Increased angiotensin-converting enzyme 2, sRAGE and immune activation, but lowered calcium and magnesium in COVID-19: association with chest CT abnormalities and lowered peripheral oxygen saturation.

Background. The characterization of new biomarkers of COVID-19 is extremely important. Few studies measured the soluble receptor for advanced glycation end product (sRAGE), angiotensin-converting enzyme 2 (ACE2), calcium and magnesium in COVID-19. Aims: To measure sRAGE, ACE2, interleukin (IL)-6, IL-10, CRP, calcium, magnesium, and albumin in COVID-19 patients in association with peripheral oxygen saturation (SpO2) and chest CT scan abnormalities (CCTA) including ground glass opacities. Methods. This study measured sRAGE, ACE2, IL-6, IL-10, CRP using ELISA techniques, and calcium, magnesium, and albumin using a spectrophotometric method in 60 COVID-19 patients and 30 healthy controls. Results. COVID-19 is characterized by significantly increased IL-6, CRP, IL-10, sRAGE, ACE2, and lowered levels of SpO2, albumin, magnesium and calcium. Neural networks showed that a combination of calcium, IL-6, CRP, and sRAGE yielded an accuracy of 100% in detecting COVID-19 patients with calcium being the most important predictor followed by IL-6, and CRP. COVID-19 patients with CCTAs showed lower SpO2 and albumin levels than those without CCTAs. SpO2 was significantly and inversely correlated with IL-6, IL-10, CRP, sRAGE, and ACE2, and positively with albumin, magnesium and calcium. Patients with positive IgG results showed a significant elevation in the serum level of IL-6, sRAGE, and ACE2 compared to the negatively IgG patient subgroup. Conclusion. The results show that immune-inflammatory and RAGE pathway biomarkers may be used as external validating criterion for the diagnosis COVID-19. Those pathways coupled with lowered SpO2, calcium and magnesium are drug targets that may help to reduce the consequences of COVID-19.


