Skip to main content
NCBI home page
Elsevier - PMC COVID-19 Collection logoLink to Elsevier - PMC COVID-19 Collection
. 2023 Feb 15;227:115152. doi: 10.1016/j.bios.2023.115152

Nanopore membrane chip-based isolation method for metabolomic analysis of plasma small extracellular vesicles from COVID-19 survivors

Qi Huang a,b,1, Wenjing Xiao a,1, Peng Chen c,1, Hui Xia a, Sufei Wang a, Yice Sun a, Qi Tan a, Xueyun Tan a, Kaimin Mao a, Han Xie b, Ping Luo d, Limin Duan a, Daquan Meng a, Yanling Ma a, Zilin Zhao a, Fen Wang a, Jianchu Zhang a,b, Bi-Feng Liu b,c,∗∗, Yang Jin a,b,∗
PMCID: PMC9928611  PMID: 36805272

Abstract

Multiple studies showed that metabolic disorders play a critical role in respiratory infectious diseases, including COVID-19. Metabolites contained in small extracellular vesicles (sEVs) are different from those in plasma at the acute stage, while the metabolic features of plasma sEVs of COVID-19 survivors remain unknown. Here, we used a nanopore membrane-based microfluidic chip for plasma sEVs separation, termed ExoSEC, and compared the sEVs obtained by UC, REG, and ExoSEC in terms the time, cost, purity, and metabolic features. The results indicated the ExoSEC was much less costly, provided higher purity by particles/proteins ratio, and achieved 205-fold and 2-fold higher sEVs yield, than UC and REG, respectively. Moreover, more metabolites were identified and several signaling pathways were significantly enriched in ExoSEC-sEVs compared to UC-sEVs and REG-sEVs. Furthermore, we detected 306 metabolites in plasma sEVs using ExoSEC from recovered asymptomatic (RA), moderate (RM), and severe/critical COVID-19 (RS) patients without underlying diseases 3 months after discharge. Our study demonstrated that COVID-19 survivors, especially RS, experienced significant metabolic alteration and the dysregulated pathways mainly involved fatty acid biosynthesis, phenylalanine metabolism, etc. Metabolites of the fatty acid biosynthesis pathway bore a significantly negative association with red blood cell counts and hemoglobin, which might be ascribed to hypoxia or respiratory failure in RM and RS but not in RA at the acute stage. Our study confirmed that ExoSEC could provide a practical and economical alternative for high throughput sEVs metabolomic study.

Keywords: Small extracellular vesicles, Metabolomics, sEVs separation and enrichment chip (ExoSEC), COVID-19

1. Introduction

Multiple metabolomic researches showed that metabolism is a critical contributor to the susceptibility and severity of COVID-19 and plays a part in the treatment and prognosis of COVID-19 (Ayres, 2020; Jia et al., 2022; Mussap and Fanos, 2021; Song et al., 2020; Wu et al., 2021). Our previous metabolomic studies also proved that COVID-19 could cause damage to various organs and systems and result in persisting mental and physical consequences, which may exacerbate metabolic disorders and thereby delay recovery (Xu et al., 2021; Zhang et al., 2021; Zhou et al., 2022).

Small extracellular vesicles (sEVs, ＜200 nm in diameter) are an essential subtype and provide promising approaches for metabolomic study (Gao et al., 2021; Williams et al., 2019). Recent studies proved that sEVs from plasma could concentrate metabolites and prevent them from degradation, allowing for circulation of metabolites that were unstable in free plasma and other biofluids, such as urine samples and malignant pleural effusion (Luo et al., 2020; Puhka et al., 2017; Song et al., 2020; Williams et al., 2019). Studies have shown that there are significant metabolomic differences between plasma and plasma sEVs of COVID-19 patients in the acute phase (Ayres, 2020). A question presents itself as to what the metabolic features of plasma sEVs are in COVID-19 survivors.

A challenge facing the metabolomic study of plasma sEVs is how to harvest sufficient pure and intact sEVs from limited samples at low cost, and in a short time, during the COVID-19 pandemic. Ultracentrifugation (UC), the gold standard of sEVs separation is labor-intensive, time-consuming, and has low yield with poor quality (Williams et al., 2019). Total Exosome Isolation Reagents (REG) based on polymer precipitation have been developed that can also precipitate other soluble materials such as lipoproteins and proteins resulting in low-purity sEVs (Janouskova et al., 2022). These limitations and the non-intactness of sEVs render them non-amenable to downstream metabolomic analysis. Therefore, clinically, an alternative method that can obtain high-quality and high-yield sEVs from the limited amount of samples at a lower cost is urgently needed.

The emerging microfluidics showed great potential for being used for sEVs isolation from biological fluids given their advantages in hydrodynamic properties, size filtration, acoustic fields, immunoaffinity, and dielectrophoretic properties. They have been successfully employed in the detection of nucleic acids and proteins from sEVs with high sensitivity and high specificity (Guo et al., 2018; Liu et al., 2017; Wang et al., 2021b). Recently, we also experimentally used unique microfluidic approaches for the efficient enrichment and purification of sEVs and have successfully achieved highly sensitive detection, functionalization, and drug delivery of sEVs (Wang et al. 2017a, 2017b, 2021a, 2021b). A pioneering technique termed ExoTIC (Liu et al., 2017), is a size-based sEVs isolation tool that shows a good prospect of clinical application. Based on the technical principles of ExoTIC, we optimized its structure and used the improved chip (referred to as ExoSEC) for the purification of sEVs.

