CD59 Protects Primary Human Cerebrovascular Smooth Muscle Cells from Cytolytic Membrane Attack Complex

Cerebral amyloid angiopathy is characterized by a weakening of the small and medium sized cerebral arteries, as their smooth muscle cells are progressively replaced with acellular amyloid β, increasing vessel fragility and vulnerability to microhemorrhage. In this context, an aberrant overactivation of the complement system would further aggravate this process. The surface protein CD59 protects most cells from complement-induced cytotoxicity, but expression levels can fluctuate due to disease and vary between cell types. The degree to which CD59 protects human cerebral vascular smooth muscle (HCSM) cells from complement-induced cytotoxicity has not yet been determined. To address this shortcoming, we selectively blocked the activity of HCSM-expressed CD59 with an antibody and challenged the cells with complement, then measured cellular viability. Unblocked HCSM cells proved resistant to all tested concentrations of complement, and this resistance decreased progressively with increasing concentrations of anti-CD59 antibody. Complete CD59 blockage, however, did not result in total loss of cellular viability, suggesting that additional factors may have some protective functions. Taken together, this implies that CD59 plays a predominant role in HCSM cellular protection against complement-induced cytotoxicity. Over-expression of CD59 could be an effective means of protecting these cells from excessive complement system activity, with consequent reduction in the incidence of microhemorrhage. The precise extent to which cellular repair mechanisms and other complement repair proteins contribute to this resistance has yet to be fully elucidated.


Background
Cerebral amyloid angiopathy (CAA) is a vascular pathology characterized by the thickening of small and medium arteries in the brain due to the deposition of amyloid β (AB) within the media and adventitia of vessel walls.Over time, the smooth muscle (SM) tissue is replaced with acellular AB plaques, accompanied by development of structural fragility and vulnerability to rupture [1].Alternatively, severe cases of microvascular AB deposition are also associated with cerebral microinfarcts [2].CAA pathology is associated with several diseases and is estimated to coincide with 75-90% of all cases of Alzheimer's disease [1,[3][4][5].Cerebrovascular dysfunction is now recognized as a risk factor for the onset of dementia.CAA-induced microhemorrhages may contribute to the progression of dementia by increasing the amount of iron present in the brain, resulting in oxidative damage and neurodegeneration [1,[6][7][8].By one statistical estimate, the average contribution of CAA to the total cognitive decline of a icted individuals is 15.7% [5].No effective treatment for CAA currently exists [9,10].
Our prior work suggested a role for the membrane attack complex (MAC) in the progression of CAA [3].
The MAC, also referred to as C5b-9, is an assembly of plasma proteins which together form a transmembrane pore structure directly connecting the internal and external cellular environments (Fig. 1a).The resultant unrestricted osmotic exchange initiates cytolysis [11][12][13].MAC formation is the convergent endpoint of the complement cascade, which is a series of protein interactions that is initiated through three distinct pathways with their speci c activation mechanisms: classical, alternative, and lectin (Fig. 1b).The classical pathway is activated by antigen-antibody complexes, the alternative pathway can be activated spontaneously, and the lectin pathway is activated via recognition of conserved pathogenic carbohydrate motifs [14].The complement cascade represents an aspect of the innate immune response, primarily targeting susceptible pathogens such as gram-negative bacteria [15].Host cell membranes are likewise vulnerable to the cytolytic activity of MAC.To protect themselves, most human cell types and tissues express surface membrane-bound complement-inhibiting molecules.One of the most important is the glycophosphatidylinositol-anchored surface protein CD59, also referred to as protectin [16][17][18][19][20][21].When an incomplete MAC complex (C5b-8) inserts itself into a host cell membrane, a functional CD59 protein binds the complex and blocks the spontaneous incorporation of multiple C9 molecules.This prevents the completion of the MAC and its characteristic transmembrane pore structure, thereby preserving membrane integrity [22][23][24].While incidental MAC insertions may still occur, causing mild membrane damage, this remains sub-lytic due to cellular repair mechanisms [25].Other surface membrane-bound complement regulatory proteins can intervene at earlier points in the complement cascade to prevent MAC formation.CD46 is a cofactor which enables the inactivation of the C3b and C4b complement proteins, and CD55 disrupts C3 and C5 convertases, which are required for C3b deposition [26].
Figure 2 The complement system protein cascade is initiated through antigen-antibody complexes (classical pathway), recognition of conserved pathogenic carbohydrate motifs (lectin pathway), or spontaneous hydrolysis of the C3 complement protein (alternative pathway).All three pathways converge by promoting C3 hydrolysis into C3a and C3b protein fragments.C3b promotes additional C3 hydrolysis, forming a self-ampli cation loop.C3b also promotes the cleavage of the C5 protein to form the protein fragment C5b, which complexes with other complement proteins to form the membrane attack complex (MAC) Vulnerability to complement-mediated cytolysis can vary across cell types according to their levels of CD59 surface expression [11,23].Low CD59 expression can sensitize cells to complement damage while high expression confers resistance [18,21,27].Normal expression levels of CD59 vary with tissue location but can also uctuate under abnormal conditions, such as Alzheimer's disease, organ transplantation, or cancer [16,18,20,[27][28][29].Regardless of expression level, CD59 protection is speciesrestricted and will not inhibit complement proteins endogenous to other species [23].A number of reports documented decreased CD59 protective potential.Genetic mutations can compromise CD59's anti-MAC functionality [27,30].Insu cient CD59 expression is believed to be associated with a number of conditions including: paroxysmal nocturnal hemoglobinuria, Alzheimer's disease, age-related macular degeneration, post-transplant organ rejection, and genetic demyelinating neuropathy in some patients [16,27,[31][32][33].Lower CD59 expression in the intracranial artery is associated with complement activation, in ammation, and possible weakening of the arterial wall [34].
Earlier work suggested that CAA-a icted cerebral blood vessels have increased MAC deposition without a compensatory upregulation of surface CD59 [3].Over time, the cumulative cytotoxic and cytolytic damage could play a role in the gradual characteristic destruction of human cerebral vascular smooth muscle cells (HCSM).However, this potential sensitivity of primary HCSM cells to complement attack has not yet been reported.In this study, the primary HCSM cells were isolated from small blood vessels of the brain, obtained during routine temporal lobe biopsies.Their surface-expressed CD59 proteins were then inhibited in a controlled dose-dependent manner, to evaluate the role of CD59 for cellular resistance against complement-dependent cytotoxicity.