Introduction
estimate the prognosis of SARS-CoV-2 infection, and to discover new drug targets to treat the consequences of the infection are an important field of study.
Recent studies showed increased levels of angiotensin-converting enzyme 2 (ACE2) following SARS-CoV-2 infection (Patel et al., 2021), and a case report showed very high ACE2 levels in a critically ill COVID-19 patient (Nagy et al., 2021). ACE2 is a common binding site for both SARS-CoV and SARS-CoV-2 (Zhu et al., 2018, Hoffmann et al., 2020. COVID-19 virus binds to ACE2 receptors in human cells through Spike proteins-S homotrimers with high affinity and leads mainly to endocytosis through the pathway of "clathrin-mediated endocytosis (Amini Pouya et al., 2020). After attachment, transmembrane serine protease 2 (TMPRSS2) and several lysosomal proteases cleave and activate the S-glycoprotein due to SARS-CoV-2 entry into the cell through endocytosis or direct membrane fusion with the host membrane (Vlachakis et al., 2020). sACE2 is cleaved from ACE2 by a disintegrin and metallopeptidase domain-17 (ADAM17) and then released into the extracellular environment (Lambert et al., 2005).
SARS-CoV-2 may cause an overzealous immune-inflammatory response and even a cytokine storm in the host associated with severe lung pathology Zumla, 2019, Huang et al., 2020a). These processes are closely linked to poor clinical outcomes (Huang et al., 2020b, Qin et al., 2020, worsening of hypercoagulation (Mehta et al., 2020), and higher mortality . For example, the severity of COVID-19 infection is associated with increased interleukin-6 (IL-6) and other proinflammatory cytokines, C-reactive protein (CRP), and IL-10, a negative immune-regulatory cytokine (Ruan et al., 2020. Mitigation of the cytokine storm may contribute to better clinical outcomes in severe cases of COVID-19. . Severe COVID-19 is also accompanied by lowered serum albumin levels or hypoalbuminemia as part of the systemic inflammatory response in patients with COVID-19   .0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
Another possible candidate biomarker of COVID-19 is a soluble form of the advanced glycation end-product receptor (sRAGE) (Yalcin Kehribar et al., 2021). Advanced glycation end products (AGEs) and high mobility group box 1 protein (HMGB1) can bind to the membranebound RAGE (mRAGE), thereby initiating intracellular signaling (Vistoli et al., 2013), which leads to activation of pro-inflammatory transcription factors, including nuclear factor (NF)-kappa B (Macaione et al., 2007, Julio et al., 2014, which in turn increased the production of, for example, IL-6 protein transcription in immune cells . sRAGE is generated through proteolytic cleavage of the RAGE extracellular domain or through alternative RNA splicing (Zhang et al., 2008b, Sterenczak et al., 2009. Several studies have documented a high prevalence of hypocalcemia in COVID-19 patients (Cappellini et al., 2020, Lippi et al., 2020, and one study found that hypocalcemia could predict hospitalization in COVID-19 patients (Di Filippo et al., 2020). Alterations of Ca2+ homeostasis may occur during viral infections (Nieto-Torres et al., 2015) (Deng et al., 2012) because viruses often utilize Ca 2+ signals to create a suitable cellular environment that meets their demands (Zhou et al., 2009). Lowered levels of calcium may interfere with calcium-regulated processes, including signal transduction networks, which are regulated by calcium's activities as a second messenger or as plasma membrane channels and pumps (Civitelli andZiambaras, 2011, Görlach et al., 2015).
Magnesium deficiency is significantly and inversely related to the severity of infection in COVID-19 patients (Quilliot et al., 2020). Magnesium shows anti-inflammatory (Abiri and Vafa, 2020) . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 26, 2021. ;https://doi.org/10.1101 doi: medRxiv preprint 8 8 and antioxidant (Güzel et al., 2019) properties and is an important cofactor for ATP, which mediates several basic enzymatic reactions (Romani and Scarpa, 1992). Nevertheless, no studies in SARS-CoV-2-infected patients with COVID-19 have examined whether a combination of the above biomarkers may be used to externally validate the diagnosis of COVID-19 and whether these pathways are associated with the presence of lowered SpO2.
Hence, the present study aimed to examine the levels of ACE2, sRAGE, IL-6, IL-10, albumin, calcium, and magnesium in patients with COVID-19 versus normal controls and to examine the association between those biomarkers and SpO2 levels and CCTAS. The specific hypotheses are that a) COVID-19 is associated with increased ACE2, sRAGE, IL-6, IL-10 levels, and lowered albumin, calcium, and magnesium; and b) that those biomarkers are associated with lowered SpO2.

Subjects
Sixty COVID-19 male patients aged 25-59 years were recruited at the Al-Sadr Teaching Hospital and Al-Amal Specialized Hospital for Communicable Diseases in Najaf governorate-Iraq between September and November 2020. All patients were admitted to these two centers, which are official quarantine centers specialized in the treatment of COVID-19. The diagnosis was made by senior physicians and virologists. All patients suffered from acute respiratory syndrome and were diagnosed with SARS-CoV-2 infection according to positive results of COVID-19 nucleic acids by reverse transcription real-time polymerase chain reaction (rRT-PCR), positive IgM, in addition to the standard symptoms of the disease including fever, breathing difficulties, cough, and loss of smell and taste. We excluded patients with premorbid medical illnesses, including diabetes . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 26, 2021. 9 9 type 1, liver, kidney and cardiovascular diseases. We also recruited 30 healthy controls, age, and sex-matched to the patient groups. All controls were free from any systemic disease. However, to enhance their immunity against COVID-19 infection, some healthy controls were taken zinc and vitamins C and D.
The "institutional ethics board of the University of Kufa" approved the study (617/2020).
All participants gave written informed consent before participation in this study. The study was

