Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Jan 1.
Published in final edited form as: Eur J Lipid Sci Technol. 2022 Dec 4;125(1):2200096. doi: 10.1002/ejlt.202200096

Detection and Semi-quantification of Lipids on High-Performance Thin-Layer Chromatography Plate using Ceric Ammonium Molybdate Staining

Kesatebrhan Haile Asressu 1, Qibin Zhang 1,2,*
PMCID: PMC9937734  NIHMSID: NIHMS1855226  PMID: 36818638

Abstract

It is desirable to quickly check the composition of lipids in small size samples, but achieving this is challenging using the existing staining methods. Herein, we developed a highly sensitive and semi-quantitative method for analysis of lipid samples with ceric ammonium molybdate (CAM) staining. The CAM detection method was systematically evaluated with a wide range of lipid classes including phospholipids, sphingolipids, glycerolipids, fatty acids (FA) and sterols, demonstrating high sensitivity, stability, and overall efficiency. Additionally, CAM staining provides a clean yellow background in high performance thin-layer chromatography (HPTLC) which facilitates quantification of lipids using image processing software. Lipids can be stained with CAM reagent regardless of their head group types, position of the carbon-carbon double bonds, geometric isomerism and the variation in the length of FA chain, but staining is mostly affected by the degree of unsaturation of the FA backbone. The mechanism of the CAM staining of lipids was proposed on principles of the reduction-oxidation reaction, in which Mo(VI) oxidizes the unsaturated lipids into carbonyl compounds on the HPTLC plate upon heating, while itself being reduced to Mo(IV). This method was applied for the separation, identification, and quantification of lipid extracts from porcine brain.

Keywords: lipids, HPTLC, CAM stain, Mo reduction-oxidation, porcine brain

Graphical abstract

graphic file with name nihms-1855226-f0005.jpg

The sensitivity of ceric ammonium molybdate staining was evaluated with several representative lipid classes on HPTLC plate. This method enabled semi-quantitative determination of the composition of complex lipid extracts from porcine brain.

1. Introduction

Lipids are a large family of hydrophobic and amphipathic biomolecules with a myriad of structural diversity.[1] They are divided into eight categories such as fatty acyls (FA), glycerolipids, glycerophospholipids, sphingolipids and sterol lipids.[2] Being one of the major categories, glycerophospholipids are critical components of all cellular membranes, which can be further classified into several classes such as phosphatidic acid (PA), phosphatidylcholine (PC), phosphatidylserine (PS), phosphatidylethanolamine (PE), phosphatidylglycerol (PG), and phosphatidylinositol (PI), based on the different head groups esterified to the sn-3 position of the glycerol backbone.[3] Sphingolipids are another major category of lipids, many of them typically have ceramide backbone composed of a long chain sphingoid base and N-linked FA. Ceramides (Cer) are precursors to the more complicated glycosphingolipids, which have important roles in cell-cell interactions. Based on the carbohydrate composition, they can be further classified into glucosylceramide (GlcCer), galactosylceramide (GalCer), lactosylceramide (LacCer), globoside and ganglioside, etc.[4, 5]

The most commonly used method for lipid separation and detection is high performance liquid chromatography (HPLC) coupled with evaporative light scattering detector or mass spectrometer.[6] Fluorescence generated from post-column derivatization could also be used, but UV detector is rarely used due to lipids lacking strong innate chromophores.[7] Although high quality separations and identifications are achievable with these methods, the equipment is expensive with high maintenance costs and it takes longer time for analysis. As a result, these methods are not convenient in routine laboratory studies for quick assessment of the compositions of a lipid extract.

Thin-layer chromatography (TLC), particularly the high performance thin-layer chromatography (HPTLC) is one of the simplest methods for separation of complex lipid mixtures.[3, 6, 8, 9] It is cost effective, has high resolving power, small amounts of sample loading and shorter separation times, and can analyze multiple samples simultaneously. But the following detection often requires special staining reagents, because the commonly used UV lamp for TLC visualization is not helpful for direct detection of lipids.[7] Several reagents have been widely used for the detection and quantification of lipids - the most common destructive reagents are cupric sulfate in aqueous phosphoric acid, boric acid in ethanol,[3, 10] sulfuric acid in ethanol (1:1, v/v),[11, 12] cupric acetate in phosphoric acid,[13] sodium or potassium dichromate in sulfuric acid,[13, 14] and phosphomolybdic acid in methanol.[15] However, these reagents are characterized by having low sensitivity (1 to 50 μg), and have limited coverages for all lipid classes. Besides, sulfuric acid and chromic acid mixture are toxic and corrosive reagents which yield weak intensity spots leading to discoloration over time. While high resolution and strong intensity can be obtained for staining of phospholipids with boric acid in ethanol, long treatment time is needed such as 3 h drying for the HPTLC plate before sample application.[3]

Lipid classes have also been visualized under UV light using nondestructive reagents such as dichlorofluorescein,[11] primuline dye in acetone/water,[16] rhodamine 6G dye and laser-excited fluorescence or iodine vapor and ninhydrine [17, 18]. Although these dyes are nondestructive, primuline staining of PE could fade when exposed to hypochlorous acid.[19] On top of that, visualization of lipid classes with dichlorofluorescein dye needs long time of treatment (up to 2.5 h).[11] Ninhydrine is limited to PE and PS because of the presence of primary amine functionality in these lipid classes. Lipid classes like PC and PS can only be weakly detected under UV after exposure to iodine vapor.[20]

Although several widely used lipid staining methods exist, it is desirable to have a more sensitive and stable staining method for detection and quantification of diverse lipid classes from biological samples. To this end, we developed a highly sensitive and semi-quantitative ceric ammonium molybdate (CAM) staining method for various lipid classes in the glycerophospholipid, sphingolipid, glycerolipid, FA and sterol categories. The developed method was applied to quick assessment of lipid compositions in biological samples such as the total lipid extract and polar lipid extract from porcine brain.

