IgG Propels Atherosclerosis by Noncanonically Activating Macrophages

77 78 Despite being a central component of adaptive immunity and a highly abundant serum protein, the 79 contribution of IgG to the milieu of atherosclerosis remains unappreciated. Here, we identify a pro-80 atherogenic role for IgG as it activates an innate immune cascade, independent of its classical antigen-81 neutralizing function. Analyses of human coronary artery plaques reveal a positive correlation between 82 IgG and cardiovascular and cerebrovascular disease severity . Integrated single-cell plaque analyses 83 localize IgG, coinciding with its recycling receptor FcRn, to pro-inflammatory and foamy macrophages. 84 Genetic ablation of FcRn in myeloid cells prevents IgG from accumulating in mouse atherosclerotic 85 lesions, diminishing plaque size and inflammation. Mechanistically, IgG acts as an endogenous ligand for 86 TLR4, triggering NF- κ B–NLRP3 inflammasome signaling without requiring its antigen-binding domain. 87 Additionally, IgG accelerates macrophage foam cell formation through upregulation of downstream 88 effector LCN2. Our work uncovers a role for previously overlooked adaptive immune molecules in the 89 pathogenesis of atherosclerosis through a noncanonical mechanism linked with innate immunity. 90 91 92 93 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 4

94 Cardiovascular diseases (CVD) are the leading cause of global mortality, a burden primarily driven by 95 atherosclerosis and its complications 1. This multifactorial disease is characterized by fatty streak 96 formation along the arterial wall 2, progressing as elevated circulating low-density lipoproteins (LDL) 97 accumulate in the subendothelial intima 3. Following oxidative modifications, oxLDL is internalized by 98 macrophages, leading to foam cell formation4. This prolonged accumulation develops into a necrotic core, 99 exacerbating plaque instability and rupture risk 5. Despite effective management of hypercholesterolemia 100 by lipid-lowering drugs, only one-third of CVD risk is controlled 6,7. This persistent residual risk 101 underscores the urgent need to deepen our understanding of mechanisms of atherogenesis and to identify 102 novel pathogenic factors for improved therapeutic and preventative strategies. 103 The immune system, comprising of innate and adaptive branches, is essential for host defense and 104 plays a critical role in the pathogenesis of diseases like atherosclerosis 8,9. The classical paradigm holds 105 that the innate immune system provides a rapid, non-specific first line of defense, which is subsequently 106 refined by the more specialized adaptive response 10. In atherosclerosis, a key research challenge is 107 identifying the endogenous damage-associated molecular patterns (DAMPs) that trigger innate immune 108 responses occurring in the arterial wall, as there is little evidence implicating pathogens as the primary 109 inducers9,11. These innate responses are not isolated; antigen-presenting cells of innate origin shape 110 subsequent adaptive responses within atherosclerotic lesions 12,13, and innate-derived cytokines (e.g., IL-1, 111 TNF-α , and IFN) promote inflammation and activate adaptive immune cells, ensuring a coordinated 112 defense14. While substantial evidence supports the classic view that innate immunity directs the adaptive 113 response, this perspective invites a crucial reciprocal question: can adaptive immune components, in turn, 114 regulate innate immunity and influence atherosclerosis progression? 115 Atherosclerotic plaques are continuously exposed to circulating proteins, including 116 immunoglobulins (Igs). Immunoglobulin G (IgG), the most abundant antibody (~80% of serum Igs), 117 comprises antigen-binding (Fab) and crystallizable fragment (Fc) regions. Traditionally, IgG is thought to 118 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 5 contribute to atherosclerosis via the formation of immune complexes (ICs) with antigens like 119 oxLDL15. These ICs engage with Fc gamma receptors (Fc γ Rs) to drive pro-inflammatory responses16. IC-120 activated macrophages exhibit gene expression profiles resembling those in vulnerable carotid plaques 17, 121 and clinical studies report positive correlations between oxLDL-IC levels and CVD severity. While 122 antigen-specific autoantibodies contribute to atherosclerosis development 26, they are unlikely to fully 123 account for the strong association between total IgG levels and CVD risk 18-21. In contrast to the less 124 abundant IgM that has gained more attention in atherosclerotic plaques 25, the role of IgG per se in 125 atherosclerosis has remained overlooked. Recently, IgG has been implicated in aging, obesity, and 126 metabolic dysfunction 22-24. Given that aging and obesity are two major atherosclerosis accelerants, we 127 hypothesized that IgG may directly contribute to atherosclerotic disease progression through mechanisms 128 beyond classical antigen recognition. 129 Moving beyond its traditional role in adaptive immunity, we position IgG as a key component of 130 atherosclerosis, particularly within the plaque microenvironment. Integrating clinical data, genetic 131 approaches and mechanistic studies, we reveal a noncanonical mechanism for IgG to act through the 132 TLR4/NF-κ B axis, to stimulate inflammasome signaling and lipocalin-2 (LCN2)-mediated macrophage 133 foam cell formation, independent of its antigen-binding Fab domain. Collectively, these findings uncover 134 a pathogenic role for IgG in atherosclerosis through a mechanism that directly bridges adaptive and innate 135 immunity, pointing to novel avenues for therapeutic intervention. 136

137 Proteomic profiling of human atherosclerotic plaques associates IgG with disease severity and 138 outcome 139 To identify potential factors directly involved in atherosclerosis, we collected atherosclerotic coronary 140 tissue from 98 coronary artery disease (CAD) patients undergoing simultaneous coronary artery bypass 141 graft (CABG) surgery and coronary endarterectomy (CE). Segments were stratified into peripheral (mild 142 plaque) and core (severe plaque) zones, yielding 91 mild and 110 advanced lesions ( Fig. 1A ). For 143 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 6 controls, 32 healthy coronary segments were obtained from 14 heart-transplant donors. In total, 233 144 specimens were subjected to data-independent acquisition (DIA) proteomic profiling ( Fig. 1B and Table 145 1). Principal component analysis (PCA) revealed a clear distinction among the three groups, validating 146 sampling strategy (Fig. S1A). Notably, all four IgG heavy-chain isoforms (IGHG1, IGHG2, IGHG3 and 147 IGHG4) were elevated in atherosclerotic plaques, with the highest in the severe core region ( Fig. 1C ). 