Measurements
The RT-PCR tests were performed using the Applied Biosystems ® QuantStudio™ 5 Real-Time PCR System (Thermo Fisher Scientific) supplied by Life Technologies Holdings Pte Ltd., Marsiling Industrial Estate, Singapore. The Lyra ® Direct SARS-CoV-2 Assay kits were supplied by Quidel Corporation, CA, USA. The procedures were exactly followed as mentioned by the manual of the kits. This is a real-time RT-PCR assay for the qualitative detection of human coronavirus SARS-CoV-2 from viral RNA extracted from nasal, nasopharyngeal or oropharyngeal swab specimens. The Assay targets the non-structural Polyprotein (pp1ab) of the SARS-CoV-2 virus. Patients had a chest X-ray and chest computed tomography scan (CT-scan) to search for lung abnormalities. CT-scans were made by using SOMATOM Definition AS, Supplied by Seimens, Munchen, Germany. The chest CT abnormalities (CCTAs) assessed in our study . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.
Accordingly, the patients' group was subdivided into two subgroups depending on the presence of CCTAs into those with (COVID9+CCTA) and without (COVID-CCTA).
The blood samples were taken in the early morning between 7.30-9.00 a.m. after awakening and before having breakfast. Five milliliters of venous blood samples were drawn and transferred into clean plain tubes. Hemolyzed samples were rejected. After ten minutes, the clotted blood samples were centrifuged for five minutes at 3000 rpm and then serum was separated and transported into three new Eppendorf tubes until assay. A qualitative ACON ® COVID-19 IgG/IgM rapid test was used to detect IgG and IgM in the sera of patients and controls. The kits have a sensitivity ≥ 99.1 % and a specificity ≥ 98.2 %. We also divided the patient's subgroups according to IgG results into negative-IgG and positive-IgG subgroups to examine the difference in the measured biomarkers between these subgroups. A C-Reactive Protein (CRP) latex slide test (Spinreact ® , Barcelona, Spain) was used for the qualitative and semi-quantitative CRP measurement in human serum. The serum titer is the reciprocal of the maximum dilution exhibiting a positive response multiplied by a positive control concentration. The approximate CRP concentration in the patient sample is calculated as follows: 6 x CRP Titer = mg/L. Serum IL-6, IL-10, sRAGE, and sACE2 were measured using ELISA kits supplied by Melsin Medical Co, Jilin, China. All kits were based on a sandwich technique and showed an inter-assay CV of less than 12%. Total calcium, albumin, and magnesium were measured spectrophotometrically by kits supplied by Biolabo ® , Maizy, France. The procedures were followed according to the manufacturer's instructions and without modifications.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 26, 2021.

Statistical analysis
Analysis of variance (ANOVA) was used to check differences in scale variables among groups, and analysis of contingency tables (χ 2 -test) was used to assess associations between nominal variables. We computed correlation matrices based on Pearson's product-moment correlation coefficients to determine correlations between biomarkers and clinical scores (e.g., duration of illness and SpO2). We used univariate generalized linear model (GLM) analysis to delineate the associations between diagnosis and the biomarkers while controlling for confounding variables, including tobacco use disorder (TUD), age, and BMI. The effect size was estimated using partial eta-squared values. We also computed model generated (GLM analysis) estimated marginal mean (SE) values and protected pairwise comparisons among treatment means. Multiple comparisons were evaluated using false discovery rate p-correction (Benjamini and Hochberg, 1995). Tests were 2-tailed, and a p-value of 0.05 was used for statistical significance.
Multilayer perceptron Neural Network models were employed to assess the more complex associations between the diagnosis of COVID-19 versus controls (entered as output variables) and biomarkers (entered as input variables). We trained the models with an automated feed-forward architecture, two hidden layers with up to 4 nodes in each layer, with minibatch training with gradient descent, and 20-50 epochs. The stopping rule was one consecutive step with no further decrease in the error term. Three samples were extracted, namely "a training sample to estimate the network parameters (46.67% of all participants), testing set to prevent overtraining (20.0%) and a holdout set to evaluate the final network (33.33%). Error, relative error, and importance and relative importance of all input variables were computed" (Moustafa et al., 2020). All statistical analyses were performed using IBM SPSS windows version 25, 2017.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 26, 2021. ; https://doi.org/10.1101/2021.03.26.21254383 doi: medRxiv preprint Table 1 demonstrates the socio-demographic data in COVID+CCTA and COVID-CTTA patients, and healthy controls (HC). No significant differences among these study groups were detected in BMI, education, residency, marital status, employment, and TUD. CCTA+ patients were somewhat older than CCTA-patients, but illness duration was not significantly different between those groups. All patients were on O2 therapy and were treated with paracetamol, bromhexine, vitamin C, vitamin D, and zinc as a continuous treatment. There were no significant differences in the use of azithromycin, enoxaparin, dexamethasone, famotidine, heparin, and meropenem between CCTA+ and CCTA-patients. Table 2 shows the different biomarkers' measurements in the three study groups with significant differences in all biomarkers, which remained significant after p-correction for FDR.

