{"paper_id":"237a8682-36ea-4b5b-b6ac-30d9a7e1d971","body_text":"1 \nTitle \nCollectin-11 regulates osteoclastogenesis and bone maintenance via a complement-\ndependent mechanism. \nShort title \n \nCollectin-11 regulates osteoclastogenesis and bone maintenance.  \n \nMark C Howard 1, Conrad A Farrar 1, Christopher L Nauser 1, Yusun Jeon 1, Anastasia Polycarpou1, \nDorota Smolarek 1, Roseanna Greenlaw 1, Peter Garred 2, Daniela A Vizitiu 1, Subhankar \nMukhopadhyay1, Steven H Sacks1 \n \n1MRC Centre for Transplantation, Peter Gorer Department of Immunobiology, School of \nImmunology & Microbial Sciences, King’s College London, United Kingdom  \n \n2Consultant Laboratory of Molecular Medicine, Department of Clinical Immunology Section 7631, \nRigshospitalet, University of Copenhagen , 2200 Copenhagen N , Denmark  \n \n \nCorresponding author:  \nProfessor Steven Sacks , MRC Centre for Transplantation , Peter Gorer Department of \nImmunobiology, School of Immunology & Microbial Sciences, King’s College London, 5th Floor Tower \nWing, Guy’s Hospital , Great Maze Pond , London, United Kingdom , SE1 9RT . E-mail: \nsteven.sacks@kcl.ac.uk. https://orcid.org/0000-0001-6361-9095. \n \nCo-authors: \nMark Howard mark.howard@sbcna.com   \nConrad Farrar conrad.farrar@kcl.ac.uk \nChristopher Nauser christopher.nauser@kcl.ac.uk \nAnastasia Polykarpou  anastasia.polycarpou@ndorms.ox.ac.uk  \nDorota Smolarek dorota.smolarek@kcl.ac.uk \nRoseanna Greenlaw roseanna.greenlaw@kcl.ac.uk \nPeter Garred Peter.Garred@regionh.dk \nDaniela Vizitiu daniela_alexandra@ymail.com \nSubhankar Mukhopadhyay subhankar.mukhopadhyay@kcl.ac.uk \nYusun Jeon yusun.jeon@kcl.ac.uk  \n \nSignificance statement  \nOur research in mice provides new insights into how mutations in the immune surveillance molecule \ncollectin-11 contribute to skeletal abnormalities in humans. Evidence from our study suggests that \nnormal osteoclasts interact with collectin-11 and complement, and that the disruption of this \ncooperation results in impaired bone main tenance in adulthood. These findings not only advance \nour understanding of osteoclast function but also highlight the therapeutic potential of targeting \ncollectin-11 in conditions associated with osteoclast dysfunction. \n  \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 2 \nAbstract \nThe human developmental disorder 3MC syndrome is characterized by skeletal deformities \nassociated with a deficiency of the pattern recognition molecule collectin-11 (CL -11), yet the \nunderlying molecular and cellular mechanisms remain unclear. Here, we demonstrate that CL -11 \ndeletion alone does not cause bone abnormalities in mice; however, combined deficiencies \ninvolving CL-11 and complement components MASP -2 (lectin pathway), CFB, or C3 (alternative \namplification pathway) lead to significant vertebral bone loss and spinal curvature by 12 weeks of \nage. Ex vivo osteoclast (OCL) differentiation from bone marrow -derived cells of these double-\nknockout (DKO) mice was markedly impaired, but differentiation capacity was substantially restored \nby supplementation wit h CL -11. Furthermore, CL -11 and membrane attack complex (C5b -9) \ndeposition were co-localized to OCLs and their precursors in normal bone tissues from embryonic \nstages to adulthood. These findings identify CL -11 as a critical osteoclastogenesis and bone \nmai ntenance regulator in conjunction with complement system-mediated signalling pathways and \nhighlight CL-11 as a potential therapeutic target in diseases involving dysregulated osteoclast \nfunction and bone remodelling. \n \nIntroduction \nCollecfn-11 (CL-11) is a soluble C-type lecfn that recognizes carbohydrate moffs on both self and \nnonself structures, acfvafng the lecfn complement pathway (LP) (Keshi et al., 2006). It plays key \nroles in innate immunity, inflammatory disorders, and embryogenesis  (Hansen et al., 2016). The \nwidespread fssue expression of CL -11 suggests its fundamental biological importance across \ndifferent cell populafons (Nauser and Sacks, 2023). \nThe basic structural unit of CL -11 is a triple -chain complex, with each chain comprising a \ncarbohydrate-recognifon domain (CRD) and a collagen -like domain (CLD), separated by a neck \nregion. These CL -11 complexes self-combine  into dimers and trimers (100 –200 kDa), enhancing \ncarbohydrate-binding avidity. Serine proteases (MASPs 1–3) physically associate with the CLD of CL-\n11, facilitafng the formafon of the classical pathway C3 convertase (C4bC2b), which leads to C3 \ncleavage (Hansen et al., 2010, Ma et al., 2013). MASP-2, besides cleaving C4 and C2, can directly \ncleave C3 (Ma et al., 2013, Schwaeble et al., 2011, Asgari et al., 2014). MASP-3 also contributes to \nthe alternafve pathway (AP) by acfvafng factor D, which converts C3bB into C3bBb, the alternafve \npathway C3 convertase (Takahashi et al., 2010). Thus, CL -11/MASP complexes can inifate both \nclassical and alternafve C3 convertases at the ligand-binding site for CL-11 (Supplemental Figure 1). \nThe subsequent cleavage of C3 and C5 generates anaphylatoxins (C3a and C5a), the opsonin C3b, \nand membrane alack complex (C5b -9, MAC), promofng cell acfvafon, damage,  and cell death. \nGiven that CL-11 is expressed in cells of ectodermal, endodermal, and mesenchymal origin (Hansen \net al., 2016, Keshi et al., 2006), this locally produced palern-recognifon molecule (PRM) may exert \nregional funcfons. Notably, mice lacking fssue CL-11 resist post-ischemic kidney injury (Farrar et al., \n2016). In humans, mutafons in COLEC11 (or the related COLEC10 gene) cause 3MC syndrome  \n(Rooryck et al., 2011, Afk et al., 2015, Munye et al., 2017), a developmental disorder characterized \nby craniofacial dysmorphia, clem palate, stunted growth, and mulforgan defects (Urquhart et al., \n2016). The MASP1 gene encodes three splice variants, MASP-1, MASP-3 and MAP-1.  