Plasma of COVID-19 Patients Does Not Alter Electrical Resistance of Human Endothelial Blood-Brain Barrier In Vitro

Abstract The pandemic of coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 instigated the most serious global health crisis. Clinical presentation of COVID-19 frequently includes severe neurological and neuropsychiatric symptoms. However, it is presently unknown whether and to which extent pathological impairment of blood-brain barrier (BBB) contributes to the development of neuropathology during COVID-19 progression. In the present study, we used human induced pluripotent stem cells-derived brain endothelial cells (iBECs) to study the effects of blood plasma derived from COVID-19 patients on the BBB integrity in vitro. We also performed a comprehensive analysis of the cytokine and chemokine profiles in the plasma of COVID-19 patients, healthy and recovered individuals. We found significantly increased levels of interferon γ-induced protein 10 kDa, hepatocyte growth factor, and interleukin-18 in the plasma of COVID-19 patients. However, blood plasma from COVID-19 patients did not affect transendothelial electrical resistance in iBEC monolayers. Our results demonstrate that COVID-19-associated blood plasma inflammatory factors do not affect BBB paracellular pathway directly and suggest that pathological remodeling (if any) of BBB during COVID-19 may occur through indirect or yet unknown mechanisms.


Introduction
The coronavirus disease 2019 (COVID-19) pandemic caused by the sev er e acute r espirator y syndr ome cor onavirus 2 (SARS-CoV-2) has resulted in significant morbidity and mortality w orldwide .2][3][4] The COVID-19 infection is associated with a wide range of neurological or neuropsychiatric symptoms, such as anosmia and ageusia, headache, dizziness, confusion, encephalopathy, isc hemic stroke , Guillain-Barr é syndr ome, cogniti v e impairment, and others. 2 , 5-8everal postmortem studies demonstrated the presence of SARS-CoV-2 vir al RN A and pr oteins in v arious r egions of the brain, [9][10][11] indicating the entrance of viral particles into the CNS.Conce ptuall y, ther e ar e sev eral possib le r outes for the SARS-CoV-2 virus invasion into the brain. 124][15] Potentially, the virus can enter the CNS by tr ansmigr ation within infected hematopoietic cells, utilizing the so-called "Trojan horse" mechanism. 12inally , the SARS-CoV -2 virus can gain an access to the brain parench yma b y par acellular tr av ersal thr ough a disrupted b loodbrain barrier (BBB).The BBB is a highly selective and dynamic interface that separates the blood from the CNS. 16 , 17It is composed of specialized brain endothelial cells (BECs) that line the brain vessels, along with astrocytes, pericytes, parenchymal and endothelial basement membranes, and peri v ascular space that mount the barrier protecting the CNS from pathogens and toxins. 18 , 19Accumulating evidence indicates that cerebral vascular dysfunction is a common feature of COVID-19: Postmortem studies on COVID-19 patients described hemorrhages and ischemic lesions in the CNS, 9 , 11 , 20 whereas imaging studies on r ecov ering patients similarly revealed abnormalities in cerebral blood flow and hypometabolism, indicating disruption of the BBB. 21he SARS-CoV-2 virus enters host cells through the angiotensin-converting enzyme 2 (ACE2) receptor, which is expr essed in v arious tissues, including BECs, 22 and ther efor e can target BBB dir ectl y.It was shown that the SARS-CoV-2 virus infects and r e plicates in cultured human inducible pluripotent stem cell (iPSC)-deri v ed BECs (iBECs), which, however, does not affect barrier function. 23At the same time, the SARS-CoV-2 spike proteins directly interact with BECs in vitro and may alter BBB function. 21 , 24However, infection of BECs with SARS-CoV-2 in vi v o has not y et been demonstr ated.
COVID-19 is accompanied by an excessi v e systemic r elease of pr oinflammator y cytokines into the b lood, a phenomenon known as cytokine storm. 25Several specific cytokine and chemokine pr ofiles hav e been linked to COVID-19 sev erity and neurological complications. 26 , 27However, the effects of cytokines and chemokines secreted during SARS-CoV-2 infection on the BBB function were not systematically investigated.Pr oinflammator y cytokines and chemokines can affect BBB by the following mechanisms: (i) dir ectl y affecting BECs, thus causing disruption of the BBB; (ii) stimulating the release of inflammatory mediators by BECs into the brain parenchyma to induce r eacti v e astr ogliosis and micr ogliosis, leading to the secondar y BBB dama ge; (iii) penetration of BBB by pr oinflammator y cytokines and chemokines, causing neuroinflammation, leading to the secondary BBB injuries; and (iv) combinations of above.
Ther efor e, it is important to understand the intricate connections between blood cytokine/chemokine profiles, BBB integrity, Current technologies enable the generation of human iBECs monocultures with transendothelial electrical resistance (TEER) in the range of 4000-5000 cm 2 , which is close to the readings obtained in vi v o. 28 , 29 In the present study, we dir ectl y monitored BBB function in response to the blood plasma from COVID-19 patients.We also c har acterized blood plasma cytokine/chemokine profiles and found significantly increased lev els of interfer on γ -induced pr otein 10 kDa (IP-10), hepatoc yte gro wth factor (HGF), and interleukin-18 (IL-18)