Differences in biomarkers between CCTA+ and CCTA-patients
There was a significant decrease in SpO2 and serum albumin in CCTA+ patients compared to CCTA-patients. The serum levels of IL-6, IL-10, sRAGE, sACE2, and CRP were higher in COVID-19 patients compared with controls. SpO2, serum albumin, calcium, and magnesium were lowered in both patient groups compared to controls, while there were no significant differences in those biomarkers between both patients' subgroups.

Neural network results
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 26, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 We have performed neural network analyses with diagnosis (COVID-19 versus controls) as output variables and the biomarkers as explanatory variables and reran the analyses with the 4 most important biomarkers, i.e., calcium, IL-6, CRP, and sRAGE. Table 3 shows the network information of the model examining the separation of COVID-19 versus controls. The network has been trained using two hidden layers with three and two units in layers 1 and 2. Hyperbolic tangent was used as the activation function in hidden layer 1 and identity in the output layer. Table   3 displays the partitioned confusion matrices showing an AUC ROC=1.000 with an accuracy of 100.0% in the holdout sample with a sensitivity of 100.0% and a specificity of 100.0%. Figure 1 shows the relative importance of the most important input variables (Ca > IL-6 > CRP > sRAGE), representing the most important determinants of the predictive power of the model.

Biomarkers results in COVID-19 subgroups with and without anti-SARS-CoV-2 IgG antibodies
The univariate GLM analysis of the biomarkers in patients' subgroups (positive versus negative anti-SARS-CoV-2 IgG antibodies showed significant elevations in serum IL-6, sRAGE, and ACE2 in patients with positive IgG results as compared with the IgG negative patient subgroup. These differences remained significant after performing p-correction for FDR, even IL-6 (p=0.042). Other biomarkers showed no significant differences between those subgroups. Table 5 shows the intercorrelation matrix of the biomarkers and duration of the infection and SpO2 levels. Duration of illness was inversely associated with SpO2, albumin, magnesium, and calcium. These p-values remained significant after FDR correction (all at p<0.003). The results showed that SpO2 was significantly and negatively correlated with duration of illness, CRP, IL-6, . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 26, 2021. ; https://doi.org/10.1101/2021.03.26.21254383 doi: medRxiv preprint IL-10, sRAGE, and ACE2, and significantly and positively with albumin, magnesium, and calcium. There were no significant associations between the biomarkers and the total CCTA score.

Discussion
The first major finding of this study is that COVID-19 is accompanied by increased IL-6, IL-10, and CRP in addition to sRAGE and ACE2 and that albumin, calcium and magnesium are significantly decreased as compared with controls. Moreover, a combination of calcium, IL-6, CRP and sRAGE was able to validate the diagnosis of COVID-19 with 100% accuracy. Duration of infection was accompanied by lowered SpO2, albumin, calcium and magnesium levels.