148 The broader IgG repertoire, encompassing both light and heavy chains, followed the same pattern ( Fig. 149 S1B). We next investigated the association of tissue IgG levels with clinical outcomes. Kaplan–Meier 150 analysis revealed elevated plaque levels of IGHG1, IGHG2, and IGHG4 to associate with a trend toward 151 shortened event-free survival, whereas higher IGHG3 was associated with a more favorable course (Fig. 152 1D). Immunohistochemical (IHC) analysis of human coronary arteries corroborated the proteomic data, 153 showing marked IgG deposition within the intima of atherosclerotic lesions (Fig. 1E, Fig. S2A, B). 154 IgG’s tissue accumulation is dependent on its recycling receptor FcRn and not B cell production 22,24. 155 Interestingly, FcRn expression mirrored the pattern found with IGHG, displaying stepwise increases in 156 mild and advanced plaques compared to controls ( Fig. 1F ). Transcriptomic interrogation of four 157 independent public data sets, spanning early and advanced lesions, stable and ruptured plaques, and 158 plaques with or without intraplaque hemorrhage (IPH), consistently showed elevated FCGRT (encoding 159 FcRn) expression in more severe plaques ( Fig. S2C). Clinically, low FcRn protein levels were protective 160 against cardiac events in our cohort ( Fig. 1G ). This association was recapitulated in the transcriptomic 161 Biobank of Karolinska Endarterectomy (BiKE) cohort, where higher plaque FCGRT correlated with an 162 elevated risk of cerebrovascular events (Fig. 1H ). Collectively, these cross-cohort human data indicate 163 that plaque IgG, likely facilitated by FcRn as it medi ates its recycling, positively correlates with the 164 progression of and adverse outcomes in CAD, highlighting its potential importance in atherosclerosis. 165 166 Single-cell resolution of IgG and FcRn in human atherosclerotic plaques 167 To explore the cell-specific roles of IgG and FcRn in human plaques, we analyzed a dataset combining 168 cellular indexing of transcriptomes and epitopes by sequencing (CITE-seq) and single-cell RNA 169 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 7 sequencing (scRNA-seq) of human carotid atherosclerotic plaques 27. Plaques were stratified by disease 170 severity (symptomatic vs. asymptomatic; Fig. 2A ), and cell populations were clustered based on 171 transcriptional profiles ( Fig. 2B ). Interrogating these clusters for FCGRT revealed a predominant 172 expression in macrophages, with lower leve ls in endothelial cells and fibroblasts ( Fig. 2C). Macrophages 173 in human plaques are heterogenous. Among the five annotated macrophage subsets (Mf1–Mf5; Fig. 174 S3A), FCGRT expression was highest in clusters 1 and 5 ( Fig. 2D , and Table 2). Cluster 1, marked by 175 complement component proteins ( C1QA, C1QB, C1QC), CD64, and MHCII, and cluster 5, marked 176 by C1QA-C, immunoglobulin light chain, MHCII, and CD32, also exhibited the highest surface IgG Fc 177 levels in the CITE-seq data (Fig. 2E and Table 2), suggesting macrophage FcRn-dependent IgG retention 178 in the plaque microenvironment. 179 While FCGRT expression did not differ between symptomatic and asymptomatic plaques ( Fig. 180 S3B–G), IgG Fc surface levels were elevated in symptomatic patients—specifically 1.76-fold higher in 181 all macrophage subsets when combined ( Fig. S3H ), with the most pronounced increase (1.97-fold) in 182 cluster 5 ( Fig. S3I-M ). To further dissect this heterogeneity, we evaluated FCGRT and IgG Fc levels 183 across phenotypic subsets ( Fig. 2F ) defined by distinct gene signatures ( Fig. S3N ). Notably, both 184 FCGRT and IgG Fc were most abundant in IL1Bhi, C1Qhi, and foamy macrophages (Fig. 2G–J and Table 185 3), suggesting that IgG accumulation is particularly associated with pro-inflammatory and lipid-laden 186 macrophages within the plaque microenvironment. 187 188 IgG accumulates in atherosclerotic lesions in a macrophage FcRn-dependent manner 189 To investigate the accumulation of IgG and its role in atherosclerosis, we utilized Ldlr -/- mice fed a 190 western diet (WD). Robust IgG deposition was revealed in atherosclerotic lesions ( Fig. 3A, B ), 191 particularly IgG1 and IgG3 subclasses ( Fig. S4A-B). In contrast, plaque IgM was minimal—likely due to 192 structural constraints of its pentameric form 28—and IgA deposits were sparse ( Fig. 3A, B ). FcRn 193 expression was apparent and colocalized with IgG-positive regions in the intima ( Fig. 3C and Fig. S4C ). 194 Interestingly, exogenous IgG purified from healthy mice preferentially accumulated in plaque-rich 195 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 8 regions within the aorta, suggesting plaque IgG is under dynamic exchange towards the arterial wall and 196 unlikely requires autoantibodies ( Fig.3D). Given that macrophages play a central role in atherogenesis 197 and highly express FcRn, we hypothesized that macrophage FcRn drives plaque IgG accumulation. To 198 test this, we employed myeloid-specific FcRn knockout mice (mKO) and transplanted their bone marrow 199 into irradiated Ldlr-/- recipients. After 16 weeks of WD feeding ( Fig. 3E ), mKO-transplanted mice 200 exhibited significantly reduced plasma IgG (but not IgM) levels (Fig. 3F) and a near-complete loss of IgG 201 in aortic arch ( Fig. 3G, H ) and root lesions ( Fig. 3I, J ). These results recapitulate the accumulation of 202 IgG seen in human plaques and demonstrate its dependence on macrophage FcRn. 203 204 Inhibiting plaque IgG accumulation protects against atherosclerosis 205 To determine whether FcRn-mediated IgG accumulation influences atherosclerosis outcome, we 206 evaluated plaque morphometrics in this cohort. Myeloid-specific deletion of FcRn reduced aortic root 207 lesion area by over 30%, as well as plaque necrosis—a hallmark of unstable, rupture-prone lesions 5—and 208 total acellular area ( Fig. 3K-N ). These benefits occurred without changes in body weight, glucose 209 tolerance or total and non-HDL cholesterol ( Fig. 3O, Fig. S4D-G ) but along with an increase in 210 circulating HDL (Fig. S4H-J). 211 To validate these findings, we employed a PCSK9-overexpression model (AAV-PCSK9) of 212 hypercholesterolemia in mKO and control mice ( Fig. S5A ). Consistently, mKO mice exhibited reduced 213 plaque and circulating IgG and a trending ~30% decrease in aortic root lesion area after 16-wk WD 214 feeding ( Fig. S5B-E ). Necrotic and acellular areas were consistently and significantly smaller, again 215 without alterations in cholesterol levels ( Fig. S5F-I). In the BMT model, mKO-transplanted Ldlr-/- mice 216 showed normal glucose tolerance and insulin sensitivity due to restrained weight gain after irradiation 217 (Fig. S4D ). In contrast, AAV-PCSK9-overexpressing mKO mice gained less weight on WD and 218 displayed improved metabolic parameters (Fig. S5J–L), recapitulating previous observations in obesity24. 219 Crucially, both models demonstrated comparable attenuation of atherosclerosis —irrespective of 220 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 9 metabolic differences—underscoring a direct pro-atherogenic role for IgG within the plaque 221 microenvironment. 