Mutafons in \nMASP1 making  MASP-3 inacfve have also been idenffied in 3MC pafents (Gardner et al., 2017, \nRooryck et al., 2011), suggesfng that disrupfon of complement acfvafon —rather than a direct \neffect of PRMs on developing fssues—may underlie the syndrome’s pathology. \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 3 \nA role for complement in bone formafon is further supported by evidence that complement factors \nB (CFB), C3, C5, and C9 are expressed in the growth plate, a specialized carflaginous region at the \nends of long bones essenfal for longitudinal growth and endochondral ossificafon (Andrades et al., \n1996, Mödinger et al., 2018). Chondrocytes also express complement components C4, C2, and C3 \n(Bradley et al., 1996). Moreover, mice lacking C3 or C5 exhibit impaired fracture healing, with C5 \ndeficiency having a greater impact than C3 deficiency, highlighfng the role of complement \nacfvafon—parfcularly its terminal effector pathway—in bone remodelling (Ehrnthaller et al., 2013). \nOsteoclast (OCL) formafon may also be directly influenced by anaphylatoxins, even in the absence \nof key differenfafon factors RANKL and M-CSF (Ignafus et al., 2011). \nTo further invesfgate the interplay between the LP and AP triggered by CL-11, we generated double \nknockout (DKO) mice with disrupfons in Colec11 and either Masp-2 (LP), C5 (AP), or C3 (common \nto both pathways). A striking proporfon of the adult DKOs developed spinal curvature \n(kyphoscoliosis), whereas most single knockout (SKO) mice lacking CL-11, MASP-2, CFB, or C3 in \nisolafon displayed no such skeletal phenotype. The present study determined  how, when, and \nwhere CL-11 contributes to this unexpected skeletal phenotype. \nResults \nDKO Mice Lacking CL-11 Exhibit Loss of Vertebral Bone Integrity \nWe generated four groups of DKO mice: Colec11-/-/C3-/-; Colec11-/-/Masp-2-/-; Colec11-/-/Cfb-/-; C3-/-\n/Masp-2-/-). By 12 weeks of age, kyphosis and/or scoliosis occurred in 23 –28% of the three DKO \ngroups lacking CL-11 (Colec11-/-/C3-/-; Colec11-/-/Masp-2-/-; Colec11-/-/Cfb-/-), compared to only 0–1% \nin wild-type (WT) mice, SKO mice, and DKO mice with intact CL -11 (C3-/-/Masp-2-/-) (Table 1). Gross \ndissection and skeletal staining confirmed the spinal deformity, revealing a marked loss of \ndistinction between bone and cartilage (Figure 1a, 1b). Thus, the absence of CL-11 predisposed mice \nto the skeletal phenotype, but only in combination with a second deletion (C3, CFB, or MASP-2). \nWhole-body µCT imaging further highlighted the skeletal abnormalities, with vertebral bodies most \naffected. These vertebral bodies appeared punctuated by numerous small lesions ranging from \nisolated defects to extensive structural disruption (Figure 2a, 2c).  Spinal processes also displayed \ndestructive lesions. Multi-angle views allowed quantification of affected vertebrae (Figure 2b), with \nthe lumbar and thoracic regions most severely impacted (Figure 2d). In contrast, scapular and pelvic \nbones exhibited only minor defects, while femurs and skull calvaria appeared unaffected \n(Supplemental Figure 2). Femur length did not significantly differ between KO and WT mice of the \nsame sex.  \nMacroscopic and histological examinations of major organs revealed no evidence of a systemic \ndisorder (Supplemental Figure 3). However, litter sizes were reduced by approximately one pup in \nDKOs lacking CL-11 compared to WT, SKO, and  C3-/-/Masp-2-/- mice, despite shared housing and \ngenetic background (Supplemental Figure 4a). Although birth weight and growth rates were lower \nacross all KO strains compared to WT, sex ratios remained balanced (Supplemental Figure 4). \nNotably, skeletal abnormalities did not worsen with aging beyond 12 weeks, as confirmed by gross \ndissection, skeletal preparation, and µCT imaging of 12 -month -old mice (n = 5/group) (data not \nshown). \n \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 4 \nVertebral Growth Plates in DKO Mice Exhibit Trabecular Bone Loss  \nHistological analysis of affected Colec11-/- DKO spines revealed significant trabecular bone loss in \nvertebral bodies (VB), accompanied by regions of increased marrow adiposity (Figure 3a -d). \nIntervertebral discs (ID), visualized with toluidine blue staining, showed degeneration in the nucleus \npulposus (NP). Thus, vertebral trabecular bone thinning, replacement by adipose tissue, and ID \ndegeneration characterize the destructive spinal lesions in Colec11-/- DKOs. \n \nEmbryonic CL-11 and C3 Deficiency Do Not Affect Bone Morphology \nTo determine whether these defects originated during embryonic development, we examined \nembryonic tissues from  Colec11-/-/C3-/-mice. Light microscopy revealed no overt morphological \nabnormalities, despite CL-11 and C3 being broadly expressed in WT embryonic vertebrae (Figure 4a, \n4b). Between  E13.5–E18.5, CL -11 and C3 staining was prominent in regions of condensing \nchondrocytes, later becoming restricted to primary ossification centres (Figure 4b, 4c). \nTo assess MAC formation, we probed embryonic tissue sections with a C9 -specific antibody. C9 \naggregates were detected dynamically between E13.5–E18.5 (Figure 5a). WT tissues showed large \npericellular C9 aggregates indicative of MAC deposition, with progressive regional restriction over \ntime, leading to an overall reduction in MAC area (Figure 5b). In contrast, Colec11-/-/C3-/- embryos \nexhibited markedly reduced C9 staining at all stages. However, SKO Colec11-/- embryos did not show \na significant decline in C9 over time. Thus, MAC formation was impaired only in the absence of both \nCL-11 and C3—but without evident morphological consequences by E18.5. \n \nOsteoclast Differentiation Requires CL-11 and Either C3 or CfB \nSince skeletal abnormalities emerged postnatally (≥ 12 weeks), we investigated a potential defect \nin bone maintenance rather than development. Given prior links between complement and OCL \nformation (Ignatius et al., 2011), we assessed bone marrow -derived stem cell (BMSC) differentiation \ninto OCLs across genotypes.  