2. Materials and methods

2.1. Chemicals and lipid standards

Chemicals such as chloroform, hexane, and methanol were from Fisher Scientific (Pittsburgh, PA), ethanol (EtOH), and diethyl ether (Et2O) from Acros Organics (Geel, Belgium), acetic acid (AcOH) triethylamine (Et3N) from Sigma-Aldrich (St. Louis, MO) and CAM staining solution from TCI America (Portland, OR). Standard lipids such as phospholipids PC 16:1(9Z)/16:1(9Z) (catalog no. 850358), PC 18:0/18:0 (catalog no. 850365), PC 18:0/18:1(9Z) (catalog no. 850467), PC 18:1(6Z)/ 18:1(6Z) (catalog no. 850374), PC 18:1(9Z)/18:1(9Z) (catalog no. 850375), PC 18:1(9E)/18:1(9E) (catalog no. 850376), PC 18:2(9Z,12Z)/ 18:2(9Z,12Z) (catalog no. 850385), PC 18:3(9Z,12Z,15Z)/18:3(9Z,12Z,15Z) (catalog no. 850395), PC 20:4(5Z,8Z,11Z,14Z)/ 20:4(5Z,8Z,11Z,14Z) (catalog no. 850397), PC 22:1(13Z)/22:1(13Z) (catalog no. 850398), PC 22:6(4Z,7Z,10Z,13Z,16Z,19Z)/22:6(4Z,7Z,10Z,13Z,16Z,19Z) (catalog no. 850400), PA 14:0/0:0 (catalog no. 857120), PA 18:1(9Z)/18:1(9Z) (catalog no. 840875), PE 18:1(9Z)/18:1(9Z) (catalog no. 850725), PS 18:1(9Z)/18:1(9Z) (catalog no. 840035), PG 18:0/18:0 (catalog no. 840465), PG 18:1(9Z)/ 18:1(9Z) (catalog no. 840475), and PI 18:1(9Z)/ 18:1(9Z) (catalog no. 850149), SM d18:1/24:1(15Z) (catalog no. 860593) and other sphingolipids like Cer d18:1/18:1(9Z) (catalog no. 860519), GlcCer d18:1/18:1(9Z) (catalog no. 860548), and LacCer d18:1/18:1(9Z) (catalog no. 860590), brain total lipid extract (catalog no. 131101), brain polar lipid extract (catalog no. 141101) and 1,2-DG 18:1(9Z)/18:1(9Z)/0:0 (catalog no. 800811) were obtained from Avanti Polar Lipids (Alabaster, AL). The other glycerolipids such as MG 18:1(6Z)/0:0/0:0 (code no. M-229), diacylglycerols such as 1,2-DG 14:0/14:0/0:0 (code no. D-141), 1,2-DG 12:1(11Z)/12:1(11Z)/0:0 (code no. D-196), 1,2-DG 22:1(13Z)/22:1(13Z)/0:0 (code no. D-301), 1,3-DG 12:1(11Z)/0:0/12:1(11Z) (code no. D-197), 1,3-DG 18:1(9Z)/0:0/18:1(9Z) (code no. D-237), and triacylglycerols such as TG 18:0/18:0/18:0 (code no. T-160) and TG 22:1(13Z)/22:1(13Z)/22:1(13Z) (code no. T-300) from Nu-Chek Prep, Inc (Elysian, MN), while FA 18:1(9E) (catalog no. 45089) and sterols namely FC (catalog no. C8667) and CE 18: 1(9Z) (catalog no. C9253) were purchased from Sigma-Aldrich.

2.2. Instrumentation

HPTLC plates (10x10 cm, normal phase silica gel HPTLC-GHL 150 μm) were obtained from Analtech (Newark, DE). TLC developing chambers with tight lids such as Latch-lid (10x10 cm) and ground joint lid (5x10 cm) and Hamilton TLC syringe (volume 10 μL) from Sigma-Aldrich and a heat gun (Proheat LCD) from Master Appliance (Racine, WI) were used for TLC development.

2.3. Preparation of lipid standard solutions

The phospholipids such as PC, PS, PA, PE, and PG (each in 10 mg/mL), brain total lipid extract (25 mg/mL) and polar lipids extract (25 mg/mL) were purchased as CHCl3 solutions. Stock solutions of PI (2.5 mg/mL in CHCl3), sphingolipids (GlcCer, LacCer, Cer and SM, 2.5 mg/mL in CHCl3/MeOH) were prepared by adding solvents directly to the respective solid reagent bottles. Stock solution (2.5 mg/mL) of each of the relatively abundant lipids such as glycerolipids (MG, DG and TG), FA, and sterol lipids (FC and CE) was prepared by mixing solvent with powders of 3 to 5 mg carefully weighed with an analytical balance capable of 0.1 mg accuracy. From the stock solution of each lipid, four dilution series (1.0, 0.50, 0.25 and 0.13 μg) were prepared to test the sensitivity of the staining method.