222 Next, we sought to evaluate the impact of IgG accumulation on plaque inflammation. In the BMT 223 model, aortic root lesions from mKO-transplanted Ldlr-/- mice exhibited reduced CD68 + macrophage area 224 and fewer Ki67 + macrophages ( Fig. S4K-M ), indicating decreased pan-macrophage activation and their 225 subsequent proliferation29. Furthermore, the inflammatory cytokine IL-1 β was markedly reduced both in 226 plaques and in circulation ( Fig. 4P-Q). Together, these findings demonstrate that preventing plaque IgG 227 accumulation curbs atherosclerosis burden. 228 229 IgG induces an inflammatory milieu in macrophages 230 Given the marked changes in plaque macrophage burden, we investigated whether IgG directly impacts 231 their inflammatory response ex vivo. Bulk RNA-seq of IgG-treated bone marrow-derived macrophages 232 (BMDMs) revealed a pronounced pro-inflammatory shift, alongside a suppression of genes involved in 233 cholesterol efflux, efferocytosis, and lysosomal activity ( Fig. 4A ). This inflammatory phenotype was 234 replicated in RAW264.7 macrophages, where IgG induced Nos2 and Il1b expression to levels comparable 235 to Lipopolysaccharide (LPS) stimulation ( Fig. 4B ). Notably, boiled IgG (b-IgG) failed to activate these 236 genes. Further, IgG—but not denatured IgG—increased IL-1 β secretion (Fig. 4C), confirming that native 237 IgG is required for the pro-inflammatory induction of macrophages. 238 NF-κ B governs macrophage inflammation by phosphorylation-dependent nuclear 239 translocation30,31, with implications in atherosclerosis 32. IgG treatment in RAW264.7 cells induced 240 phosphorylation of NF- κ B (p65) at Ser536 ( Fig. 4D) and triggered its rapid nuclear translocation within 241 15 minutes—a response delayed with LPS and absent with denatured IgG ( Fig. 4E, F and Fig. S6A ). 242 NLRP3 inflammasome activation occurs downstream of NF- κ B and contributes to the pathogenesis of 243 atherosclerosis33. IgG administration induced an upregulation of Nlrp3, Il1b, and Caspase1 transcripts, 244 similar to LPS (Fig. 4G). Of note, monocytes exhibited only a minimal response to IgG stimulation ( Fig. 245 4H), indicating that primed and differentiated macrophages are required for responding to IgG, mirroring 246 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 10 the plaque microenvironment. As an indicator of inflammasome activation 34, ASC speck formation was 247 enhanced by IgG treatment in BMDMs with LPS as a positive control, while this effect was blocked by 248 the NLRP3 inhibitor MCC950 ( Fig. 4I). Consistently, intracellular levels of NLRP3, pro-IL-1 β , and pro-249 IL-18 were induced by IgG (Fig. 4J). Nigericin (signal 2) facilitated the proteolytic cleavage of these pro-250 forms into mature forms (IL-1 β and IL-18) 35, and this process was effectively suppressed by 251 pharmacological inhibition of NLRP3 by MCC950 or NF- κ B by Licochalcone D (LD) ( Fig. 4J ). Our 252

show that IgG can effectively prime and activate the NLRP3 inflammasome. 253 To understand regulatory pathways, we pretreated BMDMs with various kinase inhibitors (MEK: 254 PD98059; PI3K: LY294002; Src/BCR-ABL: Dasatinib; BTK: Ibrutinib; Syk: R406) prior to IgG 255 exposure. None abolished IgG’s induction of inflammasome-associated genes, with some even 256 exacerbating the response ( Fig. S6B ). Furthermore, siRNA-mediated knockdown of Fc γ Rs ( Fcgr1, 257 Fcgr2b, Fcgr3, and Fcgr4 ) in BMDMs only partially or minimally attenuated pro-inflammatory gene 258 induction ( Fig. 4K, Fig. S6C ). Together, these findings demonstrate that IgG stimulates a pro-259 inflammatory state in macrophages primarily through NF- κ B, eliciting downstream activation of the 260 NLRP3 inflammasome. A similar trend was observed in THP-1 induced human macrophages ( Fig. S8H, 261 I). 262 263 IgG directly activates TLR4 in macrophages independently of its antigen-neutralization function 264 TLR4 is a prominent upstream activator of NF- κ B signaling 36,37. In RAW264.7 macrophages, IgG-265 induced NF- κ B phosphorylation and nuclear translocation were completely blocked by the TLR4 266 inhibitor TAK242, (Fig. 5A, B ; and Fig. S6D, E), as was the induction of Nos2 and Il1b (Fig. 5C), IL-1β 267 secretion (Fig. S6F), and NF-κ B phosphorylation ( Fig. 5D). Similar dependences on TLR4, NF- κ B, and 268 NLRP3 were observed in BMDMs and human THP-1 macrophages (Fig. S6G-I). Interestingly, pro-IL-1β 269 and IL-1 β levels were similarly induced by a monoclonal IgG antibody against PD1 (hPD1 mAb), and 270 this effect was also blocked by TAK242 ( Fig. 5E), implying that antigen recognition is not required for 271 TLR4 activation. To test this directly, we treated BMDMs with IgG’s antigen-binding fragment Fab or 272 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 11 constant Fc fragment. Phosphorylation and nuclear translocation of NF- κ B were largely recapitulated by 273 Fc but not Fab treatment ( Fig. 5F; Fig. S7A, B ) accompanied by upregulation of downstream targets 274 (NLRP3, pro-IL-1 β , and pro-IL-18) ( Fig. 5G; Fig. S7C ). Fc-induced activation was also abolished by 275 TAK242. Moreover, treatment of mice with an IgG Fc fragment was sufficient to enrich its presence in 276 atherosclerotic plaques ( Fig.5H). Overall, we show that IgG plays a pathogenic role in the activation of 277 macrophages and the progression of atherosclerosis in a noncanonical manner. 278 To determine whether IgG directly engages with TLR4, we employed TurboID proximity labeling in 279 293T cells, revealing positive labelling of IgG by TurboID-tagged TLR4 ( Fig. 5I ). Surface plasmon 280 resonance (SPR) confirmed a robust binding between human IgG (hIgG) and the hTLR4-MD2 co-281 receptor complex ( Fig. 5J; Fig. S8A ). In contrast, IgG failed to bind to TLR4 without MD2 ( Fig. S8B, 282 C), highlighting that an integral TLR4 complex is required. In BMDMs derived from TLR4 lps-del mice, 283 which are deficient in TLR4 signaling38, IgG failed to induce NF-κ B’s nuclear translocation (Fig. 5K, Fig. 284 S8D) or phosphorylation ( Fig. S8E-G ). Similarly, IgG's induction of pro-inflammatory genes 285 (Il1b, Nos2, Il6, Nlrp3) and repression of pro-resolving genes ( Mertk) were abolished ( Fig, 5L), and IL-286 1β secretion was unchanged in TLR4lps-del BMDMs (Fig. S8H). Collectively, these results establish TLR4 287 as a necessary mediator for IgG to activate NF-κ B-driven inflammatory signaling in macrophages. 288 289 IgG promotes macrophage foam cell formation through the TLR4/NF-κ B/LCN2 axis 290 Lipid-laden macrophages, or foam cells, are key determinants of atherosclerosis progression 39, and their 291 formation is promoted by TLR4 activation 40-42. We therefore tested whether IgG can activate TLR4 to 292 accelerate this process. BMDMs were primed with IgG treatment and subsequently loaded with Dil-293 oxLDL. IgG significantly enhanced oxLDL uptake in BMDMs, similar to LPS, and TLR4 inhibition 294 effectively abolished this effect ( Fig, 6A,B). This IgG-induced foam cell formation was also blocked by 295 the NF-κ B inhibitor Licochalcone D but not by the NLRP3 inhibitor MCC950 ( Fig, 6C,D), indicating its 296 dependence on TLR4/NF-κ B signaling but not the NLRP3 inflammasome. 