In this article, we used ExoSEC for plasma sEVs enrichment, and then compared the sEVs obtained by UC, REG, and ExoSEC in terms the time, cost, purity, and metabolic features. Our results proved that ExoSEC outperformed UC or REG, cost, sEVs yield, and metabolite identification. We also detect the metabolic alteration in plasma sEVs from COVID-19 survivors by ExoSEC. We believe that our ExoSEC could provide a useful alternative for the metabolomic analysis of sEVs.

2. Materials and methods

This section in this paper is presented in supplementary materials.

3. Results and discussion

3.1. Principle and workflow of ExoSEC

ExoSEC was a filtration method mechanistically comparable to ExoTIC. The new chip was tailored to fulfill the requirements of handling samples with limited sample amounts (Fig. S1). Different from ExoTIC, ExoSEC used the silicone pad and polycarbonate for the top gasket (layer 2) and plastic support filter (layer 5), respectively. This soft-hard materials combination effectively minimizes liquid leakage and enhances the system stability. Furthermore, to create a robust and fluid-tight seal around the periphery, the ExoSEC was secured by a ring of compressive fasteners and the silicone gasket.

Fig. 1 presented the flow of plasma sEVs isolation by ExoSEC. The concentrated sEVs and any residual sEVs bound to the filter membrane were recovered into the same tube for downstream analysis. Plasma sEVs of four healthy people were isolated by three techniques (ExoSEC, UC, or REG) and characterized by nanoparticle tracking analysis (NTA), transmission electron microscopy (TEM), Western blotting, and liquid chromatography-quadrupole time of flight-tandem mass spectrometry (LC-QTOF-MS/MS), in terms of size, concentration, morphology, expression of sEVs markers and metabolomic profiles, respectively (Fig. 1B). Finally, we determined 306 metabolites in plasma sEVs isolated by ExoSEC from 75 recovered COVID-19 (RC) patients, including 16 recovered asymptomatic (RA) patients, 29 recovered moderate (RM) patients, 30 recovered severe/critical (RS) COVID-19 patients, and 17 healthy controls (HC) by targeted metabolomics (Fig. 1C).

Fig. 1.

Fig. 1

Open in a new tab

Schematic overview of the experiment flow chart. (A) The diagram design of the ExoSEC device for small extracellular vesicle (sEVs) isolation. All materials in the device can be cleaned, sanitized, and reused, except for 50 nm and 200 nm pores. (B) sEVs were purified from the plasma of four healthy controls by ExoSEC, ultracentrifugation (UC), and sEVs isolation reagents (REG) and subjected to downstream analysis. (C) Plasma sEVs were isolated from recovered COVID-19 patients (RC), including recovered asymptomatic (RA, n = 16) patients, recovered moderate (RM, n = 29) patients, recovered severe/critical (RS, n = 30) patients, and healthy controls (HC, n = 17). Then, targeted metabolomics of isolated sEVs was performed using a Q300 kit.

3.2. Comparison among plasma sEVs isolation by ExoSEC, UC, and REG

3.2.1. ExoSEC costs less than UC and REG

Plasma sEVs from four healthy subjects were isolated, respectively, by UC, REG, and ExoSEC methods and then characterized according to the MISEV2018 guidelines (Théry et al., 2018). The steps of the three techniques are detailed in Fig. 2 A. We found that the ExoSEC costs substantially less (3 $) than UC (70 $) and REG (29 $) per mL plasma (Fig. 2A). The material cost for ExoSEC device was no more than a dollar. And the cost could be further lowered with large-scale detection since most of the material can be sterilized and reused by autoclave. Besides, the plasma sEVs separation using ExoSEC (3 h) took less time than UC (6 h). Considering the time and cost, ExoSEC might serve as an alternative tool for sEVs isolation in clinical practice.

Fig. 2.

Fig. 2

Open in a new tab

Validation and comparison of small extracellular vesicles (sEVs) by ultracentrifugation (UC), ExoSEC, and sEVs isolation reagents (REG) from four plasma of four healthy subjects. (A) The workflow, time-consuming, and cost of the sEVs isolation by three methods. (B) Representative transmission electron microscopy (TEM) images of ExoSEC-sEVs, UC-sEVs, REG-sEVs. (C) Representative images of nanoparticle tracking analysis (NTA) of ExoSEC-sEVs, UC-sEVs, REG-sEVs. (D) Mean size (a) and concentration (b) of ExoSEC-sEVs, UC-sEVs, and REG-sEVs. (E) The particle/protein ratios of sEVs by UC, ExoSEC, and REG. (F) The expression of sEVs markers (CD63, CD81, and TSG101) and non-sEVs markers (Albumin and Apolipoproteins A1) of ExoSEC-sEVs, UC-sEVs, and REG-sEVs. Abbreviations: ExoSEC-sEVs, UC-sEVs, and REG-sEVs indicated sEVs isolated by ExoSEC, UC, and REG, respectively. P-value: ∗, < 0.05; ∗∗, < 0.01; ∗∗∗, < 0.001.

3.2.2. ExoSEC achieved higher sEVs yield and purity than UC and REG

Characterization of enriched sEVs by three methods is shown in Fig. 2B–D. First, TEM images displayed that the plasma sEVs isolated by the three methods had intact membrane structures and similar morphology (Fig. 2B). Then, NTA analysis exhibited that most of the sEVs were sized 60–200 nm and all had a unimodal peak in Fig. 2C. There was no difference in the main peak of sEVs among the three approaches in Fig. S2A. The mean size of the ExoSEC-sEVs and REG-sEVs were similar, and both were smaller than UC-sEVs in Fig. 2D (a). Notably, our data proved that the ExoSEC-sEVs yield provided 205-fold (P = 0.013) and 2-fold (P = 0.034) higher than that of UC-sEVs and REG-sEVs, respectively in Fig. 2D (b).