Pa tient Informa tion and Da ta Collection
This study examined 33 patients who wer e dia gnosed with COVID-19 and were admitted to the Centre of Infectious Diseases of Vilnius Uni v ersity Hospital Santaros Klinikos between 2022 and 2023.We also assessed 30 healthy and 18 r ecov er ed subjects.Clinical information and la borator y samples wer e collected immediately after hospitalization.The permission to conduct this biomedical resear c h was issued by the Vilnius Regional Biomedical Resear c h Ethics Committee (2022/2-1407-879).
The median age of the subjects was 46 years (interquartile range (IQR) 38-58).The youngest subject was 20 and the oldest was 65 years old, 37 (46.27%) of the subjects were female and 43 (53.75%) were male.
All study participants v oluntaril y gav e informed consent to participate in the study.

Collection of Plasma Samples
Whole blood samples were collected in ethylenediaminetetraacetic acid and sodium heparin tubes.Within 30 min, samples were centrifuged at 1000 × g for 10 min at 4 • C to obtain plasma.Plasma was transferred in conical centrifuge tubes and subjected to inacti v ation at 56 • C for 30 min.Afterward, plasma was centrifuged at 1000 × g for 10 min at 4 • C, aliquoted in cryotubes, and stored at −80 • C until further use.

Maintenance of iPSCs and Differentiation to the Brain Capillary Endothelial Cells
Human exfoliated deciduous teeth stem cell (SHED)-deri v ed iPSCs (female, 7 years old) were cultured on matrigel-coated (Corning) plates with Essential 8 Flex medium (E8) in the incubator (37 • C and 5% CO 2 ).Inducible pluripotent stem cells were differentiated to iBECs according to slightly modified previousl y pub lished pr otocol. 29Briefly, 24 h after splitting, the differ entiation w as initiated by changing the E8 to the Essential 6 (E6) medium (Thermo Fisher Scientific).Essential 6 was fully r efr eshed ev er y 24 h for 4 d.On the fifth day, the E6 was changed to the human endothelial serum-free medium (hESFM, Thermo Fisher Scientific) supplemented with 20 ng/mL bFGF (Thermo Fisher Scientific), 10 μm retinoic acid (Mer c k Darmstadt, Germany), and 0.25 × B-27 (Thermo Fisher Scientific).After 48 h, the same medium w as full y r efr eshed.The next day cells were split for selection on 400 μg/mL collagen-IV and 100 μg/mL fibronectin-coated (both from Mer c k) Tr answ ell inserts (Corning) in hESFM medium supplemented with 0.25 × B-27, 50 U/mL and 50 μg/mL penicillin-str e ptomycin (Thermo Fisher Scientific, hESFM + B-27 + P/S).