Inflammation and COVID-19
The pathophysiology of COVID-19 involves a potent immune-inflammatory response, including increases in IL-6 and IL-10, which is associated with a higher risk of disease deterioration . When the immune system is activated following infection with SARS-CoV-2, an overzealous immune system may secrete many proinflammatory cytokines, and may cause a cytokine storm (Kuba et al., 2010, Jia, 2016, which is a key component in the pathophysiology of severe COVID-19 (Tetro, 2020 (Huang et al., 2020a. Sustained elevations in pro-inflammatory cytokines in COVID-19 patients indicate continuing systemic and tissue inflammation , Guan et al., 2020a. It is interesting to note that the combination of IL-6 coupled with D-dimer may be used to differentiate extreme COVID-19 cases .
Increased CRP levels have been associated with the cytokine storm and liver damage in COVID-19 patients and generally contribute to a worse prognosis . The hypoalbuminemia in our COVID-19 patients may, at least in part, be explained by the acute-phase . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 26, 2021. ; https://doi.org/10.1101/2021.03.26.21254383 doi: medRxiv preprint response downregulating negative acute-phase proteins, although increased vascular permeability and liver or kidney disease may play a role (Ronit et al., 2020). Albumin has strong antioxidant properties (Maes et al., 2011), and as such, hypoalbuminemia may play a role in the exaggerated immune responses in COVID-19. Moreover, albumin has anticoagulant and antiplatelet activities, and therefore hypoalbuminemia may be associated with hypercoagulability (Paar et al., 2017), increased D-dimers, a thrombin generation marker (Thachil et al., 2020), and enhanced risk of artery and venous thrombosis (Ronit et al., 2020). It is important to note that coagulopathies complicate the clinical course of COVID-19 (Thachil et al., 2020)   and that, therefore, hypoalbuminemia may play a role in COVID-19 associated coagulopathies.

ACE2 and COVID-19
As described in the Introduction, only a few studies reported increased ACE2 levels in patients with SARS-CoV-2 infection and COVID-19 (Nagy et al., 2021, Patel et al., 2021. ACE2 concentrations are strongly related to ACE2 catalytic activity . Increased ACE2 plasma levels are associated with an increased risk of stroke, myocardial infarction, incident diabetes, incident heart failure and, as a consequence, with total cardiovascular and noncardiovascular deaths (Narula et al., 2020). In patients with heart failure, increased plasma ACE2 activity is associated with a worsened prognosis (Oudit and Pfeffer, 2020). The most important determinants of increased ACE2 values are male sex, geographic ancestry, and body-mass index (BMI) (Narula et al., 2020). In physiological conditions, cell-bound membrane ACE2 is a counterregulatory pathway through the production of angiotensin-(1-7) and stimulation of the angiotensin 2 (AT2R) and Mas receptors, which promote anti-inflammatory and antioxidant processes, vasorelaxation, anti-fibrosis, and attenuate the renin-angiotensin system, cellular proliferation, . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 26, 2021. ; https://doi.org/10.1101/2021.03.26.21254383 doi: medRxiv preprint hypertrophy, and vasoconstriction (Narula et al., 2020). Nevertheless, the extracellular domain of the transmembrane ACE2 facilitates entry for SARS-CoV-2 into the cells via interactions with the SARS-CoV spike protein (Walls et al., 2020. This process is followed by downregulation of surface ACE2 expression that may culminate in loss of angiotensin II effects leading to increased multisystem diseases, including lung, cardiovascular and renal disease (Oudit and Pfeffer, 2020). Moreover, the downregulation of ACE2 expression through the SARS-CoV spike protein causes shedding of soluble ACE, which is therefore a consequence of SARS-CoV infection (Glowacka et al., 2010). It is hypothesized that increased sACE2 shedding may neutralize the SARS-CoV2 virus at a distance from virus entry into the cells (Brest et al., 2021). All in all, COVID-19 infection may cause part of its detrimental effect via renin-angiotensin system overactivation and increased levels of sACE2.