WT BMSCs differentiated into mature multinucleated giant cell OCLs, \nwith concurrent  CL-11 expression, OSCAR detection, and TRAP + staining (Figure 6a –c). However, \nin Colec11-/-/C3-/- BMSCs the number of multinucleated OCLs was significantly reduced as indicated \nby the microscopic area occupied by such cells (Figure 6d and 6e). MAC formation, detected via C9 \nstaining, was also impaired in  Colec11-/-/C3-/- myeloid progeny (Figure 6f, 6g). Similarly, Colec11-/-\n/Cfb-/- cells largely failed to differentiate into mature OCLs (Figure 6e; Supplemental Figure 5a). In \ncontrast, OCLs derived from  Colec11-/- SKOs were only mildly reduced in number and MAC \ndeposition (Figure 6d–g; Supplemental Figure 5b). These findings suggest that OCL differentiation is \ndependent on CL-11 facilitated by either C3 or CFB. \n \nExogenous CL-11 Restores Osteoclast Differentiation \nTo determine whether CL -11 directly supports OCL differentiation, we supplemented the \ndifferentiation medium with recombinant CL -11 (rCL -11, 0.9  µg/mL) in Colec11⁻/⁻/C3⁻/⁻ and \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 5 \nColec11⁻/⁻/Cfb⁻/⁻ BMSC cultures. This treatment partially restored CL-11 binding and significantly \nincreased the presence of multinucleated giant cell OCLs (Figure 6d and 6e; Supplemental Figure 5). \nConsistent with this observation, the average number of nuclei per OCL in rCL-11–treated \nColec11⁻/⁻/C3⁻/⁻ BMSCs increased to 12. 6, compared with 6.6 in untreated Colec11⁻/⁻/C3⁻/⁻ cells \n(p < 0.01) and 16 in WT cells (p < 0.001 vs. untreated Colec11⁻/⁻/C3⁻/⁻ cells). Cell-surface membrane \nattack complex (MAC) deposition was also partially rescued (Figure 6f, 6g). Notably, since rCL-11 \nrestored MAC formation even in the absence of added C3 (Figure 6f), this suggests an alternative, \nC3-bypass activation mechanism downstream of CL -11 that leads to activation of the terminal \ncomplement pathway. \n \nCL-11 and Complement Interact with Osteoclasts Throughout Life \nGiven these ex vivo findings, we hypothesized that CL -11–OCL interactions could activate \ncomplement and promote OCL differentiation under physiological conditions. To test this, we \nanalysed unmanipulated WT  mouse bone sections for CL -11 and MAC deposition.  Large \nmultinucleated TRAP+ OCLs were strongly positive for CL-11 and polymeric C9 (Figure 7a). Similarly, \nin normal embryonic bone at different pre-OCL developmental stages, CL-11 and C9 localized to \nregions of cellular condensation within presumptive vertebral bodies (Figure 7b). \nThese findings strongly support a functional relationship between CL-11, MAC assembly, and OCL \ndifferentiation. Our data suggest that CL -11 and complement play a  regulatory role in \nosteoclastogenesis from embryonic bone development through adult bone maintenance. We \npropose that chronic failure of this function leads to vertebral pathology by 12–13 weeks of age, as \nobserved in our mouse model. \n \nDiscussion \nOur results demonstrate that mice lacking CL-11 develop age-related vertebral bone loss, but only \nin the absence of at least one additional key component of the lectin or alternative complement \npathways. This dual deficiency was associated with impaired osteoclast (OCL) differentiation in bone \nmarrow cultures, which was largely restored by supplementation with recombinant CL -11 at \nphysiological concentration. The presence of CL-11 and the membrane attack complex (C5b -9) on \nosteoclasts and precursor cells in normal bone further supports the role of CL-11 as a physiological \nregulator of OCL differentiation, contributing to bone remodelling throughout life. \n \nNot all mice with the vulnerable double-knockout (DKO) genotype were affected, likely due to \ngenetic and environmental interplay leading to overt bone pathology in up to 28% of adults. The \nlumbo-thoracic regions of the vertebral spine were most susceptible, suggesting that repetitive \nmechanical stress from body movements during feeding and grooming may contribute to the \nphenotype. \n \nBone is a dynamic tissue maintained by bone -forming osteoblasts (OBLs) and bone-resorbing OCLs, \nboth of which interact with the complement system. OBLs, derived from mesenchymal lineage, \nsecrete bone matrix before differentiating into osteocytes (Schoengra f et al., 2013). OBLs are \npotential targets for C5a, as C5aR1 is involved in osteogenic differentiation and migration (Bergdolt \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 6 \net al., 2017, Ignatius et al., 2011). In contrast, OCLs arise from the monocyte -macrophage lineage \nand they degrade bone matrix through proteolytic enzymes and acidification (Schoengraf et al., \n2013). Deletion of CD59, an inhibitor of MAC formation, results in increased bone resorption and \ncortical bone excess due to elevated OCL differentiation, with no impact on OBL differentiation \n(Bloom et al., 2016). Our data suggest that the complement system plays a homeostatic role in bone \nmaintenance, implicating CL-11 as a critical trigger in this mechanism. While MAC gain-of-function \nappears to promote excessive long bone growth (Bloom et al., 2016), our study shows that its loss \nleads to vertebral spine disintegration. Although both gain and loss of MAC affect the skeletal \nsystem, CL -11 has multifunctional roles beyond MAC signalling, explaining the differing bone defects. \n \nThe bone abnormalities observed in