Measurement of TEER
Transendothelial electrical resistance monitoring was used as a readout of BBB integrity.Transendothelial electrical resistance of iBECs monolayers was monitored every hour using the Cel-lZscope system (NanoAnalytics, M ünster, Germany), beginning immediately after seeding cells on the insert (at −24 h) and were carried out for 24-48 h in the presence of plasma after its addition at 0-h time point.The inserts plated with iBECs wer e immediatel y loaded into a 24-well cell module of a Cel-lZscope system prefilled with hESFM + B-27 + P/S and grown in the incubator (37 • C and 5% CO 2 ).Twenty-four hours after plating, the hESFM + B-27 + P/S was refreshed, and the cells wer e tr eated with 50% plasma (v/v in the same medium) collected from three different groups: healthy volunteers, COVID-19 infected patients, and r ecov er ed COVID-19 patients.

Immunofluorescence
After TEER measurements, the culture medium was aspirated and the cells wer e w ashed with physiological buffer solution (PBS) 3 times.Cells were fixed and permeabilized using ice-cold ( −20 • C) methanol-acetone solution (1:1), for 10 min at −20 • C.Then, the membrane of the Transwell insert was cut out and placed on the parafilm.Cells were washed 3 times with PBS, blocked using 1% bovine serum albumin-PBS solution for 30 min at room temper ature , then incubated with primary antibodies [against ZO-1 (1:33), occludin (1:50), and claudin-5 (1:100), in 1% bo vine serum albumin-PBS] o vernight at 4 • C.After incubation, samples were washed 3 times with PBS and incubated with antira bbit IgG secondar y antibodies conjugated with an Alexa Fluor-594, diluted in PBS (1:1000) for 1 h at room temperature in the dark.After incubation, the cells were washed again 3 times with PBS.Membrane with cells were placed on the objecti v e slide in mounting medium with 4 ,6-diamidino-2-phenylindole (DAPI) and cov er ed with a cov erslip.Samples wer e visualized with a Leica TCS SP8 confocal microscope using Diode 405 nm, diodepumped solid-state (DPSS) 561 nm, lasers.Photos were taken with a 63 × immersion lens.
Cytokine concentrations were calculated using the standard curve generated by 5-parameter logistic r egr ession method.

Sta tistical Anal ysis
Statistical anal ysis w as performed using Graph Pad Prism R softw ar e 8.0.2 (GraphPad Softw ar e, Inc., USA).Differences between the groups were compared by 2-way ANOVA (following Tukey's multiple comparisons test) or nonparametric Kruskal-Wallis test (following Dunn's multiple comparisons test).Results were considered as significant when P < 0.05.

Plasma of COVID-19 Patients Does Not Affect TEER
To investigate the potential effect of patient-deri v ed b lood plasma on the integrity of the BBB, iBECs were seeded on Transwell inserts and treated with patient-derived blood plasma, obtained fr om thr ee differ ent gr oups: healthy v olunteers, COVID-19 patients, and r ecov er ed COVID-19 patients.The treatment inv olv ed appl ying 50% (v/v) plasma to the cells.
We first tested the effects of exposure of endothelial monolayer to pooled plasma administered at the apical, basolateral or both sides of the Tr answ ell inserts.Our results show that plasma contacting only basolateral or apical and basolateral sides completely inhibited TEER of iBEC monolayers, whereas iBECs exposed to the plasma only from the apical side demonstrated normal TEER readings ( Figure 1 A).
Immediately after TEER measurements, cells were fixed for immunostaining with antibodies against TJ proteins claudin-5, occludin, and zonula occludens-1 (ZO-1).Confocal microscopy images showed that exposure to plasma from basolateral and both sides disrupted TJs in iBEC monolayers as evidenced by mor e irr e gular and fr a gmented staining patterns of TJ pr oteins when compared to controls or cells treated from the apical side ( Figure 1 B).
Next, we investigated the effects of plasma from healthy, r ecov er ed subjects and COVID-19 patients administered to the apical side of the monolayer on its electrical resistance.Dynamic changes in TEER values show a similar trend for all groups ( Figure 2 B), starting to rise around 12 h after seeding the cells on Tr answ ell inserts and reaching the values of ∼2500 cm 2 after 24 h after seeding when the plasma was added (zero time point).Immediately after the addition of plasma, TEER dropped slightly pr oba b l y r eflecting injection artefact.Following this dr op, TEER continued to rise steadily and reached its peak exceeding 4000 cm 2 around 12 h after the addition of plasma.After reaching the peak, the TEER slowly started to decline but remained above 2000 cm 2 at the end of the 48-h measurement period.Analysis of TEER data did not r ev eal any significant differences between experimental groups at any time point during the 48-h observation period after plasma addition.