sRAGE and COVID-19
As described in the Introduction, increased levels of plasma sRAGE were reported previously (Yalcin Kehribar et al., 2021). In another study, sRAGE was significantly increased in COVID-19 patients with diabetes mellitus versus normal controls (Dozio et al., 2020a).
Nevertheless, in our study, patients with diabetes mellitus were excluded indicating that the increased sRAGE concentrations in COVID-19 are not due to diabetes mellitus. These findings extend those of Dozio et al. (2020) who reported that in COVID-19 patients, the RAGE pathway may be modulated regardless of glycemic control (Dozio et al., 2020b). It is important to note that increased sRAGE in bronchoalveolar lavage and plasma indicates acute pulmonary injury and type 1 alveolar epithelial cell injury (Uchida et al., 2006). Nevertheless, in our study, sRAGE levels were not significantly increased in COVID-19 patients with CCTA than those without CCTA.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Calcium, magnesium and COVID-19
In our study, the most important biomarker of COVID-19 was lowered calcium. Low calcium is common in severe COVID-19 patients and is associated with the severity of illness , and lowered calcium levels may detect the most critical COVID-19 patients with high specificity (Yang et al., 2021). However, hypocalcemia is also detectable in COVID-19 patients with non-severe illness (Pal et al., 2020). In addition, patients with very low calcium values have an increased prevalence of acute respiratory distress syndrome (ARDS) . Interestingly, inflammation appears to be higher in COVID-19 patients with extremely low blood calcium levels compared to patients with normal calcium levels (Di Filippo et al., 2020, di Filippo et al., 2021. This is not an invariant finding, however, as hypocalcaemia is sometimes detectable in COVID-19 patients with non-severe illness rather arguing against inflammation as a universal cause of the phenomenon (Pal et al., 2020). Hypocalcemia causes an increased secretion of PTH, hypoproteinemia, hypomagnesemia (Kelly and Levine, 2013) and more detrimental effects as described in the Introduction.
SARS-CoV and MERS-CoV are known to disrupt intracellular calcium homeostasis to foster membrane fusion and increase infectivity (Millet andWhittaker, 2018, Straus et al., 2019) . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 26, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 and the hypocalcemia seen in patients with severe COVID-19 is also evident in patients following SARS-CoV infection (Booth, 2003). This suggests that there may be a common mechanism driving the development of calcium dyshomeostasis following beta coronavirus infections. Several mechanisms are involved in calcium dyshomeostasis mediated by a plethora of viruses following cell invasion including activation of Ca 2+ ATPase pumps, the transient receptor potential canonical (TRPC) channels with the receptor-operated Ca 2+ currents (ROCs), and the storeoperated calcium entry (SOCE) channels . The latter may be of particular interest as one potential mechanism explaining beta coronavirus-mediated calcium dyshomeostasis is that SOCE may be activated by endoplasmatic-reticulum (ER) stress (Chen et al., 2015, which is an invariant consequence of the replication strategies employed by these viruses (Fung andLiu, 2014, Morris et al., 2020a). The plausibility of SOCE activation as a vehicle to explain SARS-CoV2 mediated decreases in plasma calcium levels is increased by the presence of in vivo data confirming that SARS-CoV ORF8 and ORF3 have the capacity to activate SOCE (Hyser and Estes, 2015). Importantly, SOCE is activated in conditions of inflammation and oxidative stress (Nunes and Demaurex, 2014), which is characteristic of SARS-CoV-2 infection (Morris et al., 2020b). SOCE activation also upregulates nuclear factor-KB potentially resulting in a positive feedback mechanism amplifying inflammation, oxidative stress and calcium dyshomeostasis (Berry et al., 2018, Dresselhaus andMeffert, 2019).
Lowered serum magnesium is associated with increased thrombosis risk, decreased fibrinolysis, endothelial dysfunction, mitochondrial dysfunctions, increased inflammatory and oxidative stress, and increased fatty acid production (Çiçek et al., 2016, Zheltova et al., 2016, Gromova et al., 2018, Nielsen, 2018. Furthermore, in vivo studies have demonstrated that . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