As expected, the ExoSEC and ExoTIC devices for sEVs isolation showed the same trend that enrichment of plasma-derived sEVs using chips is more efficient than UC and REG. Our results of NTA analysis were partially different from previously reported findings (Liu et al., 2017). In terms of the peak, ExoTIC-sEVs and UC-sEVs had a similar unimodal peak, but they were both smaller than REG-sEVs. In terms of the mean size, their results revealed that the mean size of ExoTIC-sEVs was nearly 100 nm larger than that isolated by the other two sEVs isolation reagents. In brief, the sEVs yield of several microfluidic chips including our ExoSEC and two previous other chips ExoTIC and Exodisc were generally significantly higher than other methods the contrast UC method particularly (Dong et al., 2020; Liu et al., 2017). The discrepancy could be partly ascribed to the fact that their sEVs were isolated from the cell culture medium rather than plasma. Besides, the difference in subjects, reagents used, or isolation mechanisms of the three approaches also needed to be considered. Overall, ExoSEC successfully isolated sEVs from plasma and attained higher sEVs yield and purity but cost less than UC and REG.

To estimate the purity of plasma sEVs by UC, ExoSEC, and REG, the protein concentrations of sEVs samples were detected. The results showed that the protein concentrations were higher overall in plasma sEVs by REG, followed by ExoSEC and UC, which was in line with the Coomassie blue staining (Figs. S2B–C). Moreover, we calculated the particle/protein ratio to certify the purity of sEVs. ExoSEC-sEVs obtained a higher ratio than REG-sEVs and UC-sEVs (Fig. 2E), implying that ExoSEC-sEVs achieved higher purity than UC and REG. The results of our study are different from those reported by the previously published study, which showed that the particle/protein ratios of plasma sEVs isolated by another microfluidic device, Exodisc were lower than UC (Dong et al., 2020). Perhaps the purity may be associated with the different chip structures. Next, sEVs markers (CD63, CD81, and TSG101) and non-sEVs markers (Albumin and Apolipoproteins A1) were assessed using Western blotting. Our data revealed the presence of CD63, CD81, and TSG101 (Fig. 2F). Expression of Apolipoproteins A1 and Albumin in plasma sEVs was significantly lower than in plasma from the same individuals, but also minimally examined in sEVs by three methods (Fig. 2F).

In a previous study of a comprehensive evaluation of methods for sEVs separation from human plasma, they found that the sEVs samples were positive for Apolipoproteins A1 (Apo-A1) expression and Exodisc had the highest level than UC (Dong et al., 2020). The currently published literature suggests that albumin and lipoprotein expression of plasma-derived sEVs are in varying degrees, indicating that no single technique can completely remove the contamination of plasma albumin and lipoproteins (Wei et al., 2020).

3.2.3. Metabolomics of plasma sEVs isolated by ExoSEC, UC, and REG

To confirm the feasibility of ExoSEC, we first compared the metabolomic features of ExoSEC-sEVs with UC-sEVs and REG-sEVs by using LC-QTOF-MS/MS. Quality control (QC) samples verified the stability and reproducibility of the sample analysis (Fig. S3A). Orthogonal Partial Least Squares-Discriminant Analysis (OPLS-DA) was utilized for multiclass classification and identification of differential metabolites. The result revealed that the metabolites of ExoSEC-sEVs could be easily distinguished from those of UC-sEVs and REG-sEVs (Figs. S3B–D). Venn diagram showed that 2038, 2007, and 1826 metabolites were identified in ExoSEC-sEVs, UC-sEVs, and REG-sEVs, respectively, while no significant difference was found in species and distribution of the sEVs metabolites among the three groups (Fig. 3 A and Fig. S4). All the differential metabolites in the three groups were visualized in the heatmap (Fig. 3B). The volcano plots showed that 241 and 101 metabolites were significantly increased, while 91 and 52 metabolites were significantly decreased in the ExoSEC-sEVs compared to UC-sEVs and REG-sEVs (Fig. 3C and D), respectively. The top 50 differential metabolites are shown in violins, most of which were higher in ExoSEC-sEVs than in the other two groups (Fig. S5). Then, we performed a KEGG enrichment analysis to identify the altered metabolic pathways among the three groups. Our results demonstrated that the signaling pathways involved in the metabolism of phenylalanine, alanine, aspartate, glutamate, glyoxylate, and dicarboxylate were significantly enriched in the ExoSEC-sEVs compared to the other two groups (Fig. 3E and F). Collectively, our results showed that several differential metabolic signaling pathways were enriched using ExoSEC compared to UC and REG.

Fig. 3.

Fig. 3

Open in a new tab

Evaluation of metabolomic features of small extracellular vesicles (sEVs) isolated by ultracentrifugation (UC), ExoSEC, and sEVs isolation reagents (REG) from plasma of four healthy subjects. (A) Venn diagram showed identified metabolites in ExoSEC-sEVs, UC-sEVs, and REG-sEVs. (B) Cluster heatmap of differential metabolites among ExoSEC-sEVs, UC-sEVs, REG-sEVs. The level of metabolites was normalized by the Z-score. (C–D) Volcano plots of differential metabolites in ExoSEC-sEVs vs. UC-sEVs and ExoSEC-sEVs vs. REG-sEVs. The X-axis represents log2FC, and the Y-axis is -log10(P-value). FC (fold change) is the level of metabolites in ExoSEC-sEVs relative to the level of UC-sEVs or REG-sEVs. Significant metabolites with log2FC > 0.25 are represented by red dots, and those with log2FC < −0.25 are represented by green dots. (E–F) KEGG pathway analysis of differentiating metabolites in ExoSEC-sEVs vs. UC-sEVs (E) and ExoSEC-sEVs vs. REG-sEVs (F). The abbreviations: ExoSEC-sEVs, UC-sEVs, and REG-sEVs indicated sEVs isolated by ExoSEC, UC, and REG, respectively.

