HDL proteome and apolipoproteins concentrations in severe ICU COVID-19 patients

doi:10.21203/rs.3.rs-4978039/v1

Download PDF

Research Article

HDL proteome and apolipoproteins concentrations in severe ICU COVID-19 patients

https://doi.org/10.21203/rs.3.rs-4978039/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background: SARS-CoV-2 infection affects both lipid metabolism and lung function. The severity of the disease has been associated with reduced levels of both high-density lipoprotein (HDL) and low-density lipoprotein cholesterol. Despite the crucial role that these nanoparticles play in SARS-CoV-2 infection, few studies have examined their structure during COVID-19 beyond HDL quantity.

The study aimed to assess apolipoprotein levels in COVID-19 patients who either survived or died following ICU admission. In addition, ICU survivors and non-survivors were compared for HDL particle size and proteome.

Methods: Between February and April 2020, our study enrolled 37 COVID-19 patients upon their intensive care unit admission. Among them, 18 survived the disease, while 19 succumbed to it. We used mass spectrometry to assess plasma levels of 14 apolipoproteins and LCAT. Additionally, we analyzed HDL subpopulation distribution by utilizing native polyacrylamide gel electrophoresis. HDL particles were isolated from both surviving and non-surviving patients using ultracentrifugation, followed by characterization of their proteomes with NanoLC-MS/MS.

Results: Plasma apolipoproteins, including Apo A-II, Apo Cs (I, II, III), Apo H, Apo J, Apo M, and LCAT, were decreased in patients who did not survive COVID-19. However, no alterations were noted in the distribution of HDL subpopulations in relation to mortality. HDL composition was further altered based on mortality, displaying a decline in Apo H and paraoxonase 3.

Conclusion: In conclusion, we have shown an alteration in plasma apolipoproteins and HDL composition between surviving COVID-19 patients and non-survivors. Some markers, such as Apo H, are more predictive than baseline lipid concentrations such as HDL-C.

COVID-19

mortality

HDL particle

apolipoproteome

apolipoprotein

proteomic

Intensive care unit

Coronavirus disease 2019 (COVID-19) pandemic, which started in Wuhan, China in December 2019, is caused by a Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) infection and has caused more than 6 million deaths worldwide [1, 2]. COVID-19 presents a broad range of symptoms and gravity, varying from ambulatory mild disease (including asymptomatic forms) to severe cases, that may result in systemic failure and death [3, 4]. Severe forms of COVID-19 are associated with a hyperinflammatory response marked by a cytokine storm resulting in excessive production of pro-inflammatory cytokines like monocyte chemotactic protein-3 (MCP-3), interleukin − 1 (IL-1), and interferon α/β (IFNα/β). These three cytokines are closely associated with the severe form of COVID-19, although the cytokine profile of COVID-19 patients varies [5–7].

Lipoproteins are macromolecular complexes composed of apolipoproteins and lipids. Apolipoproteins regulate the lipoprotein structure, lipid transport, enzyme co-activation, and provide anti-inflammatory and antioxidant effects [8]. The apolipoproteins are classified into subfamilies, including Apo A, B, C, D, E, L, F, H, M, and (a). These proteins are distributed across various lipoproteins such as low-density lipoproteins (LDL) and high-density lipoproteins (HDL). [9, 10].

During a SARS-CoV-2 infection, the concentrations of total cholesterol, LDL-cholesterol (LDL-C), and HDL-cholesterol (HDL-C) are reduced and are linked to the severity of the disease. [11–14]. Interestingly, numerous studies have also shown that lipid and lipoprotein metabolism are disrupted during the acute phase of SARS-CoV-2 infection, along with other illnesses [15].

Some apolipoproteins, including Apo A-I and Apo B100, can be dosed routinely like HDL-C, LDL-C, TC. Moreover, Apo A-I, A-II, and B are present in low levels in COVID-19 patients and are linked to an increased likelihood of developing severe forms of the disease [16]. In a previous study, our team found lower plasmatic concentrations of nine apolipoproteins in COVID-19 patients compared to controls [13].

Beyond quantitative changes, few studies have focused on the structure and function of HDL during COVID-19. Nonetheless, HDL possesses pleiotropic properties, including antioxidant and anti-inflammatory effects, in addition to its primary role in reverse cholesterol transport [11, 15, 17–19]. In a recent study of our team comparing COVID-19 patients versus controls, with an enrichment of acute phase proteins and a depletion of apolipoproteins, HDL particles proteome was profoundly altered in COVID-19 patients [20]. Moreover, in our in vitro analyses, we underlined that HDL from COVID-19 patients was much less endothelioprotective [20].

Whereas previous studies mostly compared COVID-19 patients with controls, the aim of the present work was to compare apolipoprotein concentrations of deceased patients with those of patients who survived an episode of COVID-19 leading to severe form of the disease. The second objective of this work was to characterize HDL particles (size and proteome) at admission to ICU between deceased patients and survivors.

2.1. Study population

We conducted an analysis with data from the prospective French COVID cohort (NCT04262921) [21]. The French COVID cohort includes hospitalized patients with a PCR-confirmed COVID‐19 and who gave written consent. The study was sponsored by INSERM, conducted in accordance with the Declaration of Helsinki, and Ethics approval was given by the French Ethics Committee CPP Ile‐de‐France VI (ID RCB: 2020‐A00256‐33).

We selected adult patients from the French COVID cohort study, from February 1, 2020 until July 31, 2020 (corresponding to the first wave of pandemic in France). Pregnant women were excluded. None of the patients received corticosteroids. In this study, patients selected from the French COVID cohort had to meet the following criteria: - admitted in ICU; - at least 1 ml of plasmas for lipoprotein isolations and analyze; - not be on long-term statin therapy.