297 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 12 The mediators underlying the transformation of inflammatory macrophages into foam cells are under 298 active investigation. Lipocalin-2 (LCN2) is a secretory protein implicated in immune responses 43,44 and 299 atherosclerosis45-47. Noting its upregulation by IgG treatment in BMDMs by RNA-seq, we validated that it 300 was robustly induced by IgG in a TLR4-dependent manner ( Fig, 6E,F ). Recombinant LCN2 protein 301 facilitated Dil-oxLDL uptake in BMDMs, with no additive effect from IgG, and this process was 302 abrogated by the LCN2 inhibitor ZINC ( Fig, 6G,H), suggesting that IgG acts through LCN2 to promote 303 macrophage foam cell formation. Interestingly, IgG-induced foam cell formation was also recapitulated in 304 human THP-1 macrophages and blocked by Nipocalimab, a clinically used monoclonal antibody against 305 FcRn ( Fig, 6I, J ). We further investigated the clinical relevance of this axis. In the proteomic Fuwai 306 cohort, LCN2 protein levels were increased in both mild (periphery) and advanced (core) coronary artery 307 plaques compared to controls ( Fig, 6K), with higher plaque LCN2 trending toward shorter event-free 308 survival (P=0.224) (Fig, 6L). In the larger BiKE carotid plaque transcriptomic cohort, high plaque LCN2 309 expression was associated with significantly shorter event-free survival (P=0.0194) ( Fig, 6M ). In 310 conclusion, we identify LCN2 as a downstream effector of IgG-induced TLR4/NF- κ B signaling that 311 promotes macrophage foam cell formation, suggesting that the robust changes in plaque burden with IgG 312 accumulation inhibition may be occurring through this milieu. 313 314

315 The vascular endothelium is perpetually bathed in abundant immunoglobulins—so ubiquitous that their 316 pathogenic potential is often overlooked. Here, we identify IgG as a critical pro-atherogenic factor, one 317 that accumulates in plaques under dyslipidemia conditions—a finding with direct translational relevance 318 as it correlates positively with at herosclerosis burden in humans. This accumulation is governed by 319 macrophage FcRn-mediated IgG recycling, which not only sustains intraplaque IgG levels but also 320 amplifies its inflammatory impact via TLR4/NF- κ B signaling. Inhibiting this accumulation reduces 321 plaque inflammation and attenuates atherosclerosis in mice. Our findings redefine atherosclerosis 322 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 13 pathogenesis through the lens of an abundant yet underappreciated mediator, pos itioning IgG as a novel 323 and potential therapeutic node. 324 While IgG is indispensable for immune defense, its unchecked accumulation in plaques—akin to its 325 recently described role in thrombosis via platelet complement activation48—can exacerbate cardiovascular 326 pathophysiology. Notably, pathogenic IgG may arise not only from increased abundance but also 327 from altered specificity or post-translational modifications (e.g., glycosylation 49), potentially converting 328 IgG from a protective agent to a driver of maladaptive inflammation. Further, different IgG subtypes exert 329 divergent effects. We observed that elevated plaque levels of IgG1, IgG2 and IgG4 were positively 330 correlated with CAD-related events, whereas IgG3 was associated with a more favorable outcome. 331 Intriguingly, IgG3 is the least abundant IgG subtype and has the shortest half-life due to its weak binding 332 affinity to FcRn among all IgG isotypes 50. Although oxLDL-specific IgG autoantibodies have been 333 documented in plaques 15,51, regular IgG isolated from healthy mice, which unlikely contain such 334 atherogenic autoantibodies, was actively enriched in plaques, and IgG-Fc, even without IgG’s antigen-335 recognizing Fab domain, was also able to enter the plaque, confirming that plaque IgG deposition is not 336 necessarily antigen-specific. 337 The canonical paradigm holds that IgG binds antigen via its Fab to form ICs, which then trigger 338 Fcγ R-mediated responses 52. Our results demonstrate that IgG can bypass this classic pathway. IgG 339 directly interacts with TLR4, and its constant Fc domain is sufficient to activate NF- κ B in macrophages, 340 indicating antigen-binding is not necessary. This pathway does require the native IgG structure and an 341 intact TLR4/MD2 complex, as neither denatured IgG nor TLR4 alone could replicate this effect. TLR4 is 342 a sentinel of innate immunity, known for its response to bacterial LPS but also to endogenous ligands like 343 S100A853, heat shock proteins (HSPs), LDL and viral proteins 54,55. Herein, we reveal a non-canonical 344 function of IgG as a direct activator of innate immunity, forging a molecular bridge between the adaptive 345 and innate immune systems. This mechanism provides a streamlined explanation for chronic 346 inflammation in atherosclerosis, with LCN2 identified as a critical downstream mediator promoting 347 plaque foam cell formation. Collectively, our study establishes a new paradigm in which IgG, acting as a 348 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 14 pathogenic factor driving atherosclerosis through a non-canonical mechanism that may also apply to other 349 chronic inflammatory diseases. 350 Systemic IgG homeostasis relies more on FcRn-mediated recycling than de novo B cell production 56. 351 This study pioneers the exploration of IgG recycling in atherosclerosis. By establishing that plaque IgG 352 deposition is not antigen-binding-dependent but rather FcRn-dependent, we unveil FcRn as a previously 353 unrecognized arbiter of disease progression. Clinically, our proteomic analysis links both IgG and FcRn 354 to atherosclerosis severity, a connection also confirmed in mice, where macrophage-355 specific FcRn deletion reduces plaque IgG depots and attenuates disease outcome. We further endow a 356 dual role for IgG in driving pro-inflammatory macrophage responses and promoting foam cell formation. 357 Our previous studies reveal that mKO mice prevent IgG from accumulating in the serum and in visceral 358 adipose tissue in aging and obesity, but not at basal conditions 57,58. Therefore, strategically targeting 359 FcRn offers distinct advantages over B-cell depletion, which compromises all antibody production 59,60, as 360 it selectively reduces IgG while preserving immune competence. Indeed, we show that the FDA-approved 361 FcRn antibody—Nipocalimab works effectively to suppress foam cell formation in activated human THP-362 1 macrophages. By identifying IgG as a key proatherogenic factor, our work unveils a novel therapeutic 363 avenue to curb cardiovascular risk, particularly in high-risk conditions like aging and obesity. 364 365 366 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 15