Although studies are mounting on sEVs metabolomics, sEVs isolation from plasma by UC remains the mainstay method (Williams et al., 2019). Previous studies proved that ExoTIC and UC methods for sEVs isolation from prostate cancer cell lines yielded comparable microRNA profiles, and they also found that some microRNAs were higher, but others were lower in ExoTIC-sEVs compared with UC-sEVs. Their results further demonstrated that a more significant number of proteins were identified in ExoTIC-sEVs compared to UC-sEVs (Liu et al., 2017). We also found a similar phenomenon in plasma sEVs metabolomics using ExoSEC. All these differences in micro-RNA, protein expression, and metabolomic profiles can be attributed to the different mechanisms of the separation methods. Comparing the similarities and differences between the two chips (ExoSEC and ExoTIC), we found that they were based on a similar principle, are more efficient, and cost less for plasma sEVs isolation. They also showed higher sEVs yield compared to UC and REG. When it comes to differences, we stated that ExoSEC was innovative in that it is also suitable for sEVs metabolomic studies, and more metabolites were identified within ExoSEC-sEVs as compared with UC-sEVs and REG-sEVs. Besides, we first used our optimized ExoTIC for clinical plasma sEVs isolation of recovered COVID-19 patients (RA, RM, RS) without underlying diseases 3 months after discharge to the HC. This innovative nature was not mentioned by the ExoTIC in the study by Liu et al. Both the two chip opens up new avenues for clinical batch precious samples with a small sample volume of the plasma, although we did not directly compare the sEVs isolation effects of ExoTIC and ExoSEC in this study.

3.3. Application of ExoSEC in the plasma sEVs metabolomics of COVID-19 survivors

3.3.1. Clinical features of recovered COVID-19 patients

The clinical characteristics, including hematological, biochemical, and coagulation parameters of RC at two-time points (admission and recovery at 3 months), are displayed in Table S1 and Fig. S4. There were no significant differences between the three groups and HC in terms of age, sex, or body mass index (BMI) (P > 0.05). Table S1 shows that all the RC and HC were negative for SARS-CoV-2 RNA at recovery. The IgG seropositivity rates were 87.5%, 89.7%, and 90.0% in the RA, RM, and RS groups, respectively. IgM seropositivity rates were 12.5%, 17.2 and 13.3%, respectively. These results indicated that all the RC had a past SARS-CoV-2 infection and were in recovery. Multiple clinical indicators, elevated at admission, restored during recovery, such as C-reactive protein, liver function profile (alkaline phosphatase, alanine aminotransferase, aspartate aminotransferase, lactate dehydrogenase), the renal function (creatinine), cardiac function (creatine kinase-myocardial band), in the RM or RS group. Nonetheless, a few measures, such as lymphocytes, α-hydroxybutyrate dehydrogenase, blood urea nitrogen, prothrombin time, and glucose, failed to return to normal, implying that part of the renal and cardiac functions, immune and coagulation system remained impaired (Table S1 and Fig. S6). The clinical characteristics in our cohort were similar to our previous findings (Zhang et al., 2021).

3.3.2. The validation of sEVs from RA, RM, RS groups, and HC

All plasma sEVs from RC and HC were isolated by ExoSEC and evaluated by TEM, NTA, and Western blotting (Fig. 4 ). As expected, TEM images revealed that the sEVs isolated in the four groups had an intact membrane architecture, and similar morphology (Fig. 4A). Moreover, NTA analysis exhibited that there existed no differences in sEVs amount, mean size and size distribution, percentage of the main peak among the four groups (Fig. 4B and C and Tables S2–3). Subsequently, Western blotting identified sEVs positive markers and non-sEVs markers expression in four groups (Fig. 4D). To the best of our knowledge, it was the first study to use optimized size-based sEVs separation microfluidic chip for plasma sEVs isolation from COVID-19 survivors. Most plasma sEVs isolation methods in these studies used exosome isolation reagents, such as the MEKit (ME-020p-Kit), ExoQuick, Exo-Spin exosome purification reagents or size-exclusion chromatography (Alzahrani et al., 2021; Barberis et al., 2021; Cocozza et al., 2020). The sEVs downstream metabolomic analysis of different separation methods warrants further investigation.

Fig. 4.

Fig. 4

Open in a new tab

Verification of small extracellular vesicles isolated by ExoSEC (ExoSEC-sEVs) from healthy controls (HC), recovered asymptomatic (RA) patients, recovered moderate (RM) patients, and recovered severe/critical (RS) patients. (A) Representative Transmission electron microscopy (TEM) images of ExoSEC-sEVs from four groups. Scale bar: 100 nm. (B–C) nanoparticle tracking analysis (NTA) showed the diameter and size distribution of sEVs from HC, RA, RM, and RS patients (n = 5). (D) Representative immunoblot showed the enrichment level of CD63, CD81, and TSG101, Albumin, and Apolipoproteins A1.

3.3.3. Metabolic profiles of plasma sEVs from RC and HC by targeted metabolomics

Previous data proved that recovered SARS-COV1 survivors and SARS-COV2 survivors developed significant metabolic disorders twelve years and 1-month after infection, respectively (Wu et al. 2017, 2021). Our previous untargeted metabolomics revealed that recovered COVID-19 patients presented different plasma metabolomic profiles from HC. In this study, we first compared the metabolomic profiles of recovered COVID-19 patients (RA, RM, RS) to that of HC to identify differential metabolites and metabolic pathways by targeted metabolomics using the Q300 Kit.