2.2. Quantification of lipoprotein subpopulation by tube gel electrophoresis (Quantimetrix Lipoprint)

Plasma HDL subpopulations were determined by polyacrylamide gel electrophoresis (PAGE) using Quantimetrix Lipoprint HDL (Quantimetrix Corporation, Redondo Beach, CA, USA). Lipoprint was performed and analyzed according to the manufacturer's protocol. Briefly, 25 µL EDTA plasma was mixed with 300 µL loading gel, and this sample gel was applied to the Lipoprint HDL gel tube before photopolymerization for 35 minutes at RT. Gel tube electrophoresis was performed at 3 mA/gel tube until albumin reached 1.5 cm at the end of the gel tube (± 50 min). After resting in the dark for 30 min, the tubes were scanned and analyzed using Lipoware HDL software. After tube analysis, three main subpopulations were determined from the 10 subpopulations measured: large HDL (HDL-1 to HDL-3), intermediate HDL (HDL-4 to HDL-7) and large HDL (HDL-8 and HDL-10).

2.3. Lipoprotein isolation from plasma

All subsequent steps were performed under a Class II microbial safety bench. Lipoproteins were isolated from EDTA plasma by three-step ultracentrifugation. All ultracentrifugation steps were performed at 252,000 g for 20 hours at 10°C using ultracentrifuge tubes placed in a 50.4 Ti rotor in a Beckman Coulter Optima L-80 XP ultracentrifuge. For the first ultracentrifugation step, plasma was weighed to a density of 1.019 using a potassium bromide (KBr) solution with a density of 1.35. The weighted plasma was covered with a 1.019 density KBr solution in ultracentrifugation tubes. After ultracentrifugation, the upper fraction corresponding to VLDL was collected. For the second ultracentrifugation step, plasma was collected and weighted to a density of 1.063 using a KBr solution with a density of 1.35. The weighted plasma was covered with a 1.063 density KBr solution in ultracentrifuge tubes. After ultracentrifugation, the upper fraction corresponding to LDL was collected. For the final ultracentrifugation step, plasma was weighted to a density of 1.21 using a KBr solution with a density of 1.21. The weighted plasma was covered with a 1.21 density KBr solution in ultracentrifugation tubes. After ultracentrifugation, the upper fraction corresponding to HDL was collected. After isolation, lipoproteins were washed five times with saline EN buffer (0.15 M NaCl, 1 mM EDTA, 0.025% NaN3, pH 7.3) using a centrifugal filtration tube (Amicon Ultra, 3 kDa cut-off, UFC500396). Tubes were centrifuged at 14,000g for 15 minutes at 4°C, and the volume was concentrated to 100 µL. Lipoprotein proteins were then assayed by the bicinchoninic acid method (BCA, Sigma), and HDL purity was verified by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) with Coomassie blue staining.

2.4. Fast protein Liquid Chromatography (FPLC)

HDL isolated from each patient was purified in an ÄKTA Pure FPLC system (GE Healthcare) on a SuperdexTM 200 Increase 10/300 GL column (10 x 300 mm; Cytiva) in EN saline buffer. HDL was loaded into a 500 µL loop and injected at a flow rate of 0.250 mL/min. Elution was then performed with Saline EN buffer at a constant flow rate of 0.300 mL/min, and fractions were collected every 500 µL. Finally, the column was washed with two column volumes at a flow rate of 0.700 mL/min.

2.5. Label-free Quantification by Nano LC-MS/MS analysis

Label-free protein quantification was performed on HDL at the time of ICU admission from patients who survived their COVID-19 (n = 14) and on HDL from patients who did not survive their severe COVID-19 (n = 15).

2.5.1. HDL sample preparation and proteolytic digestion

The concentration of total HDL protein was measured using the BCA assay, with bovine serum albumin serving as the standard. As previously described [22], 5 µg HDL was combined with 25 µL digestion buffer (6 M urea, 50 mM ammonium bicarbonate) and denatured at 80°C for 10 minutes for sample preparation. Dithiothreitol was used for sample reduction, followed by iodoacetamide alkylation. Upon alkylation, the samples were diluted with four volumes of 50 mM ammonium bicarbonate to reduce the urea concentration. The HDLs were digested with 3 µg of ditosylphenylalanyl chloromethyl ketone trypsin (TPCK-treated Sigma) overnight at 37°C. Tryptic digestion was halted with 2% TFA and incubated at 4°C for 15 minutes. Subsequently, samples were centrifuged at 14,000 g for 15 minutes at 10°C and the desalted peptides in the supernatant were collected on peptide desalting spin columns (Thermo Fisher Scientific Pierce, part number 2162704). Peptides were eluted using a solution of 50% acetonitrile (ACN) and 0.1% trifluoroacetic acid (TFA), followed by drying with a Speed Vacuum. Then, 20 µL of 40% ACN with 0.1% TFA was used to elute the peptide, which was subsequently vacuum-dried at 60°C prior to mass spectrometric analysis.

2.5.2. Nano LC-MS/MS method

The method was carried out with two solvents, solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in 80% ACN). Digested peptides were solubilized in 20 µL of 4% ACN with 0.1% TFA. Mass spectrometry analysis was performed by nano-LC on a Thermo Fisher Ultimate 3000 Series NCS-3500 RS coupled to a nanospray ionization (NSI)-Q-Orbitrap mass spectrometer (Q Exactive Plus, Thermo Fisher Scientific, Bremen, Germany) [20]. Briefly, 5 µL of samples were injected into a C18 LC-EASY-spray column (2 µm, 100 Å, 75 µm x 50 cm, Thermo Fisher Scientific). Peptides were eluted on two linear elution gradients starting from 4–25% solution B (0.1% formic acid in 80% ACN) until 103 min, and then from 25–40% solution for 20 min. The column was then washed for 5 min with 90% solution B and equilibrated with 4% solution B for 20 min. Solution A was composed of 0.1% formic acid in water. To minimize contamination between samples, two blank runs containing solution A were conducted after every sample run. The flow rate was maintained at 0.300 µL/min, and the column temperature was maintained at 40°C. The NSI was set at 1.8 kV, the capillary temperature was set at 275°C, and the S lens RF level was set at 30% for mass spectrometry parameters.