Cell isolation and culture
Brain tissues were obtained from a routine temporal lobe biopsy, performed at Loma Linda University Medical Center, as approved by the institutional IRB (#5170023).Small blood vessels were manually isolated with forceps, washed, and sonicated to obtain primary human cerebral vascular (HCV) cells.The cells were incubated, expanded, and passaged in SM cell media supplemented with 2-5%V/V FBS, 100 units/mL penicillin, 100 ug/mL streptomycin, and SM cell growth supplement (50 µL/mL).Cells were cryopreserved in complete SM cell media supplemented with 40% v/v FBS and 10% v/v glycerol.Active cultures were kept incubated at 5% CO 2 at 37 O C in complete SM cell media, which was changed every two or three days.
The HCV cell phenotype was identi ed to be HCSM by visual con rmation of co-expression of alpha smooth muscle actin (αSMA) and desmin.Short tandem repeat analysis was carried out by the University of Arizona's Genetics Core ((Facility RRID:SCR_012429) https://azgc.arizona.edu))and con rmed that the HCSM cells were genetically distinct from available cell lines.

Immuno uorescence microscopy
Collagen-coated glass coverslips were placed in 6-well plates at 2-3 coverslips per well.HCV cells were seeded into the wells at a density that approximated 2x10 4 cells per coverslip.Coverslips were xed for 20 minutes with 4% paraformaldehyde in DPBS with calcium and magnesium (+ Ca ++ /+Mg ++ ) at 18 O C. Some coverslips were stored in DPBS + Ca ++ /+Mg ++ at 4 O C for 1-2 weeks and the cells subsequently permeabilized with 0.2% triton X.All cells were blocked with 1% BSA in DPBS + Ca ++ /+Mg ++ .Then, the cover slips were rinsed three times.Primary and secondary antibodies were applied as directed by the supplier and incubated at 18 O C for 1 hour each.Coverslips were mounted on glass microscope slides, allowed to cure overnight, and sealed with nail polish.