Cytokine Profiles of Plasma of COVID-19 Patients, Reco vered Pa tients, and Healthy Controls
We compared concentrations of cytokines, chemokines, and growth factors in the plasma of hospitalized COVID-19 patients with healthy and r ecov er ed indi viduals.We first used a 45anal yte m ultiplex imm unoassay panel to measur e plasma lev els of 45 cytokines in COVID-19 patients ( N = 18), r ecov er ed ( N = 13), and healthy ( N = 9) subjects ( Figure 2 ).Analysis of cytokine profiles showed increased plasma levels of IP-10, HGF, VEGF-A, IL-7, IL-18, and MCP-1/CCL2 in COVID-19 patients.However, statistically significant differences between tested groups were found only for IP-10 and HGF.Concentrations of IP-10 were statistically higher in blood plasma of COVID-19 patients compared to recovered subjects ( P < 0.05).Statistically significantly higher levels of HGF were detected in the plasma of COVID-19 patients compared to healthy ( P < 0.05) and r ecov er ed ( P < 0.001) subjects.We also observed that SCF levels in COVID-19 patients were statisticall y significantl y lower than in healthy individuals ( P < 0.05).Eotaxin/CCL11, BDNF , LIF , PIGF-1, PDGF-BB, MIP-1 α/CCL3, MIP-1 β/CCL4, RANTES/CCL5, SDF-1 α, and VEGF-D were detected in the plasma of all 3 groups but showed no significant differences between the tested groups.From the 45 analytes assessed, 15 were below the lower limit of quantification in 90% of all samples (irr especti v e of gr oup) and wer e ther efor e excluded fr om further analysis.A total of 12 analytes were observed at detectable levels in some samples of all 3 groups but showed no significant differences between the tested groups and were also excluded from further analysis.
The final analysis of the cytokine assays showed that the levels of IP-10, HGF, IL-18, eotaxin, RANTES, and MIP-1 β were statisticall y significantl y differ ent between the tested gr oups ( Figure 3 ).Concentrations of IP-10 and HGF were statistically higher in blood plasma of COVID-19 patients compared to recovered ( P < 0.0001) and healthy ( P < 0.05 and P < 0.001) subjects.Statisticall y significantl y higher lev els of IL-18 wer e detected in the plasma of COVID-19 patients compared to recovered subjects ( P < 0.05).Meanwhile, the levels of eotaxin in healthy subjects wer e statisticall y significantl y higher than in COVID-19 patients and r ecov er ed indi viduals ( P < 0.05).Plasma MIP-1 β levels were significantly higher in healthy subjects than in recover ed indi viduals ( P < 0.05), while RANTES levels were significantly higher in recovered than in healthy subjects ( P < 0.05).
Thus, our results showed that COVID-19 significantly incr eases the lev els of IP-10, HGF, and IL-18 in the plasma of COVID-19 patients compared to healthy and/or r ecov er ed individuals.As all COVID-19 patients in our study were hospitalized,