The mass spectra were obtained in positive ion mode using data-dependent acquisition within the m/z range of 350 to 2,000. Full scan mass spectra were acquired at a resolving power of 70,000 (at m/z 400) using automatic gain control (AGC) at a set value of 5e5 and a maximum injection time (Max IT) of 100 ms. MS/MS spectra of the top ten most intense precursor ions were acquired at a resolving power of 35,000 (at m/z 400) with an AGC set to 1e5 and a maximum IT of 100 ms. Peptide fragmentation was performed by setting the normalized collision energy to 28. Precursor ions were excluded dynamically for 60 seconds to prevent peptide selection repetition. The Orbitrap's performance was assessed weekly, and the mass spectrometer underwent external calibration before analysis using the LTQ ESI positive ion calibration solution (Pierce™). XCalibur 4.2.47 software (Thermo Fisher Scientific) was used to collect the data.

2.5.3. Protein identification and quantification

Protein identification and quantification were performed automatically from the raw files with Proteome discoverer software (version 2.2.2.0). MS and MS/MS spectra were compared to the Uniprot Human reference proteome database containing only canonical sequences (26 559 sequences; 01 July 2021) using the SEQUEST HT search engine. The following features were defined in SEQUEST HT for identification: variable modifications are methionine oxidation and N-terminal acetylation of proteins, then, a fixed modification sets the carbamidomethylation of proteins and the specific enzyme selected is trypsin with two miscleavage allowed. A mass tolerance of 10 ppm for precursor ions and 0.02 Da for-product ions was used in SEQUEST HT. A decoy database search strategy with target inversion was applied with a false discovery rate (FDR) set at 1% for proteins and peptides. Then, only proteins with at least one unique peptide with a Sequest Score (Xcorr) ≥ 1.5 were used and for Xcorr < 1.5, manual verification of peptide identification was performed on the MS/MS spectrum. Identified peptides were quantified by label-free using the peak intensity of a validated peptide (unique and razor) for a given protein. After quantification, all proteins with more than 60% missing value were not considered in the analysis. For proteins with ≤ 60% missing values, they were replaced by the IQR of the corresponding proteins in the respective groups. The list of identified and quantified proteins and peptides is available in the supplemental table.

2.5.4. Data processing

Proteome Discoverer (version 2.2.2.2.0) was utilized for the computations. Protein abundance refers to the mean of the protein intensities observed in all samples among each group (COVID-19 survivors versus non-survivors). Protein intensity was defined as the combination of protein and peptide intensities measured for each individual. All protein intensity values were subjected to logarithmic transformation (base 2). The log2 (Fold Change) is determined by taking the base 2 logarithm of the disparity between the average intensities of a specific protein in COVID-19 patients who did not survive and the average intensities of the same protein in COVID-19 patients who did survive.

2.6. Statistical analysis

The Shapiro-Wilk test was used to test the normality of the data distributions. Mann-Whitney test or unpaired t-test was then performed, as appropriate, to compare patients who survived COVID-19 with those who did not. A volcano plot was generated by Prism to evaluate differential proteins on the basis of two indicators: the -log10 p-value of the Mann-Whitney test and the log2 (protein abundance fold change of non-survivors and survivors). Marker performance was measured using receiver operating characteristic (ROC) curves. A p-value < 0.05 is considered to indicate a statistically significant result. Statistical analyses were performed using Prism software (GraphPad Software Inc., San Diego, CA, USA).

3.1 Study population

37 patients were included in this analysis. Survivors (n = 18) and non-survivors (n = 19) were matched based on age and gender. General characteristics of the patient population are presented in Table 1 and a flow chart is expressed on supplemental figure S1.

Table 1

Characteristics and outcomes of the study population
Characterisitics	Study population (n = 37)	Survivors (INC) (n = 18)	Non-survivors (n = 19)	Survivors (INC) vs Non-survivors
Characterisitics	Study population (n = 37)	Survivors (INC) (n = 18)	Non-survivors (n = 19)	p-values
Age, years, median [IQR]	61 [54; 68]	59 [51; 65]	61 [57; 68]	0.3047 (ns)
Male sex, n (%)	37 (89%)	16 (89%)	17 (89%)	> 0.9999 (ns)
Obesity	13 (35%)	6 (33%)	7 (37%)	> 0.9999 (ns)
BMI, kg/m2, median [IQR]	28 [24; 31]	26 [24; 30]	29 [25; 33]	0.3677 (ns)
High Blood pressure, n (%)	15 (41%)	5 (28%)	10 (53%)	0.1837 (ns)
Diabetes mellitus, n (%)	13 (35%)	5 (28%)	8 (42%)	0.4951 (ns)
Timing between fist symptoms and hospitalization (days), median [IQR]	8 [5; 9]	8 [5; 11]	7 [5; 9]	0.6781 (ns)
Medication
Statin, n (%)	0	0	0	> 0.9999 (ns)
Corticosteroides, n (%)	0	0	0	> 0.9999 (ns)
Outcomes
Time between hospitalization and death (days), median [IQR]	13 [8; 27]	N.A.	13 [8; 27]	N.A.
Clinical features
TC, mmol/L, median [IQR]	3.45 [3.09; 3.91]	3.28 [2.81; 3.62]	3.62 [3.19; 4.24]	0.0426 (*)
HDL-C, mmol/L, median [IQR]	0.49 [0.40; 0.58]	0.50 [0.43; 0.58]	0.48 [0.40; 0.63]	0.8381 (ns)
LDL-C, mmol/L, median [IQR]	1.64 [1.37; 2.15]	1.98 [1.51; 2.26]	1.46 [1.34; 2.06]	0.1100 (ns)
TG, mmol/L, median [IQR]	1.74 [1.51; 2.06]	1.71 [1.52; 2.17]	1,75 [1.46; 1.95]	0.0554 (ns)
BMI: body mass index; TC: total cholesterol; HDL-C: high-density lipoprotein-cholesterol; LDL-C: low-density lipoprotein-cholesterol; TG: triglycerides

3.2 Comparison of plasma apolipoprotein concentrations between COVID-19 survivors and non-survivors

We used mass spectrometry to measure 14 apolipoproteins and LCAT directly from the plasma of COVID-19 survivors and non-survivors. The following apolipoproteins were measured: Apo(a), Apo As (A-I, A-II, A-IV), Apo B100, Apo Cs (C-I, C-II, C-III), Apo D, Apo E, Apo H, Apo J, Apo L1, and Apo M. The concentration of seven apolipoproteins was significantly reduced in patients who did not survive compared with those who did. These apolipoproteins include Apo A-II (4.68 ± 3.76 vs. 2.30 ± 1.51; p = 0.0147), Apo Cs (C-I (2.29 ± 1.71 vs. 0.87 ± 0.62; p = 0.0060), C-II (1.64 ± 0.99 vs. 1.00 ± 0.58; p = 0.0269) and C-III (0.21 ± 0.18 vs. 0.09 ± 0.08; p = 0.0045), Apo H (11.45 ± 3.45 vs. 9.28 ± 2.52; p = 0.0351), Apo J (13.55 ± 3.99 vs. 11.03 ± 2.81; p = 0.0318) and Apo M (2.90 ± 1.09 vs. 2.12 ± 0.63; p = 0.0112). Furthermore, the concentration of LCAT (p = 0.0137) is also significantly diminished in patients who did not survive COVID-19. Results are expressed in Fig. 1.