QC samples were used for the verification of the stability and reproducibility of the sample analysis. Principal component analysis (PCA) without supervision provided a snapshot of the metabolite characteristics of the four groups, and the results indicated that the differentiating power was low among the four groups (Fig. 5 A). Therefore, OPLS-DA was used for multiclass classification and identification of differentially-altered metabolites (Figs. S7A–B). We focused on the differences between the RA, RM, RS, and HC. A pie chart was used to illustrate the percentage in each group in terms of the changing pattern (Figs. S8A–C). The results indicated that over 50% of these differential metabolites were diacylglycerol and more than 10% were organic acids in RA, RM, and RS relative to HC. In addition, the proportions of amino acids, fatty acids, organic acids, and carbohydrates among differential metabolites in the three recovery groups (RA, RM, and RS) incrementally went up with the severity of COVID-19 compared to the HC. On the other hand, the proportions of diacylglycerol gradually dropped (Figs. S8A–C). The results showed apparent metabolomic differences in RM and RS compared to HC, illustrating that plasma sEVs metabolic alteration becomes more significant as the condition exacerbates.

Fig. 5.

Fig. 5

Open in a new tab

Metabolic profiles of plasma-derived small extracellular vesicle isolated by ExoSEC (ExoSEC-sEVs) from healthy controls (HC), recovered asymptomatic (RA) patients, recovered moderate (RM) patients, and recovered severe/critical (RS) patients. (A) Score plots of principal component analysis (PCA) of 306 metabolites detected in ExoSEC-sEVs from four groups. (B–D) Volcano plots of differential metabolites in ExoSEC-sEVs from four groups. The X-axis is the value of log2(FC), and the Y-axis is −log (P-value). FC is the ratio of the mean level of the metabolites in the RA, RM, or RS to the mean level of HC, respectively. Red and blue dots represent the upregulation and downregulation of metabolites with P < 0.05, and grey dots denotes no change. (E) Venn diagram of differential metabolites in RA vs. HC, RM vs. HC, and RS vs. HC. (F) Heat map of significant differential sEVs metabolites in RA, RM, and RS patients relative to HC. Only differential metabolites with P < 0.05 and |log2FC| > 0.25 are displayed. And the shade of red indicates an increase of metabolites, and blue indicates a decrease. The abbreviation ExoSEC-sEVs means sEVs isolated by ExoSEC.

For RC patients of various severity, |log2FC| > 0.25 and P < 0.05 were considered significant (Fig. 5B–E). All the differential metabolites in the four groups are given in Fig. 5F and Table S4. The volcano plots show that, compared to the HC group, only 2 differential metabolites were significantly increased, and 8 decreased in the RA group, and 18 rose, and 8 dropped in the RM group, while the numbers were even more significant in the RS group, with 32 being upregulated and 12 downregulated (Fig. 5B–D and Table S4). We found that four differential metabolites overlapped in the three RC subgroups compared to the HC (Fig. 5E). Three of them, including benzoic acid, 3-hydroxyisovaleric acid, and 10Z-heptadecanoic acid, were correlated with disease severity (Fig. 6 A). Benzoic acid was sharply reduced. The other two compounds increased gradually as the disease deteriorated (Fig. 6A). Among these differential metabolites, 8 metabolites showed comparable changes both in the RM and in RS but not in RA versus HC. These metabolites included hydrocinnamic acid, oleylcarnitine, butyric acid, myristic acid, palmitoleic acid myristoleic acid, palmitelaidic acid, and oleic acid (Fig. S9).

Fig. 6.

Fig. 6

Open in a new tab

KEGG enrichment analysis and correlations between clinical indexes and differential ExoSEC-sEVs metabolites in the recovered moderate (RM) patients and recovered severe/critical (RS) patients relative to healthy controls (HC). (A) The dot graphs indicate the dose intensity change of 10Z-Heptadecenoic acid, 3-Hydroxyisovaleric acid, and benzoic acid in four groups. Each dot represents an individual subject: RA (green), RM (red), RS (blue), or HC (black). P-value: ∗, < 0.05; ∗∗, < 0.01; ∗∗∗, < 0.001. (B–C) The related KEGG pathway analysis of differential metabolites RM vs. HC and RS vs. HC. The color of the bubbles represents the value of P value, and the size of bubbles represents the number of counts (sorted by metabolites ratio). (D–F) The correlation analysis between the differential metabolites with clinical parameters in recovered asymptomatic (RA) patients, RM, and RS relative to HC. Only differential metabolites with P < 0.05 and r > 0.3 are displayed; blue circus indicates a positive correlation, and red indicates a negative. The abbreviation ExoSEC-sEVs means sEVs isolated by ExoSEC.

To further analyze the differential metabolites of plasma sEVs in the four groups, KEGG functional enrichment analysis was performed to annotate the potential pathways. The differential metabolic pathways were mainly related to sulfur/propanoate/glycine, serine, threonine/glyoxylate, and dicarboxylate metabolism in RM compared to HC. Both fatty acid biosynthesis and phenylalanine metabolism were enriched in the RM and RS compared with HC. And butanoate/alanine, aspartate, glutamate/arginine, and proline metabolism were enriched in the RS compared with HC (Fig. 6B and C).

