures and pH values. Alpha-crystallin in the native state (pH 7.5 and room temperature) exhibited a single minimum at 217 nm, representing a characteristic feature of a beta-sheeted rich protein (Figure 4A). The peaks at 217 nm were unchanged during heat stress (20–94 ◦C), indicating that the alpha-crystallin beta-sheeted core structure was preserved during heat stress. Moreover, the ellipticity at 217 nm was unchanged between 20 ◦C and 50 ◦C, indicating that the secondary structure was intact over this temperature range. Interestingly, above 50 ◦C, a sharp increase in the ellipticity minima at 203 nm was observed without altering the 217 nm ellipticity, indicating the formation of a random coil-like structure while maintaining the original beta-sheeted structure (Figure 4A, inset).

‐

‐ ‐

‐ ‐ ‐ ‐ − ‐ − ‐ **Figure 4.** Far-UV circular dichroism (CD) spectra of camel lens alpha-crystallin at different temperatures and pH values. Alpha-crystallin (0.2 mg mL−<sup>1</sup> ) was heat-denatured at a constant rate (1 ◦C min−<sup>1</sup> ) at pH 7.5 (**A**), pH 4.0 (**B**), pH 2.0 (**C**), and pH 1.0 (**D**). Far-UV CD spectra were collected from 200 to 250 nm at intervals of 1 ◦C from 20 ◦C to 94 ◦C. In the inset figure, the blue line shows the changes at 203 nm and the pink dots at 217 nm, with respect to temperature.

‐ ‐

‐ ‐

‐ ‐ ‐ ‐ ‐

‐

‐

‐ ‐ ‐ ‐ Our results also demonstrated that far-UV CD ellipticity at 217 nm was unchanged at pH 7.5 and 2.0 (Figure 4A,C). By contrast, a slight increase in the ellipticity minima at 217 nm at pH 4.0 and 1.0 (Figure 4B,D) was detected during thermal denaturation. These results indicated that the core beta-sheeted structure of alpha-crystallin remained intact during thermal denaturation (Figure 4). Moreover, the far-UV CD spectra of thermally stressed (>80 ◦C) alpha-crystallin at pH 7.5 were similar to those of alpha-crystallin at pH 2.0 at all temperatures (20 ◦C–94 ◦C). These data showed the presence of random-coiled and beta-sheeted structures under two conditions: alpha-crystallin above 50 ◦C at pH 7.5 and alpha-crystallin at all temperatures at pH 2.0 (Figure 4A,C). The far-UV CD spectra of alpha-crystallin at pH 4.0 and 1.0 were similar to native-like alpha-crystallin, and these conformations did not undergo any major structural transitions during heat stress, except a slight gain of ellipticity minima at 217 nm (Figure 4B,D). Moreover, thermal stress at pH 1.0, 2.0, 4.0, and 7.5, respectively was reversible, and no aggregation was detected.

The thermal transition midpoints (*T*m) and enthalpy of alpha-crystallin at pH 1.0, 2.0, 4.0, and 7.5, respectively, were determined (Table 1) using Global 3 analysis software

provided by Applied Photophysics Ltd., UK. The three-dimensional model of the thermal transitions in alpha-crystallin at pH 1.0, 2.0, 4.0, and 7.5, respectively, was generated using Global 3 analysis software (Figure 5). ‐

**Table 1.** Thermal transition midpoints ™ and enthalpies of alpha-crystallin at pH 1.0, 2.0, 4.0, and 7.5, respectively.


‐ ‐ ‐ ‐ ‐ ‐ − **Figure 5.** Calculated temperature, wavelength, and far-UV CD signal of alpha-crystallin at different pHs. The three-dimensional model of alpha-crystallin at pH 7.5 (**A**), 4.0 **(B**), 2.0 (**C**), and 1.0 (**D**), respectively, was calculated using Global 3 software and the far-UV CD signal obtained during temperature ramping (1 ◦C min−<sup>1</sup> ).

#### *‐ 3.5. Changes in the Surface Hydrophobicity of Alpha-Crystallin at Different Temperatures and pH Values*

‐ ‐ ‐ ‐ ‐ ‐ ‐ Extrinsic fluorophore ANS was used to monitor the exposure of hydrophobic patches in alpha-crystallin in response to thermal stress. ANS fluorescence of alpha-crystallin at pH 1.0, 2.0, 4.0, and 7.5, respectively, was evaluated at different temperatures, from 20 ◦C to 90 ◦C (Figure 6). When alpha-crystallin at pH 7.5 was heat stressed from 20 ◦C, a slight redshift in the wavelength maxima was observed above 35 ◦C, and a sharp redshift was observed above 55 ◦C (Figure 6A), indicating exposure of hydrophobic residues at the surface of alpha-crystallin in response to heat stress. Alpha-crystallin at pH 7.5 exhibited low ANS fluorescence intensity and displayed poor ANS binding with native state alphacrystallin at pH 7.5 (Figure 6A). Because ANS is a temperature-sensitive probe, a gradual decrease in ANS fluorescence intensity was observed at all pH values. At acidic pH (4.0, 2.0,