2.4. HPTLC analysis of lipid classes

The HPTLC plate was cut to the appropriate size using a glass cutter, and then activated by heating at 100 °C in oven for 10 min and cooling to room temperature (rt) in a desiccator. Then it was marked with a pencil at 1.5 cm from the bottom edge of the plate. Next, the samples were uniformly applied on the marked spots using a Hamilton TLC syringe and blow dried with air. Subsequently, the HPTLC plate was immersed in a developing chamber pre-equilibrated with 1 cm deep solvent and the solvent was allowed to migrate to the top edge of the plate by capillary action. Two solvent systems, CHCl3/EtOH/Et3N/H2O (3/3.5/3.5/0.7, v/v)[3] and hexane/Et2O/AcOH (8.0/2.0/0.3, v/v)[21], were used depending on the polarity of samples. After the TLC development completed in 10 to 40 min, the HPTLC plate was taken out from the chamber and allowed to dry at 70 °C in oven for 5 min and cooled to rt under desiccation.

To visualize the location of the sample spots, the dried HPTLC plate was dipped into the CAM solution for 10 sec, then taken out and heated for 5 to 10 min using a heat gun, with longer heating time for lower concentration samples. Lipids were detected as dark blue spots against a yellow background. Once cooled, the HPTLC plate was scanned. Semi-quantification of the detected lipids was conducted on the scanned images using the free ImageJ software.[11]

3. Results and Discussion

Several destructive reagents have been developed for staining lipids either by spraying or dipping in solutions followed with charring. The most common ones include cupric sulfate, boric acid,[3, 11] cupric acetate, [13] phosphomolybdic acid,[15] and 50% sulfuric acid.[11, 14] CAM or Hanessian’s staining has been extensively used for visualization of carbohydrates (polyhydroxylated compounds)[22, 23] and carbonyl compounds[24]. The application of CAM staining for detection of lipids chemically synthesized or extracted from biological samples has been documented in the literature.[25–29] However, most of the applications to lipids detection were for qualitative analysis only, and to our knowledge there is no report exploring the sensitivity or quantitative aspects of CAM staining with lipids. To this end, we evaluated CAM staining for its quantitative detection of diverse lipid classes, focusing on the lower limit of quantitation. Based on the drastic differences in staining sensitivity between saturated and unsaturated lipids, we also explored the mechanism of staining lipids with CAM.

3.1. Factors affecting CAM staining

3.1.1. Lipid head group

To investigate whether CAM staining is affected by lipid head groups, phospholipids having variable head groups (PC, PA, PS, PE, PG and PI) but fixed FA composition were tested under the reported mobile phase (CHCl3/EtOH/Et3N/H2O (3.0/3.5/3.5/0.7, v/v)).[3] HPTLC result revealed that all phospholipids were detectable with CAM staining, generating dark blue colored spots against a yellow background when heated with a heat gun, although the intensities of PA and PI were weaker than that of PC, PS, PE and PG (Figure 1A) with equal amount (1 μg) loaded for each lipid.

Figure 1.

Figure 1.

Open in a new tab

HPTLC chromatograms of representative lipid classes stained with CAM. A) Phospholipids with variable head groups: PC 18:1(9Z)/18:1(9Z), PA 18:1(9Z)/18:1(9Z), PS 18:1(9Z)/18:1(9Z), PE 18:1(9Z)/18:1(9Z), PG 18:1(9Z)/18:1(9Z) and PI 18:1(9Z)/18:1(9Z). B) PCs with variable degrees of unsaturation: PC 18:0/18:1(9Z), PC 18:1(9Z)/18:1(9Z), PC 18:2(9Z,12Z)/18:2(9Z,12Z), PC 18:3(9Z,12Z,15Z)/18:3(9Z,12Z,15Z), PC 20:4(5Z,8Z,11Z,14Z)/20:4(5Z,8Z,11Z,14Z), and PC 22:6(4Z,7Z,10Z,13Z,16Z,19Z)/22:6 (4Z,7Z,10Z,13Z,16Z,19Z). C) PCs with different C=C bond position and geometry: PC 18:1(6Z)/18:1(6Z), PC 18:1(9E)/18:1(9E) and PC 18:1(9Z)/18:1(9Z). D) Unsaturated PCs with variable lengths of FA: PC 16:1(9Z)/16:1(9Z), PC 18:1(9Z)/18:1(9Z) and PC 22:1(13Z)/22:1(13Z). E) Unsaturated 1,2-DG and 1,3-DG with variable lengths of FA: 1,2-DG 12:1(11Z)/12:1(11Z)/0:0, 1,3-DG 12:1(11Z)/0:0/12:1(11Z), 1,2-DG 18:1(9Z)/18:1(9Z)/0:0, 1,3-DG 18:1(9Z)/0:0/18:1(9Z) and 1,2-DG 22:1(13Z)/22:1(13Z)/0:0. Amount of samples spotted: 1 μg each for A and E, and 2 μg for each B to D. Solvent system CHCl3/EtOH/Et3N/H2O (3.0/3.5/3.5/0.7, v/v) for A to D and hexane/Et2O/AcOH (8.0/2.0/0.3, v/v) for E. HPTLC development time: 20 min for A and D, 40 min for B, 24 min for C and 14 min for E. Heating time for detection: 5 min for all.

3.1.2. Degree of fatty acyl unsaturation

While studying the influence of head groups in lipid staining, we found that the intensities of phospholipids containing unsaturated FA moieties were higher than those of the saturated ones. This encouraged us to investigate the effect of the degree of FA unsaturation in CAM staining of lipids. To this end, we compared CAM staining of PC 18:0/18:1, PC 18:1/18:1, PC 18:2/18:2, PC 18:3/18:3, PC 20:4/20:4 and PC 22:6/22:6 at 2 μg level. In general, the spot intensity increased as the number of C=C bonds increased (Figure 1B). But the increase in intensity from 1 to 2 C=C was more pronounced than from 4 to 12. While we cannot fully understand the phenomenon, we suspect it might be related to the steric hindrance effect in oxidation (see 3.4 Staining mechanism). The observation that the intensities of spots positively correlated with the number of C=C bonds is in accordance with the report of Baron and Coburn who detected neutral lipids using cupric acetate reagent.[30, 31]

We also examined the influences of the C=C position and geometry and found neither the position nor the cis (Z) or trans (E) configuration of C=C affected the staining intensity (Figure 1C).