3.3 Modulation of HDL subpopulations size between survivors and non-survivors COVID-19 patients

The different HDL subpopulations size (large – intermediate – small) of COVID-19 patients were determined using the Lipoprint HDL tube gel electrophoresis technique. No difference was found between groups (Fig. 2).

3.4 HDL proteome analysis of surviving versus non-surviving patients of their COVID-19

HDL was isolated and prepared for shotgun proteomics analysis from 14 COVID-19 survivor patients and 15 non-survivor COVID-19 patients. Digested peptides obtained from 5 µg of HDL were prepared and analyzed using mass spectrometry combined with NanoLC. A first analysis of the MS/MS spectra by Proteome Discoverer software from the SEQUEST HT database identified 98 proteins that may be associated with HDL from either surviving or non-surviving patients (see Supplemental Table S2). After verification, our inclusion criteria (see Materials and Methods) indicate that 72 proteins are actually associated with patient HDL. To compare COVID-19 survivor and non-survivor groups, we conducted a fold change analysis of intensities associated with HDL proteins prior to log base-2 transformation. Out of the 72 proteins identified, the relative abundance of 11 proteins was lower in non-survivor HDL as compared to control HDL. These proteins included alpha-1 antitrypsin (AAT), cathelicidin antimicrobial peptide, complement C4 (A, B), serotransferrin, apolipoprotein C-I, heat shock protein HSP90-beta, protein-glutamine gamma-glutamyltransferase E, centrosomal protein, and vitronectin. In contrast, HDL from non-survivors showed an increased abundance of nine proteins, including serum amyloid A-2, CD99 antigen, mitochondria-associated proteins, and cellular debris (Fig. 2).

The intensities of the 72 proteins associated with HDL in the patients were logarithmically transformed to base 2 to facilitate statistical analyses. The results indicate that HDL protein profile displays more significant alterations in non-surviving COVID-19 patients when compared to those who survived. Specifically, there were significant differences in the levels of 24 proteins in the HDL of non-survivors relative to COVID-19 survivors (Fig. 4.A). The levels of 19 of these proteins were significantly lower in the HDL of the non-survivors. The mentioned proteins comprise ribosomal protein 60S L7a (RL7A), adenosylhomocysteinase (ALDH1L1), alpha-1-antitrypsin (AAT), alpha-2-antiplasmin (SERPINF2), aminopeptidase N (ANPEP), apolipoprotein (a) (LPA), apolipoprotein H (APOH), 135 kDa centrosomal protein (CEP135), complement C4-A (CO4A), complement component C9 (CO9), haptoglobin-related protein (HPTR), heat shock protein HSP 90-beta (HS90B), prenylcysteine oxidase 1 (PCYOX1), protein S100-A8, protein-glutamine gamma-glutamyltransferase E, paraoxonase serum/lactonase 3 (PON3) and vitronectin (VTNC). However, five proteins are significantly elevated in HDL from the non-surviving group. The proteins that have increased are ATP synthase subunit beta mitochondrial (ATP5E), beta-2-microglobulin (B2M), CD99 antigen (CD99), cytosolic 10-formyltetrahydrofolate dehydrogenase (ALDH1L1) and secretoglobin family member 1 3A (SCGB3A2) (Fig. 4.B).

3.5 Diagnostic performance via receiver characteristic (ROC) curve

ROC curves were used to assess the diagnostic accuracy of diverse lipid markers (Fig. 5). Except for TC (AUC = 0.71 ± 0.09, p = 0.0418), the regular lipid markers had areas under the curve (AUC) lower than 0.7 and a p-value over 0.05 (Fig. 5A-D). Plasma apolipoproteins and LCAT provide a AUC score of greater than 0.7 for all except Apo H (AUC = 0.68 ± 0.09, p = 0.0056) and Apo J (AUC = 0.69 ± 0.09, p = 0.0483), which are unreliable indicators of patient outcomes (Fig. 5E-L). Among the potential markers identified in HDL particles from COVID-19 patients are Apo H (AUC = 0.94 ± 0.06, p < 0.0001) and the 135 kDa CEP135 protein (AUC = 0.91 ± 0.05, p = 0.0001), all of which have shown significant effectiveness in predicting survival, with an AUC greater than 0.90. When using patient HDL proteins as markers, ATP5E, CO4A, PON3 and ITIH4 demonstrate effectiveness in predicting patient survival, while CO9 display lesser effectiveness with a lower AUC (Fig. 5M-T).

In this present study, we have therefore shown that while lipid profiles on admission to ICU for COVID-19 cannot be used to stratify patients who died from those who survived, plasma analysis of certain apolipoproteins and proteomic analysis of HDL seem to be able to do so. Especially, APOH and CEP135 are potentially novel biomarkers for COVID-19 severity linked to HDL proteins.

Lipid metabolism plays a key role in SARS-CoV-2 infection. Several hypotheses have been proposed to explain the changes in lipid metabolism, including viral use and modification of lipids to support replication, hemodilution, and increased vascular permeability. [23, 24]. Lowering cholesterol levels in lipoproteins is also aided by the cytokine storm [25]. However, while in bacterial sepsis HDL-C concentrations are associated with patient outcome and in particular mortality, these links were not found in COVID-19

[13, 14, 20, 26–30]. In this present study, we confirmed that HDL-C, LDL-C, and TG concentrations are similar between survivors and non-survivors COVID-19 patients. Similarly, Sun et al. reported no disparity in HDL-C, LDL-C, and TC levels between surviving and non-surviving patients with COVID-19, although TG levels were higher in non-surviving patients.

