Qualitative and Quantitative Analysis Method of Recombinant Collagen in Complex Matrix Based on HPLC-MS/MS

Abstract

The purpose of this study is to achieve the quantitative detection of recombinant type III collagen (rh-COL-III) in dressings with complex matrix. First of all, the marker peptide (GEAGIPGVPGAK) of rhCOL-III was identified with HPLC-MS/MS. Then, a qualitative and quantitative method based on marker peptides was established and validated. In order to obtain higher sensitivity, a pretreatment method of liquid, gel, and ointment dressings was optimized. The reference material for quantification was combined using rhCOL-III and blank matrix of each dressing. The results indicated that the relative standard deviation (RSD) of the quantitative method was 2.77%, and the RSD of intraday and interday precision was 2.76% and 2.31%, respectively. The spiking recovery rate was between 80% and 90%. The optimal pretreatment method was Tris-HCl solvent replacement. The optimal trypsin concentration for the dressing solution was 20 μg in 500 μL. The method of preparing standard substances with a blank matrix can effectively eliminate the influence of the matrix effect on the quantitative results. The average spiking recovery rates of 50 μg/mL, 100 μg/mL, and 200 μg/mL in three different dressings ranged from 80% to 120%. The quantitative detection of limit (LOD) of rhCOL-III was 1 ng/mL, 2 ng/g, and 1 ng/g in liquid, ointment, and gel dressings.

1. Introduction

Recombinant collagen, with low immunogenicity, good water solubility, and stable quality compared to tissue-derived collagen [1,2,3], is increasingly used in cosmetics, tissue-engineered medical products, and other medical devices. In the medical devices field, rhCOL is widely used in medical dressings as a highlighted component [4,5], which is applied in the treatment of superficial or non-chronic wounds and the surrounding skin, such as small wounds and abrasions [6,7,8]. In addition to rhCOL, these medical dressings also contain a complex matrix, co-solvents, and a small amount of preservatives [9,10]. Specifically, the content of rhCOL is the most important target in assessing the effectiveness of medical dressing products. However, it is difficult to determine the accurate content of rhCOL in these products. One challenge is the interference from the complex matrix. The matrix used in the dressing is very complex, most of which exists in the form of oil in water [11]. The rhCOL is encapsulated in the matrix and is difficult to release, which creates certain difficulties in the quantitative analysis of rhCOL. On the other hand, most dressings only contain trace amounts of rhCOL, making it difficult to achieve accurate quantitative analysis using conventional methods. Currently, there are no appropriate methods for the qualitative analysis and accurate quantitative determination of rhCOL in medical dressings.

At present, hydroxyproline quantification and ELISA methods are usually applied to determine collagen contents. However, there are some disadvantages to these methods. Mammalian cells like CHO and HEK293 have similar translation mechanisms to human cells, which could produce rhCOL with correct modifications [12,13]. The rhCOL from other cells did not contain hydroxyproline due to a lack of hydroxylase. Thus, the quantitative detection method involving hydroxyproline is not suitable for determining rhCOL content [14]. The ELISA method assay is based on the binding reaction between the antigen and the antibody [15,16]. Currently, some types of rhCOL on the market are not full-length collagen but rather multiple repeats of a segment of the amino acid sequence [17]. If the amino acid sequence of rhCOL does not contain antigen–antibody binding sites, it can easily lead to false-negative results when using the ELISA method [18]. There are no available specific antibodies that are designed for a particular sequence of rhCOL on the market. Additionally, the content of rhCOL in different dressings varies greatly, and traces of rhCOL in complex matrices are difficult to detect. The optimization of dressing pretreatment methods and the establishment of a highly sensitive method for the detection of Re-COL in dressings are of great significance for the quality control and market regulation of dressings. However, studies on the quantification of rhCOL in dressings have rarely been reported.