‐

and 1.0), the ANS fluorescence of alpha-crystallin exhibited increased fluorescence intensity resulting from the exposure of hydrophobic patches (Figure 6B–D). A slight redshift in emission was observed only above 65 ◦C when alpha-crystallin was at pH 1.0, 2.0, and 4.0, respectively (Figure 6E). ‐ ‐

‐ ‐ λ ‐ **Figure 6.** Differential scanning fluorometry of alpha-crystallin using an extrinsic fluorophore at four different pH values. ANS fluorescence was monitored by incubating alpha-crystallin at pH 7.5, 4.0, 2.0, and 1.0, respectively, from 20 ◦C to 90 ◦C by stepwise temperature increases of 5 ◦C. (**A**) pH 7.5, (**B**) pH 4.0, (**C**) pH 2.0, and (**D**) pH 1.0. (**E**) Effect of temperature on the λmax of alpha-crystallin at pH 7.5, 4.0, 2.0, and 1.0, respectively.

#### **4. Discussion**

‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ Ocular proteins are exposed to environmental stress (solar radiation and ambient temperature), making them susceptible to unfolding and aggregation. Alpha-crystallin naturally maintains ocular proteins in a soluble state. The Arabian camel has evolved to live in a stressful desert climate of intense heat, solar radiation, and dryness. The camel possesses several unique anatomical, physiological, and biochemical features to survive and thrive in the extreme desert climate [44–48]. The camel eye lens has two modifications with respect to crystallin proteins: it contains levels of zeta-crystallin (a taxon-specific crystallin) [49] and an extended N-terminal domain in the alphaB-crystallin protein [7]. Camel alphaA-crystallin (NCBI Reference Sequence: XP\_010998042.1) is identical in length and shares 93% homology with human alphaA-crystallin, whereas camel alphaB-crystallin (NCBI Reference Sequence: XP\_010984284.1) contains 44 additional residues at its Nterminus and shows an overall 78% similarity with human alphaB-crystallin [7].

‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ The alpha-crystallin architecture is highly complex, and its quaternary structure changes (forms larger functional oligomers) as the temperature increases. Alpha-crystallin presents challenges for structural biologists to study. Despite several attempts, the crystal structure of the alpha-crystallin has not been solved. Its three-dimensional structure has been reconstituted using multiple techniques, including cryo-electron microscopy, NMR, dynamic light scattering, analytical ultracentrifugation, and structural modeling [50,51]. Because the hetero-oligomeric quaternary structure of alpha-crystallin changes in response to temperature, interpreting the structural changes in alpha-crystallin using spectroscopic

techniques precisely is challenging. We used multiple spectroscopic methods (DMS, intrinsic, and ANS fluorescence) to characterize the changes in the secondary and tertiary structure and surface hydrophobicity of camel lens alpha-crystallin in response to changes in pH and temperature.

At pH 7.5, camel lens alpha-crystallin exhibited a single minimum at 217 nm, indicating the presence of a beta-sheeted dominant structure. Earlier studies reported a beta-sheeted dominant structure in bovine alpha-crystallin at pH 7.5 [32,34]. The effect of pH from 1.0 to 7.5 on the secondary structure of camel lens alpha-crystallin revealed that the single negative minimum remained at 217 nm between pH 7.5 and 4.0. During the pH scanning experiment, multiple conformational changes were detected in the secondary structure of alpha-crystallin. Initially, loss of the beta-sheeted structure occurred up to pH 5.0. A further reduction of pH to 4.0 induced a beta-sheeted structure, resulting in the formation of native-like alpha-crystallin. A subsequent reduction of pH resulted in major conformational changes, as shown in Figure 1. Below pH 4.0, alpha-crystallin contained native-like betasheeted and random-coiled structures. Interestingly, at pH 1.5 and 1.0, the far-UV CD spectra nearly overlapped with alpha-crystallin at pH 7.5, indicating a restoration of the native-like beta-sheeted structure and a loss of the random-coiled structure.