CAM staining was also performed on saturated lipids such as PC 18:0/18:0, PG 18:0/18:0, lyso PA 14:0/0:0, DG 14:0/14:0/0:0 and TG 18:0/18:0/18:0. These saturated lipids were detectable when applied in 6 μg or greater quantity upon either heating the HPTLC plate for a little longer time or at higher temperature (data not included). Under such conditions, however, the HPTLC plate is usually damaged, leading to the formation of dark blue background particularly on the solvent front which interferes strongly with the detection of lipids. The lack of reactivity of saturated lipids with staining reagents has also been described in the literature, for example saturated lipid standards such as PC 16:0/0:0, PC 18:0/0:0 and FA 18:0 showed no staining with either dichlorofluorescein or sulfuric acid in ethanol (1:1, v/v).[11, 30]

3.1.3. Chain length of unsaturated fatty acids

Next, we examined whether the chain length of unsaturated FA can affect CAM staining. For polar lipids, we selected PC 16:1/16:1, PC 18:1/18:1 and PC 22:1/22:1, which have the same head group and the same number of C=C bonds. HPTLC analysis showed that all these PC species were detected with nearly the same intensity regardless of the variation in the chain length of the FA chain (Figure 1D).

We also evaluated the effect of chain length of unsaturated FAs on CAM staining of neutral glycerolipids (Figure 1E). To this end, 1,2-DG (DG 12:1, DG 18:1, and DG 22:1) and 1,3-DG (DG 12:1, and DG 18:1) were used as representative neutral lipids and HPTLC was performed under the solvent system hexane/Et2O/AcOH (8.0/2.0/0.3, v/v). Except 1,2-DG 12:1 and 1,3-DG 12:1 which contain terminal C=C bond, all the other DGs have internal C=C bond. The results clearly showed that the intensities of DG spots after CAM staining were not affected by the chain length of the unsaturated FAs in glycerolipids. As shown in Figure 1E, the retardation factor (Rf) is determined by the total number of carbons, positional isomerization and C=C bond position. Clearly, Rf of 1,2-DG 22:1 and 1,2-DG 18:1 followed the order of carbon chain length. But with the same FA chain length, 1,3-DG 18:1 had larger Rf than 1,2-DG 18:1 as expected from the polarity differences. On the contrary, the isomers 1,2-DG 12:1 and 1,3-DG 12:1 which contain terminal C=C bonds had similar Rf values. Similar Rf values of 1,2-DG 12:1 and 1,2-DG 18:1 can only be attributed to the variation in the position of the C=C bonds (terminal versus internal C=C bond) under the reported hexane/Et2O/AcOH (8.0/2.0/0.3, v/v) eluent system.[21]

3.2. Sensitivity of CAM staining

The senstivities of established lipid staining methods have a range of 1 to 50 μg.[3, 10, 11, 31] It is always desirable to detect lipids from lower amount of sample. To evaluate the sensitivity of CAM staining, we selected four phospholipid classes bearing different head groups and similar FA distribution and unsaturation, and analyzed them on one HPTLC plate in four different loading quantities (1.0, 0.50, 0.25 and 0.13 μg). PC, PE and PG were detectable at 0.13 μg (Figure 2A), but PS was at 0.25 μg. Phospholipids PE and PG were detected with a high sensitivity compared to PC and PS. For sphingolipids, we selected SM, LacCer, GlcCer and Cer for analysis and found that LacCer, GlcCer and SM were detectable at 0.13 μg while Cer at 0.25 μg, and LacCer and GlcCer had higher intensity spots relative to SM and Cer (Figure 2B & C).

Figure 2.

Figure 2.

Open in a new tab

Sensitivity of CAM staining method using standard lipids with four dilution series (1.0, 0.50, 0.25 and 0.13 μg). A) HPTLC chromatograms of representative phospholipids: PC 18:1(9Z)/18:1(9Z), PS 18:1(9Z)/18:1(9Z), PE 18:1(9Z)/18:1(9Z), PG 18:1(9Z)/18:1(9Z). B-C) HPTLC chromatograms of representative sphingolipids: SM d18:1/24:1(15Z), LacCer d18:1/18:1(9Z), GlcCer d18:1/18:1(9Z) with CHCl3/EtOH/Et3N/H2O (3.0/3.5/3.5/0.7, v/v) and C) Cer d18:1/18:1(9Z) with hexane/Et2O/AcOH (5.0/5.0/0.3, v/v) solvent system. D) HPTLC chromatograms of representative glycerolipids: MG 18:1(6Z)/0:0/0:0, 1,3-DG 18:1(9Z)/0:0/18:1(9Z), TG 22:1(13Z)/22:1(13Z)/22:1(13Z), and FA 18:1(9E) with hexane/Et2O/AcOH (8.0/2.0/0.3, v/v). E) HPTLC chromatograms of representative sterols: FC and CE 18:1(9Z) with hexane/Et2O/AcOH (9.0/1.0/0.3, v/v) solvent system. HPTLC development time: 20 min for A and B, 10 min for C and E, 11 min for D. Amount of sample spotted shown in X axis. Heating time for detection: 10 min for all. Some spots may be too faint to be distinguishable in the reproduced images.