367 1. A. E. Moran et al. , Temporal trends in ischemic heart disease mortality in 21 world 368 regions, 1980 to 2010: the Global Burden of Disease 2010 study. Circulation 129, 1483-369 1492 (2014). 370 2. J. F. Bentzon, F. Otsuka, R. Virmani, E. Falk, Mechanisms of plaque formation and 371 rupture. Circ Res 114, 1852-1866 (2014). 372 3. K. Skålén et al. , Subendothelial retention of atherogenic lipoproteins in early 373 atherosclerosis. Nature 417, 750-754 (2002). 374 4. P. Libby, A. H. Lichtman, G. K. Hansson, Immune effector mechanisms implicated in 375 atherosclerosis: from mice to humans. Immunity 38, 1092-1104 (2013). 376 5. L. Badimon, G. Vilahur, Thrombosis formation on atherosclerotic lesions and plaque 377 rupture. J Intern Med 276, 618-632 (2014). 378 6. B. B. Adhyaru, T. A. Jacobson, Safety and efficacy of statin therapy. Nat Rev Cardiol 15, 379 757-769 (2018). 380 7. G. G. Schwartz et al. , Alirocumab and Cardiovascular Outcomes after Acute Coronary 381 Syndrome. N Engl J Med 379, 2097-2107 (2018). 382 8. L. Sun et al. , Innate-adaptive immunity interplay and redox regulation in immune 383 response. Redox biology 37, 101759 (2020). 384 9. J. L. Witztum, A. H. Lichtman, The influence of innate and adaptive immune responses 385 on atherosclerosis. Annu Rev Pathol 9, 73-102 (2014). 386 10. J. Č erný, I. St ř íž, Adaptive innate immunity or innate adaptive immunity? Clinical 387 science (London, England : 1979) 133, 1549-1565 (2019). 388 11. D. Wolf, K. Ley, Immunity and Inflammation in Atherosclerosis. Circulation research 389 124, 315-327 (2019). 390 12. S. C. Eisenbarth, R. A. Flavell, Innate instruction of adaptive immunity revisited: the 391 inflammasome. EMBO molecular medicine 1, 92-98 (2009). 392 13. E. K. Koltsova et al. , Dynamic T cell-APC interactions sustain chronic inflammation in 393 atherosclerosis. J Clin Invest 122, 3114-3126 (2012). 394 14. R. Wang et al., The interaction of innate immune and adaptive immune system. 395 MedComm 5, e714 (2024). 396 15. S. Ylä-Herttuala et al., Rabbit and human atherosclerotic lesions contain IgG that 397 recognizes epitopes of oxidized LDL. Arterioscler Thromb 14, 32-40 (1994). 398 16. H. W. Schroeder, Jr., L. Cavacini, Structure and function of immunoglobulins. J Allergy 399 Clin Immunol 125, S41-52 (2010). 400 17. M. R. Lennartz et al. , Ligation of macrophage Fc γ receptors recapitulates the gene 401 expression pattern of vulnerable human carotid plaques. PLoS One 6, e21803 (2011). 402 18. S. R. Khan et al. , The Association of Serum Immunoglobulins with Risk of 403 Cardiovascular Disease and Mortality: the Rotterdam Study. J Clin Immunol 43, 769-779 404 (2023). 405 19. A. Ravandi et al., Relationship of IgG and IgM autoantibodies and immune complexes to 406 oxidized LDL with markers of oxidation and inflammation and cardiovascular events: 407

from the EPIC-Norfolk Study. J Lipid Res 52, 1829-1836 (2011). 408 20. P. T. Kovanen, M. Mänttäri, T. Palosuo, V. Manninen, K. Aho, Prediction of myocardial 409 infarction in dyslipidemic men by elevated levels of immunoglobulin classes A, E, and G, 410 but not M. Arch Intern Med 158, 1434-1439 (1998). 411 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 16 21. R. Y. Khamis et al. , High Serum Immunoglobulin G and M Levels Predict Freedom 412 From Adverse Cardiovascular Events in Hypertension: A Nested Case-Control Substudy 413 of the Anglo-Scandinavian Cardiac Outcomes Trial. EBioMedicine 9, 372-380 (2016). 414 22. L. Yu et al. , IgG is an aging factor that drives adipose tissue fibrosis and metabolic 415 decline. Cell Metab 36, 793-807.e795 (2024). 416 23. S. Ma et al. , Spatial transcriptomic landscape unveils immunoglobin-associated 417 senescence as a hallmark of aging. Cell 187, 7025-7044.e7034 (2024). 418 24. L. Yu et al. , FcRn-dependent IgG accumulation in adipose tissue unmasks obesity 419 pathophysiology. Cell Metab 37, 656-672.e657 (2025). 420 25. T. Kyaw, P. Tipping, A. Bobik, B. H. Toh, Protective role of natural IgM-producing B1a 421 cells in atherosclerosis. Trends Cardiovasc Med 22, 48-53 (2012). 422 26. S. Tsimikas et al. , Relationship of IgG and IgM autoantibodies to oxidized low density 423 lipoprotein with coronary artery disease and cardiovascular events. J Lipid Res 48, 425-424 433 (2007). 425 27. A. C. Bashore et al. , High-Dimensional Single-Cell Multimodal Landscape of Human 426 Carotid Atherosclerosis. Arterioscler Thromb Vasc Biol 44, 930-945 (2024). 427 28. B. A. Keyt, R. Baliga, A. M. Sinclair, S. F. Carroll, M. S. Peterson, Structure, Function, 428 and Therapeutic Use of IgM Antibodies. Antibodies (Basel) 9, (2020). 429 29. S. Ren et al., Expression of NF-κ B, CD68 and CD105 in carotid atherosclerotic plaque. J 430 Thorac Dis 5, 771-776 (2013). 431 30. M. Mussbacher, M. Derler, J. Basílio, J. A. Schmid, NF- κ B in monocytes and 432 macrophages - an inflammatory master regulator in multitalented immune cells. Front 433 Immunol 14, 1134661 (2023). 434 31. M. G. Dorrington, I. D. C. Fraser, NF- κ B Signaling in Macrophages: Dynamics, 435 Crosstalk, and Signal Integration. Front Immunol 10, 705 (2019). 436 32. M. P. de Winther, E. Kanters, G. Kraal, M. H. Hofker, Nuclear factor kappaB signaling 437 in atherogenesis. Arterioscler Thromb Vasc Biol 25, 904-914 (2005). 438 33. P. Kong et al. , Inflammation and atherosclerosis: signaling pathways and therapeutic 439 intervention. Signal Transduct Target Ther 7, 131 (2022). 440 34. A. Stutz, G. L. Horvath, B. G. Monks, E. Latz, ASC speck formation as a readout for 441 inflammasome activation. Methods Mol Biol 1040, 91-101 (2013). 442 35. S. Paik, J. K. Kim, P. Silwal, C. Sasakawa, E. K. Jo, An update on the regulatory 443 mechanisms of NLRP3 inflammasome activation. Cell Mol Immunol 18, 1141-1160 444 (2021). 445 36. M. R. Zinatizadeh et al. , The Nuclear Factor Kappa B (NF-kB) signaling in cancer 446 development and immune diseases. Genes Dis 8, 287-297 (2021). 447 37. Y. C. Lu, W. C. Yeh, P. S. Ohashi, LPS/TLR4 signal transduction pathway. Cytokine 42, 448 145-151 (2008). 449 38. A. Poltorak et al. , Defective LPS signaling in C3H/HeJ and C57BL/10ScCr mice: 450 mutations in Tlr4 gene. Science 282, 2085-2088 (1998). 451 39. K. J. Moore, I. Tabas, Macrophages in the pathogenesis of atherosclerosis. Cell 145, 341-452 355 (2011). 453 40. Y. W. Yin et al., TLR4-mediated inflammation promotes foam cell formation of vascular 454 smooth muscle cell by upregulating ACAT1 expression. Cell Death Dis 6, 1659 (2015). 455 41. Y. I. Miller, S. H. Choi, L. Fang, R. Harkewicz, Toll-like receptor-4 and lipoprotein 456 accumulation in macrophages. Trends Cardiovasc Med 19, 227-232 (2009). 457 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 17 42. S. H. Choi et al. , Lipoprotein accumulation in macrophages via toll-like receptor-4-458 dependent fluid phase uptake. Circ Res 104, 1355-1363 (2009). 459 43. X. Xiao, B. S. Yeoh, M. Vijay-Kumar, Lipocalin 2: An Emerging Player in Iron 460 Homeostasis and Inflammation. Annu Rev Nutr 37, 103-130 (2017). 461 44. S. A. Jaberi et al. , Lipocalin-2: Structure, function, distribution and role in metabolic 462 disorders. Biomed Pharmacother 142, 112002 (2021). 