In this study, the marker peptides of recombinant type III collagen (rhCOL-III) were identified with a high-performance liquid chromatography–tandem mass spectrometer (HPLC-MS/MS). Then, a sensitive quantitative detection method for rhCOL-III was established and validated. Afterwards, liquid, gel, and ointment dressings containing rhCOL-III were selected as the research objects. Specific pretreatment methods for three different dressings, the amount of trypsin, and the standard solution preparation method were optimized. Additionally, the recovery rates of rhCOL-III in three kinds of dressings with different dosages were analyzed to verify the accuracy of the pretreatment method. The content of rhCOL-III in the actual products of three kinds of dressings was further analyzed. The quantitative determination of rhCOL-III in three different dressings was achieved using this method.

2. Materials and Methods

2.1. Materials and Instruments

Materials: Liquid dressing, ointment dressing, gel dressing, and recombinant type III collagen raw material were provided by Xian Giant Biogene Co., Ltd., Xi’an, Shanxi, China. Chromatographic pure formic acid was obtained from Merck, Kennett Square, NJ, USA. Sequence-grade trypsin was obtained from Promega Corporation, Madison, WI, USA. Marker peptides (GEAGIPGVPGAK) were synthetized by QYAOBIO (ChinaPeptides Co., Ltd.), Shanghai, China, and the HPLC purity was 99.10%. All other reagents were commercially analytically pure.

Instruments: Orbitrap mass spectrometer (Exploris 480, Thermo Fisher Scientific, Waltham, MA, USA) was used to identify the marker peptide of rhCOL-III. Triple quadrupole mass spectrometer (TSQ, Quantum ACCESS MAX, Thermo Fisher Scientific, Waltham, MA, USA) was used for rhCOL-III quantification.

2.2. Design of Marker Peptide Specific for rhCOL-III

The selection principles of the marker peptide are as follows: Sequence blast alignment was performed on human type III collagen. The peptide, which appears only in rhCOL-III sequences and has no effect on its abundance in terms of chemical modification or cross-linking, can be used as the reference peptide. The reference peptide should have higher resolution in the process of chromatographic separation. The reference peptide of rhCOL-III used in this study was designed as “GEAGIPGVPGAK”. It was used as reference peptide and detection target of test sample for qualification and quantitation of rhCOL III in dressings.

2.3. Marker Peptides Identification by HPLC-MS/MS

2.3.1. Sample Preparation of rhCOL-III Enzymatic Digestion

The rhCOL-III with different concentrations was prepared with 0.1 mol/L Tris-HCl (pH 8.0) solution. Then they were centrifuged at 12,000× g for 10 min. The supernatant was collected for analysis.

Sample digestion: The rhCOL-III (5 mg) was dissolved in 10 mL 0.1 mol/L Tris-HCl (pH 8.0) solution. The rhCOL-III solution (0.5 mL) was mixed with 100 μL trypsin solution (0.2 mg/mL). The mixture was incubated at 37 °C for 24 h. And 10% formic acid solution (60 μL) was added to the rhCOL-III digestion to terminate the enzymatic reaction. The enzymatic digestion was centrifuged at 12,000× g for 10 min. The supernatant was collected for marker peptide identification.

2.3.2. Marker Peptides Identification

The marker peptides in rhCOL-III digestion were identified by HPLC/MS/MS. The on-line chromatographic separation was performed by reversed-phased chromatography on a Peptide BEH C18 column (2.1 × 150 mm, 1.7 μm) (Waters, USA) by using the UHPLC (Vanquish, Thermo Fisher, USA). The mobile phase consisted of water with 0.1% formic acid (A) and acetonitrile–water solution (v/v, 6:4) with 0.1% formic acid (B). The gradient elution procedure was performed as follows: 0–60 min, 5–40% B; 60–85 min, 40–90% B; 85–98 min, 90% B; 98−100 min, 90−5% B; 100−110 min, 5% B. The flow rate was 0.2 mL/min, the injection volume was 10 μL, and the column temperature was held at 60 °C. The outlet of the column was introduced into an orbitrap mass spectrometer (Exploris 480, Thermo Fisher Scientific, Waltham, MA, USA). Electron spray ionization (ESI) in positive mode was used to perform the orbitrap mass spectrometry. The spray voltage was set to 3.5 kV. The capillary temperature and vaporizer temperature were 320 °C and 300 °C, respectively. The sheath gas was 19.8 mL/min. The aux gas was 5 psi. The MS scan range was set from m/z 300 to 2000 and the resolution was set to 60,000. The RF lens was 45%. The normalized AGC Target was 300%. The maximum ion injection time was 100 ms. The scan event 2 was data-dependent MS/MS, and the resolution was set to 15,000. The isolation window was m/z 1.6. The maximum ion injection time was 200 ms, and the normalized AGC Target was 100%. The MS/MS collision energy was 30%. The rhCOL-III marker peptides were identified through the SEQUEST algorithm in Protein Discoverer 2.4 software (Thermo Fisher, USA) [19].