These pathways, involving the metabolism of propanoate, aspartate, arginine/proline, glycine/serine, and phenylalanine, were also dysfunctional at the acute stage of COVID-19 and during early recovery stage (when COVID-19 test yielded negative results twice and isolation was lifted) as revealed by untargeted and targeted metabolomic analysis of plasma but not by plasma-derived sEVs (Jia et al., 2022), implying that similar pathway dysfunction exists in plasma and plasma sEVs. Both fatty acid biosynthesis and phenylalanine metabolism were enriched in the RM and RS compared with HC. A previous study demonstrated that the essential enzymes for fatty acid synthesis were also elevated in the acute stage, providing indirect support to the conclusion reached in this study (Taneri et al., 2020; Tanner and Alfieri, 2021). Increasing studies on plasma metabolomic profiles proved that phenylalanine metabolism was significantly abnormal in the acute stage of COVID-19 (Luporini et al., 2021). Phenylalanine is a marker indicative of the severity of COVID-19. Serum levels of phenylalanine were recently found to be associated with aggravated inflammation, higher sequential organ failure assessment scores, ICU admission, and higher mortality rates among non-COVID-19 patients (Luporini et al., 2021). All these data suggested that the abnormal metabolic profiles in COVID-19 survivors failed to fully return to normal even 3 months after discharge from the hospital, although most of their blood and biochemical parameters were practically restored to normal and clinical manifestations improved in RA, RM, and RS. These differential metabolomic pathways in COVID-19 survivors in this study were partly consistent with our previous findings, which suggested that RA recovered well, but some clinical indicators and plasma pathways in RM and RC remained abnormal compared with HC. An untargeted metabolomic study showed that these abnormal pathways involved the tricarboxylic cycle, purine and glycerophospholipid metabolism, and phenylalanine metabolism in plasma (Zhang et al., 2021). Although the final analysis showed that some differential metabolites and metabolic pathways overlapped, such as phenylalanine metabolism, other new differential metabolites, and metabolic pathways were found in plasma sEVs in this study, suggesting that metabolomic pathways are different in plasma and plasma sEVs of RC.

3.3.4. Associations between differential metabolites and clinical parameters

We further examined the differential metabolites of plasma sEVs with respect to the clinical indicators of RC patients. As expected, Spearman correlation analysis revealed a complicated relationship between the discriminative clinical indexes and the differential metabolites. We found that some metabolites bore a stronger relationship with clinical indexes in the RM and RS versus HC. Moreover, most of the associations between metabolites and clinical indexes were consistent in (or across all groups) both groups, except for methylmalonic acid, palmitoleic acid, and palmitelaidic acid (Fig. 6D–F).

Inflammatory markers, such as C-reactive protein, white blood cells, lymphocytes, and neutrophils, are related to oleylcarnitine, 3-hydroxyisovaleric acid, and 10Z-heptadecanoic acid in RC. Several differential metabolites were closely associated with red blood cells (RBC) and hemoglobin (Hb) in RM and RS but not in RA. Five differential metabolites, i.e., myristic acid, myristoleic acid, palmitoleic acid, palmitelaidic acid, and oleic acid in the signaling pathway of fatty acid biosynthesis were significantly negatively related to RBC and Hb in both RM and RS groups but not in RA group. 3-Hydroxyisovaleric acid was positively associated with RBC and Hb, and 10Z-heptadecanoic acid was negatively associated with RBC and Hb, but no association was found between benzoic acid and the three makers (Fig. 6D and E). The renal function markers (creatinine, blood urea nitrogen, and urine acid) were also linked to differential metabolites in RA, RM, and RS (Fig. 6D–F). Benzoic acid and myristoleic acid were positively associated with creatinine in RM and RS groups but not in RA. 3-Hydroxyisovaleric acid was positively, and 10Z-heptadecanoic acid was negatively associated with creatinine in RA, RM, and RS.

Anemia reduces oxygen delivery to cells, tissues, and organs and thus plays a vital role in the development of multi-organ failure. Besides, compared with moderate COVID-19 patients, critical patients had lower RBC counts and higher red blood cell volume distribution width (Thomas et al., 2020). A great many COVID-19 patients suffer from shortness of breath and persistent dry cough, some of which persist during recovery. RBC may play a role in the deterioration of hypoxemia in COVID-19 patients. Previous research revealed that SARS-CoV-2 infection results in significantly dysregulated RBC metabolism by elevating the level of free fatty acids, despite the minor increase in the mean corpuscular volume and no significant changes in the RBC count, hematocrit, or other hematological indicators (Thomas et al., 2020). Hb levels were lower in the elderly, and a higher percentage of COVID-19 patients had diabetes, hypertension, and systemic conditions and tended to be admitted to ICU (Taneri et al., 2020). Our data revealed that five differential metabolites, i.e., myristic acid, myristoleic acid, palmitoleic acid, palmitelaidic acid, and oleic acid implicated in fatty acid biosynthesis signaling pathway were significantly negatively related to RBC and Hb in both RM and RS groups but not in RA group. We found that 3-hydroxyisovaleric acid was positively associated with RBC and Hb, and the result is in line with the previous finding that 3-hydroxyisovaleric acid was increased in the plasma of mild and severe COVID-19 patients as revealed by gas chromatography-mass spectrometry (Páez-Franco et al., 2021). All these results suggest that RBC at the metabolomic level functionally does not restore to normal in COVID-19 survivors, and the mechanism needs to be further studied in the future. This evidence implies that RC survivors still have to be examined in future studies for any multiorgan damage, and targeting these abnormal pathways could help patients recover from SARS-CoV2 infection.