463 45. J. Ni et al., Serum lipocalin-2 levels positively correlate with coronary artery disease and 464 metabolic syndrome. Cardiovasc Diabetol 12, 176 (2013). 465 46. J. Gan et al., Serum Lipocalin-2 Levels Are Increased and Independently Associated 466 With Early-Stage Renal Damage and Carotid Atherosclerotic Plaque in Patients With 467 T2DM. Front Endocrinol (Lausanne) 13, 855616 (2022). 468 47. H. H. Yang et al. , Lipocalin family proteins and their diverse roles in cardiovascular 469 disease. Pharmacol Ther 244, 108385 (2023). 470 48. K. Stark et al. , Antibodies and complement are key drivers of thrombosis. Immunity 57, 471 2140-2156.e2110 (2024). 472 49. F. Nimmerjahn, G. Vidarsson, M. S. Cragg, Effect of posttranslational modifications and 473 subclass on IgG activity: from immunity to immunotherapy. Nat Immunol 24, 1244-1255 474 (2023). 475 50. G. Vidarsson, G. Dekkers, T. Rispens, IgG subclasses and allotypes: from structure to 476 effector functions. Front Immunol 5, 520 (2014). 477 51. J. P. Rhoads, A. S. Major, How Oxidized Low-Density Lipoprotein Activates 478 Inflammatory Responses. Crit Rev Immunol 38, 333-342 (2018). 479 52. P. Bruhns, F. Jönsson, Mouse and human FcR effector functions. Immunol Rev 268, 25-480 51 (2015). 481 53. T. Vogl et al. , Mrp8 and Mrp14 are endogenous activators of Toll-like receptor 4, 482 promoting lethal, endotoxin-induced shock. Nat Med 13, 1042-1049 (2007). 483 54. S. W. Brubaker, K. S. Bonham, I. Zanoni, J. C. Kagan, Innate immune pattern 484 recognition: a cell biological perspective. Annu Rev Immunol 33, 257-290 (2015). 485 55. M. Mehmeti et al. , Wnt5a is a TLR2/4-ligand that induces tolerance in human myeloid 486 cells. Commun Biol 2, 176 (2019). 487 56. J. Kim, W. L. Hayton, J. M. Robinson, C. L. Anderson, Kinetics of FcRn-mediated 488 recycling of IgG and albumin in human: pathophysiology and therapeutic implications 489 using a simplified mechanism-based model. Clin Immunol 122, 146-155 (2007). 490 57. L. Yu et al. , FcRn-dependent IgG accumulation in adipose tissue unmasks obesity 491 pathophysiology. Cell Metab 37, 656-672 e657 (2025). 492 58. L. X. Yu et al. , IgG is an aging factor that drives adipose tissue fibrosis and metabolic 493 decline. Cell Metab 36, (2024). 494 59. H. Ait-Oufella et al., B cell depletion reduces the development of atherosclerosis in mice. 495 J Exp Med 207, 1579-1587 (2010). 496 60. T. Kyaw et al. , Conventional B2 B cell depletion ameliorates whereas its adoptive 497 transfer aggravates atherosclerosis. J Immunol 185, 4410-4419 (2010). 498 61. M. I. Love, W. Huber, S. Anders, Moderated estimation of fold change and dispersion for 499 RNA-seq data with DESeq2. Genome Biol 15, 550 (2014). 500 62. T. Zahr et al. , PPAR γ (Peroxisome Proliferator-Activated Receptor γ ) Deacetylation 501 Suppresses Aging-Associated Atherosclerosis and Hypercholesterolemia. Arterioscler 502 Thromb Vasc Biol 43, 30-44 (2023). 503 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 18 63. R. Li, Y. Tang, Z. Chen, Y. Liu, Screening TLR4 Binding Peptide from Naja atra Venom 504 Glands Based on Phage Display. Toxins (Basel) 16, (2024). 505 506 507 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 19 ACKNOWLEDGMENTS 508 Funding 509 This research was supported by the China Noncommunicable Chronic Diseases-National Science and 510 Technology Major Project (2023ZD0507900, L.Q., 2023ZD0504701 to W.F.), the National Natural 511 Science Foundation of China (32430047 to L.Q., 32471224, HY2022-8 to L.W.), National Institutes of 512 Health (NIH) R01DK134471 (L.Q.), the Russell Berrie Foundation (L.Q.), and the American Heart 513 Association (AHA) predoctoral fellowship (24PRE1198199 to T.Z.). We gratefully acknowledge the 514 NBDC Histology Core, and NYNORC for their support. We also thank George Kuriakose and Dr. Ira 515 Tabas' Atherosclerosis Phenotyping Core at Columbia University for blinded plaque morphometric 516 analyses. The content is solely the authors' responsi bility and does not necessarily represent the official 517 views of the AHA, NIH, or NSFC. 518 Author contributions 519 T.Z., K.Z., L.W. and L.Q. conceptualized the study, designed the experiments, and wrote the manuscript. 520 T.Z., K.Z., S.H., L.Y., Z.Y., D. L., Q.W., and QF.W. performed the experiments. T.Z., K.Z., S.H., C.X., 521 B.L., X.L., M.S., and A.R.K. performed data analyses. S.H., Z.H. and W.F. were in charge of human 522 sample collection and analyses; F.Y. and M.P.R. helped with resources and reagents. L.Q. is the primary 523 overseer of this study, and as such, has full access to all the data in the manuscript and takes responsibility 524 for the integrity of the data and the accuracy of the data analysis. 525 526 Competing interests 527 The authors declare no competing interests. 528 529 Data, code, and materials availability 530 /circle6 Data that support the findings of this study are available from the corresponding author upon 531 reasonable request. 532 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 20 /circle6 Clinical proteomic data will be made publicly available upon publication. 533 /circle6 Mouse RNA-seq data can be found in the GSA with accession number PRJCA02063722. 534 /circle6 Human scRNA-seq and CITE-seq data can be found in the Gene Expression Omnibus with 535 persistent ID GSE25390427. 536 /circle6 This paper does not contain original code. 537 /circle6 The research materials in this study are available upon request from the lead contact. 538 539 540 Supplementary Materials 541 Experimental model and study participant details. 542 Figures. S1 to S8 543 Graphic abstract. 544 545 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 21 Figure Legends 546 547 Figure 1. Elevated IgG and FcRn in advanced coronary artery plaques correlate with adverse548 clinical outcomes. 549 sis 21 se (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 22 A, Representative image of endarterectomized coronary intimal tissue collected during coronary artery 550 bypass graft surgery in patients with coronary artery disease (CAD). B, Schematic illustration of the 551 experimental design and cohort stratification. C, Violin plots showing the protein abundance of the four 552 IgG heavy chain isoforms (IGHG1–IGHG4) among the three groups. D, Kaplan–Meier curves illustrating 553 the association between IgG levels and the incidence of major adverse cardiac and cerebrovascular events 554 (MACCE) over a median follow-up period of 44 months. E, Representative immunohistochemical 555 staining of IgG on endarterectomized coronary intimal tissue sections collected during coronary artery 556 bypass grafting (CABG) from patients with coronary artery disease (CAD), (P1, P2, P3 = patient 1, 2 and 557 3). Scale bars, 50, 300 or 400 μm as indicated. F, Violin plots showing FcRn protein expression in plaque 558 cores, peripheral regions, and normal controls. G-H, Kaplan–Meier analyses depicting correlations 559 between FCGRT(FcRn) expression and the incidence of MACCE over a median follow-up period of 44 560 months ( G, Fuwai cohort) or 3000 days ( H, BiKE cohort). Data are presented as mean ± SEM. All 561 datasets were assessed for normality and equal variances. Kruskal-Wallis test combined with Dunn's post-562 hoc test was used for group comparisons. Survival analyses were performed using Kaplan–Meier method. 