We have also shown that a crude concentration of HDL-C does not appear to be sufficient to stratify patients during COVID-19. Thus, the first step was to characterize the plasma apolipoproteins which are key regulators of lipid metabolism and have anti-inflammatory, antioxidant and endothelial protective properties [9, 31]. The mechanisms between apolipoproteins and the severity of COVID-19 are not yet fully understood. However, multiple studies have demonstrated a negative correlation between Apo A-I and Apo B levels and the severity of COVID-19 [16, 32, 33]. In our previous studies, we demonstrated that COVID-19 patients exhibit diminished levels of nine Apo forms, such as Apo A-II, C (I, II), H, J, M, as well as LCAT, alongside elevated levels of Apo E as compared to controls [13]. Our study confirms the previous one and highlights several apolipoproteins that are significantly more reduced on admission to ICU in patients who will die secondarily, such as Apo A-II, Apo Cs (C-I, C-II and C-III), Apo H, Apo J and Apo M. Most of these apolipoproteins are linked to reduced macrophage function [34]. Cholesterol efflux by macrophages induces production of Apo C through Liver X Receptor (LXR) activation [35]. Reduced Apo C concentrations indicate a decrease in macrophage cholesterol efflux function.

Interestingly, we observed a decreased in Apo A-II but not in Apo A-I concentrations among non-survivors of COVID-19 when compared to survivors. This observation differs from most of the studies reporting essentially a decrease in Apo A-I concentration with an association with worth outcome [32, 33]. This difference can be attributed to the sample size, which is significantly larger than ours. Moreover, corticosteroid therapy modifies lipidi profile. Our findings show a trend towards reduced Apo A-II concentrations. Our previous studies, however, observed apo A-II concentrations in both HDL and plasma of COVID-19 patients [13, 20]. Additionally, other research has also seen lower Apo A-II levels in COVID-19 patients [34, 36]. This decrease in Apo A-II suggests a potential alteration in cholesterol efflux, as well as a decrease in the anti-inflammatory and antioxidant properties of HDL [37, 38].

It is interesting to note that ApoB100 and LCAT were associated with mortality in our last study [13] but not in this work. The role of Apo H, also known as beta-2-glycoprotein 1, is not well understood. According to earlier research [39], it has been observed that Apo H levels decrease in COVID-19 patients. This decrease in Apo H could potentially create a prothrombotic environment in COVID-19, leading to antiphospholipid syndrome. This decrease in Apo H could potentially create a prothrombotic environment in COVID-19, due to an antiphospholipid syndrome. Furthermore, Apo H seems to capture the SARS-CoV-2 virus, leading to the hypothesis that the virus utilizes Apo H [40]. Clusterin, also known as Apo J, acts as an inhibitor of the complement system [41]. Reductions in the concentration of Apo J led to complement system activation, which then causes a prothrombotic environment [42, 43]. Our study indicates also that Apo M levels were decreased in COVID-19 patients who did not survive. A study conducted by Marfia et al [44] showed that the concentration of sphingosine-1-P (S1P) was closely related to the concentration of apo M and was able to accurately predict the prognosis of COVID-19 patients. Our findings reaffirm that Apo M can serve as a potential marker for assessing COVID-19 severity. Furthermore, Apo M depletion is associated with loss of anti-inflammatory and anti-atherogenic functions of S1P/ApoM complex [45, 46]. Therefore, deficiency of these apolipoproteins promotes an overall thromboinflammatory environment [47].

Our study revealed a decrease in 19 proteins and an increase in five proteins in the HDL of non-surviving COVID-19 patients. Among the proteins increased in the HDL of COVID-19 patients, the CD99 antigen plays a key role in the T-cell adhesion process. CD99 is mainly localized on T lymphocytes and endothelial cells. The level of this particular protein is lower in COVID-19 patients with severe symptoms, indicating reduced endothelial integrity and recruitment of monocytes, neutrophils and T cells [48]. A hypothesis to explain its increase in HDL from non-survivors of COVID-19 is that it is captured by the latter. HDL levels in non-survivors are found to be enriched with mitochondrial ATP synthase subunit beta, which interacts with the ecto-F1-ATPase receptor on the cell surface [49]. SARS-CoV-2 is able to interact with ecto-F1-ATPase and ATP synthase [50, 51]. An elevated beta-2-microglobulin has been identified as a potential biomarker of COVID-19 severity [52, 53], which may explain the elevated levels in the HDL of non-surviving COVID-19 patients. Interestingly, the HDL proteins of non-survivors of COVID-19 exhibit a decrease in AAT, I'TIH4, Complement C4A, and C9, which are normally increased in plasma as COVID-19 severity increases. This increase could be explained by an increase in the compounds and proteins of the acute phase complement and a decrease in the proteins involved in the coagulation cascade [54, 55]. Among the coagulation-related proteins, the levels of beta 2 glycoprotein 1 (Apo H) are reduced in the HDL of COVID-19 non-survivors, suggesting a disturbance in the coagulation cascades [34, 55]. The remodeling of high-density lipoprotein in patients during the COVID-19 pandemic suggests a loss of high-density lipoprotein (HDL) functionality [20].

Our study is the first to investigate LDL and HDL distribution changes during COVID-19 using tube gel electrophoresis (Lipoprint) among patients who survived or did not survive the disease. The distribution of lipoproteins was similar regardless of patient outcome. COVID-19 brings about a modification in the lipoprotein subpopulations' distribution. This change is characterized by a reduction in the concentrations of small, intermediate, and large subpopulations of HDL and large LDL, accompanied by an escalation in small and dense LDL concentrations. [56].

Finally, routine lipid markers in our study demonstrated poor accuracy in assessing disease severity. It is surprising to observe this in light of the association between HDL-C and disease severity [57]. In contrast, apolipoproteins have superior performance to routine lipid assays in assessing patient outcome. This is expected since various studies have shown that apolipoproteins are decreased in COVID-19 [13, 23, 32–34]. However, it seems that HDL-associated Apo H is a more accurate indicator of COVID-19 severity than total plasma Apo H. Nevertheless, these results merit further analysis with more patients included.