For a detailed investigation of the spectroscopic and thermodynamic properties of alpha-crystallin, four different pH values were selected: pH 7.5, native alpha-crystallin; pH 4.0, native-like alpha-crystallin; pH 2.0, alpha-crystallin with random coils; and pH 1.0, native-like alpha-crystallin. DMS based on far-UV CD spectroscopy was used to obtain thermodynamic and spectroscopic data. DMS is an information-rich technique that precisely determines changes in the secondary structure of proteins under different conditions over the entire temperature range [43,52]. Figure 4A–D and Table 1 show the thermodynamic and spectroscopic data obtained by DMS. Below 50 ◦C (pre-transition) and at pH 7.5, alpha-crystallin exists as a beta-sheeted dominant protein (Figure 4A). It retained a native beta-sheeted secondary structure up to 50 ◦C at pH 7.5. However, above 50 ◦C, the minima shifted toward 203 nm and maintained a shoulder at 217 nm. The ellipticity at 217 nm in the pre- and post-transition spectra remained the same. Chemical denaturants (urea and single minimum) induce a decrease in ellipticity at 217 nm (loss of beta-sheeted secondary structure) and subunit dissociation in the bovine alpha-crystallin [53]. Our study and earlier studies showed that the core of the beta-sheeted structures in alpha-crystallin at pH 7.5 remained intact under thermal denaturation temperatures up to 94 ◦C [28,32,34,35]. However, thermal stress above 50 ◦C induced a random-coil-like structure at pH 7.5.

Large conformational changes in alpha-crystallin occurred over the pH range of 1.0 to 7.5 (Figure 1). Initially, a loss of ellipticity at 217 nm occurred after shifting the pH from 7.5 to 5.0. Further reduction of the pH resulted in the formation of a beta-sheeted core structure in alpha-crystallin. Interestingly, the temperature did not affect the beta-sheeted secondary structure at all the pH values (7.5, 4.0, 2.0, and 1.0) tested; a slight gain of a beta-sheeted structure occurred at pH 4.0 and 1.0 (Figure 4). The formation of a random-coiled structure in alpha-crystallin occurred under two conditions: (i) >50 ◦C at pH 7.5 and (ii) 20 ◦C at pH 2.0.

Camel lens alpha-crystallin at pH undergoes a single thermal transition between 50 ◦C and 80 ◦C. After the thermal transition, whether alpha-crystallin was fully folded, partially folded, or fully denatured was unclear. However, the folding species was reversible. In the present study, the mid-point of the thermal transition (*T*m) was 60.9 ± 0.1 ◦C, and the enthalpy of denaturation was 237.0 ± 1.9 kJ/mol. In earlier studies, the thermal stability and structural changes of eye lens alpha-crystallin from other sources were determined [28,35,54]. Alpha-crystallin underwent a single thermal transition (*T*m) at approximately 61–64 ◦C [27–29]. The minor ambiguity in the *T*<sup>m</sup> may be due to the difference in the detection methods (far-UV CD, DSC, or FTIR), buffer pH, ionic strength, or experimental conditions. The enthalpy of bovine lens alpha-crystallin was 235 kJ/mol by DSC [32]. Usually, the unfolding enthalpy of monomeric globular proteins of similar sizes is approximately 2.92 kJ/mol per residue [55]. Therefore, the calculated unfolding enthalpy of the

alpha-crystallin subunits was more than 500 kJ/mol. Moreover, oligomerization increased the unfolding enthalpy of the proteins [56,57]. Although the crystal or NMR structure of alpha-crystallin is unavailable, data from multi-technique investigations have led to a consensus that subunit assembly of alpha-crystallin is controlled by the hydrophobicity of the N-terminal domains [50,51,58]. Therefore, dissociation of the oligomeric structure would result in the exposure of the buried hydrophobic patches, leading to an endothermic effect, which increases the overall unfolding enthalpy. Therefore, the calculated unfolding enthalpy of oligomeric alpha-crystallin at pH 7.5 may be much higher than 500 kJ/mol. Less than half of the enthalpic changes were determined in this study and earlier studies [32]. Accordingly, the DMS data presented in this work and an earlier study revealed that alpha-crystallin retained a secondary structure above the thermal transition (*T*m). These data also indicate that the thermal transition of alpha-crystallin does not dissociate its subunits. Denaturation without subunit dissociation was also observed in the Arc repressor [59]. The presence of a secondary structure in the alpha-crystallin or Arc repressor enabled a persistent interaction among subunits, maintaining the oligomeric structure during thermal stress.

Alpha-crystallin retained a native-like beta-sheeted secondary structure under thermal stress at pH 7.5, 4.0, 2.0, and 1.0, respectively. The unfolding enthalpies of alpha-crystallin at acidic pH values (4.0, 2.0, and 1.0, respectively) were less marked compared with those at pH 7.5 (Table 1). Thus, camel alpha-crystallin cannot be fully denatured at high temperatures (94 ◦C) and low pH values (e.g., pH 1.0).