563 *p < 0.05, **p < 0.01, ***p < 0.001. See also Figures S1 and S2. 564 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 23 565 Figure 2. Immune profiling of FcRn and IgG expression in human carotid atherosclerotic plaques. 566 A, Schematic overview of the experimental design used to obtain scRNA-seq (n=21) and CITE-seq (n=6)567 datasets of human carotid atherosclerotic plaques for profiling FCGRT and IgG Fc expression. B,568 Uniform manifold approximation and projection (UMAP) of all cell types ide ntified within the collected569 plaque samples. C, UMAP showing FCGRT expression levels across all cell types identified within the570 collected plaque samples. The arrow highlights macrophage populations defined in the referenced study.571 D, Violin plot comparing median FCGRT expression across all macrophage clusters. E, Violin plot572 comparing median surface IgG Fc expression across all macrophage subclusters. F, UMAP of573 macrophages and dendritic cells (DCs) clusters colored by their phenotypic and functional chara cteristics574 sis 23 6) , ed he y. lot of ics (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 24 as depicted in the referenced study. pDC, plasmacytoid dendritic cell; cDC, conventional dendritic cell; 575 the T cell cluster represents residual cells. G, UMAP of the same clusters indicating the range of FCGRT 576 expression from scRNA-seq data. H, Violin plots showing median FCGRT expression across all 577 macrophage phenotype subclusters. I, UMAP of macrophage phenotype clusters (encircled) indicating the 578 range of surface IgG Fc expression from CITE-seq data. J, Violin plots showing median IgG Fc 579 expression across all macrophage phenotype subclusters. See also Figures S3. 580 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 25 581 sis 25 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 26 Figure 3. Myeloid FcRn drives plaque IgG accumulation and atherosclerosis progression. 582 A, Representative immunohistochemistry (IHC) images showing IgG, IgM, and IgA staining in aortic 583 root lesions from Ldlr-/- mice fed a western diet (WD) for 16 weeks. Scale bar, 50 μm. Black dashed lines 584 delineate the intima encompassing a plaque-rich leaflet from the medial layer of the aortic root. B, 585 Quantification of the percentage of immunoglobulin-positive (Ig+) area within the intimal layer of the 586 aortic root (n=7, 7, 7). C, Immunofluorescent staining (IF) for FcRn and IgG in atherosclerotic lesions of 587 Ldlr-/- mice. Nuclei were stained with DAPI. Scale bars, 50 μm (left) and 25 μ m (right). White dashed 588 lines separate the intima (right) encompassing a plaque-rich leaflet from the medial layer (left) of the 589 aortic root. D, Ex vivo fluorescence imaging showing mouse IgG-Cy5 distribution in the aorta of 590 atherosclerotic mice and normocholesterolemic controls. E, Schematic of the bone marrow 591 transplantation (BMT) atherosclerosis model. Bone marrow cells from FcRn myeloid knockout (mKO) or 592 control mice were transplanted into Ldlr-/- recipients, which were then fed a WD for 16 weeks to induce 593 atherosclerosis after recovery. F, Plasma levels of IgG and IgM in control and mKO mice as determined 594 by western blotting (WB); LC = loading control (Coomassie Brilliant Blue). G-H, IgG protein levels in 595 the aortic arch of BMT atherosclerotic mice, as measured by WB ( G) and quantified in ( H) (n=4, 4). 596 HSP90 was used as a loading control. I, Representative IHC images of IgG staining in the aortic root 597 plaques from BMT mice. Scale bar, 100 μ m. J, Quantification of the IgG-positive area within the intima, 598 encircled in black (n=5, 5). K, Representative H&E staining of aortic root sections from mKO and control 599 BMT Ldlr-/- mice with atherosclerosis. Scale bar, 100 μm. L, Quantification of aortic root lesion area 600 (n=11, 10). M-N, Quantification of the necrotic core area ( M) and acellular area ( N) within plaques, as 601 delineated by the dashed circle in ( L) (n=11, 10). O, Total plasma cholesterol levels in mKO and control 602 BMT mice with atherosclerosis at sacrifice (n=11, 10). P, Immunostaining of IgG and IL-1 β in the intima 603 region (encircled in white) of aortic root lesions from BMT mice. Scale bar, 100 μm. DAPI was used for 604 nuclear staining. Q, Quantification of total IL-1 β + area out of the total intima region (n=8, 8). Data 605 represent mean ± SEM of independent biological replicates. All datapoints were assessed for normality 606 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 27 and equal variances. Statistical analyses were performed using One-way ANOVA in ( B). Two- tailed607 Student’s t-tests were used for comparisons between two groups. Mann-Whitney U test was used for608 lesion analyses that did not pass normality. **p < 0.01, ***p < 0.001. See also Figures S4-S5. 609 610 611 Figure 4. IgG activates NF-κB and the NLRP3 inflammasome in macrophages. 612 sis 27 ed for (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 28 A, Heatmap showing differentially expressed genes in BMDMs after 16-hr 100 μg/ml IgG treatment 613 (n=4, 4). Gene expression was calculated as log2FPKM and scaled as Z-scores. B, QPCR analysis of 614 inflammatory genes Il1b and Nos2 in RAW264.7 cells treated with IgG (100 μg/mL), boiled IgG (b-IgG, 615 100 μg/mL), or LPS (50 μg/mL) for 24 hours (hrs) (n=3/group). C, ELISA quantification of the amount 616 of secreted IL-1 β from cells in ( B) (n=7/group). D, WB of phosphorylated and total NF- κB in treated 617 cells as in ( B) and quantification of the ratio of phosphorylated NF- κB to total NF- κB (n=3/group). 618 HSP90 was used as the loading control. E-F, RAW264.7 cells were treated with IgG (100 μg/mL), b-IgG 619 (100 μg/mL), or LPS (50 μg/mL) for 15 minutes. Cells were fixed and stained with NF- κB (p65) and 620 CD68 as shown in ( E), and Quantification of the percentage of cells with nuclear NF- κB localization 621 based on images from (F) (n=5/group). Scale bar, 50 μm. G, QPCR analysis of Caspase1, Il1b and Nlrp3 622 in BMDMs treated with IgG (100 μg/mL), or LPS (50 μg/mL) for 24 hrs (n=5/group). H, QPCR analysis 623 of Nlrp3, Il1b, Tnfa, Inos and macrophage marker Adgre1(F4/80) in bone marrow-derived monocytes 624 treated with IgG (100 μg/mL) for 24 hrs (n=5/group). I, Representative IF images of BMDMs treated 625 with IgG (100 μ g/mL), LPS (50 μg/mL) and MCC950 (100 nM) for 24 hrs, followed by incubation with 626 Nigericin (10 μ M) for 30 minutes, and stained for ASC. Scale bar, 20 μ m, and the quantification of the 627 number of ASC specks per cell (n=8/group). DAPI was used for nuclear staining and cell counting. J, 628 Representative WB analysis of BMDMs treated with IgG (100 μ g/mL), LPS (50 μ g/mL), MCC950 (100 629 nM), or Licochalcone D (10 μ M) for 24 hrs, followed by Nigericin (10 μ M) for 30 minutes. Blots were 630 probed for NLRP3, IL-1 β , and IL-18. K, QPCR analysis of Nos2, Il1b, and Nlrp3 mRNA levels in WT 631 BMDMs transfected with siRNAs against mouse Fcgr1, Fcgr2b, Fcgr3, Fcgr4 or sham control(siNC). 