2.4. Investigation of Quantitative Method

2.4.1. Detection Procedure of Marker Peptide in rhCOL-III

The reference peptide and trypsin-digested rhCOL-III samples were separated with a Zorbax C18 column (2.1 × 150 mm, 5 μm) (Agilent Technologies, Santa Clara, CA, USA) using an HPLC (U3000, Thermo Fisher Scientific, Waltham, MA, USA). The mobile phase consisted of water with 0.1% formic acid (A) and acetonitrile with 0.1% formic acid (B). The gradient elution procedure was performed as follows: 0–0.5 min, 5–45% B; 0.5–2 min, 45–55% B; 2–8 min, 55–65% B; 8–8.1 min, 65–5% B; 8.1–10 min, 5% B. The flow rate was 0.2 mL/min, the injection volume was 5 μL, and the column temperature was held at 30 °C. The outlet of the column was introduced into TSQ mass spectrometer (Quantumn Access MAX, Thermo Fisher Scientific, Waltham, MA, USA). The ESI positive mode was used to perform the TSQ mass spectrometry. The spray voltage was set to 3.5 kV. The capillary temperature and vaporizer temperature were 320 °C and 300 °C, respectively. The sheath gas was 19.8 mL/min. The aux gas was 5 psi. The selected reaction monitoring (SRM) was used. The monitored target ions of rhCOL-III were m/z 526.71–371.97, 625.22 (GEAGIPGVPGAK).

2.4.2. Quantitative Methodological Validation

The rhCOL-III raw material was used for quantitative methodological validation. The marker peptide with different concentrations (0.1 μg/mL–500 μg/mL) were prepared with 0.1 mol/L Tris-HCl (pH 8.0). The marker peptide reference solution was analyzed 6 times to verify repeatability. The rhCOL-III raw material was prepared at the concentration of 0.5 mg/mL for three parallel samples. Each sample was digested respectively and analyzed 6 times to determine intraday precision. One sample was used for the interday precision test, which was tested on three different days, six replicates per day. The reference peptides (0.1, 0.25, 0.5 μg) were added to rhCOL-III digestion solution for analysis of spiking recovery rate. Each sample was analyzed three times in parallel, and the recovery rate was calculated. The limit of detection (LOD) and limit of quantitation (LOQ) for rhCOL-III raw material were determined from the calibration curve as 3 S/N and 10 S/N, respectively, where S is the signal of the target peak and N is the baseline noise.

2.5. Establish Qualitative and Quantitative Analysis Methods of rhCOL-III Dressings

2.5.1. Simulated rhCOL-III Dressing Sample Preparation

The 2.0 g ointment blank dressing matrix, 2.0 g gel blank dressing matrix, and 2 mL liquid blank dressing matrix were prepared as 12 parallel samples, respectively. Each kind of dressing was divided into 4 groups, with 3 parallels per group. The 0.01 mL, 0.05 mL, 0.1 mL, and 0.2 mL of rhCOL-III raw material solutions (prepared in Section 2.3.1) were added to the four groups of blank dressings matrices. The final concentration of rhCOL-III in each group was 10 μg/mL (group A), 50 μg/mL (group B), 100 μg/mL (group C), and 200 μg/mL (group D), respectively. Then, they were mixed well.

2.5.2. Investigation of Sample Pretreatment Methods

Three kinds of dressing samples from Group C, prepared in Section 2.5.1 were used for optimization of pre-processing methods, as described below.