The primary obstacle we encountered was the small number of available samples, which hindered our ability to conduct experiments on a representative sample of the population under study. Nevertheless, our preliminary results are promising. This study is focused on a modest sample of 37 patients during the initial wave of COVID-19, providing supplementary information to the current knowledge on COVID-19. Another limitation is the low levels of HDL volume of patients included in our study. It would be pertinent to evaluate the lipoproteins' functionality to ascertain the effect of changes in their proteome. Further comparative studies carried out in various centers can help confirm and validate the conclusions presented in our research.

In conclusion, in a cohort of patients hospitalized in ICU for severe COVID-19, we demonstrated major differences in plasma apolipoproteins and in protein composition of HDL by comparing surviving and non-surviving patients. These markers, measured on admission to ICU, also appear to be more predictive of mortality than BASIC lipid concentrations such as HDL-C. Nevertheless, these results should be interpreted with caution given the small number of patients included in this study. In this context, it is important to conduct a high-powered multicenter study in the future.

Conflicts of interests

The authors declare that they have no conflict of interest to disclose.

Funding

FB is supported by a grant from the Région Réunion for their PhD (European Union-Reunion Region). This research was funded by FEDER "LIPICOR" RE0027600 and "TFORCE-COVIR" RE0027601 (EU—Région Réunion—French State national counterpart), and by the BIOST Fédération.

Author Contribution

OM and ST designed the study. FB, BV acquired the data. FB, OM, ST analyzed the data. FB, ST, OM wrote the manuscript. PR, OM, ST, AF and SJ critically reviewed the manuscript. All authors reviewed the manuscript.

Acknowledgement

The sponsor of the LIPICOR and T-FORCE COVIR studies were respectively Assistance Publique - Hopitaux de Paris (Délégation à la Recherche Clinique et à l'Innovation) and FEDER Région Réunion.