Alternative complement pathway activity test
Aliquots (2mL) of normal human serum (NHS) were thawed in a < 4 O C water bath and kept on ice to minimize complement protein degradation.Heat-inactivated serum (HIS) was created by heating thawed NHS in a 56 O C bead bath for 30 minutes.Serum dilutions ranging from 1:8 to 1:16 were established by adding volumes of serum and gelatin veronal buffer to rabbit RBCs (1.5x10 7 cells/mL).Gelatin veronal buffer or double distilled H 2 0 were added to rabbit RBCs to establish blank and total lysis controls, respectively.All RBC suspensions were subsequently incubated at 37 O C in 5% CO 2 for 30 minutes with a gentle inversion halfway through.Cells suspensions were centrifuged at 1,500 x g for 5 minutes to sediment the RBCs, and the supernatants were transferred to a at-bottomed 96-well plate at 100 µL/well as technical triplicates.To each well was added 100 µL of double-distilled H 2 0 and the absorbance values were measured at 540 nm.The percentage of total RBC lysis from each serum dilution was calculated with the following equation: where L = % lysis, A t(av) = average test absorbance, A b(av) = average blank absorbance, and A T(av) = average absorbance after total lysis.

Complement-dependent cytotoxicity assays
HCSM cells were seeded into 96-well plates at 16,700 cells/well and incubated at 37 O C in 5% CO 2 to establish full con uence overnight.The following day, concentrations of normal or heat-inactivated human serum, ranging from 0-100% V/V, in SM cell media were applied to the wells.The wells were incubated at 37 O C in 5% CO 2 for one hour, then washed three times with DPBS + Ca ++ /+Mg ++ .Cells were subsequently incubated with a 10% V/V solution of resazurin reagent in culture media for a period of two hours.The metabolic reduction of resazurin to resoru n was quenched by addition of SDS (3%V/V nal), and the resultant solutions were stored in the dark at 4 O C, and analyzed within three days.Resoru n uorescence intensity values were measured at 560/590 nm, outlier values were excluded, and average uorescence values were calculated for each condition.The uorescence values of cultures incubated with NHS were normalized to their corresponding HIS-incubated cultures to calculate the percentage difference in cellular viability.All assays were performed in triplicate and the resultant percent averages and standard errors were calculated.One-way ANOVA analysis and Tukey's Honest Signi cant Difference post-hoc tests were applied to evaluate statistical signi cance.

L = 100 , (A t(av) − A b(av) ) (A T (av) − A b(av) )
In the experiment which examined CD59's contribution to complement resistance, HCSM cells were seeded as described above.Function-blocking anti-CD59 primary antibody YTH53.1 (αCD59) was applied to the wells at 0-50 µg/mL in 100 µL DPBS + Ca ++ /+Mg ++ per well and incubated at 18 O C for 10 minutes before removal.Human serum was diluted to 80% V/V in SM cell media and added to the wells in biological triplicate per antibody concentration.Cells were incubated, washed, and the cellular viability was assessed with resazurin as described in the paragraph above.Outlier uorescence values were excluded, and average uorescence values were calculated for each condition.The average uorescence value of each condition was normalized to that of the 0 µg/ml condition to calculate the percentage differences in cellular viability.These assays were performed in triplicate and the resultant percent averages and standard errors were calculated.

Statistical Analysis
One-way ANOVA analysis was performed using the Microsoft Excell software's Data Analysis Tool.The subsequent Tukey's Honest Signi cant Difference post-hoc test was performed using GraphPad Prism version 9.3.1.

Figures
Illustrations were adapted from "Formation of the Membrane Attack Complex" and "Roles of the complement Cascade in Innate Immunity" by BioRender.com(2024).Retrieved from https://app.biorender.com/biorender-templates.Microscopy images were adjusted using the Fiji distribution of ImageJ (NIH).Graphs were constructed using Excel (Microsoft).