our study required the loss of a second complement activation \ncomponent alongside CL -11. Redundancy is a hallmark of the complement system, ensuring \neffective conversion of C3 and C5 into their active forms through multiple pathways. This complexity \nmay explain why an additional activation-pathway defect (e.g., CFB, MASP-2, C3) was necessary for \nthe bone repair mechanism to fail. Conversely, MAC formation was restored by CL -11 \nsupplementation, even in the absence of C3. Like thrombin —a protease capable of cleaving C5 to \nbypass C3 (Huber-Lang et al., 2006)—CL-11-driven LP activation may directly cleave C5, initiating \nMAC formation independently of C3, albeit less efficiently. \n \nThe mouse phenotype in our model is more specific than that of most of the reported 3MC \nsyndrome patients, who present with multiple developmental defects. Besides common features \nsuch as hypertelorism and craniosynostosis  (Munye et al., 2017) skeletal abnormalities in 3MC \ninclude malformed ears (Talenti et al., 2018), hypoplastic scapulae (Gardner et al., 2017), skull \nasymmetry (Basdemirci et al., 2019), and various cleft palate forms (Leal et al., 2008, Munye et al., \n2017). Mechanistic studies have focused on neural crest cell (NCC) patterning (Nauser and Sacks, \n2023, Rooryck et al., 2011), which gives rise to craniofacial mesenchyme and contributes to skull \nand pharyngeal bone development (Minoux and Rijli, 2010). The restricted phenotype observed in \nour mice, manifesting only around 12 weeks of age, suggests distinct pathogenic mechanisms \nbeyond embryonic NCC migration. Notably, no skull abnormalities were detected in affected mice \nby CT measurement. Our focus on OCLs reveals a different CL-11-regulated cellular mechanism. The \nobserved trabecular bone loss, increased marrow adiposity, and nucleus pulposus degradation of \nintervertebral discs align with dysregulation in both mesenchymal (e.g., OBLs, chondrocytes, \nadipocytes) and hematopoietic (e.g., OCLs) lineages. \n \nComplement activation in wild-type (WT) embryos was evident between E13.5 and E18.5, with MAC \nformation detected on chondrocytes from E13.5 —coinciding with the cartilaginous anlagen \nformation before primary ossification at E18.5. This period marks the transition of proliferating \nchondrocytes to hypertrophic chondrocytes, followed by invasion by OCLs and endothelial cells, and \nsubsequent bone matrix deposition by differentiating OBLs (Berendsen and Olsen, 2015, Egawa et \nal., 2014). The reduced MAC formation observed in DKO embryos from E13.5 to E18.5 may have \nimpaired endochondral ossification, predisposing them to mechanical stress-induced bone decay in \nadult life. \n \nBone marrow stromal cells and monocyte-macrophage lineages produce complement components \n(e.g., CL -11, C3, CFB) that regulate the local microenvironment. Our differentiation experiments \nincorporated no exogenous complement, except for rCL-11 in the reconstitution study. Given  this \nCL-11 gene expression and detected complement activation products (C3d, MAC), we propose that \ncomplement proteins produced locally regulate OCL differentiation from myeloid precursors. Thus, \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 7 \nOCLs and their progenitors likely regulate by secreting CL -11 and other complement factors, \ncontributing to OCL renewal and bone maintenance. \n \nIn summary, we identify a novel function of CL-11 in regulating OCL differentiation from myeloid \nprecursors. This function depends on synergy with other lectin or alternative pathway components; \nwithout it, bone integrity deteriorates with age. Our model offers insight into how CL-11 mutations \nand lack of complement signalling contribute to human skeletal abnormalities and suggest \ninvolvement of 3MC syndrome mechanisms beyond embryonic NCC migration . Conversely, \nexcessive OCL activity is a hallmark of diseases such as renal bone disease, osteoporosis and erosive \nosteoarthritis, where monoclonal antibody drugs targeting the OCL differentiation factor RANKL are \nnow in clinical use or being evaluated (Cummings et al., 2009, Wittoek et al., 2024) . Our study \nsuggests CL-11 too as a potential target in this context. \n \nMaterials and Methods \nAnimals \nColec11-/- mice (Farrar et al., 2016; Howard et al., 2020) were backcrossed to C57BL/6 for four \ngenerations. Masp-2-/- mice were kindly provided by W. Schwaeble, Cambridge, UK (Asgari et al., \n2014). C3-/- mice  were used as previously described (Zhou et al., 2000; Wessels et al., 1995). Cfb-/- \nmice were a generous gift from R. Wetsel, University of Texas (Matsumoto et al., 1997; Watanabe \net al., 2000). \nDKO-homozygote mice were generated by crossbreeding Colec11-/- mice with C3-/-, Cfb-/- and Masp-\n2-/- strains and genotyped by PCR. Kyphosis and scoliosis were identified through visual inspection \nand spine palpation by two independent observers in both KO and WT mice of the same age. All \nexperiments adhered to the Animals (Scientific Procedures) Act 1986. \nDissection and Micro-CT \nA dorsal incision from the scapula to the pelvis allowed direct inspection of the spines. Computed \ntomography was performed using a GE Explore Locus SP µCT scanner at the KCL Craniofacial \nRegenerative Biology Centre (Tabler et al., 2013). Three-dimensional isosurfaces were quantified \nusing MicroView software (GE), enabling the identification of abnormal vertebrae with interruptions \nin smooth surfaces. The affected scapular area was calculated as a proportion of the total area. \nSkeletal Preparations \nWhole-mount skeletal staining followed the method described by Rigueur and Lyons (2014). \nDehydrated skeletons were fixed in 95% ethanol overnight at room temperature (RT) twice, \nfollowed by fixation in 100% acetone at