Camel lens alpha-crystallin loses tertiary structure below pH 4.0 at room temperature (Figure 2). The tryptophan fluorescence spectra exhibited a 16 nm redshift at pH 2.0 and 20 ◦C. AlphaA-crystallin contains one tryptophan residue, whereas alphaB-crystallin contains two tryptophan residues [7]. These tryptophan residues are partially buried and located at the N-terminal domains (Supplementary Figure S2). Acid denaturation at pH 2.0 leads to partial unfolding of the N-terminal domain of alpha-crystallin. The complete unfolding of alpha-crystallin by chemical denaturants (6 M GdnHCl) resulted in a 27 nm redshift in tryptophan fluorescence (Figure 2). The far-UV CD data showed that camel alpha-crystallin retained a native-like beta-sheeted secondary structure at pH 2.0 but lost tertiary structure at this pH. Thus, alpha-crystallin forms a molten-like, globular structure at pH 2.0. Changes in the tertiary structure below pH 4.0 caused an increase in hydrophobicity (Figure 3). An increase in protonation below pH 2.0 caused charge– charge repulsion and forced alpha-crystallin to attain native-like secondary and tertiary structures (Figures 1 and 2). Moreover, the unfolding enthalpy at pH 1.0 was closer to that of alpha-crystallin at pH 7.5 (Table 1).

Alpha-crystallin at pH 7.5 exhibited a slight increase in hydrophobicity above 35 ◦C and a large exposure of hydrophobic patches above 55 ◦C. Alpha-crystallin at acidic pH values (4.0, 2.0, and 1.0, respectively) resulted in little exposure of hydrophobic patches above 65 ◦C (Figure 6). We recently reported that the chaperoning effect of alpha-crystallin was activated in a stepwise manner and correlated with the biphasic changes in the tertiary structure and surface hydrophobicity of alpha-crystallin during thermal stress [7]. Alphacrystallin even retained its chaperone activity above the post-transition temperature (89 ◦C) at pH 7.5 [7]. These data suggest that above the thermal transition state (denatured state), alpha-crystallin is not denatured. It retains a core beta-sheeted structure, maintains an oligomeric state, and performs chaperone activity efficiently (~90% protection at 89 ◦C when the alpha-crystallin to substrate *w*/*w* ratio was 0.87:1) [7]. This state is possible only when each monomeric subunit of alpha-crystallin retains a substrate-binding site (i.e., remains in the functional state). Therefore, a higher temperature does not denature alpha-crystallin at pH 7.5 but results in conformational changes required to activate its chaperone activities.

#### **5. Conclusions**

In this study, the thermodynamic parameters obtained during the thermal transition of camel lens alpha-crystallin and those determined in earlier studies for other alphacrystallins represent less than half of the theoretically calculated values for complete protein denaturation. Even at acidic pH values, the enthalpies were much lower than those at pH 7.5. This finding showed that alpha-crystallin was never completely denatured at an extreme pH or temperature or both. Folding species formed post-transition were neither dissociated nor aggregated and were reversible. Interestingly, the folding species formed post-transition at pH 7.5 remained fully active (i.e., they retained almost a 1:1 substrate binding site). Maintaining a functional state post-thermal transition is a feature that is incongruent with the unfolded state. This phenomenon may be due to the natural selection of alpha-crystallin to suppress aggregation in the lens under stress and maintain clarity in the lens throughout life. These tasks may be more challenging to perform in the camel eye lens, which is exposed to high temperatures, solar radiation, and dryness from the desert climate. To overcome these larger tasks, camel lens alpha-crystallin has the natural ability to retain secondary and oligomeric structures and maintain solubility and activity at extreme temperatures.

**Supplementary Materials:** The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/gels8050273/s1, Figure S1. Changes in far-UV CD signal at 217 nm at different pH values are plotted. With a de-crease in pH, the far-UV CD ellipticities initially decreased between pH 7.5 and 5.0. The far-UV CD ellipticities at 217 nm rapidly increased to a nativelike structure below pH 5.0; Figure S2. Modeled 3D structure of camel lens alpha-crystallin. The tryptophan residue in the A- and B-chain of alpha-crystallin is shown in blue: (A) alphaA-crystallin and (B) alphaB-crystallin.

**Author Contributions:** Conceptualization, A.M.; methodology, A.M. and J.M.K.; software, A.M.; formal analysis, A.M. and M.S.O.; resources, A.S.A. and M.S.O.; writing—original draft preparation, A.M. and J.M.K.; writing—review and editing, A.M., A.S.A. and J.M.K.; funding acquisition, J.M.K. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by [King Saud University] grant number [RSP-2021/360] and The APC was funded by [RSP-2021/360].

**Informed Consent Statement:** Not applicable.

**Data Availability Statement:** Not applicable.

**Acknowledgments:** The authors are grateful to the Researchers Supporting Project Number (RSP-2021/360), King Saud University, Riyadh, Saudi Arabia.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