4. Conclusion and limitations

In this paper, we optimized a nanopore membrane-based microfluidic chip for plasma sEVs separation. We made a comprehensive comparison in plasma metabolomic analysis between ExoSEC and UC or REG methods. The results indicated that ExoSEC achieved higher sEVs yield and purity and identified more sEVs metabolites but cost less than UC or REG, which provides a practical and economical alternative to sEVs metabolomic study. Notably, we found a significant metabolic alteration in plasma sEVs from COVID-19 survivors using the ExoSEC. It has two major limitations in this study. First, we did not directly compare the metabolomic features of plasma vs plasma ExoSEC-sEVs and failed to detect plasma sEVs metabolites in the acute phase of COVID-19 patients. Second, we did not directly compare the sEVs isolation effects of ExoTIC and ExoSEC in this study. For our future work, the downstream multi-omics validation and application in other diseases of ExoSEC need further investigation.

CRediT author statement

Yang Jin and Bi-Feng Liu designed the study and were responsible for the overall study. Daquan Meng, Yanling Ma, Sufei Wang, Fen Wang, Jianchu Zhang, and Limin Duan collected the clinical data and blood samples. Wenjing Xiao, Hui Xia, Qi Tan, Qi Huang, Peng Chen, Kaimin Mao, Xueyun Tan, and Xie Han performed experiments. Wenjing Xiao, Ping Luo, Qi Huang, and Yice Sun conducted the metabolite and bioinformatic analysis. Qi Huang organized and edited the manuscript, and Wenjing Xiao, Zilin Zhao, and Hui Xia revised the manuscript. All authors participated in the discussion of the initial draft and gave constructive suggestions for revision. Yang Jin and Bi-Feng Liu revised the final manuscript. All authors reviewed and approved the final version.

Funding

This work was supported by the National Natural Science Foundation of China (81900095, 82070099, 22074047, 21775049); Major Projects of the National Science and Technology (2019ZX09301001); Ministry of Science and Technology of the People's Republic of China (CN) under Grant (2020YFC0844300); National Science and Technology Major Project of the Ministry of Science and Technology of China (2022YFF1203300); the National Key Research and Development Program of China (2021YFA1101500); the Natural Science Foundation of Hubei Province, China (2020CFB809); the Fundamental Research Funds for the Central Universities, HUST (2020kfyXGYJ011, 2020kfyXGYJ103); Tongji-Rongcheng Centre for biomedicine, Huazhong University of Science and Technology.

Declaration of competing interest

All the authors declare that they have no known conflict of interest in this paper.

Acknowledgements

The authors thank all the participants, including COVID-19 survivors and HC.

Footnotes

Appendix A

Supplementary data to this article can be found online at https://doi.org/10.1016/j.bios.2023.115152.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

Multimedia component 1
mmc1.docx (3.2MB, docx)

Fig. S1.

Fig. S1

Open in a new tab

Fig. S2.

Fig. S2

Open in a new tab

Fig. S3.

Fig. S3

Open in a new tab

Fig. S4.

Fig. S4

Open in a new tab

Fig. S5.

Fig. S5

Open in a new tab

Fig. S6.

Fig. S6

Open in a new tab

Fig. S7.

Fig. S7

Open in a new tab

Fig. S8.

Fig. S8

Open in a new tab

Fig. S9.

Fig. S9

Open in a new tab

Data availability

The authors are unable or have chosen not to specify which data has been used.