RT to remove adipose tissue. Specimens were stained with \nAlcian Blue (0.03% w/v in 80% ethanol and 20% glacial acetic acid) for three days, washed in 75% \nethanol, destained overnight in 95% ethanol, and precleared in 1% potassium hydroxide (w/v) \novernight. Alizarin Red staining (0.005% w/v; MP  Biochemicals) in 1% potassium hydroxide was \nperformed for five days. Samples were cleared in 1% potassium hydroxide overnight and stored in \n100% glycerol (Sigma). Imaging was conducted using an SL1 digital camera (Canon) with specimens \nilluminated from behind. GNU Image Manipulation Program (GIMP) software was used to \nstandardize light levels across specimens. \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 8 \nHistology \nKidney, liver, heart, and pituitary tissues were fixed overnight in 4% paraformaldehyde (PFA) in PBS, \ntransferred to 70% ethanol, and paraffin embedded. Bone specimens were dissected to remove soft \ntissue and fixed overnight in 4% PFA. Long bones were decalcified in 14% EDTA (w/v; Sigma) for 14 \ndays (pH 7.2), with daily EDTA replacement. Samples were rinsed four times in dH2O, dehydrated \nthrough graded ethanol (30%-70%), and paraffin-embedded. Embryonic tissue was fixed overnight \nin 4% PFA and transferred to 70% ethanol before embedding. \nSections (4-7 µm) were stained with haematoxylin -eosin, toluidine blue, or Masson-Goldner stain \n(Sigma -Aldrich; 1.00485). Toluidine blue was used to detect proteoglycans/glycosaminoglycans in \ncartilage, while Masson-Goldner staining identified muscle fibres, collagenous fibres, fibrin, and \nerythrocytes. Haematoxylin staining was used for vertebral body visualization. \nImmunohistochemistry and Immunofluorescence \nFor immunohistochemistry, antigen retrieval was performed on deparaffinized tissue sections using \n100 mM sodium citrate (pH 6.0), followed by transfer to PBS. Tissue sections were blocked for 1 \nhour before incubation with primary antibodies (Table 2) overnight at 4°C. A biotinylated secondary \nantibody was applied for 1 hour at RT, followed by ABC kit and DAB (Vector Labs), counterstaining \nwith haematoxylin, and dehydration for mounting with DPX (Sigma).  \nFor immunofluorescence, a biotinylated anti-rabbit secondary antibody (Vector Labs; BA-1000) was \napplied for 1 hour at RT, followed by FITC -conjugated streptavidin (GeneTex; GTX30950) or \nstreptavidin-594 (Vector Labs; SA -5594). Nuclear staining was performed using DAPI (1:10,000 in \nPBS; Life Technologies) before mounting with PermaFluor (LabVision). Mouse anti -human CL -11 \nstaining employed a mouse -on-mouse kit (Vector Labs; BMK -2202). Spinal sections were stained \nwith TRAP (Sigma -Aldrich; 387A) for osteoclast identification. \nBone Marrow Extraction \nBone marrow (BM) was extracted from femurs and tibias of euthanized mice following Maridas et \nal. (2018). Bones were cleared of soft tissue, washed in 70% ethanol for 1 minute, and transferred \nto PBS. BM was flushed out using a 25G needle, filtered through a 70 µm strainer, centrifuged (10 \nmin at 1200 rpm), resuspended at 2 × 10 6 cells/mL in BM medium (10% FBS, 1% Pen/Strep in α -\nMEM [Gibco]), and preincubated for 2 hours on a 13  mm glass coverslips coated with 1% gelatin \n(w/v) (Howard et al., 2020). Media was changed after 72 hours and subsequently every 48 hours \nuntil confluency. \nOsteoclast Differentiation \nOsteoclast differentiation was induced using BM medium supplemented with 0.05 µg/mL murine \nRANKL (PeproTech; 315 -11C) and 0.015 µg/mL murine CSF (Life Technologies; RP8615), replaced \nevery 48 hours for seven days. Rescue experiments included 2 mM CaCl 2 (Sigma -Aldrich) and 0.9 \nµg/mL human rCL -11 (Howard et al., 2020; Venkatraman Girija et al., 2015) at each medium \nreplacement. Staining was performed for TRAP or Alexa Fluor 488-conjugated phalloidin (Thermo-\nFisher Scientific; A12379)  and photographs were taken using  an Olympus BX51 fluorescent \nmicroscope.  \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 9 \nQuantitative Real-Time PCR \nRNA was extracted from BM -derived cells using the miRNeasy kit (Qiagen; 217084), followed by \ncDNA synthesis (GoScript Reverse Transcription System, Promega; A500). Gene expression was \nquantified using the 2-ΔΔCt method, with GAPDH as the reference gene. Re sults are expressed as \nfold-change in expression of test genes in differentiated versus undifferentiated cells. \nQuantification of Staining \nImageJ (NIH) software (Howard et al., 2020; Farrar et al., 2016) in conjunction with Ilastik 1.4.1rc2 \nsoftware were used to generate pixel-based segmentation image s of TRAP-stained OCLs and \ncalculate the area of OCLs in microscopic images  —see Supplemental Fig 5c for details. Percent C9 \nstaining was normalized to DAPI. Vertebral adiposity was estimated by counting circular white areas \nin at least three vertebrae across six sections per vertebral region. Nucleus pulposus area was \nmeasured using NDP Viewer (Hamamatsu Photonics). Osteoclast quantification was based on TRAP-\nstained cultures at 100× magnification. \nStatistics \nData are presented as mean ± SEM. Comparisons between two groups were analysed using an \nunpaired two-tailed Student’s t-test (p < 0.05 considered significant). One-way or two-way ANOVA \nwith multiple comparisons was used for three or more groups (p < 0.05 considered significant). \nStatistical analysis was conducted using GraphPad Prism v9. \nSummary of Supplemental Material \nSupplemental figures include CT images of extra-vertebral bones (Supplemental Fig 2), histological \nsections of DKO soft tissues (Supplemental Fig 3), litter size and growth rates (Supplemental Fig 4), \nand differentiation study results (Supplemental Fig 5). A graphical abstract is also provided. \nData Availability Statement \nAll relevant data are included in the main manuscript and/or Supplemental files. \nAcknowledgements \nWe thank Christopher Howard for his guidance on image analysis and Professor Paul Morgan for \ndonating the C9 antibody used in this study. We are grateful to Professor Russel Wallis for providing \nthe rCL-11 clone. Additionally, we appreciate the valuable advice on manuscript preparation from \nProfessor Karen Liu, Professor Agamemnon Grigoriadis, and Dr William Barrell. This work was \nsupported by the Medical Research Council grant MR/M012263/1: Collectin-11 as a Trigger of the \nInnate Immune Response in Renal Transplantation. \nAuthor Contributions \nM. Howard, C. Nauser, C. Farrar, S. Mukhopadhyay, and S. Sacks conceived and designed the \nexperiments. M. Howard, A. Polycarpou, D. Smolarek, R. Greenlaw, C. Nauser, Y. Jeon, D. Vizitiu, and \nC. Farrar performed the experiments. M. Howard conducted the statistical analysis. Peter Garred \nprovided a specific antibody and contributed to manuscript preparation. C. Farrar and S. Sacks \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 10 \nassisted with result interpretation. M. Howard led the manuscript writing, with substantial input \nfrom S. Sacks. All authors provided critical feedback and contributed to shaping the research, \nanalysis, and manuscript. \n \nDisclosure of Conflicts of Interest \nNone. \n \n \nReferences \n \nANDRADES, J. A., NIMNI, M. E., BECERRA, J., EISENSTEIN, R., DAVIS, M. & SORGENTE, N. 1996. \nComplement proteins are present in developing endochondral bone and may mediate \ncartilage cell death and vascularization. Exp Cell Res, 227, 208-13. \nASGARI, E., FARRAR, C. A., LYNCH, N., ALI, Y. M., ROSCHER, S., STOVER, C., ZHOU, W., SCHWAEBLE, \nW. J. & SACKS, S. H. 2014. Mannan-binding lectin-associated serine protease 2 is critical for \nthe development of renal ischemia reperfusion injury and mediates tissue injury in the \nabsence of complement C4. Faseb j, 28, 3996-4003. \nATIK, T., KOPARIR, A., BADEMCI, G., FOSTER, J., 2ND, ALTUNOGLU, U., MUTLU, G. Y., BOWDIN, S., \nELCIOGLU, N., TAYFUN, G. A., ATIK, S. S., OZEN, M., OZKINAY, F., ALANAY, Y., KAYSERILI, H., \nTHIEL, S. & TEKIN, M. 2015. Novel MASP1 mutations are associated with an expanded \nphenotype in 3MC1 syndrome. Orphanet J Rare Dis, 10, 128. \nBASDEMIRCI, M., SEN, A. & CEYLANER, S. 2019. Novel mutation in MASP1 gene in a new family \nwith 3MC syndrome. Clin Dysmorphol, 28, 91-93. \nBERENDSEN, A. D. & OLSEN, B. R. 2015. Bone development. Bone, 80, 14-18. \nBERGDOLT, S., KOVTUN, A., HÄGELE, Y., LIEDERT, A., SCHINKE, T., AMLING, M., HUBER-LANG, M. & \nIGNATIUS, A. 2017. Osteoblast-specific overexpression of complement receptor C5aR1 \nimpairs fracture healing. PLoS One, 12, e0179512. \nBLOOM, A. C., COLLINS, F. L., VAN'T HOF, R. J., RYAN, E. S., JONES, E., HUGHES, T. R., MORGAN, B. \nP., ERLANDSSON, M., BOKAREWA, M., AESCHLIMANN, D., EVANS, B. A. J. & WILLIAMS, A. S. \n2016. Deletion of the membrane complement inhibitor CD59a drives age and gender-\ndependent alterations to bone phenotype in mice. Bone, 84, 253-261. \nBRADLEY, K., NORTH, J., SAUNDERS, D., SCHWAEBLE, W., JEZIORSKA, M., WOOLLEY, D. E. & \nWHALEY, K. 1996. Synthesis of classical pathway complement components by \nchondrocytes. Immunology, 88, 648-56. \nCUMMINGS, S. R., SAN MARTIN, J., MCCLUNG, M. R., SIRIS, E. S., EASTELL, R., REID, I. R., DELMAS, \nP., ZOOG, H. B., AUSTIN, M., WANG, A., KUTILEK, S., ADAMI, S., ZANCHETTA, J., LIBANATI, \nC., SIDDHANTI, S. & CHRISTIANSEN, C. 2009. Denosumab for prevention of fractures in \npostmenopausal women with osteoporosis. N Engl J Med, 361, 756-65. \nEGAWA, S., MIURA, S., YOKOYAMA, H., ENDO, T. & TAMURA, K. 2014. Growth and differentiation \nof a long bone in limb development, repair and regeneration. Dev Growth Differ, 56, 410-\n24. \nEHRNTHALLER, C., HUBER-LANG, M., NILSSON, P., BINDL, R., REDEKER, S., RECKNAGEL, S., RAPP, A., \nMOLLNES, T., AMLING, M., GEBHARD, F. & IGNATIUS, A. 2013. Complement C3 and C5 \ndeficiency affects fracture healing. PLoS One, 8, e81341. \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 11 \nFARRAR, C. A., TRAN, D., LI, K., WU, W., PENG, Q., SCHWAEBLE, W., ZHOU, W. & SACKS, S. H. 2016. \nCollectin-11 detects stress-induced L-fucose pattern to trigger renal epithelial injury. J Clin \nInvest, 126, 1911-25. \nGARDNER, O. K., HAYNES, K., SCHWEITZER, D., JOHNS, A., MAGEE, W. P., URATA, M. M. & \nSANCHEZ-LARA, P. A. 2017. Familial Recurrence of 3MC Syndrome in Consanguineous \nFamilies: A Clinical and Molecular Diagnostic Approach With Review of the Literature. Cleft \nPalate Craniofac J, 54, 739-748. \nHANSEN, S., SELMAN, L., PALANIYAR, N., ZIEGLER, K., BRANDT, J., KLIEM, A., JONASSON, M., \nSKJOEDT, M. O., NIELSEN, O., HARTSHORN, K., JØRGENSEN, T. J., SKJØDT, K. & HOLMSKOV, \nU. 2010. Collectin 11 (CL-11, CL-K1) is a MASP-1/3-associated plasma collectin with \nmicrobial-binding activity. J Immunol, 185, 6096-104. \nHANSEN, S. W., OHTANI, K., ROY, N. & WAKAMIYA, N. 2016. The collectins CL-L1, CL-K1 and CL-P1, \nand their roles in complement and innate immunity. Immunobiology, 221, 1058-67. \nHUBER-LANG, M., SARMA, J. V., ZETOUNE, F. S., RITTIRSCH, D., NEFF, T. A., MCGUIRE, S. R., \nLAMBRIS, J. D., WARNER, R. L., FLIERL, M. A., HOESEL, L. M., GEBHARD, F., YOUNGER, J. G., \nDROUIN, S. M., WETSEL, R. A. & WARD, P. A. 2006. Generation of C5a in the absence of C3: \na new complement activation pathway. Nat Med, 12, 682-7. \nIGNATIUS, A., SCHOENGRAF, P., KREJA, L., LIEDERT, A., RECKNAGEL, S., KANDERT, S., BRENNER, R. \nE., SCHNEIDER, M., LAMBRIS, J. D. & HUBER-LANG, M. 2011. Complement C3a and C5a \nmodulate osteoclast formation and inflammatory response of osteoblasts in synergism \nwith IL-1β. J Cell Biochem, 112, 