Acetone precipitation method: Three dressing samples were mixed with cold acetone in the ratio of 1:5 (v/v). Then, they were shaken well and stored at −20 °C overnight. The samples were centrifuged (12,000× g, 20 min). Finally, the supernatants were removed, and the precipitates were redissolved with 0.5 mL 0.1 mol/L Tris-HCl (pH 8.0).

Ultrafiltration centrifugal method: The ointment and gel dressing samples (2 g) were first mixed with 0.7 mL 4 mol/L NaOH solution and centrifuged (12,000× g, 20 min). Then, the syringe was used to draw up the lower layer of clear liquid in the ointment dressing. The supernatant of the gel dressing sample was collected. The collected solutions of ointment dressing, gel dressing, and the liquid dressing samples were mixed with 1 mL ultrapure water, respectively. Subsequently, the samples were transferred to ultrafiltration centrifuge tubes with molecular weight cutoffs of 3 kDa, and centrifuged (8400× g, 20 min). The above operation was repeated 3 times. The supernatant was collected and lyophilized. At last, the samples were redissolved with 2 mL 0.1 mol/L Tris-HCl (pH 8.0) and magnetically stirred for 20 min.

Organic solvent extraction methods: The ointment and gel dressing samples (2 g) were first mixed with 0.5 mL 2 mol/L HCl solution and centrifuged (12,000× g, 20 min). Then, the lower layer of the clear liquid in the ointment dressing was drawn up using syringes. The supernatant of the gel dressing sample was collected. The collected solutions of ointment dressing, gel dressing, and liquid dressing samples were mixed with dichloromethane (v/v, 1:5), respectively. The solutions were mixed well and left for 2 h. Then, the upper liquid layer was collected and lyophilized. At last, the samples were redissolved with 2 mL 0.1 mol/L Tris-HCl (pH 8.0) and magnetically stirred for 20 min.

Solvent displacement methods: Firstly, three kinds of dressing samples were lyophilized. The lyophilized ointment and gel dressing samples were redissolved with 4 mL 0.1 mol/L Tris-HCl (pH 8.0). The lyophilized liquid dressing sample was redissolved with 2 mL 0.1 mol/L Tris-HCl (pH 8.0). All samples were magnetically stirred for 20 min.

2.5.3. Investigation of Trypsin Dosage

The samples (0.5 mL) prepared in Section 2.5.2 were mixed with 5 μg or 20 μg trypsin (100 μL), respectively. Then, they were incubated at 37 °C for 24 h. In order to terminate the enzymatic reaction, 10% formic acid solution (60 μL) was added to rhCOL-III digestion. The enzymatic hydrolysate was centrifuged at 12,000× g for 10 min, and the supernatant was collected.

2.5.4. Preparation of rhCOL-III Reference Working Solution

The rhCOL-III raw material was mixed with dressing blank matrices. The mixture was used as working solution for quantitation of rhCOL-III in different dressings. Preparation of blank matrix solutions was conducted as follows: First, 5 g of blank ointment dressing matrix and blank gel dressing matrix and 5 mL of liquid dressing blank matrix were lyophilized. Then, the blank ointment dressing matrix and blank gel dressing matrix were redissolved by 10 mL 0.1 mol/L Tris-HCl (pH 8.0). The liquid dressing blank matrix was redissolved by 5 mL 0.1 mol/L Tris-HCl (pH 8.0). All the blank matrices were magnetically stirred for 20 min and centrifuged at 12,000× g for 10 min. The supernatants of the blank gel and liquid dressing matrix solutions was collected, and the lower layer of the blank ointment dressing matrix solution was drawn up by syringes.

Preparation of working solution of rhCOL-III reference: The rhCOL-III enzymatic digestion solution prepared in Section 2.3.1 was diluted to different concentrations by three blank matrix solutions, respectively. Then, they were used for quantitation of rhCOL-III in different dressings, ensuring that the marker peptide to be detected in the reference sample and test samples have the same matrix background.