Li Q, Guan X, Wu P, et al. Early Transmission Dynamics in Wuhan, China, of Novel Coronavirus–Infected Pneumonia. N Engl J Med. 2020;382:1199–207. https://doi.org/10.1056/NEJMoa2001316.
Zhu N, Zhang D, Wang W, et al. A Novel Coronavirus from Patients with Pneumonia in China, 2019. N Engl J Med. 2020;382:727–33. https://doi.org/10.1056/NEJMoa2001017.
Zhang X, Tan Y, Ling Y, et al. Viral and host factors related to the clinical outcome of COVID-19. Nature. 2020;583:437–40. https://doi.org/10.1038/s41586-020-2355-0.
WHO Working Group on the Clinical Characterisation and Management of COVID-19 infection. A minimal common outcome measure set for COVID-19 clinical research. Lancet Infect Dis. 2020;20:e192–7. https://doi.org/10.1016/S1473-3099(20)30483-7.
Vaninov N. In the eye of the COVID-19 cytokine storm. Nat Rev Immunol. 2020;20:277–277. https://doi.org/10.1038/s41577-020-0305-6.
Hu B, Huang S, Yin L. The cytokine storm and COVID-19. J Med Virol. 2021;93:250–6. https://doi.org/10.1002/jmv.26232.
Yang Y, Shen C, Li J et al. (2020) Exuberant elevation of IP-10, MCP-3 and IL-1ra during SARS-CoV-2 infection is associated with disease severity and fatal outcom.
Jonas A, Phillips MC. Lipoprotein structure. Biochemistry of Lipids, Lipoproteins and Membranes. Elsevier; 2008. pp. 485–506.
Dominiczak MH, Caslake MJ. Apolipoproteins: metabolic role and clinical biochemistry applications. Ann Clin Biochem. 2011;48:498–515. https://doi.org/10.1258/acb.2011.011111.
Alaupovic P. (1996) [2] Significance of apolipoproteins for structure, function, and classification of plasma lipoproteins. In: Methods in Enzymology. Academic Press, pp 32–60.
(2020) Hypolipidemia is associated with the severity of COVID-19. J Clin Lipidol 14:297–304. https://doi.org/10.1016/j.jacl.2020.04.008
Almas T, Malik J, Alsubai AK, et al. Effect of COVID-19 on lipid profile parameters and its correlation with acute phase reactants: A single-center retrospective analysis. Annals Med Surg. 2022;78:103856. https://doi.org/10.1016/j.amsu.2022.103856.
Begue F, Chemello K, Veeren B, et al. Plasma Apolipoprotein Concentrations Are Highly Altered in Severe Intensive Care Unit COVID-19 Patients: Preliminary Results from the LIPICOR Cohort Study. Int J Mol Sci. 2023;24:4605. https://doi.org/10.3390/ijms24054605.
Tanaka S, Tymowski CD, Assadi M, et al. Lipoprotein concentrations over time in the intensive care unit COVID-19 patients: Results from the ApoCOVID study. PLoS ONE. 2020;15:e0239573. https://doi.org/10.1371/journal.pone.0239573.
Li Y, Zhang Y, Lu R, et al. Lipid metabolism changes in patients with severe COVID-19. Clin Chim Acta. 2021;517:66–73. https://doi.org/10.1016/j.cca.2021.02.011.
Ulloque-Badaracco JR, Hernandez-Bustamante EA, Herrera-Añazco P, Benites-Zapata VA. Prognostic value of apolipoproteins in COVID-19 patients: A systematic review and meta-analysis. Travel Med Infect Dis. 2021;44:102200. https://doi.org/10.1016/j.tmaid.2021.102200.
Tran-Dinh A, Diallo D, Delbosc S, et al. HDL and endothelial protection. Br J Pharmacol. 2013;169:493–511. https://doi.org/10.1111/bph.12174.
Wang G, Zhang Q, Zhao X, et al. Low high-density lipoprotein level is correlated with the severity of COVID-19 patients: an observational study. Lipids Health Dis. 2020;19:1–7. https://doi.org/10.1186/s12944-020-01382-9.
Tanaka S, Couret D, Tran-Dinh A, et al. High-density lipoproteins during sepsis: from bench to bedside. Crit Care. 2020;24:134. https://doi.org/10.1186/s13054-020-02860-3.
Begue F, Tanaka S, Mouktadi Z, et al. Altered high-density lipoprotein composition and functions during severe COVID-19. Sci Rep. 2021;11:2291. https://doi.org/10.1038/s41598-021-81638-1.
Yazdanpanah Y, French COVID, cohort investigators and study group. Impact on disease mortality of clinical, biological, and virological characteristics at hospital admission and overtime in COVID-19 patients. J Med Virol. 2021;93:2149–59. https://doi.org/10.1002/jmv.26601.
Tanaka S, Begue F, Veeren B, et al. First Recombinant High-Density Lipoprotein Particles Administration in a Severe ICU COVID-19 Patient, a Multi-Omics Exploratory Investigation. Biomedicines. 2022;10:754. https://doi.org/10.3390/biomedicines10040754.
Song J-W, Lam SM, Fan X, et al. Omics-Driven Systems Interrogation of Metabolic Dysregulation in COVID-19 Pathogenesis. Cell Metabol. 2020;32:188–e2025. https://doi.org/10.1016/j.cmet.2020.06.016.
Tian S, Hu W, Niu L, et al. Pulmonary Pathology of Early-Phase 2019 Novel Coronavirus (COVID-19) Pneumonia in Two Patients With Lung Cancer. J Thorac Oncol. 2020;15:700–4. https://doi.org/10.1016/j.jtho.2020.02.010.
Sorokin AV, Karathanasis SK, Yang Z-H, et al. COVID-19—Associated dyslipidemia: Implications for mechanism of impaired resolution and novel therapeutic approaches. FASEB J. 2020;34:9843–53. https://doi.org/10.1096/fj.202001451.
Sun JT, Chen Z, Nie P et al. (2020) Lipid Profile Features and Their Associations With Disease Severity and Mortality in Patients With COVID-19. Front Cardiovasc Med 7.
Sampedro-Nuñez M, Aguirre-Moreno N, García-Fraile Fraile L, et al. Finding answers in lipid profile in COVID-19 patients. Endocrine. 2021;74:443–54. https://doi.org/10.1007/s12020-021-02881-0.
Chien J-Y, Jerng J-S, Yu C-J, Yang P-C. Low serum level of high-density lipoprotein cholesterol is a poor prognostic factor for severe sepsis*. Crit Care Med. 2005;33:1688–93. https://doi.org/10.1097/01.CCM.0000171183.79525.6B.
Tanaka S, Stern J, Bouzid D, et al. Relationship between lipoprotein concentrations and short-term and 1-year mortality in intensive care unit septic patients: results from the HIGHSEPS study. Ann Intensiv Care. 2021;11:11. https://doi.org/10.1186/s13613-021-00800-0.
Barlage S, Gnewuch C, Liebisch G, et al. Changes in HDL-associated apolipoproteins relate to mortality in human sepsis and correlate to monocyte and platelet activation. Intensive Care Med. 2009;35:1877–85. https://doi.org/10.1007/s00134-009-1609-y.
Mehta A, Shapiro MD. Apolipoproteins in vascular biology and atherosclerotic disease. Nat Reviews Cardiol. 2022;19:168–79. https://doi.org/10.1038/s41569-021-00613-5.
Poynard T, Deckmyn O, Rudler M, et al. Performance of serum apolipoprotein-A1 as a sentinel of Covid-19. PLoS ONE. 2020;15:e0242306. https://doi.org/10.1371/journal.pone.0242306.
Zhu Z, Yang Y, Fan L, et al. Low serum level of apolipoprotein A1 may predict the severity of COVID-19: A retrospective study. J Clin Lab Anal. 2021;35:e23911. https://doi.org/10.1002/jcla.23911.
Shen B, Yi X, Sun Y, et al. Proteomic and Metabolomic Characterization of COVID-19 Patient Sera. Cell. 2020;182:59–e7215. https://doi.org/10.1016/j.cell.2020.05.032.
Mak PA, Laffitte BA, Desrumaux C, et al. Regulated Expression of the Apolipoprotein E/C-I/C-IV/C-II Gene Cluster in Murine and Human Macrophages: A CRITICAL ROLE FOR NUCLEAR LIVER X RECEPTORS α AND β*. J Biol Chem. 2002;277:31900–8. https://doi.org/10.1074/jbc.M202993200.
Kimhofer T, Lodge S, Whiley L, et al. Integrative Modeling of Quantitative Plasma Lipoprotein, Metabolic, and Amino Acid Data Reveals a Multiorgan Pathological Signature of SARS-CoV-2 Infection. J Proteome Res. 2020;19:4442–54. https://doi.org/10.1021/acs.jproteome.0c00519.
Melchior JT, Street SE, Andraski AB, et al. Apolipoprotein A-II alters the proteome of human lipoproteins and enhances cholesterol efflux from ABCA1. J Lipid Res. 2017;58:1374–85. https://doi.org/10.1194/jlr.M075382.
Zvintzou E, Xepapadaki E, Kalogeropoulou C, et al. Pleiotropic effects of apolipoprotein A-Ⅱ on high-density lipoprotein functionality, adipose tissue metabolic activity and plasma glucose homeostasis. J Biomed Res. 2020;34:14–26. https://doi.org/10.7555/JBR.33.20190048.
Serrano M, Espinosa G, Lalueza A, et al. Beta-2-Glycoprotein-I Deficiency Could Precipitate an Antiphospholipid Syndrome-like Prothrombotic Situation in Patients With Coronavirus Disease 2019. ACR Open Rheumatol. 2021;3:267–76. https://doi.org/10.1002/acr2.11245.
Kang J, Jang H, Kim TH, et al. Accurate Diagnosis of COVID-19 from Self-Collectable Biospecimens Using Synthetic Apolipoprotein H Peptide-Coated Nanoparticle Assay. Anal Chem. 2022;94:17186–94. https://doi.org/10.1021/acs.analchem.2c03813.
Tschopp J, Chonn A, Hertig S, French LE. Clusterin, the human apolipoprotein and complement inhibitor, binds to complement C7, C8 beta, and the b domain of C9. J Immunol. 1993;151:2159–65.
de Bont CM, Boelens WC, Pruijn GJM. NETosis, complement, and coagulation: a triangular relationship. Cell Mol Immunol. 2019;16:19–27. https://doi.org/10.1038/s41423-018-0024-0.
Zhou Y, Xu Z, Liu Z. Impact of Neutrophil Extracellular Traps on Thrombosis Formation: New Findings and Future Perspective. Front Cell Infect Microbiol. 2022;12:910908. https://doi.org/10.3389/fcimb.2022.910908.
Marfia G, Navone S, Guarnaccia L, et al. Decreased serum level of sphingosine-1-phosphate: a novel predictor of clinical severity in COVID-19. EMBO Mol Med. 2021;13:e13424. https://doi.org/10.15252/emmm.202013424.
Wolfrum C, Poy MN, Stoffel M. Apolipoprotein M is required for prebeta-HDL formation and cholesterol efflux to HDL and protects against atherosclerosis. Nat Med. 2005;11:418–22. https://doi.org/10.1038/nm1211.
Yao Mattisson I, Christoffersen C. Apolipoprotein M and its impact on endothelial dysfunction and inflammation in the cardiovascular system. Atherosclerosis. 2021;334:76–84. https://doi.org/10.1016/j.atherosclerosis.2021.08.039.
Martins-Gonçalves R, Hottz ED, Bozza PT. Acute to post-acute COVID-19 thromboinflammation persistence: Mechanisms and potential consequences. Curr Res Immunol. 2023;4:100058. https://doi.org/10.1016/j.crimmu.2023.100058.
Siwy J, Wendt R, Albalat A, et al. CD99 and polymeric immunoglobulin receptor peptides deregulation in critical COVID-19: A potential link to molecular pathophysiology? Proteomics. 2021;21:e2100133. https://doi.org/10.1002/pmic.202100133.
Martinez LO, Najib S, Perret B, et al. Ecto-F1-ATPase/P2Y pathways in metabolic and vascular functions of high density lipoproteins. Atherosclerosis. 2015;238:89–100. https://doi.org/10.1016/j.atherosclerosis.2014.11.017.
Panfoli I, Esposito A. Beneficial effect of polyphenols in COVID-19 and the ectopic F1FO‐ATP synthase: Is there a link? J Cell Biochem. 2022;123:1281–4. https://doi.org/10.1002/jcb.30306.
Ramachandran K, Maity S, Muthukumar AR, et al. SARS-CoV-2 infection enhances mitochondrial PTP complex activity to perturb cardiac energetics. iScience. 2022;25:103722. https://doi.org/10.1016/j.isci.2021.103722.
Conca W, Alabdely M, Albaiz F, et al. Serum β2-microglobulin levels in Coronavirus disease 2019 (Covid-19): Another prognosticator of disease severity? PLoS ONE. 2021;16:e0247758. https://doi.org/10.1371/journal.pone.0247758.
Calvet J, Berenguer-Llergo A, Gay M, et al. Biomarker candidates for progression and clinical management of COVID-19 associated pneumonia at time of admission. Sci Rep. 2022;12:640. https://doi.org/10.1038/s41598-021-04683-w.
Medjeral-Thomas NR, Troldborg A, Hansen AG, et al. Protease inhibitor plasma concentrations associate with COVID-19 infection. Oxf Open Immunol. 2021;2:iqab014. https://doi.org/10.1093/oxfimm/iqab014.
Geyer PE, Arend FM, Doll S, et al. High-resolution serum proteome trajectories in COVID-19 reveal patient-specific seroconversion. EMBO Mol Med. 2021;13:e14167. https://doi.org/10.15252/emmm.202114167.
Ballout RA, Kong H, Sampson M, et al. The NIH Lipo-COVID Study: A Pilot NMR Investigation of Lipoprotein Subfractions and Other Metabolites in Patients with Severe COVID-19. Biomedicines. 2021;9:1090. https://doi.org/10.3390/biomedicines9091090.
Hu X, Chen D, Wu L, et al. Declined serum high density lipoprotein cholesterol is associated with the severity of COVID-19 infection. Clin Chim Acta. 2020;510:105–10. https://doi.org/10.1016/j.cca.2020.07.015.