CCTA and consequences
The second major finding of the present study is that CCTAs are accompanied by lowered albumin and SpO2 levels and that the latter is associated with all biomarkers described above.
Chest imaging, especially by computed tomography scan (CT-scan), is of great importance for the diagnosis, management, and follow-up of patients with COVID-19 infection . For example, the appearance of GGOs in COVID-19 indicates severe or persistent inflammation in the lungs, and more severe symptoms including bronchiolitis and pneumonia in both lungs, and lung fibrosis (Sadhukhan et al., 2020). These lung infection sites may cause the recruitment of various immune cells, and consequently, a pro-inflammatory response is mounted, including increased levels of IL-6 (Sadhukhan et al., 2020). Our results also indicate that lung lesions as detected with chest-CT scan may lead to decreased oxygenation. In fact, many of the patients included here (88.3%) showed hypoxemia as indicated by SpO2 < 92%. Moreover, hypoxia may induce inflammation (Eltzschig et al., 2014) and may upregulate ACE2 gene expression and protein levels in lung and kidney tissues, which may further contribute to the severity of COVID-19 (Shenoy et al., 2020). There are also reports that patients with higher GGO CT-scores have lower calcium levels and that SPO2 is correlated with serum calcium with those COVID-19 patients staying longer in ICU (Yang et al., 2021).

Biomarkers and IgG positivity
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 26, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021 Another important result of our study is that patients with positive IgG results showed higher IL-6, sRAGE and ACE2 levels as compared to those with negative IgG results. In our study, 66.7% of the COVID-19 patients showed IgG positivity which is in agreement with previous studies showing that around 77.9% of plasma samples were IgG positive .
Following exposure to the SARS-CoV-2 infection, a humoral immune response is mounted which starts with IgM formation after 3-7 days, indicating an acute or continuing infection and those levels peak after three weeks (Xiao et al., 2020). Increased IgG levels may be detected 14 days after symptom onset (Guo et al., 2020a). In fact, the antibody dynamics in COVID-19 are quite similar to those in other viral infections, with IgG levels increasing when IgM levels start to decrease (Zhou et al., 2020c). It is well known that IL-6 may drive IgG formation by B-cells (Maeda et al., 2010). Increased IgG titers are more prevalent in severe than in mild clinical status, especially in female patients (Zeng et al.). As such, the association between positive IgG titers and ACE2 and sRAGE could be explained by the increased severity of illness.

Acknowledgment
We thank the staff of Al-Sadr Teaching Hospital and Al-Amal Specialized Hospital for Communicable Diseases in Najaf governorate-Iraq for their help in the collection of samples. We also thank the high-skilled staff of the hospital's internal labs, especially Mr. Murtadha Al-Hilo, for their help in the estimation of biomarkers levels.

Funding
There was no specific funding for this specific study.
. CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

(which was not certified by peer review)
The copyright holder for this preprint this version posted March 26, 2021. ;https://doi.org/10.1101https://doi.org/10. /2021  . CC-BY 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)    Shown are Pearson's product moment correlation coefficients (p-value). IL: interleukin, CRP: C-reactive protein, sRAGE: soluble receptor for advanced glycation end products, sACE2: soluble angiotensin-converting enzyme 2, SPO2 %: peripheral oxygen saturation percentage. Figure 1. Results of neural network (importance chart) with diagnosis of COVID-19 as output variables and biomarkers (in z-scores) as input variables. IL-6: interleukin-6, CRP: C-reactive protein, sRAGE: soluble receptor for advanced glycation end product.