We also investigated the sensitivity of glycerolipids (MG, DG, TG), FA and sterols (FC and CE) by CAM staining, with all lipid species having at least one C=C. Interestingly, lipid classes such as MG, DG, FA, and FC were detected at 0.13 μg amount with high intensity (Figure 2D & E), indicating the high sensitivity of CAM staining. However, TG and CE were detected up to 0.25 μg quantity. It was observed that the intensity of the sample spot decreased as the sample amount decreased. Kessler and his co-workers also obtained similar results with diverse lipid classes using dichlorofluorescein and 50% sulfuric acid as staining reagents with a sensitivity of 2.5 μg.[11] Among the previously reported staining methods, cupric sulfate was reported to have high sensitivity (1.7 to 2.0 μg) in the detection of lipids.[3, 10] Compared to the literature reports, the CAM detection method demonstrated a higher sensitivity, stability and overall efficiency for a wide range of lipid classes such as phospholipids, sphingolipids, glycerolipids, FA and sterols.

3.3. Semi-quantification of lipids

As reported in the literature,[11] we used ImageJ software to measure the spot intensities for quantification of lipids on the HPTLC plate. Clearly, different lipid species had different response factors. Among phospholipids, PG had the highest signal, which was followed by PE > PS > PC (Figure 3). For sphingolipids, GlcCer was found to have the highest signal followed by LacCer > Cer > SM (Figure 3). As expected, the signal was proportional to the amount of sample loaded on the plate, and the coefficient of determination (R2) of the linear regression curves were all > 0.9.

Figure 3.

Figure 3.

Open in a new tab

Calibration curves of standard lipids obtained from HPTLC plate stained with CAM.

We also evaluated whether a linear relationship can be obtained for the other lipid species, such as glycerolipids (MG, DG, TG), FA and sterol lipids (FC and CE) (Figure 3). TG responded to CAM staining better than DG and MG, but due to its sharp slope of response curve, the lowest loading amount 0.13 μg was not detectable when compared to MG and DG. Similarly, CE had higher response than FC, but CE also had sharp response slope compared to FC. While it is not a focus of this work, we also tested the upper limit of detection for selected classes of lipids, and found that the calibration curves were still linear at 40 μg for both PG and FC (data not included).

3.4. CAM staining mechanism

It has been reported that the C=C bond of lipids and olefins could undergo transition metal-based catalytic oxidative cleavage to form a variety of products containing carbon-oxygen double bonds such as aldehydes, ketones and carboxylic acids as a result of over oxidation.[32, 33] The mechanism of CAM staining of lipids has not been reported before, we suspected similar molybdenum (Mo)-based oxidative cleavage could occur under a homogenous catalysis system.

CAM staining of lipids involves dipping the HPTLC plate in CAM solution followed by heating to form a blue spot on a yellow background. The observed blue color is likely resulted from the reduction-oxidation reaction between the CAM and lipids, which reduces Mo(VI) to a lower oxidation state, Mo(V) or Mo(IV). We propose that the reaction of ammonium molybdate and cerium sulfate in the presence of 10% sulfuric acid (H2SO4) in water (H2O) converts molybdate in the CAM reagent to molybdic acid (H2MoO4) (Scheme 1). Next, the H2MoO4 oxidizes the unsaturated lipid classes into carbonyl compounds on the HPTLC plate upon heating while itself being reduced to Mo(IV). As depicted in Scheme 1, the molybdate ion reacts with the unsaturated lipids to generate the dihydroxylated intermediates as the main metal diester intermediate, which in turn undergoes intramolecular ring opening reaction to yield the aldehydes. Additionally, the dihydroxylated intermediates could be hydrolyzed to form diols, which then further oxidized to aldehydes. The color intensifies with the increasing number of C=C bonds in the lipid backbone being stained. We observed that unsaturated lipid classes bearing additional hydroxyl groups such as PG, GlcCer and LacCer had a high response to CAM staining. This might be attributed to the oxidation of the aliphatic vicinal diols into two aldehyde molecules by CAM.[34]

Scheme 1.

Scheme 1.

Open in a new tab

Plausible mechanism of lipid staining with CAM

The CAM staining mechanism is similar to the reaction of Osmium(VIII) oxide (OsO4) with unsaturated lipids.[35] OsO4 oxidizes C=C bond to form cyclic osmate(VI) diester intermediate, which subsequently subjected to hydrolysis reaction to give vicinal diols and the reduced Os(VI).[36] Like the CAM staining mechanism, this reaction is based on the principle of reduction-oxidation reaction which involves the reduction of Os(VIII) to Os(VI) or Os(IV). [37] Unlike CAM, the OsO4 reaction requires sodium periodate (NaIO4) to further oxidize the diol intermediate to carbonyl compounds as the final products.[38] Furthermore, the OsO4 reaction usually takes place in the presence of solvent mixture such as H2O and dioxane, THF, acetone, acetic acid, t-butanol or pyridine.[37] The direct oxidation of C=C bond with OsO4 needs hydrogen peroxide or tert-butyl hydrogen peroxide and oxone (monopotassium peroxysulfate salt) as co-oxidants in DMF.[39] However, CAM staining oxidizes the C=C bond directly into carbonyl compounds and carboxylic acids [32, 33] without the involvements of additives and solvents.