References

  1. Alzahrani F.A., Shait Mohammed M.R., Alkarim S., et al. Untargeted metabolic profiling of extracellular vesicles of SARS-CoV-2-infected patients shows presence of potent anti-inflammatory metabolites. Int. J. Mol. Sci. 2021;22(19) doi: 10.3390/ijms221910467. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ayres J.S. A metabolic handbook for the COVID-19 pandemic. Nat Metab. 2020;2(7):572–585. doi: 10.1038/s42255-020-0237-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Barberis E., Vanella V.V., Falasca M., et al. Circulating exosomes are strongly involved in SARS-CoV-2 infection. Front. Mol. Biosci. 2021;8 doi: 10.3389/fmolb.2021.632290. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Cocozza F., Névo N., Piovesana E., et al. Extracellular vesicles containing ACE2 efficiently prevent infection by SARS-CoV-2 Spike protein-containing virus. J. Extracell. Vesicles. 2020;10(2) doi: 10.1002/jev2.12050. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Dong L., Zieren R.C., Horie K., et al. Comprehensive evaluation of methods for small extracellular vesicles separation from human plasma, urine and cell culture medium. J. Extracell. Vesicles. 2020;10(2) doi: 10.1002/jev2.12044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Gao Y., Qin Y., Wan C., et al. Small extracellular vesicles: a novel avenue for cancer management. Front. Oncol. 2021;11 doi: 10.3389/fonc.2021.638357. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Guo S.C., Tao S.C., Dawn H. Microfluidics-based on-a-chip systems for isolating and analysing extracellular vesicles. J. Extracell. Vesicles. 2018;7(1) doi: 10.1080/20013078.2018.1508271. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Janouskova O., Herma R., Semeradtova A., et al. Conventional and nonconventional sources of exosomes-isolation methods and influence on their downstream biomedical application. Front. Mol. Biosci. 2022;9 doi: 10.3389/fmolb.2022.846650. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Jia H., Liu C., Li D., et al. Metabolomic analyses reveal new stage-specific features of COVID-19. Eur. Respir. J. 2022;59(2) doi: 10.1183/13993003.00284-2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Liu F., Vermesh O., Mani V., et al. The exosome total isolation chip. ACS Nano. 2017;11(11):10712–10723. doi: 10.1021/acsnano.7b04878. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Luo P., Mao K., Xu J., et al. Metabolic characteristics of large and small extracellular vesicles from pleural effusion reveal biomarker candidates for the diagnosis of tuberculosis and malignancy. J. Extracell. Vesicles. 2020;9(1) doi: 10.1080/20013078.2020.1790158. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Luporini R.L., Pott-Junior H., Di Medeiros Leal M.C.B., et al. Phenylalanine and COVID-19: tracking disease severity markers. Int. Immunopharm. 2021;101(Pt A) doi: 10.1016/j.intimp.2021.108313. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Mussap M., Fanos V. Could metabolomics drive the fate of COVID-19 pandemic? A narrative review on lights and shadows. Clin. Chem. Lab. Med. 2021;59(12):1891–1905. doi: 10.1515/cclm-2021-0414. [DOI] [PubMed] [Google Scholar]
  14. Páez-Franco J.C., Torres-Ruiz J., Sosa-Hernández V.A., et al. Metabolomics analysis reveals a modified amino acid metabolism that correlates with altered oxygen homeostasis in COVID-19 patients. Sci. Rep. 2021;11(1):6350. doi: 10.1038/s41598-021-85788-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Puhka M., Takatalo M., Nordberg M.E., et al. Metabolomic profiling of extracellular vesicles and alternative normalization methods reveal enriched metabolites and strategies to study prostate cancer-related changes. Theranostics. 2017;7(16):3824–3841. doi: 10.7150/thno.19890. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Song J.W., Lam S.M., Fan X., et al. Omics-driven systems interrogation of metabolic dysregulation in COVID-19 pathogenesis. Cell Metabol. 2020;32(2):188–202.e185. doi: 10.1016/j.cmet.2020.06.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Taneri P.E., Gómez-Ochoa S.A., Llanaj E., et al. Anemia and iron metabolism in COVID-19: a systematic review and meta-analysis. Eur. J. Epidemiol. 2020;35(8):763–773. doi: 10.1007/s10654-020-00678-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Tanner J.E., Alfieri C. The fatty acid lipid metabolism nexus in COVID-19. Viruses. 2021;13(1) doi: 10.3390/v13010090. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Théry C., Witwer K.W., Aikawa E., et al. Minimal information for studies of extracellular vesicles 2018 (MISEV2018): a position statement of the International Society for Extracellular Vesicles and update of the MISEV2014 guidelines. J. Extracell. Vesicles. 2018;7(1) doi: 10.1080/20013078.2018.1535750. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Thomas T., Stefanoni D., Dzieciatkowska M., et al. Evidence of structural protein damage and membrane lipid remodeling in red blood cells from COVID-19 patients. J. Proteome Res. 2020;19(11):4455–4469. doi: 10.1021/acs.jproteome.0c00606. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Wang J., Chen P., Dong Y., et al. Designer exosomes enabling tumor targeted efficient chemo/gene/photothermal therapy. Biomaterials. 2021;276 doi: 10.1016/j.biomaterials.2021.121056. [DOI] [PubMed] [Google Scholar]
  22. Wang J., Li W., Lu Z., et al. The use of RGD-engineered exosomes for enhanced targeting ability and synergistic therapy toward angiogenesis. Nanoscale. 2017;9(40):15598–15605. doi: 10.1039/c7nr04425a. [DOI] [PubMed] [Google Scholar]
  23. Wang J., Li W., Zhang L., et al. Chemically edited exosomes with dual ligand purified by microfluidic device for active targeted drug delivery to tumor cells. ACS Appl. Mater. Interfaces. 2017;9(33):27441–27452. doi: 10.1021/acsami.7b06464. [DOI] [PubMed] [Google Scholar]
  24. Wang J., Ma P., Kim D.H., et al. Towards microfluidic-based exosome isolation and detection for tumor therapy. Nano Today. 2021;37 doi: 10.1016/j.nantod.2020.101066. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Wei R., Zhao L., Kong G., et al. Combination of size-exclusion chromatography and ultracentrifugation improves the proteomic profiling of plasma-derived small extracellular vesicles. Biol. Proced. Online. 2020;22:12. doi: 10.1186/s12575-020-00125-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Williams C., Palviainen M., Reichardt N.C., et al. Metabolomics applied to the study of extracellular vesicles. Metabolites. 2019;9(11) doi: 10.3390/metabo9110276. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Wu J., Zhao M., Li C., et al. The SARS-CoV-2 induced targeted amino acid profiling in patients at hospitalized and convalescent stage. Biosci. Rep. 2021;41(3) doi: 10.1042/BSR20204201. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Wu Q., Zhou L., Sun X., et al. Altered lipid metabolism in recovered SARS patients twelve years after infection. Sci. Rep. 2017;7(1):9110. doi: 10.1038/s41598-017-09536-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Xu J., Zhou M., Luo P., et al. Plasma metabolomic profiling of patients recovered from coronavirus disease 2019 (COVID-19) with pulmonary sequelae 3 Months after discharge. Clin. Infect. Dis. 2021;73(12):2228–2239. doi: 10.1093/cid/ciab147. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Zhang S., Luo P., Xu J., et al. Plasma metabolomic profiles in recovered COVID-19 patients without previous underlying diseases 3 Months after discharge. J. Inflamm. Res. 2021;14:4485–4501. doi: 10.2147/JIR.S325853. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Zhou M., Tan X., Luo P., et al. Changes in glomerular filtration rate and metabolomic differences in severely ill coronavirus disease survivors 3 months after discharge. Biochim. Biophys. Acta, Mol. Basis Dis. 2022;1868(1) doi: 10.1016/j.bbadis.2021.166289. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Multimedia component 1
mmc1.docx (3.2MB, docx)

Data Availability Statement

The authors are unable or have chosen not to specify which data has been used.

Articles from Biosensors & Bioelectronics are provided here courtesy of Elsevier

RESOURCES