Generation of primary HCSM cells from human cerebral vasculature
To explore how CD59 plays a role in primary HCSM cellular resistance to complement-dependent cytotoxicity, we rst generated primary cells from a brain sample of a temporal lobe resection patient as described in materials and methods.As Fig. 3a shows, the visible morphology of the HCV cells resembles the synthetic phenotype of SM cells as opposed to the contractile phenotype, as de ned earlier [35].All (100%) of the examined cells express external plasma membrane-bound CD59 (Fig. 3b).Brighter uorescence observed at the HCSM cell margins indicated higher CD59 expression in those areas.Cytoskeletal elements αSMA (Fig. 3c), and desmin (Fig. 3d) were both expressed and extensive overlap was observed between both proteins (Fig. 3e) [36].Functional evaluation of CD59 in HCSM cell resistance to complement attack CD59 contributes to complement attack resistance by preventing the formation of a complete MAC structure.To quantify the level of protection that CD59 imparts to primary HCSM cells, a source of complement proteins capable of generating functional MACs was required.To verify that the human serum used in these experiments can generate functional MACs, rabbit erythrocytes were utilized because they are vulnerable to cytolysis through spontaneous activation of the alternative complement pathway [37,38].Heat-inactivated human serum acted as a negative control since complement proteins are sensitive to thermal denaturation [39,40].Rabbit erythrocytes were subjected to a series of human serum concentrations, normal or heat-inactivated, and absorbance values were measured for liberated hemoglobin to determine the percentage of lysed cells.As Fig. 4 shows, standard curves were established demonstrating that the complement cascade is active in normal human serum, whereas minimal complement activity is detected in the heat-inactivated serum.The concentration of serum required to lyse 75% of the rabbit erythrocytes was calculated to be approximately 7.4% V/V.
Once the NHS was con rmed to be a source of functional complement, then the general HCSM cellular resistance was tested against complement-dependent cytotoxicity.Cultured HCSM cells were challenged with up to the maximum concentration of normal human serum or heat-inactivated human serum.The relative viabilities of the challenged cell cultures were subsequently measured by their metabolic conversion of resazurin to uorescent resoru n.As Fig. 5 shows, HCSM cells proved resistant to complement-induced cytotoxicity.The apparent minor drop in HCSM viability observed under the 60-100% V/V NHS conditions did not reach statistical signi cance according to Tukey's post-hoc test.
Figure 5 Endogenous HCSM cellular resistance to complement-dependent cytotoxicity.Cultured cell viabilities were measured by resazurin assay and then normalized to their respective HIS serum conditions to obtain relative percentages.Three separate assays were performed with each condition examined in triplicate.One-way ANOVA testing and Tukey's HSD post-hoc analysis showed no signi cant differences in outcome between conditions.Error bars are +/-SE The degree to which endogenous CD59 surface expression protects HCSM cells from complementdependent cytotoxicity was evaluated by blocking the CD59 with increasing concentrations of αCD59 [41,42].Then, the blocked HCSM cells were challenged with normal human serum (80% V/V) and the relative viabilities were subsequently measured by their metabolic conversion of resazurin to uorescent resoru n to establish the dose-response curve seen in Fig. 6.The non-linear regression t of the data produced the following equation: The cytotoxic EC 50 of αCD59 was calculated to be 6.1 µg/ml.

Identi cation HCSM cells
We identi ed our isolated primary human cerebral vascular cells as HCSM cells by their visible morphology and endogenous expression of both αSMA and desmin (Fig. 3c,d).The visible morphology appeared to be consistent with the SM cell synthetic phenotype.In contrast with the spindle-shaped contractile phenotype characteristic of SM cells in healthy blood vessels, the synthetic phenotype is a less differentiated form that is associated with SM cell migration, proliferation, and post-insult vessel repair [35].While αSMA is also present in myo broblasts, a cell type which could hypothetically be extracted and cultured by accident, the muscle cell marker desmin is only weakly expressed in one subtype of myo broblast [43,44].Strong expression of both cell markers is consistent with these cells being physiologically SM.

Con rmation of complement activity in normal human serum
Nucleated cells are not lysed by the MAC unless multiple complete channels are formed across the plasma membrane.Complement-induced damage is resisted through increased cell proliferation, inhibition of apoptosis, and the elimination of terminal complexes from the plasma membrane [45,46].By contrast, non-nucleated erythrocytes are relatively vulnerable to action of the terminal complement complex, as only a single completed complement channel is required to initiate cytolysis [46,47].
Measurement of complement-driven hemolysis is a standard method to determine the relative levels complement pathway activity present in human serum.Rabbit erythrocytes are vulnerable to spontaneous activation of the alternative complement pathway and therefore are used to determine the percentage of human serum required to lyse 75% of the suspended rabbit erythrocytes.Comparison of the respective percent hemolysis of NHS and HIS (Fig. 4) con rms that the normal human serum used in these studies contains the active set of complement proteins required to form functional membrane attack complexes.