2.5.5. Methodological Validation of Pretreatment Method

Sample preparation: Four groups of the samples prepared in Section 2.5.1 and three actual dressing products were used to validate the pretreatment method. First of all, they were treated using the solvent displacement methods listed in Section 2.5.2. Then, they (0.5 mL) were mixed with 20 μg trypsin (100 μL) and incubated at 37 °C for 24 h, respectively. At last, 10% formic acid solution (60 μL) was added to the digestion. The enzymatic digestion was centrifuged at 12,000× g for 10 min, and the supernatant was collected.

LOD of the method in different dressings: The LOD values of rhCOL-III in different dressings subjected to pretreatment methods were determined.

3. Results

3.1. Marker Peptide Identification of rhCOL-III

Firstly, the marker peptides of rhCOL-III were identified using HPLC-MS/MS. Figure 1 shows the total ion chromatogram of digested rhCOL-III. Subsequently, the mass spectrometric data were analyzed using BioFinder software. In order to retrieve more peptides and exclude peptides undergoing modifications in marker peptides screening, the common modifications, such as oxidation and dethiomethylation of methionine, deamidation of asparagine and glutamine, pyroglutamylation of glutamine and glutamic acid, ammonia-loss, and acetylation of the nitrogen terminal were set up in BioFinder software. The peptides identified in rhCOL-III with no modification sites and which do not exist in other animals can be used as the target peptides. Figure 2 shows the MS spectrum of m/z 526.79071 with two charges. Therefore, the molecular weight of m/z 526.79071 was 1051.566 Da, which was consistent with the molecular weight of the peptide (GEAGIPGVPGAK). Figure 3 shows the MS/MS spectrum of m/z 526.79071. The MS/MS ions in Figure 3 are match well with the theoretical product ions of the peptide (GEAGIPGVPGAK). The peptide (GEAGIPGVPGAK) had no modification sites and only existed in rhCOL-III. Therefore, the peptide (GEAGIPGVPGAK) could be used as the marker peptide of rhCOL-III.

3.2. Establishment and Validation of Quantitative Method for Marker Peptide in rhCOL-III

Marker peptides with concentrations from 0.1 μg/mL to 500 μg/mL were used as the reference solutions. Figure 4 shows that the retention time of the marker peptide (GEAGIPGVPGAK) was 3.99 min. The abscissa was the concentration of the marker peptide working solutions. The peak area of the marker peptide in working solutions was used as the ordinate. The regression equation and correlation coefficient were y = 995631x + 543070 and R² = 0.9999 (GEAGIPGVPGAK). In order to verify the repeatability of the quantitative method, the maker peptide reference solution was repeatedly analyzed six times. The RSD of the peak area was 2.77%, indicating the higher repeatability of this method. The intraday precision was analyzed with three rhCOL-III samples, and each sample was detected six times (

Table 1. The intraday precision of quantification method for marker peptide of rhCOL-III.

Sample	Concentration (μg/mL)	Average Concentration (μg/mL)	SD	RSD (%)
S1	10.97	10.75	0.21	1.98
	10.74
	10.42
	10.81
	10.62
	10.97
S2	10.12	10.30	0.13	1.26
	10.36
	10.33
	10.32
	10.17
	10.47
S3	10.30	10.41	0.29	2.83
	10.40
	10.12
	10.98
	10.31
	10.34
Average concentration (μg/mL)			10.49
SD			0.29
RSD			2.76

). The results showed that the RSD of intraday precision was 2.76%. Then, one sample was used for the validation of interday precision and analyzed on different days (

Table 2. The interday precision of quantification method for marker peptide of rhCOL-III.

Sample	Concentration (μg/mL)	Average Concentration (μg/mL)	SD	RSD (%)
D1	10.97	10.75	0.21	1.98
	10.74
	10.42
	10.81
	10.62
	10.97
D2	10.65	10.55	0.29	2.75
	10.61
	10.54
	10.04
	10.50
	10.93
D3	10.31	10.59	0.22	2.04
	10.77
	10.35
	10.63
	10.85
	10.59
Average concentration (μg/mL)			10.63
SD			0.25
RSD			2.31

). The results showed that the RSD of the interday precision was 2.31%. Thus, the quantification method has good intraday and interday precision.