No competing interests reported.

Download PDF

Editorial decision: Revision requested
10 Sep, 2024
Reviews received at journal
10 Sep, 2024
Reviews received at journal
09 Sep, 2024
Reviews received at journal
06 Sep, 2024
Reviews received at journal
02 Sep, 2024
Reviewers agreed at journal
30 Aug, 2024
Reviewers agreed at journal
30 Aug, 2024
Reviewers agreed at journal
29 Aug, 2024
Reviewers agreed at journal
29 Aug, 2024
Reviewers agreed at journal
29 Aug, 2024
Reviewers agreed at journal
28 Aug, 2024
Reviewers invited by journal
28 Aug, 2024
Editor assigned by journal
26 Aug, 2024
Submission checks completed at journal
26 Aug, 2024
First submitted to journal
26 Aug, 2024

You are reading this latest preprint version

HDL proteome and apolipoproteins concentrations in severe ICU COVID-19 patients

Status:

Version 1

Abstract

Figures

1. Introduction

2. Materials and methods

2.1. Study population

2.2. Quantification of lipoprotein subpopulation by tube gel electrophoresis (Quantimetrix Lipoprint)

2.3. Lipoprotein isolation from plasma

2.4. Fast protein Liquid Chromatography (FPLC)

2.5. Label-free Quantification by Nano LC-MS/MS analysis

2.5.1. HDL sample preparation and proteolytic digestion

2.5.2. Nano LC-MS/MS method

2.5.3. Protein identification and quantification

2.5.4. Data processing

2.6. Statistical analysis

3. Results

3.1 Study population

3.2 Comparison of plasma apolipoprotein concentrations between COVID-19 survivors and non-survivors

3.3 Modulation of HDL subpopulations size between survivors and non-survivors COVID-19 patients

3.4 HDL proteome analysis of surviving versus non-surviving patients of their COVID-19

3.5 Diagnostic performance via receiver characteristic (ROC) curve

4. Discussion

Declarations

Conflicts of interests

Funding

Author Contribution

Acknowledgement

References

Additional Declarations

Supplementary Files

Status:

Version 1