Discussion
In the present study, we investigated the effects of COVID-19 patient-deri v ed b lood plasma on the in vitr o model of human BBB.The apical side of iBECs monolayers was directly exposed to the heat-inacti v ated patient or healthy subject-deri v ed plasma and TEER w as contin uousl y monitor ed during a 48-h period.Using this experimental model, w e w er e a b le to achiev e peak TEER values exceeding 4000-5000 cm 2 , thus showing compatibility of BECs to the exposure with human plasma.We also found that exposure of iBECs to the plasma either from basolateral or basolateral and apical sides completely disrupted barrier function; at the same time plasma administration at the apical side only did not decrease TEER ( Figure 1 ).Apicobasal polarization is crucial for the proper barrier function of the BECs. 30 , 31xtrav asated b lood plasma components ar e dir ectl y toxic to the br ain parenc hyma and also trigger massi v e secondar y disruption of the BBB, leading to the brain oedema and tissue necrosis. 32Our in vitro model demonstrates the importance of apicobasal polarity for BEC barrier function and could be used for the modeling of different pathological conditions related to the secondary BBB injury.
We found no changes in the TEER values between the experimental groups consisting of SARS-CoV-2-infected patients and healthy or r ecov er ed indi viduals.We suggest that our results could be explained by the following: (i) experimental design and (ii) heat-insensiti v e COVID-19-associated b lood plasma inflammatory factors do not affect BEC barrier functions directly.
In this study, we used heat-inacti v ated b lood plasma (56 • C for 30 min).Heat inacti v ation w as necessar y to pr ev ent complement-mediated lysis in cell cultures and as a safety precaution for the effecti v e inacti v ation of SARS-CoV-2 viral particles.However, heat inactivation can affect quantities and biological activity of soluble factors present in plasma.Comparison of pr oteomic pr ofiles r ev ealed downr egulation of many cytokines, chemokines, and growth factors in heat-inactivated plasma. 33t was also demonstrated that plasma from patients hospitalized with acute SARS-CoV-2 infection decreased transendothelial resistance of human lung microvascular endothelial cells, but these destructi v e effects wer e susce ptib le to heat inactiv ation. 34Finall y, complement components can by themselv es damage BBB.Various injuries may initiate the inflammatory response of BECs by promoting activation and binding of circulating complement components to the a b luminal membranes, stim ulating the secr etion of pr oinflammator y factors and instigating BBB leakage. 35 , 36We, ther efor e, cannot exclude the possibility that heat-sensiti v e components pr esent in the plasma of SARS-CoV-2-infected patients may affect the electrical resistance of iBEC monolayers.Further studies are needed to address this question.
Our data demonstrate that heat-insensiti v e COVID-19associated blood plasma inflammatory factors do not affect BEC barrier functions dir ectl y.Howev er, we do not exclude the possibility that COVID-19-associated plasma factors can trigger BECs to r elease pr oinflammator y factors into the br ain parenc hyma leading to secondary BBB damage.For instance, infection of human iPSCs-deri v ed BECs with SARS-CoV-2 did not affect barrier properties but upregulated IFN-γ signaling, and these r esults wer e consistent with histopathological studies showing upr egulated IFN-γ pathw ay in COVID-19 human neur ov ascular unit. 23Finally, BBB can selectively transport several proinflammatory cytokines from the peripheral circulation into the brain par enchyma and pr omote secondar y BBB injur y. 37 In the futur e, After analyzing 45 cytokines in the plasma of COVID-19 patients, we found that only the levels of IP-10, HGF, and IL-18 wer e significantl y higher in COVID-19 patients compared to healthy and/or r ecov er ed subjects.IL-18 is known as a proinflammator y cytokine inv olv ed in host defense against infections and regulating innate and acquired immune responses. 38Our r esults ar e in a gr eement with other studies showing correlations of IL-18 serum levels with other markers of inflammation and disease severity. 8 , 39-41nterfer on-γ -induced pr otein (IP-10), also known as small inducible cytokine B10, belongs to the CXC chemokine family (also known as CXCL10).This 10-kDa cytokine r ecruits imm une cells, including T cells, natural killer cells, and macr opha ges, to the inflamed tissue in inflammatory diseases.IP-10 induces T cells and is ther efor e important for anti vir al defense .It is upregulated in the blood of hospitalized COVID-19 patients. 42hese data corroborate our study, where IP-10 was significantly higher in the plasma of COVID-19 patients than in the plasma of healthy ( P < 0.05) and r ecov er ed ( P < 0.0001) subjects.Comparison of cytokine expression profiles between critically ill, severe, and moderate COVID-19 cases r ev ealed a significant association of IP-10 with disease severity. 43Previous studies demonstrated that IP-10 is one of the most abundant and the earliest chemokines associated with BBB damage in various viral infections. 44Besides COVID-19, IP-10 was also associated with the severity of diseases caused by other viruses such as MERS-CoV and influenza. 45 , 468][49] In J apanese ence phalitis, for example, IP-10 promoted BBB damage by inducing TNF-α pr oduction thr ough the JNK-c-Jun signaling pathw ay; in turn, TNF-α affected the expression and distribution of tight junctions in brain micr ov ascular endothelial cells r esulting in BBB damage . 44How ever, in our study, TNF-α levels were below the lower limit of quantification in 95% of all tested samples.This possib l y reflects an inappropriate inflammatory response that may occur during COVID-19. 50e also found that plasma HGF lev els wer e significantl y higher in COVID-19 patients than in r ecov er ed ( P < 0.001) and healthy ( P < 0.05) subjects.Hepatocyte growth factor produced by stromal and mesenchymal cells regulates epithelial cell proliferation, motility, morphogenesis, and ang iog enesis. 51e patocyte gr owth factor is a pleiotropic cytokine with antiinflammator y pr operties that plays a key role in lung tissue r e pair and can modulate the adapti v e imm une r esponse. 52 , 53ncr eased HGF pr oduction induced by lung injury promotes tissue r e pair. 51In general, incr eased circulation of gr owth factors such as HGF is associated with r e pair mechanisms following acute lung injury during SARS-CoV-2 infection. 54However, a marked increase in HGF was also significantly correlated with disease severity. 51 , 54He patocyte gr owth factor is also a marker of neutr ophil acti v ation, it w as consider ed as one of the str ongest indicators of critical illness in COVID-19. 55Mor eov er, upr egulated HGF in intensi v e car e unit patients is elevated in chronic inflammatory diseases. 56In our study, HGF levels in the blood of COVID-19 patients were 11 times higher than in r ecov er ed subjects ( P < 0.001) and 5 times higher than in healthy subjects ( P < 0.05).Our findings are in a gr eement with studies showing an association of HGF with the severity of COVID-19. 43 , 52hroughout the course of the COVID-19 pandemic, various strains of the SARS-CoV-2 virus have emerged, exhibiting different c har acteristics and potential pathological effects. 57It is worth noting that most of the pub lished r esear c h investigating cytokine profiles and viral impact on the BBB was conducted during the early and middle stages of the pandemic.However, for our study, blood plasma samples were collected during the period fr om Februar y 2022 to April 2023, which r e pr esents a later stage of the pandemic.It is plausible to speculate that the novel cir culating vir al str ains, whic h emerged during this period, may have undergone genetic variations and evolutionary changes, potentially leading to alterations in cytokine profiles and their influence on BBB.
In conclusion, our study demonstrates that blood plasma from COVID-19 patients, although enriched with pathologically r elev ant cytokines and chemokines, does not affect BBB electrical resistance.Our findings warrant further resear c h to explore possib le indir ect mechanisms of pathological BBB r emodeling during COVID-19.