632 These cells were treated with IgG (100 μ g/mL) for 24 hrs (n=4/group). *Compared with siNC group, 633 #compared with siNC+IgG group. Data represent mean ± SEM. All datasets were assessed for normality 634 and equal variances. One-way ANOVA with multiple comparisons and two-tailed Student’s t-tests for 635 pairwise group comparisons were used for statistical analysis. * /#p < 0.05, **/##p < 0.01, ***/###p < 0.001. 636 See also Figures S6. 637 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 29 638 Figure 5. IgG stimulates TLR4 signaling upstream of NF-κB. 639 A, Representative IF images of NF- κB (p65) and CD68 in RAW264.7 cells pretreated with TLR4640 inhibitor TAK242 (1 μ M), followed by IgG (100 μ g/mL) stimulation for 2 hrs. Scale bar, 50 μm. B,641 Quantification of percentage of nuclear NF-kB cells based on images from ( A) (n=5, 5, 3). C, QPCR642 sis 29 R4 , R (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 30 analysis of Il1b and Nos2 expression in cells treated as in (A) for 24 hrs (n=3/group). D, WB analysis of 643 phosphorylated and total NF-κ B in RAW264.7 cells treated as in ( C) and quantification of 644 phosphorylated NF-κB levels (n=3/group). HSP90 was used as the loading control. E, WB analysis of 645 pro-IL-1β , IL-1β in lysates from human THP-1 cells pretreated with TAK242 (1 μM) for 1 hr, followed 646 by 12-hour treatment with vehicle, IgG (100 μ g/mL) or hPD1 mAb (100 μg/mL). F, Quantification of 647 phospho-NF-κB intensity and the NF- κB p65 nuclear translocation based on IF images of BMDMs 648 pretreated with TAK242 (1 μM) for 1 hr, followed by 1-hr treatment with vehicle, IgG (100 μg/mL), IgG 649 Fab domain (20 μ g/mL), IgG Fc domain (100 μg/mL) as shown in Extended Data Figure 9A and 9B 650 (n=8/group). G, WB analysis of NLRP3, pro-IL-1 β , pro-IL-18 in lysates from BMDMs pretreated with 651 TAK242 (1 μM) for 1 hr, followed by 1-hr treatment with vehicle, IgG (100 μg/mL), IgG Fab domain (20 652 μg/mL), IgG Fc domain (100 μg/mL). H, Ex vivo fluorescence imaging showing mouse IgG Fc-Cy5 653 distribution in the aorta of control and atherogenic mice. I, WB analysis of streptavidin pull-downs from 654 TurboID-TLR4/MD2 co-transfected 293T cells. Cells were treated with vehicle or 100 μg/mL IgG for 12 655 hrs, 1 day post transfection. J, Surface plasmon resonance analysis of kinetic binding of human IgG 656 (hIgG) with human TLR4-MD2 complex. The titration of native IgG at concentrations from 10nM – 657 625nM. K, Quantification of percentage of cells with nuclear NF-kB localization from IF images shown 658 in (J) (n=6, 6, 7, 6). L, QPCR analysis of inflammatory gene expression in WT and TLR4 lps-del BMDMs 659 (n=4/group). *Compared with WT group, # compared with TLR4lps-del group. Data represent mean ± SEM. 660 All datasets were assessed for normality and equal variances. One-way ANOVA followed multiple 661 comparisons was used for statistical analysis. */#p < 0.05, **/##p < 0.01, ***/###p < 0.001. See also Figures 662 S7-S8. 663 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 31 664 Figure 6. IgG promotes macrophage foam cell formation through a TLR4/NF- κ B/LCN2 axis. 665 A, Representative IF images of Dil-oxLDL staining in BMDMs treated with Vehicle, IgG (100 μg/mL),666 LPS (50 μ g/mL) in the presence or absence of TAK242 (1 μM) for 18 hrs, followed by incubation with667 Dil-oxLDL (10 μg/mL) for 6 hrs. Cells were fixed and stained for imaging. Scale bar, 20 μ m. B,668 Quantification of Dil-oxLDL fluorescence intensity per cell from images in ( A) (n = 8 per group). C,669 sis 31 , ith B, C, (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 32 Representative IF images of Dil-oxLDL in BMDMs treated with IgG (100 μg/mL) , LPS (50 μg/mL) , 670 MCC950 (100 nM), or Licochalcone D (10 μ M) for 18 hrs, followed by Dil-oxLDL (10 μg/mL) for 6 hrs. 671 scale bar, 20 μm. D, Quantification of Dil-oxLDL fluorescence intensity per cell from ( C) (n = 8 per 672 group). E, QPCR analysis of Lcn2 expression in BMDMs treated with oxLDL (50 μg/mL), followed by 673 IgG (100 μ g/mL), LPS (50 μg/mL) or TAK242 (1 μM) for 24 hrs (n=5, 5, 5, 3, 5, 5). F, Representative 674 WB analysis of BMDMs from ( E). LCN2 was probed. G, Representative IF images of Dil-oxLDL in 675 BMDMs treated with IgG (100 μg/mL), the LCN2 inhibitor ZINC00640089 (1 μ M), or recombinant 676 LCN2 (10 μ g/mL) for 18 hrs, followed by Dil-oxLDL (10 μ g/mL) for 6 hrs. Scale bar, 20 μ m. H, 677 Quantification of Dil-oxLDL fluorescence intensity per cell from images shown in ( G) (n = 8 per group). 678 I, Representative IF images of Dil-oxLDL in THP-1 cells treated with Vehicle, IgG (200 μg/mL), in the 679 presence or absence of Nipocalimab (10 μg/mL) for 18 hrs, followed by Dil-oxLDL (10 μg/mL) for 6 hrs. 680 J, Quantification of Dil-oxLDL fluorescence intensity per cell from ( I) (n = 6 per group). K, Proteomic 681 analysis of LCN2 protein levels in normal arteries, plaque periphery and core regions from the Fuwai 682 cohort. L, Kaplan–Meier curves showing the association between LCN2 expression and the incidence of 683 MACCE over a median follow-up period of 44 months from the Fuwai cohort. M, Kaplan–Meier curves 684 depicting the association between LCN2 and the incidence of MACCE over a median follow-up period of 685 3000 days from the BiKE cohort. Data represent mean ± SEM. All datasets were assessed for normality 686 and equal variances. Statistical comparisons were performed using One-way ANOVA with multiple 687 comparisons. *p < 0.05, **p < 0.01, ***p < 0.001. 688 689 690 691 692 Tables 693 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint Zahr et al. IgG propels atherosclerosis 33 Table 1. Baseline characteristics of coronary heart disease (CHD) patients and non-CHD controls. 694 Characteristics CHD (n=98)a Non-CHD (n=14) a P-value b Age, y 65 (36-79) 45 (7-63) <0.001 Sex, male, n (%) 67 (68%) 9 (64%) 0.8 BMI, y 25.8 (18.9-37.4) 21.5 (13.8-27.5) <0.001 Hypertension, n (%) 74 (76%) 3 (21%) <0.001 Diabetes, n (%) 45 (46%) 5 (36%) 0.5 695 aMedian (Min-Max); n (%) 696 bWilcoxon rank sum test; Fisher’s exact test; Pearson’s Chi-squared test 697 698 Table 2. Median expression levels of FCGRT and IgG in macrophage (Mf) clusters 1-5. 699 Feature Mf1 Mf2 Mf3 Mf4 Mf5 FCGRT 2.18 1.73 0.00 1.45 2.03 IgG_Fc_ADT 2.03 1.52 1.09 1.14 2.01 700 701 Table 3. Median expression levels of FCGRT and IgG Fc in macrophage phenotype clusters. 702 Feature C1Qhi IL1Bhi Foamy 1 Apoptotic Foamy 2 Proliferative FCGRT 2.19 1.79 1.77 1.28 2.31 1.02 IgG_Fc_ADT 2.07 1.61 1.72 1.23 1.89 1.40 703 (which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted March 17, 2026. ; https://doi.org/10.64898/2026.03.13.711718doi: bioRxiv preprint