Table 3. Spiking recovery rate of quantification method for marker peptide of rhCOL-III.

NO.	Concentration (μg/mL)	Recovery Quality (μg)	Additive Quality (μg)	Spiking Recovery Rate (%)	Average Spiking Recovery Rate (%)
T-1-1	2.315	0.084	0.1	83.67	82.14
T-1-2	2.308	0.082	0.1	82.18
T-1-3	2.300	0.081	0.1	80.56
T-2-1	2.919	0.204	0.25	81.72	84.39
T-2-2	2.921	0.205	0.25	81.94
T-2-3	3.016	0.224	0.25	89.50
T-3-1	3.977	0.416	0.5	83.21	82.93
T-3-2	3.947	0.410	0.5	81.99
T-3-3	3.987	0.418	0.5	83.58
Average spiking recovery rate (%)					83.15
SD					2.58
RSD (%)					3.10

shows that the spiking recovery rate at different spiking levels was between 80% and 90%. The average spiking recovery rate was 83.15%. The RSD between the different spiking levels was 3.10%. This meets the requirements for spiking recoveries in methodological validation. The LOQ and LOD are very important in the analysis of trace substances. In this study, the marker peptide was gradually diluted and detected. The results showed that the LOQ and LOD of the marker peptide were 0.5 × 10⁻³ μg/mL and 0.2 × 10⁻³ μg/mL, respectively.

3.3. Establishment of Quantitative Method of rhCOL-III in Simulated rhCOL-III Dressing

3.3.1. Effect of Pretreatment Method on Recovery Rate of rhCOL-III

The dressing matrix was very complex and varied in different dressings. In this study, group C (prepared in Section 2.5.1) was used as the simulated dressing samples. In order to exclude the influence of matrix on the quantitative results, four different pretreatment methods, including acetone precipitation, centrifugal ultrafiltration, dichloromethane extraction, and solvent displacement, were compared. Figure 5 shows the effects of different pretreatment methods on the recoveries in simulated rhCOL-III dressings. The results showed that the recoveries of the centrifugal ultrafiltration method, acetone precipitation method, and dichloromethane extraction method were lower. In these three simulated rhCOL-III dressings, the recovery of the solvent displacement method was higher than the other three methods.

3.3.2. Effect of Trypsin Dosage on Recovery Rate of rhCOL-III

The carbomer in the three dressings can inhibit the activity of trypsin. Therefore, trypsin dosage was an important factor which could affect recovery rates. To ensure that rhCol-III in the dressing sample can be fully enzymatically hydrolyzed, excessive trypsin was added to the samples. In this study, 0.5 mL of the three simulated rhCOL-III dressing samples was mixed with 5 μg or 20 μg trypsin, respectively. The effect of enzyme dosage on the recovery of rhCOL-III was analyzed. Figure 6 shows the effect of different trypsin dosages on the recovery of rhCOL-III in simulated rhCOL-III dressings. The results showed that the recoveries of rhCOL-III using 20 μg trypsin digestion in three rhCOL-III dressings were all significantly higher than those of the 5 μg group. The results indicate that the increase in trypsin dosage could effectively increase the recovery rate of rhCOL-III in simulated rhCOL-III dressings.

3.3.3. Effect of Different Reference Working Solutions on the Recovery of rhCOL-III

The matrix effect is an important factor in mass spectrometry quantification. The contents and properties of the matrix will affect the ionization efficiency of the target compounds, resulting in the enhancement, weakening, or deviation of the mass spectrum signal. Thus, an rhCOL-III reference working solution was prepared with a blank matrix, and the recovery was compared with the results of the Tris-HCl method. Figure 7 shows the recovery of rhCOL-III in three simulated dressings obtained by the working solution curve method of the blank matrix and Tris-HCl prepared as a reference. The recoveries of rhCOL-III obtained by the method of blank matrix prepared as a reference working solution curve were all between 80% and 120%, which met the requirement of the recovery rate in the methodology.