Figure 2 .
Figure 2. The effect of plasma from healthy volunteers, COVID-19 patients, and recovered subjects on TEER of human iBECs monolayers.(A) Experimental protocol.(B) Time course of TEER ( cm 2 ) changes of iBECs grown on transwell inserts and treated with plasma from apical side; data are presented as a mean ± SD.Healthy volunteers ( N = 20), COVID-19 patients ( N = 24), and recovered subjects ( N = 17).a systematic comparison of molecular responses between BECs exposed to COVID-19 and control plasma may r esolv e this issue.After analyzing 45 cytokines in the plasma of COVID-19 patients, we found that only the levels of IP-10, HGF, and IL-18 wer e significantl y higher in COVID-19 patients compared to healthy and/or r ecov er ed subjects.IL-18 is known as a proinflammator y cytokine inv olv ed in host defense against infections and regulating innate and acquired immune responses.38Our r esults ar e in a gr eement with other studies showing correlations of IL-18 serum levels with other markers of inflammation and disease severity.8 , 39-41 Interfer on-γ -induced pr otein (IP-10), also known as small inducible cytokine B10, belongs to the CXC chemokine family (also known as CXCL10).This 10-kDa cytokine r ecruits imm une cells, including T cells, natural killer cells, and macr opha ges, to the inflamed tissue in inflammatory diseases.IP-10 induces
and the severity of neurological manifestations of COVID-19.This may facilitate the development of potent diagnostic tools, ena b ling earl y pr ediction and pr ev ention of neur ologic complications in COVID-19 patients.