3.5. Application of CAM staining to biological samples

We demonstrated the utility of CAM staining method for identification and semi-quantification of lipids in two commercially available complex lipid samples, namely total lipid extract (TLE) and polar lipid extract (PLE) from porcine brain. To this effect, 7 standard lipid classes, i.e. SM, PC, PS, PA, PE, PG and FC were used as the reference standards (Figure 4A). After HPTLC separation achieved with CHCl3/EtOH/Et3N/H2O (3.0/3.5/3.5/0.7, v/v) solvent system, lipids were visualized using CAM staining upon gentle heating with heat gun for about 5 min. Dark blue spots appeared on all samples against the yellow background of the HPTLC plate and identification was referenced to the standard lipids. It was observed that FC, PE, and PC were the dominant lipids in both the TLE and PLE, whereas PG, PS, PA and SM were present in minor amounts (Figure 4). Relatively higher level of PS was found in PLE than in TLE.

Figure 4.

Figure 4.

Open in a new tab

Application of developed method to analysis of biological samples. A) HPTLC chromatograms of porcine brain lipid extract stained with CAM. TLE: Total lipid extract, PLE: Polar lipid extract. Standard lipid classes: SM d18:1/24:1(15Z), PC 18:1(9Z)/18:1(9Z), PS 18:1(9Z)/18:1(9Z), PA 18:1(9Z)/18:1(9Z), PE 18:1(9Z)/18:1(9Z), PG 18:1(9Z)/18:1(9Z) and FC. Amount of sample spotted: 0.4 μl (1 μg) of each standard lipid and 1.6 μl the lipid extract. Solvent system: CHCl3/EtOH/Et3N/H2O (3.0/3.5/3.5/0.7, v/v). B) Semi-quantification of the compositions of TLE and PLE in A) using calibration curves presented in Figure 3. HPTLC development time: 27 min and the heating time for detection: 5 min.

After HPTLC separation and identification of the component lipid classes, the area density of each lipid was quantified using the established calibration curves in Figure 3. We found that FC (2.42 μg), PE (1.13 μg), PC (1.08 μg), PG (0.76 μg), and PS (0.53 μg) existed as the major lipids in the TLE, and FC (1.62 μg), PE (1.44 μg), PG (1.21 μg), PS (0.82 μg), and PC (0.63 μg) were also present in high levels in PLE. Compared to TLE, higher amount of PS (0.82 μg) was observed in PLE. The quantified components of the porcine brain extracts were compared with the lipid compositions of these extracts reported in the literature, agreeing well that PE, PS, and PC are the major phospholipids in both the TLE[40] and PLE.[41]

In addition to the phospholipids of the brain lipid extracts, we identified high level of FC (1.62 to 2.42 μg) and minor amount of SM as shown in Figure 4B. Our finding is supported by previous reports of the presence of high quantity of FC in porcine brain lipid extract.[42] While the high amount of FC in these two samples can largely explain the high percentage of “unknown” as listed in the composition of these two samples, we need to note that the summed amount of all of the detected lipid classes did not match the amount of sample spotted on the plate, although the amount of phospholipids we detected correlated well with the weight percentage (wt%) listed by the manufacturer. This is likely the result of the dominance of saturated phospholipids in the sample. As reported by others [42], saturated lipids exist in larger percentage in porcine brain. They will co-elute with the unsaturated lipids in the same class under our TLC conditions (the Rfs and shapes of the phospholipids in the lipid extracts were the same as the reference phospholipid standards) but have much lower sensitivity, therefore their presence was largely unaccounted for in our analysis.

4. Conclusions

We have developed a CAM staining method for detection of diverse lipid classes such as phospholipids, sphingolipids, glycerolipids and sterols, with sensitivity down to 0.13 μg on plate. The staining is mainly influenced by the degree of the C=C bond unsaturation in the FA backbone, while the position of the C=C bond, cis (Z) or trans (E) geometry and the variation in the length of FA chain do not seem to affect the signal intensity. The staining intensity increases linearly with the amount of unsaturated lipids on plate. The established method was applied for analysis of the total lipid extract and polar lipid extract from porcine brain and good agreement with the literature was achieved. Although the mechanism we proposed for CAM staining cannot fully explain all observations, it does not diminish the usefulness of our method in identification and semi-quantitative determination of the lipid composition of complex lipid extracts, particularly when the sample amount is low. It is of note that to get more accurate quantitation, relatively well matching in composition (unsaturation level) between the reference standards and the sample is required.

Practical applications.

The CAM staining method that we developed in this work showed a high sensitivity for diverse lipid classes following HPTLC separation. HPTLC with CAM staining is a promising method for quick assessment of the identity and quantity of diverse lipid classes in lipid extracts of small size biological samples.

Funding:

This work was supported by the National Institutes of Health grants R01DK123499 and R01AI150095.

Abbreviations:

CAM

ceric ammonium molybdate

HPTLC

high performance thin-layer chromatography

FA

fatty acid/acyl

PA

phosphatidic acid

PC

phosphatidylcholine

PS

phosphatidylserine

PE

phosphatidylethanolamine

PG

phosphatidylglycerol

PI

phosphatidylinositol

GlcCer

glucosylceramide

LacCer

lactosylceramide

MG

monoacylglycrol

DG

diacylglycerol

TG

triacylglycerol

FC

free cholesterol = cholesterol

CE

cholesteryl ester

Cer

ceramide

SM

sphingomyelin

TLE

total lipid extract

PLE

polar lipid extract

Footnotes

Conflict of Interest

The authors have declared no conflict of interest.