3.3.4. Methodology Validation of Pretreatment Method in Simulated Dressing Samples

To further verify the accuracy of the method, rhCOL-III was added to three kinds of dressing matrices to final concentrations of 10 μg/mL (group A), 50 μg/mL (group B), 100 μg/mL (group C), and 200 μg/mL (group D). The optimized method was used to analyze the content of rhCOL-III in the simulated dressing samples.

Table 4. Recoveries of rhCOL-III in simulated dressing samples with different additive levels of rhCOL-III.

NO.	Liquid Dressing		Ointment Dressing		Gel Dressing
	Recovery (%)	Average Recovery (%)	Recovery (%)	Average Recovery (%)	Recovery (%)	Average Recovery (%)
A-1	80.21	87.85	73.06	70.46	83.29	80.02
A-2	80.63		70.91		80.71
A-3	79.95		67.40		76.06
B-1	85.87	80.26	87.22	81.87	88.42	92.01
B-2	89.32		81.14		95.59
B-3	88.34		77.24		92.03
C-1	85.68	83.07	81.04	80.46	111.98	102.59
C-2	79.94		78.47		90.83
C-3	83.59		81.87		104.95
D-1	82.89	82.78	84.23	82.69	89.94	93.66
D-2	81.73		85.49		99.34
D-3	83.73		78.37		91.72

shows the recovery of rhCOL-III in dressing samples with different spiking levels. The results showed that the average recoveries of rhCOL-III in liquid dressings and gel dressings with spiked amounts of 10 μg/mL, 50 μg/mL, 100 μg/mL, and 200 μg/mL were all between 80% and 120%. The average recoveries of 50 μg/mL, 100 μg/mL, and 200 μg/mL in the ointment dressings were all between 80% and 120%. The average recovery of the ointment dressing with 10 μg/mL was 70.46%. The main reason for these results may be due to the semi-solid state of the ointment dressing and the uneven mixing of trace amounts of rhCOL-III in the ointment dressing.

3.3.5. Quantification and Quantitation of rhCOL-III in Actual rhCOL-III Dressing Products

The rhCOL-III dressing products were further analyzed using the optimized method. Figure 8 shows that the MS spectra of marker peptide in dressing products were the same as in the rhCOL-III reference samples. This finding indicates that the three different dressing products contained rhCOL-III.

The theoretical rhCOL-III content in three dressing products was 100 μg/mL.

Table 5. The content of rhCOL-III in different dressing products.

Product Name	The Content of rhCOL-III (μg/mL)	Average Content (μg/mL)
rhCOL-III liquid dressing	104.29	106.61
	104.44
	111.11
rhCOL-III ointment dressing	82.49	82.81
	80.77
	85.17
rhCOL-III gel dressing	93.53	91.95
	90.32
	92.00

showed that the contents of rhCOL-III detected in three dressing products ranged from 80 to 112 μg/mL. The recoveries were all between 80% to 120% compared to theoretical contents.

The LOQ of the pretreatment method is an important indicator that influences the application of quantitative methods. Currently, the content of rhCOL in dressing products on the market varies widely. Some dressings only contain traces of rhCOL. In this research, we added a series of trace amounts of rhCOL-III to the blank matrices of the three dressings. Then, they were analyzed with the optimized method. The LOQ of rhCOL-III in liquid, ointment, and gel dressings with the optimized method was 1 ng/g, 2 ng/g, and 1 ng/g, respectively.

4. Discussion

In this study, a quantitative method for rhCoL based on marker peptide detection with HPLC/MS/MS was first established. Then, the accuracy of the quantitative method was verified. The RSD of repeatability was 2.77%. The RSD of intraday and interday precision was 2.76% and 2.31%, respectively. The spiking recovery rate was between 80% and 90%. Afterwards, this method was applied to the quantitative analysis of rhCOL-III contained in different dressings. Dressing matrices were complex and varied widely from one dressing matrix to another. Collagen in the dressing was mostly coated by matrices in form of oil in water, which caused difficulties in quantitative detection [20]. Therefore, pretreatment is essential for the detection of rhCOL in dressing products.