2594-605. \nKESHI, H., SAKAMOTO, T., KAWAI, T., OHTANI, K., KATOH, T., JANG, S. J., MOTOMURA, W., \nYOSHIZAKI, T., FUKUDA, M., KOYAMA, S., FUKUZAWA, J., FUKUOH, A., YOSHIDA, I., SUZUKI, \nY. & WAKAMIYA, N. 2006. Identification and characterization of a novel human collectin CL-\nK1. Microbiol Immunol, 50, 1001-13. \nLEAL, G. F., SILVA, E. O., DUARTE, A. R. & CAMPOS, J. F. 2008. Blepharophimosis, blepharoptosis, \ndefects of the anterior chamber of the eye, caudal appendage, radioulnar synostosis, \nhearing loss and umbilical anomalies in sibs: 3MC syndrome? Am J Med Genet A, 146a, \n1059-62. \nMA, Y. J., SKJOEDT, M. O. & GARRED, P. 2013. Collectin-11/MASP complex formation triggers \nactivation of the lectin complement pathway--the fifth lectin pathway initiation complex. J \nInnate Immun, 5, 242-50. \nMINOUX, M. & RIJLI, F. M. 2010. Molecular mechanisms of cranial neural crest cell migration and \npatterning in craniofacial development. Development, 137, 2605-21. \nMÖDINGER, Y., LÖFFLER, B., HUBER-LANG, M. & IGNATIUS, A. 2018. Complement involvement in \nbone homeostasis and bone disorders. Semin Immunol, 37, 53-65. \nMUNYE, M. M., DIAZ-FONT, A., OCAKA, L., HENRIKSEN, M. L., LEES, M., BRADY, A., JENKINS, D., \nMORTON, J., HANSEN, S. W., BACCHELLI, C., BEALES, P. L. & HERNANDEZ-HERNANDEZ, V. \n2017. COLEC10 is mutated in 3MC patients and regulates early craniofacial development. \nPLoS Genet, 13, e1006679. \nNAUSER, C. L. & SACKS, S. H. 2023. Local complement synthesis-A process with near and far \nconsequences for ischemia reperfusion injury and transplantation. Immunol Rev, 313, 320-\n326. \nROORYCK, C., DIAZ-FONT, A., OSBORN, D. P., CHABCHOUB, E., HERNANDEZ-HERNANDEZ, V., \nSHAMSELDIN, H., KENNY, J., WATERS, A., JENKINS, D., KAISSI, A. A., LEAL, G. F., \nDALLAPICCOLA, B., CARNEVALE, F., BITNER-GLINDZICZ, M., LEES, M., HENNEKAM, R., \nSTANIER, P., BURNS, A. J., PEETERS, H., ALKURAYA, F. S. & BEALES, P. L. 2011. Mutations in \nlectin complement pathway genes COLEC11 and MASP1 cause 3MC syndrome. Nat Genet, \n43, 197-203. \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 12 \nSCHOENGRAF, P., LAMBRIS, J. D., RECKNAGEL, S., KREJA, L., LIEDERT, A., BRENNER, R. E., HUBER-\nLANG, M. & IGNATIUS, A. 2013. Does complement play a role in bone development and \nregeneration? Immunobiology, 218, 1-9. \nSCHWAEBLE, W. J., LYNCH, N. J., CLARK, J. E., MARBER, M., SAMANI, N. J., ALI, Y. M., DUDLER, T., \nPARENT, B., LHOTTA, K., WALLIS, R., FARRAR, C. A., SACKS, S., LEE, H., ZHANG, M., IWAKI, \nD., TAKAHASHI, M., FUJITA, T., TEDFORD, C. E. & STOVER, C. M. 2011. Targeting of mannan-\nbinding lectin-associated serine protease-2 confers protection from myocardial and \ngastrointestinal ischemia/reperfusion injury. Proc Natl Acad Sci U S A, 108, 7523-8. \nTAKAHASHI, M., ISHIDA, Y., IWAKI, D., KANNO, K., SUZUKI, T., ENDO, Y., HOMMA, Y. & FUJITA, T. \n2010. Essential role of mannose-binding lectin-associated serine protease-1 in activation of \nthe complement factor D. J Exp Med, 207, 29-37. \nTALENTI, G., PINELLI, L., DAVIES, B., WYATT, M., NASH, R. & D'ARCO, F. 2018. Petrous Bone CT \nFindings in Patient With 3MC Syndrome. Otol Neurotol, 39, e743-e745. \nURQUHART, J., ROBERTS, R., DE SILVA, D., SHALEV, S., CHERVINSKY, E., NAMPOOTHIRI, S., \nSZNAJER, Y., REVENCU, N., GUNASEKERA, R., SURI, M., ELLINGFORD, J., WILLIAMS, S., \nBHASKAR, S. & CLAYTON-SMITH, J. 2016. Exploring the genetic basis of 3MC syndrome: \nFindings in 12 further families. Am J Med Genet A, 170a, 1216-24. \nWITTOEK, R., VERBRUGGEN, G., VANHAVERBEKE, T., COLMAN, R. & ELEWAUT, D. 2024. RANKL \nblockade for erosive hand osteoarthritis: a randomized placebo-controlled phase 2a trial. \nNat Med, 30, 829-836. \n \n \n  \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 13 \nFigures (Supplemental figures – please see separate PDF) \n \nFigure 1. Phenotype of mice with single and double deletions of complement genes.  \n(A) Representative images of the spinal phenotype in Colec11-/-/Cfb-/- and Colec11-/-/C3-/- and \ndouble-knockout (DKO) mice, showing kyphosis (upper panel) and scoliosis (lower panel). Lesions \nare exposed through skin incisions (n = 3 mice per group). Arrows indicate regions of enhanced \nspinal curvature. Magnified insert highlights increased fat deposits (black arrow). \n(B) Skeletal preparations of representative mice at different ages. Vertebral bodies (VBs) are \nstained with Alizarin Red, and intervertebral discs (IDs) with Alcian Blue (n = 5–7 mice per group). \nStars indicate regions where VBs have distinct outlines and IDs are separate from VBs in WT and \nyoung DKO mice. In older DKO mice, VBs lose defined edges and blend with adjacent IDs. \nMagnified inserts highlight these regions. \n  \nColec11-/-/C3-/-\nWT Colec11-/-/C3-/-\n7 weeks 13 weeks9 weeks\nColec11-/-/Cfb-/-a\nb\n***\nVB\nID\nColec11-/-/Cfb-/-\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 14 \nFigure 2. Micro-computed tomography (micro-CT) scans of affected and unaffected \nmice. \n(A) Vertebrae in WT and DKO mice, with severity ranging from mild abnormality to severe spinal \ndisorganization. Arrow 1: normal vertebra; Arrow 2: small hole and deformed transverse \nprocesses; Arrow 3: \"peppering\" from multiple small holes; Arrow 4: complete vertebral \nbreakdown. \n(B) Number of vertebrae affected by structural abnormalities (n = 6–11 mice per group).  \n(C) Multiple views of the spine in a Colec11-/-/Cfb-/- mo use, showing vertebral and spinous process \ndamage (yellow arrowheads), with the anterior surface of the vertebrae spared. \n(D) Distribution of vertebral regions affected by abnormalities (as in 2A, C). Each region is \nrepresented as a percentage of vertebrae containing structural abnormalities. Error bars = SEM. *p \n< 0.05, **p < 0.01, ****p < 0.0001. \n  \na\nColec11-/-/Cfb-/- iiColec11-/-/C3-/-\nColec11-/-/Cfb-/- i Colec11-/- /MASP-2-/-\nWT\n1\n2\n3 4\nb\nWT\nE\nColec11-/-/Cfb-/-\nc\nd\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 15 \n \nFigure 3. Histological characterization of bone phenotype and intervertebral disc \nlesions in affected mice. \n(A, B) Longitudinal sections of vertebral bodies (VB) and intervertebral discs (ID) stained with \ntoluidine blue. \n(A) Images of WT and DKO mice at 10 weeks (upper panel) and 14–20 weeks (lower panel) (scale \nbar = 500 µm).  \n(B) Higher magnification images from (A) of older mice, showing nucleus pulposus (NP) \ndegradation (black arrow), loss of structural organization, and increased adiposity in DKO mice \n(white arrows). \n(C) Quantification of NP degradation and increased VB adiposity. \n(D) Masson-Goldner staining of VBs, showing similar osteoid thickness (white arrows) (n = 3–5 \nmice per group, scale bar = 250 µm). Error bars = SEM. *p < 0.05. \n10 weeks\nWT\n>13 weeks\na\nb\nNP\nNP\nVB\nIDVB VB\nCo le c11-/-/Cf b-/-\nCo le c11-/-/C3-/-WT\nWT\nc\nd\nCo le c11-/-/C3-/-WT\nColec11-/-/C3-/- Colec11-/-/Cfb-/-\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 16 \n \nFigure 4. Phenotype and complement staining for CL-11 and C3 in developing \nembryos. \n(A) Sagittal sections of E18.5 WT and DKO embryos stained with H&E and toluidine blue, showing \nno overt embryonic phenotype (scale bar = 1000 µm). \n(B) Immunostaining for C3 and CL -11 in specific E13.5 WT tissues, including pituitary, heart and \npresumptive spinal column (scale bar = 100 µm). \"No 1°\" indicates control without primary \nantibody. \n(C) CL-11 staining of long bones at E13.5, E14.5, and E18.5. CL-11 localizes to chondrocytes (Ch) at \nE13.5 and E14.5, appears in the perichondrium  (P) at E14.5, and concentrates in the primar y \nossification centre (POC) at E18.5. Colec11-/-/C3-/- tissues serve as antibody controls, showing \nautofluorescence in blood cells but no specific staining in bone tissue (scale bar = 100 µm). \n  \nE14.5 E18.5\nWT\nb\nE13.5\nC3 CL-11No 1o\nPituitary\nVertebrae\nHeart\nc\nChP\nCh\nP\nPOC\nP\nColec11-/-/C3-/-\nCL-11\nCh\nP\nCh P\nE18.5\nWT\nColec11-/-/C3-/-\na\nE13.5\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 17 \nFigure 5. Membrane Attack Complex (MAC) detection in embryonic bone.  \n(A, B) C9 staining in long bones at different embryonic stages (scale bar = 100 µm). Clumped C9 \nstaining (white arrows) suggests polymeric C9 incorporation into MAC (scale bar = 50 µm). \nColec11-/-/C3-/- mice  show significantly reduced C9 deposition at E13.5 and E14.5, while Colec11-/- \nmice  show no major reduction (n = 3 –4 mice per group). The last row shows WT embryos stained \nwithout primary antibody to control for secondary antibody specificity. Error bars = SEM. *p < \n0.05, **p < 0.01. \n  \nb\nWT\nE13.5 E14.5 E18.5\nC9\nCellular \nlocation\nC9\nC9\nColec11-/-\nColec11-/-/C3-/-\na\nWT\nNo 1o\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 18 \n \nFigure 6. Osteoclast (OCL) differentiation from adult bone marrow stem cells and CL-\n11 impact. \n(A) qPCR analysis of OSCAR and CL-11 expression in undifferentiated and differentiated WT cells. \nBars represent means from two technical replicates per gene (n = 12 coverslips). Data is \nrepresentative of two independent experiments.   \n(B) Agarose gels of qPCR products with GAPDH as a loading control. \n(C) Microscopy of cells stained for TRAP and F-actin (scale bar = 100 µm). \n(D) Representative images of OCL differentiation in WT, Colec11-/-, Colec11-/-/C3-/-and Colec11-/-\n/Cfb-/-cells ± rCL-11 addition. Arrows indicate multinucleated OCLs with size changes in DKO \ncultures (scale bar = 100 µm). Each row of images in each treatment group represents a single \nmouse.  \n(E) Quantification of OCL area in square pixels. Each dot represents one image (n=6 WT, n=3 \nColec11-/-, n=1 Colec11-/-/C3-/-and n=2 Colec11-/-/Cfb-/- mouse ). Data were analysed using two-way \nANOVA with Sidak’s multiple comparisons. \na b c\nd\ne\nf\ng\nUndifferentiated Di ff erenti ate d\nTRAP\nF-actin\nWT\nCo le c11-/-\nCo le c11-/-/C3-/-\nCo le c11-/-/C3-/-\n  + rCL-11\nCL-11 C3 C9\n15 0\n50\n15 0\n50\nbon e\nma rr ow\nda y 7  \ndiff .\nda y 0  \nund if f.\nbon e\nma rr ow\nda y 7  \ndiff .\nda y 0  \nund if f.\nOSCAR ( 68 b p) CL-11 (149 bp)\nGAPDH (96 bp)\ncDNA\ncDNA\nosteoclast osteoclast\nbon e\nma rr ow\nda y 7  \ndiff .\nda y 0  \nund if f.\nosteoclast\ne\nOSCAR CL-11\nWT Co le c11-/-/C3-/- Co le c11-/-/Cf b-/-Co le c11-/-\n-  rCL-11\n+  rCL-11\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 19 \n(F) Immunostaining of cultures for CL-11, C3, or C9, showing C9 aggregates on OCLs (scale bar = \n250 µm).  \n(G) Quantification of C9 staining in (F). Error bars = SEM. *p < 0.05, ***p < 0.001, ****p < 0.0001. \n  \n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint \n\n 20 \n \nFigure 7. Localization of osteoclasts, CL-11, and MAC in adult and embryonic spines. \n(A) TRAP-stained adult vertebral bodies, showing multinucleated OCLs (red arrows), CL-11 (green \narrows), and C9 (yellow arrows) (n = 3). \n(B) Confocal images of embryonic vertebral bodies at E13.5 and E18.5 (white arrows). Adult spines \ncontain differentiated OCLs with CL-11 and MAC deposition. By E18.5, vertebral condensation \nbegins, with CL-11 and MAC-positive progenitor cells present. \n \nE13.5\nE18.5\n× 20\n × 40\n× 20 × 40\n× 20\n × 40\n× 20 × 40\nC9CL-11\nEmbryo\nFigure 7\nAdult spine\nTRAP\nC9\n× 40 × 60\n× 60\nCL-11\n× 20\n × 40× 40\n × 60\n× 40\na\nb\n.CC-BY 4.0 International licensemade available under a \n(which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is \nThe copyright holder for this preprintthis version posted May 13, 2025. ; https://doi.org/10.1101/2025.05.08.652605doi: bioRxiv preprint","source_license":"CC-BY-4.0","license_restricted":false}