References

  • 1.Khoury S, Canlet C, Lacroix MZ, Berdeaux O, Jouhet J, Bertrand-Michel J Quantification of Lipids: Model, Reality, and Compromise. Biomolecules 2018, 8, 174. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Fahy E, Subramaniam S, Brown HA, Glass CK, Merrill AH Jr., Murphy RC, Raetz CR, Russell DW, Seyama Y, Shaw W, Shimizu T, Spener F, Meer G, VanNieuwenhze MS, White SH, Witztum JL, Dennis EA A comprehensive classification system for lipids. J. Lipid Res 2005, 46, 839–861. [DOI] [PubMed] [Google Scholar]
  • 3.Pinault M, Guimaraes C, Dumas J-F, Servais S, Chevalier S; Besson P; Goupille CA 1D High Performance Thin Layer Chromatography Method Validated to Quantify Phospholipids Including Cardiolipin and Monolysocardiolipin from Biological Samples. Eur. J. Lipid Sci. Technol 2020, 122, 1900240. [Google Scholar]
  • 4.Reza S, Ugorski M, Suchankski J Glucosylceramide and galactosylceramide, small glycosphingolipids with significant impact on health and disease. Glycobiology. 2021, 31, 1416–1434. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Nakamura K, Suzuki Y, Goto-Inoue N, Yoshida-Noro C, Suzuki A Structural characterization of neutral glycosphingolipids by thin-layer chromatography coupled to matrix-assisted laser desorption/ionization quadrupole ion trap time-of-flight MS/MS. Anal. Chem 2006, 78, 5736–5743. [DOI] [PubMed] [Google Scholar]
  • 6.Muthing J, Distler U Advances on the compositional analysis of glycosphingolipids combining thin-layer chromatography with mass spectrometry. Mass Spectrom. Rev 2010, 29, 425–479. [DOI] [PubMed] [Google Scholar]
  • 7.Ibrahim H, Caudron E, Kasselouri A Prognon P Interest of fluorescence derivatization and fluorescence probe assisted post-column detection of phospholipids: A short review. Molecules 2010, 15, 352–373. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Muthing J High-resolution thin-layer chromatography of gangliosides. J Chromatogr A. 1996, 720, 3–25. [DOI] [PubMed] [Google Scholar]
  • 9.Cheng YS, Zheng Y, VanderGheynst JS Rapid quantitative analysis of lipids using a colorimetric method in a microplate format. Lipids 2011, 46, 95–103. [DOI] [PubMed] [Google Scholar]
  • 10.Churchward MA Brandman DM, Rogasevskaia T, Coorssen JR Copper (II) sulfate charring for high sensitivity on-plate fluorescent detection of lipids and sterols: quantitative analyses of the composition of functional secretory vesicles. J. Chem. Biol 2008, 1, 79–87. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Laggai S, Simon Y, Ranssweiler T, Kiemer AK, Kessler SM Rapid chromatographic method to decipher distinct alterations in lipid classes in NAFLD/NASH. World J. Hepatol 2013, 5, 558–567. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Ogawa K, Fujiwara Y, Sugamata K, Abe T Thin-layer chromatography of neutral glycosphingolipids: an improved simple method for the resolution of GlcCer, GalCer, LacCer and Ga2Cer. J. Chromatogr. B: Anal. Technol. Biomed. Life Sci 1988, 426, 188–193. [DOI] [PubMed] [Google Scholar]
  • 13.Macala LJ, Yu RK, Ando S Analysis of brain lipids by high performance thin-layer chromatography and densitometry. J. Lipid Res 1983, 24, 1243–1250. [PubMed] [Google Scholar]
  • 14.Freeman CP, West D Complete separation of lipid classes on a single thin-layer plate. J. Lipid Res 1966, 7, 324–327. [PubMed] [Google Scholar]
  • 15.Zarzycki PK, Bartoszuk M,A, Radziwon AI Optimization of TLC detection by phosphomolybdic acid staining for robust quantification of cholesterol and bile acids. J. Planar Chromatogr. Mod. TLC. 2006, 19, 52–57. [Google Scholar]
  • 16.Fuchs B, Schiller J, Süß R, Nimptsch A, Schürenberg M, Suckau D Capabilities and disadvantages of combined matrix-assisted laser-desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) and high-performance thin-layer chromatography (HPTLC): Analysis of egg yolk lipids. J. Planar Chromatogr. Mod. TLC. 2009, 22, 35–42. [Google Scholar]
  • 17.Wagener H Detection and Documentation of Lipids after Thin-Layer Chromatography. Nature 1965, 205, 386–387. [DOI] [PubMed] [Google Scholar]
  • 18.Kishimoto K, Urade R, Ogawa T, Moriyama T Nondestructive quantification of neutral lipids by thin-layer chromatography and laser-fluorescent scanning: Suitable methods for “lipidome” analysis. Biochem. Biophys. Res. Commun 2001, 281, 657–662. [DOI] [PubMed] [Google Scholar]
  • 19.Richter G, Schober C, Suss R, Fuchs B, Muller M, Schiller J The reaction between phosphatidylethanolamines and HOCl investigated by TLC: Fading of the dye primuline is induced by dichloramines. J. Chromatogr. B. 2008, 867, 233–237. [DOI] [PubMed] [Google Scholar]
  • 20.Fuchs B, Popkova Y, Süß R, Schiller J Separation of (Phospho)Lipids by Thin-Layer Chromatography. In Instrumental Thin-Layer Chromatography, Elsevier, Boston: 2015, 375–405 [Google Scholar]