HCSM cellular resistance to complement-dependent cytotoxicity
Cultured HCSM cells exhibiting the synthetic phenotype cells appear mostly resistant against high levels of complement until a su cient proportion of their surface CD59 molecules are bound and inactivated by an anti-CD59 antibody (Fig. 5-6).It is worth noting the possibility that complement attack on the HCSM cells may have proceeded through different pathways depending on the assay.HCSM cells that were not blocked with αCD59 could only have been affected through the alternative pathway because antibodies are required to activate the classical pathway, whereas HCSM cells blocked with αCD59 may have been affected through both the alternative and classical pathways [14].The IgG 2 antibody subclass to which αCD59 belongs is known to be a weak activator of complement, requiring high epitope concentrations in order to activate the classical pathway [49].Thus, under normal physiological conditions, the endogenous CD59 expression levels of HCSM cells are su cient to protect these cells from complement-driven cytolysis.Because our experiments utilized human serum as the complement protein source, levels of complement did not exceed the range found in whole blood.Therefore, we currently cannot predict whether excessive complement levels might overwhelm the endogenous CD59 or whether the HCSM cells can upregulate CD59 to compensate.
Apparent saturation of CD59 with αCD59 did not produce uorescence values that indicate zero cellular viability, suggesting that complete inhibition of CD59 is not su cient to eliminate all HCSM cells within a culture through complement-mediated cytotoxicity.This may be due to the protective action of other plasma membrane-bound complement regulators, such as CD55 and CD46, in conjunction with the selfrepair mechanisms employed by nucleated cells.
Taken together, these data indicate that CD59 expression and functionality are critical aspects of endogenous HCSM cellular defense against complement-mediated cytotoxicity.Further, this suggests that arti cial enhancement of CD59 expression may be an effective means of protecting HCSM cells if the normal expression declines or if complement system activity becomes aberrantly high.This approach would leave the upstream complement cascade intact, avoiding potential issues such as increased susceptibility to bacterial infection and interference with certain neurologic processes [25].Insertion of additional CD59 gene copies into the nuclear DNA, for example, has been shown to increase resistance to complement-induced cytolysis in a number of cell types [32,50,51].The degree to which alternative complement regulatory proteins and cellular repair mechanisms mitigate complement damage to HCSMs remains to be further elucidated.

Declarations Data Availability
Data are available upon request.
Authorship: All authors contributed to the study conception and design.Initial primary cell cultures were isolated from post-surgical specimens by YN.Material preparation, data collection and initial analysis were performed by CW.Subsequent analyses were performed by CW, SS and DSB.All authors read and approved the nal manuscript, with the exception of WMK who passed away.The complement system protein cascade is initiated through antigen-antibody complexes (classical pathway), recognition of conserved pathogenic carbohydrate motifs (lectin pathway), or spontaneous hydrolysis of the C3 complement protein (alternative pathway).All three pathways converge by promoting C3 hydrolysis into C3a and C3b protein fragments.C3b promotes additional C3 hydrolysis, forming a self-ampli cation loop.C3b also promotes the cleavage of the C5 protein to form the protein fragment C5b, which complexes with other complement proteins to form the membrane attack complex (MAC)  Endogenous HCSM cellular resistance to complement-dependent cytotoxicity.Cultured cell viabilities were measured by resazurin assay and then normalized to their respective HIS serum conditions to obtain relative percentages.Three separate assays were performed with each condition examined in triplicate.
One-way ANOVA testing and Tukey's HSD post-hoc analysis showed no signi cant differences in outcome between conditions.Error bars are +/-SE

Figure 6
Figure 6 HCSM cellular resistance to complement decreases following treatment with αCD59.Panel 1: The residuals of measured cell viability compared to the predicted curve values.Panel 2: The αCD59 dose-response curve created by logarithmic four-factor regression.Error bars are +/-SE

Figures Figure 1
Figures

Figure 4 Alternative
Figure 4