Four different pretreatment methods were compared. The recoveries of the samples prepared by centrifugal ultrafiltration method were the lowest. The main reason for this was that ointment dressings and gel dressings needed to be centrifuged to remove insoluble particulate matter to avoid clogging the ultrafiltration tube before ultrafiltration. In these two dressings, rhCOL-III was encapsulated in a matrix material in the form of oil in water. Centrifugation had difficulty breaking down this form, which resulted in part of the collagen being removed before ultrafiltration. Additionally, a part of rhCOL-III may be adsorbed on the ultrafiltration membrane during the ultrafiltration process, resulting in lower recovery. The recovery of ointment and gel dressings processed using acetone precipitation did not differ significantly from other ultrafiltration methods. This may be due to the fact that rhCOL-III in ointment and gel dressings was encapsulated by the matrices in the form of oil in water, which could not be emulsified using the acetone precipitation method. The recovery result of liquid dressings obtained by acetone precipitation was significantly higher than that of ultrafiltration. This was because the liquid dressing was in an aqueous state, and the acetone precipitation method could precipitate part of the rhCOL-III. However, the recovery rates remained relatively low. This may be due to the fact that the content of rhCOL-III in the dressings was only 100 μg/mL in general, and it was difficult for the acetone precipitation method to adequately precipitate trace amounts of this protein. The recoveries of rhCOL-III in these three dressings obtained by organic solvent extraction were significantly higher than via acetone precipitation and ultrafiltration, but the recoveries were still lower. On the one hand, the molecular weight of rhCOL-III is lower. It is soluble in organic solvents and a portion of the protein was carried away by the organic phase. On the other hand, the water-soluble matrix cannot be removed by organic solvents. Some water-soluble matrix may affect enzyme activity. Thus, rhCOL-III cannot be completely digested, resulting in a lower recovery rate. The sample with the highest recovery rate was obtained using the Tris-HCl solvent replacement method. In this method, the oil-in-water structure was destroyed through the lyophilization process to release rhCOL-III. Then, the freeze-dried sample was resolved by Tris-HCl. This was a simple step to reduce the loss of rhCOL-III caused by the pretreatment process.

In order to improve the recovery rate, the amount of trypsin was optimized. For the 0.5 mL sample solution, both 5 μg and 20 μg of trypsin were excessive. The 20 μg dosage of trypsin significantly improved the recovery. The matrices of three dressings could potentially inhibit the activity of trypsin. The content of matrix in the sample was mainly related to the sample volume. Therefore, there was no significant correlation between the amount of trypsin and the protein content in the sample; rather, it mainly related to the sample volume.

In addition to the pretreatment process, matrices are an important factor influencing quantitative detection via mass spectrometry [21,22]. During mass spectrometry ionization, matrix components in dressings may produce higher ionization energy, which inhibit the ionization of the rhCOL-III marker peptide, resulting in low detection recoveries. In addition, the matrix and rhCOL-III marker peptide may co-evaporate during ionization, forming a mixed gas stream that interferes with the signal of the target. The internal standard method is an effective method to exclude matrix effects [23]. However, the target compound was a peptide in this study. The chemically synthesized, isotopically labeled peptide was not stable in the aqueous solution and therefore degraded in a very short period of time. The method of reference working solution preparation using blank matrix can both exclude the interference of matrices and ensure the stability and accuracy of the results. The reason for this may be that the test sample and reference working solution have the same matrix background.

5. Conclusions

In this study, we first established a quantitative method of rhCOL-III based on HPLC/MS/MS. Afterwards, the quantitative method was used to analyze the rhCOL-III content in liquid, ointment, and gel dressings. We found that the pretreatment method was an important factor in the quantitative analysis of rhCOL-III dressings. Furthermore, we optimized and validated the pretreatment method of different dressings. The recoveries of different dressings treated by this method meet the quantitative methodological requirements and enable the analysis of trace collagen in different dressings. This study established an accurate quantitative method that could be used for the detection of recombinant collagen in dressings with complex matrices. We believe that this method is of great significance for the quality control and market supervision of dressing products.