  • 21.Massart J, Zierath JR, Chibalin AV A simple and rapid method to characterize lipid fate in skeletal muscle. BMC Research Notes. 2014, 7, 391. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Asressu KH, Chang CW, Lam S, Wang CC Donor-Reactivity-Controlled Sialylation Reactions. Eur. J. Org. Chem 2021, 4525–4530. [Google Scholar]
  • 23.Weldu WD, Wang CC Rationalizing the Stereoelectronic Influence of Interglycosidic Bond Conformations on the Reactivity of 1,4-O-Linked Disaccharide Donors. J. Org. Chem 2021, 86, 17906–17917. [DOI] [PubMed] [Google Scholar]
  • 24.Fruhmann P, Hametner C, Mikula H, Adam G, Krska R, Fröhlich J Stereoselective luche reduction of deoxynivalenol and three of its Acetylated derivatives at C8. Toxins (Basel). 2013, 6, 325–336. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Tamura T, Fujisawa A, Tsuchiya M, Shen Y, Nagao K, Kawano S, Tamura Y, Endo T, Umeda M, Hamachi I Organelle membrane-specific chemical labeling and dynamic imaging in living cells. Na Che Biol. 2020, 16, 1361–1367. [DOI] [PubMed] [Google Scholar]
  • 26.Burugupalli S, Richardson MB, Williams SJ Total synthesis and mass spectrometric analysis of a Mycobacterium tuberculosis phosphatidylglycerol featuring a two-step synthesis of (R)-tuberculostearic acid. Org. Biomol. Chem 2017, 15, 7422–7429. [DOI] [PubMed] [Google Scholar]
  • 27.Kain L, Webb B, Anderson BL, Deng S, Holt M, Costanzo A, Zhao M, Self K, Teyton A, Everett C, Kronenberg M, Zajonc DM, Bendelac A, Savage PB, Teyton L The Identification of the Endogenous Ligands of Natural Killer T Cells Reveals the Presence of Mammalian α-Linked Glycosylceramides. Immunity 2014, 41, 543–554. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Mve-Obiang A, Lee RE, Portaels F, Small PL Heterogeneity of Mycolactones Produced by Clinical Isolates of Mycobacterium ulcerans: Implications for Virulence. Infect. Immun 2003, 71, 774–783. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.George KM, Chatterjee D, Gunawardana G, Welty D, Hayman J, Lee R, Small PL Mycolactone: A Polyketide Toxin from Mycobacterium ulcerans Required for Virulence. Science 1999, 283, 854–857. [DOI] [PubMed] [Google Scholar]
  • 30.Kurantz M, Maxwell R, Kwoczak R, Taylor F Rapid and sensitive method for the quantitation of non-polar lipids by high-performance thin-layer chromatography and fluorodensitometry. J. Chromatogr. A. 1991, 549, 387–399. [Google Scholar]
  • 31.Baron CB, Coburn RF Comparison of Two Copper Reagents for Detection of Saturated and Unsaturated Neutral Lipids by Charring Densitometry. J. Liq. Chromatogr 1984, 7, 2793–2801. [Google Scholar]
  • 32.Kerenkan AE, Béland F, Do TO Chemically catalyzed oxidative cleavage of unsaturated fatty acids and their derivatives into valuable products for industrial applications: A review and perspective. Catal. Sci. Technol 2016, 6, 971–987. [Google Scholar]
  • 33.Spannring P, Bruijnincx PCA, Weckhuysen BM, Klein Gebbink RJM Transition metal-catalyzed oxidative double bond cleavage of simple and bio-derived alkenes and unsaturated fatty acids. Catal. Sci. Technol 2014, 4, 2182–2209. [Google Scholar]
  • 34.Li J, Lutz M, Klein Gebbink RJM A Cp-based Molybdenum Catalyst for the Deoxydehydration of Biomass-derived Diols. ChemCatChem. 2020, 12, 6356–6365. [Google Scholar]
  • 35.Hayes TL, Lindgren FT, Gofman JW A Quantitative Determination of The Osmium Tetroxide-Lipoprotein Interaction. J. Cell Biol 1963, 19, 251–255. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Belazi D, Solé-Domènech S, Johansson B, Schalling M, Sjövall P Chemical analysis of osmium tetroxide staining in adipose tissue using imaging ToF-SIMS. Histochem. Cell Biol 2009, 132, 105–115. [DOI] [PubMed] [Google Scholar]
  • 37.Cairns JF, Roberts HL Osmium Tetroxide-catalysed Oxidation of Olefins. J. Chem. Soc C. 1968, 640–642. [Google Scholar]
  • 38.Pappo R, Allen D, Lemieux R Jr, Johnson W Osmium Tetroxide-Catalyzed Periodate Oxidation of Olefinic Bonds. J. Org. Chem 1956, 21, 478–479. [Google Scholar]
  • 39.Travis BR, Narayan RS, Borhan B Osmium Tetroxide-Promoted Catalytic Oxidative Cleavage of Olefins: An Organometallic Ozonolysis. J. Am. Chem. Soc 2002, 124, 3824–3825. [DOI] [PubMed] [Google Scholar]
  • 40.Sandrino B, Jochelavicius K, Volpati D, Sandrino B, Barbosa SC, Nobre TM, Oliveira ON Jr. The prion fragment PrP106-127 adopts a secondary structure typical of aggregated fibrils in langmuir monolayers of brain lipid extract. Chem. Phys. Lipids 2020, 230, 104930. [DOI] [PubMed] [Google Scholar]
  • 41.Kotapati HK, Bates PD Normal phase HPLC method for combined separation of both polar and neutral lipid classes with application to lipid metabolic flux. J. Chromatogr. B 2020, 1145, 122099. [DOI] [PubMed] [Google Scholar]
  • 42.Chanted J, Panpipat W, Panya A, Phonsatta N, Cheong LZ, Chaijan M Compositional Features and Nutritional Value of Pig Brain: Potential and Challenges as a Sustainable Source of Nutrients. Foods 2021, 10, 2943. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES