Single-nucleus RNA sequencing reveals the cellular diversity of human cerebrospinal fluid in the context of intraventricular hemorrhage

The cellular diversity of human cerebrospinal fluid following intraventricular hemorrhage revealed by single-nucleus RNA sequencing | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results The cellular diversity of human cerebrospinal fluid following intraventricular hemorrhage revealed by single-nucleus RNA sequencing View ORCID Profile Sonia Malaiya , View ORCID Profile Riccardo Serra , Marcia Cortes-Gutierrez , View ORCID Profile Bradley E. Wilhelmy , View ORCID Profile Emily Jusuf , Mayur Somalinga , Dave Peprah , Hirschel Nambiar , Kevin T. Kim , View ORCID Profile Jordan R. Saadon , View ORCID Profile Parantap D. Patel , View ORCID Profile Steven K. Yarmoska , View ORCID Profile Maureen Rakovec , John Kim , Catherine Lei , Shreyas Panchagnula , View ORCID Profile Reina Ambrocio , Meagan Cherney , View ORCID Profile Nicholas A. Morris , Natarajan Ayithan , Xiaoxuan Fan , Volodymyr Gerzanich , J. Marc Simard , Neeraj Badjatia , View ORCID Profile Gary Schwartzbauer , View ORCID Profile Gunjan Y. Parikh , View ORCID Profile Seth Ament , View ORCID Profile Prajwal Ciryam doi: https://doi.org/10.1101/2025.06.05.657834 Sonia Malaiya 1 Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, USA . 2 Program in Neuroscience, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Sonia Malaiya Riccardo Serra 3 Department of Neurosurgery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Riccardo Serra Marcia Cortes-Gutierrez 1 Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Bradley E. Wilhelmy 4 Department of Neurosurgery, The Ohio State University , Columbus, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Bradley E. Wilhelmy Emily Jusuf 5 Program in Trauma, University of Maryland Shock Trauma Center , Baltimore, USA . 6 Renaissance School of Medicine, Stony Brook University , Stony Brook, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Emily Jusuf Mayur Somalinga 5 Program in Trauma, University of Maryland Shock Trauma Center , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Dave Peprah 5 Program in Trauma, University of Maryland Shock Trauma Center , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Hirschel Nambiar 5 Program in Trauma, University of Maryland Shock Trauma Center , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Kevin T. Kim 3 Department of Neurosurgery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Jordan R. Saadon 3 Department of Neurosurgery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Jordan R. Saadon Parantap D. Patel 3 Department of Neurosurgery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Parantap D. Patel Steven K. Yarmoska 3 Department of Neurosurgery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Steven K. Yarmoska Maureen Rakovec 3 Department of Neurosurgery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Maureen Rakovec John Kim 3 Department of Neurosurgery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Catherine Lei 3 Department of Neurosurgery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Shreyas Panchagnula 3 Department of Neurosurgery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Reina Ambrocio 7 Department of Neurology, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Reina Ambrocio Meagan Cherney 8 University of Maryland Medical Center , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Nicholas A. Morris 5 Program in Trauma, University of Maryland Shock Trauma Center , Baltimore, USA . 7 Department of Neurology, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Nicholas A. Morris Natarajan Ayithan 9 Department of Microbiology and Immunology, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Xiaoxuan Fan 9 Department of Microbiology and Immunology, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Volodymyr Gerzanich 3 Department of Neurosurgery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site J. Marc Simard 3 Department of Neurosurgery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Neeraj Badjatia 5 Program in Trauma, University of Maryland Shock Trauma Center , Baltimore, USA . 7 Department of Neurology, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site Gary Schwartzbauer 3 Department of Neurosurgery, University of Maryland School of Medicine , Baltimore, USA . 5 Program in Trauma, University of Maryland Shock Trauma Center , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Gary Schwartzbauer Gunjan Y. Parikh 5 Program in Trauma, University of Maryland Shock Trauma Center , Baltimore, USA . 7 Department of Neurology, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Gunjan Y. Parikh Seth Ament 1 Institute for Genome Sciences, University of Maryland School of Medicine , Baltimore, USA . 10 Maryland Psychiatric Research Center, Department of Psychiatry, University of Maryland School of Medicine , Baltimore, USA . 11 UM-MIND Institute for Neuroscience Discovery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Seth Ament For correspondence: sament{at}som.umaryland.edu pciryam{at}som.umaryland.edu Prajwal Ciryam 5 Program in Trauma, University of Maryland Shock Trauma Center , Baltimore, USA . 7 Department of Neurology, University of Maryland School of Medicine , Baltimore, USA . 11 UM-MIND Institute for Neuroscience Discovery, University of Maryland School of Medicine , Baltimore, USA . Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Prajwal Ciryam For correspondence: sament{at}som.umaryland.edu pciryam{at}som.umaryland.edu Abstract Full Text Info/History Metrics Supplementary material Data/Code Preview PDF Abstract Background Intraventricular hemorrhage (IVH) is a common and severe complication of hemorrhagic brain injury. Current treatments offer limited improvement in long-term neurological outcomes. Inflammatory responses in the cerebrospinal fluid (CSF) after IVH are thought to drive secondary injury, but the cellular mechanisms underlying this inflammation remain poorly defined. Methods We performed single-nucleus RNA sequencing of leukocytes isolated from CSF collected through external ventricular drains in subjects with intracerebral ( n = 6) or subarachnoid ( n = 1) hemorrhage. We characterized transcriptionally distinct subpopulations of neutrophils, monocytes, and lymphocytes by comparison to reference datasets. Cell–cell signaling networks were analyzed to infer cytokine-mediated communication, and a flow cytometry panel was developed to validate transcriptomic findings in independent CSF samples. Results We obtained 11,191 high-quality nuclei comprising neutrophils (53.8%), monocytes (26.1%), lymphocytes (17.8%), and non-immune cells (2.4%). Neutrophils segregated into Nascent, Quiescent, and Interferon-Activated states. Monocytes exhibited classical phenotypes that include interferon-activated states (characterized by expression of VCAN or PROK2 ) and CXC-chemokine expressing states (characterized by expression of CXCL5 or CXCL8 ). Lymphocytes were mainly naïve and central memory CD4⁺ T cells. Cell–cell signaling analysis predicted strong CXC chemokine signaling from monocytes to neutrophil subsets and IL-1 family–driven inflammatory responses across multiple populations. Type I and III interferon signaling defined a neutrophil population not previously described in the central nervous system. Conclusion This study delineates the diverse cellular immune landscape of CSF after IVH. Transcriptomic profiles reveal interferon, IL-1, and CXC chemokine signaling networks as potential therapeutic targets to mitigate secondary injury. Introduction Bleeding into the cerebral ventricles, known as intraventricular hemorrhage (IVH), is an ominous complication of hemorrhagic brain diseases, including intracerebral hemorrhage (ICH), subarachnoid hemorrhage (SAH), and traumatic brain injury (TBI). In adults, it results in death or severe disability in over 50% of cases 1 . The development of elevated intracranial pressure and hydrocephalus contribute to morbidity and mortality 2 – 4 . Treatment options are limited, and while around 50% of patients require temporary drainage of cerebrospinal fluid (CSF) with an external ventricular drain (EVD) 5 , 6 , lifelong CSF shunting via ventriculoperitoneal shunting is needed in up to 40% of subjects 7 , 8 . Intraventricular injection of tissue plasminogen activator does not improve functional outcomes or reduce the rate of shunting 7 . As new prognostic and treatment approaches are sought for this morbid disease, promising strategies may arise from the critical role of inflammation in IVH. The deposition of blood into the cerebral ventricles stimulates a profound inflammatory response that recruits leukocytes into the central nervous system 9 , 10 . In the first several days after injury, this response is dominated by myeloid cells: first neutrophils, then monocytes and macrophages. These cells interact with CNS resident cells, with both pathological and reparative consequences. On one hand, inflammation after IVH can cause neuronal injury, cerebral edema, and hydrocephalus; on the other, it may stimulate neurogenesis and clearance of hemorrhagic debris 11 – 14 . Because of the numerous and pleiotropic contributors to inflammation in IVH, comprehensive approaches are needed to characterize this process. Here, we sought to characterize the CSF inflammatory response in patients after IVH, using single-nucleus RNA sequencing. Our approach identified distinct states of neutrophils and monocytes in the CSF, including a population of interferon-induced neutrophils not previously described in the central nervous system. These findings illustrate the complex cellular inflammatory response to IVH and point to potential novel therapeutic targets for further investigation. Results A snRNA-seq atlas for cellular diversity of ventricular CSF in the context of intraventricular hemorrhage Previous scRNA-seq studies of CSF samples from patients with IVH did not capture its full cellular diversity. Based on prior CSF cell count data in IVH 15 , we anticipated that the CSF samples we collected would be highly represented in neutrophils and other fragile populations that could be depleted during sample processing. We reasoned that single-nucleus RNA sequencing (snRNA-seq) could provide a means to limit these processing-induced biases and might yield a more representative cellular atlas of IVH. We obtained ventricular CSF of seven study participants who presented with IVH, either directly to the University of Maryland Medical Center or after emergent transfer from another institution ( Fig. 1A ; Table 1 ). Six subjects presented after ICH, with etiologies including hypertension and coagulopathy/anticoagulation-related hemorrhage, with one subject whose etiology was unknown. One subject presented with aneurysmal SAH. Of the seven subjects, five presented with severe deficits in level of consciousness as quantified by Glasgow Coma Scale (GCS) between 3-8 and two presented with a mild deficit (GCS 13-15). Standardized scales of disease severity were consistent with moderate to severe disease. For subjects with ICH, hemorrhages originated in deep structures of the brain in a majority of cases. All subjects had IVH, either at presentation or subsequently in their hospital course. When CSF is obtained for clinical purposes, excess fluid that would otherwise be discarded is sometimes available. We collected this CSF immediately after it was aspirated. As these are convenience samples, they were obtained at a range of time points after disease onset. Non-contrast CT scans of the head from these subjects demonstrate the presence of IVH and show the EVD tip in the lateral ventricles near the foramen of Monro ( Fig. 1B ). CSF is therefore sampled directly from the ventricles, unlike lumbar puncture, which samples CSF from the lumbar cistern in the lower back. CytoSpin cytology data available from the electronic health record for sample aliquots from five of the seven patients indicated extensive infiltration of leukocytes into the CSF, primarily comprised of neutrophils, monocytes/macrophages, and T cells ( Fig. 1C ). Download figure Open in new tab Figure 1. Clinical characteristics of study participants. A. Study design (Created in https://BioRender.com ). B. Computed tomography (CT) scans. Axial images (slice thickness=3 mm) from the first CT study obtained after EVD placement for a subject with ICH (i,ii) and SAH (iii, iv). i. Thalamic ICH with IVH most prominent in right lateral ventricle (RLV) and ventriculomegaly consistent with hydrocephalus (HC), with EVD visible in left lateral ventricle. ii. More caudal slice from same CT as in (i) showing IVH in the third ventricle (3V) and bilateral lateral ventricles (LVs). iii. Aneurysmal SAH with small amount of IVH in bilateral LVs, with EVD visible at the foramen of Monro. iv. More caudal slice from same CT as in (iii) showing blood in the basal cisterns in classical aneurysmal SAH pattern, with dilated temporal horns consistent with HC. C. Cell proportions in ventricular CSF assessed by cytospin columns. snRNA-seq of CSF from these seven study participants yielded 11,191 high-quality nuclear transcriptomes ( Fig. 2A, B ). Neutrophils (n = 6,017; S100A8 + , CXCR1 + , FCGR3A + , FCGR3B + ) were the most abundant cell type, followed by monocytes (n = 2,920; VCAN + , CXCL8 + , CCL2L1 + ), and lymphocytes (n = 1,990; CD247 + ) ( Table S1 ). We also identified a small number of cells expressing glial or neuronal markers (n = 264; e.g., PTPRD, RSPO3, FLT1, CLDN10, RYR2, RBFOX1, DLG2 ) ( Fig 2C ). These cells may derive from damage to the cerebral grey matter during or after the placement of the EVD and were not analyzed further. Cell type proportions identified by snRNA-seq were in the same range as those obtained using CytoSpin columns ( Fig. 1C ), although our pooling strategy does not allow us to match individual samples across the two technologies. Overall, our data represent a relatively unbiased atlas for the cellular diversity of the CSF in the context of IVH. Download figure Open in new tab Figure 2. snRNA-seq of ventricular CSF after IVH. A. Uniform manifold approximation and projection (UMAP) plot for snRNA-seq of ventricular CSF (11,191 nuclei from seven study participants). B. Quality metrics comparing cell processing methods. C. Dot plot showing expression of canonical marker genes across cell types. Participants’ cells were preserved for snRNA-seq processing by two different methods, comparing distinct cell processing protocols in which nuclei were isolated directly from frozen cell pellets vs. cells that had been cryopreserved with DMSO and briefly thawed. snRNA-seq of nuclei derived from pelleted vs. cryopreserved cells yielded a larger number of nuclei from the cryopreservation method, but this method also produced lower RNA counts per nucleus and higher mitochondrial counts per nucleus ( Fig. 2B , Fig. S1 A-C ). The library derived from cryopreserved cells contained a smaller fraction of neutrophils and larger fraction of lymphocytes. It is unclear whether these differences correspond to biological variation among the participants, or whether they reflect a greater vulnerability of neutrophils during cryopreservation and thawing. Nevertheless, these data suggest that all of the cell types in the CSF can be obtained for snRNA-seq by either cell processing strategy. Neutrophil diversity in the ventricular CSF Next, we sought to characterize leukocyte populations in greater detail. Re-integration of the 6,017 nuclei identified as neutrophils revealed eight transcriptionally-distinct neutrophil sub-clusters. All eight of these sub-clusters were detected in both preservation methods ( Fig. S2A-D ), but with varying proportions. We annotated these neutrophil sub-clusters via comparisons to scRNA-seq of neutrophil states in the blood and in lumbar CSF 16 , enabling us to merge the eight sub-clusters into three broad neutrophil populations ( Fig. 3A , Table S4 ). The majority of neutrophils form a continuum from Nascent Neutrophils (21.29%) (neu4, neu6, neu8), which expressed S100A8 , S100A9 , and SELL , to Quiescent Neutrophil states (65.43%) (neu1, neu2, neu5, and neu7), which expressed markers such as MALAT1 and CXCL8 ( Table S2 ). In addition, a third population expressed genes that are activated during interferon (IFN) responses (13.28%) (neu3), including IFIT1 , IFIT2 , IFIT3 , and FCGR3B . These IFN-Activated Neutrophils are a cell population with diverse but not fully characterized functions 17 , 18 . To our knowledge, they have never been described in the CSF previously. Download figure Open in new tab Figure 3. Neutrophil diversity in CSF after IVH. A. UMAP indicating neutrophil subtypes. B. Annotation of neutrophil subtypes by scoring with marker genes from neutrophil subtypes identified in the blood 16 . C. Gene co-expression modules differentially expressed across neutrophil subtypes. Wilcoxon rank-sum test (one-sided for reference marker scoring, two-sided for gene co-expression networks) with Bonferroni correction used for statistical significance testing (see Tables S3, S4). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. To further characterize these neutrophil populations, we scored them on the activity of marker genes for neutrophil populations identified in the blood in two previous scRNA-seq studies by Wigerblad et al. 16 ( Fig. 3B , Tables S3, S8 ) and Kwok et al. 19 ( Fig. S3A , Tables S3, S8 ). We also scored neutrophils for the activity of 12 gene co-expression modules (N1-N12) computed within the neutrophils from our snRNA-seq data to describe distinct patterns of activity across the neutrophil populations ( Fig. 3C , S4A Table S4 ). Nascent Neutrophils were defined by elevated expression of markers for the “Nh0” immature neutrophil population from the Wigerblad dataset (adjusted p = 2.86×10 -153 ), as well as markers for two immature neutrophil populations from the Kwok dataset (S100A8-9 hi neutrophil markers (adjusted p = 6.10×10 -219 ) and IL1R2 + immature neutrophil markers (adjusted p = 2.55×10 108 ). Nascent Neutrophils also displayed high expression of gene co-expression module N7 (adjusted p = 7.27×10 -99 ). Genes in this module were over-represented for NF-κΒ activation genes (GO:0051092) and signaling pathways for pattern recognition protein receptors (GO:0002221), Toll-Like receptors (GO:0002224), and cytokine receptors (GO:0019221, GO:0004896), including cytokine-signaling pathways related to IL-12 (GO:0032655), and IL-1 (GO:0032651, GO:0005149). Quiescent Neutrophils were defined by the expression of long noncoding RNAs that modulate with inflammatory responses (e.g., MALAT1 , NEAT1 , and LUCAT1 20 – 23 ) and genes that are associated with neutrophil migration and chemotaxis (e.g., SSH2 , DOCK5 , CXCL8 24 – 26 ) . While markers of the Wigerblad ‘Nh2’ “transcriptionally inactive” cluster had overall low expression in our dataset, Quiescent Neutrophils had a modestly increased expression of these markers relative to the other neutrophil populations (log 2 FC = 0.190, adjusted p = 4.94×10 -37 ). Several gene co-expression modules were upregulated in Quiescent Neutrophils, including N5 (adjusted p = 3.02×10 -16 ), N3 (adjusted p = 1.61×10 -39 ), N10 (adjusted p = 1.24×10 -27 ), and N11 (adjusted p = 1.88×10 -64 ). MALAT1 and NEAT1 were members of N5, suggesting a role for this module in the epigenetic regulation of inflammatory responses. N5, N3, and N11 were over-represented in genes associated with serine/threonine kinase activity (GO:0004674), with the strongest upregulation of SIK3 , STK4 , and CAMK1D . Serine/threonine kinases, including STK4 and CAMK1D , have been reported to play a variety of roles in neutrophils 27 – 29 , but their functional significance in the Quiescent Neutrophil population is unclear. Interestingly, CAMK1D and MALAT1 have also been found to be co-expressed in a subset of granulocytes in the blood 30 . The neutrophil module N12 consists of several genes that play roles in cell adhesion and migration (ADGRG3, ADGRE2, MYADM, ADAM8). The elevated expression of this module in nascent (adjusted p = 5.03×10 -74 ) and quiescent neutrophils (adjusted p = 9.55×10 -06 ) may suggest an active recruitment of these cells at the time of sample collection. IFN-Activated Neutrophils were defined by elevated expression of many canonical interferon-stimulated genes, genes. Markers of the ‘Nh3’ “IFN inducible” neutrophil population in the Wigerblad dataset were strongly induced (adjusted p < 2.23×10 -308 ), as was gene co-expression module N8 (adjusted p = 1.31×10 -269 ), whose hub genes included canonical interferon-stimulated genes such as IFI6, IFI30, IFI44 , IFI44L, IFIT1, IFIT2 and IFIT3 . Other marker genes annotated as activated in response to symbionts (GO:0140546), viruses (GO:0051607), cytokines (GO:0034097), and interferons (GO:0034341, GO:0034340, GO:0035456) included MX1, IFITM3, IFIT5, IFI6, CD74, CXCR4, and SELL . This cell state also had decreased expression of the IL1B -containing module N10 (adjusted p = 1.24×10 -63 ), consistent with the suppressive effects of interferons on IL-1 production 31 . Monocyte/Macrophage diversity in the ventricular CSF We performed similar analyses to characterize monocyte lineages in ventricular CSF ( Fig. 4 ). By sub-clustering the 2,920 monocyte-lineage cells, we identified 6 transcriptionally distinct Seurat clusters annotated based on gene expression markers in the literature as classical monocytes (four clusters, 90.6%), macrophages (5.2%), and dendritic cells (4.2%) ( Fig. 4A , Fig. S5A-D , Table S5 ). We further characterized these clusters of monocytes/macrophages based on comparison to publicly available gene expression data for monocyte clusters in the MoMac data set 32 , which integrated monocytes in health and disease states from 41 studies in 13 different tissues ( Fig. 4B , Fig. S6A , Table S6 ). We also performed genes co-expression module detection and obtained 11 transcriptionally distinct modules (M1-M11), which helped us better define the cell states we observed ( Fig. 4C , Fig. S7A , Table S7 ). Download figure Open in new tab Figure 4. Monocyte diversity in CSF after IVH. A. UMAP indicating monocyte subtypes. B. Annotation of monocyte subtypes by scoring with marker genes from select subtypes identified in a meta-analysis of monocytes and macrophages from healthy and diseased samples (MoMac) 32 . C. Gene co-expression modules differentially expressed across monocyte subtypes. Wilcoxon rank-sum test (one-sided for reference marker scoring, two-sided for gene co-expression networks) with Bonferroni correction used for statistical significance testing (see Tables S6, S7). *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. Classical monocyte markers ( SELL, CSF1R, ITGAX, CD163, CSF3R, S100A8 , S100A9 , S100A12, ALOX5AP ) were split between four distinct clusters that also expressed markers associated with monocyte activated or inflammatory states ( HLA-DRA, HLA-DRB1, CD163, IL1B, CXCL8, CCL3, CCL4, CXCL5, members of the TNF family); hence we have labelled them Activated Classical Monocytes (AC Mono). We defined each of the four AC Mono subtypes by its most differentially expressed genes encoding secreted proteins: ACMono-VCAN (36.68% of all monocyte/macrophage cells), ACMono-PROK2 (36.1%), ACMono-CXCL8 (15.24%), and ACMono-CXCL5 (2.53%). The first two subtypes were the most abundant and had in common a pro-inflammatory signature including interferon-induced genes, while the latter two were defined by their high expression of neutrophil-attracting chemokines. ACMono-VCAN was the most abundant monocyte subtype. Its eponymous marker VCAN (avg log 2 FC = 2.64, adjusted p = 1.53×10 -268 ), encodes versican, a chondroitin sulfate proteoglycan that modulates inflammatory responses in many cell types 33 , 34 . It has been reported that VCAN activation in monocytes/macrophages is dependent on Type I interferon signaling 35 . Consistent with this, we observed elevated expression of the interferon-induced genes IFI44 and IFI44L. ACMono-VCAN also had upregulation of genes in co-expression modules M2 (adjusted p = 1.35×10 -1 18 ) and M3 (adjusted p = 1.05×10 -11 3 ), enriched for viral defense responses and endocytic/phagocytic genes, respectively ( Table S7 ). For instance, scavenger protein genes such as STAB1 (avg log 2 FC = 1.83, adjusted p = 1.18×10 -117 ) and CD163 (avg log 2 FC = 1.76, adjusted p = 2.25×10 -83 ) were upregulated, along with components of the THBS pathway ( Fig. 5C ), suggesting that this is an interferon-activated cell state responding through modification of the cell matrix. By contrast, ACMono-VCAN appeared not to be activated by other cytokines such as interleukin-1, as it had decreased expression of M5 (adjusted p = 7.42×10 -164 ), a gene co-expression module enriched in genes involved in cellular response to interleukin 1 (GO:0071347), apoptotic process (GO:0006915), and NF-κB signaling (GO:0007249) module. Download figure Open in new tab Figure 5. Lymphocyte diversity in the ventricular CSF. A. UMAP showing reference annotations in blood from the Azimuth data set 51 . B. Annotation of lymphocyte subtypes in CSF after IVH based on label transfer analysis. (C.,D.) Dot plots showing expression of marker genes in the reference set (C) and our lymphocyte population in CSF after IVH (D). ACMono-PROK2 was characterized by the expression of PROK2 (avg log 2 FC = 1.76, adjusted p = 1.68×10 -38 ), encoding prokineticin-2, a peptide with chemokine–like activity that is highly expressed in many classical monocytes 36 with roles in cell migration 37 , 38 and differentiation 39 , 40 and neurogenesis 41 , 42 . More broadly, the ACMono-PROK2 gene signature suggests a pro-inflammatory state, with elevated expression of gene co-expression module M8 (adjusted p = 3.52×10 -115 ), including the cytokines TNF , IL6 , and IL1B , and genes related to NF-κB signaling. Comparisons to MoMac revealed overlap with MoMac clusters 1 (adjusted p = 2.35×10 -126 ) and 8 (adjusted p = 4.13×10 -101 ), both of which are subtypes of CD16 + monocytes with high expression of inflammatory genes. ACMono-CXCL8 was defined by elevated expression of pro-inflammatory ligands such as CXCL8 , CCL3 , and CCL4 , markers of activated M1-like monocytes that act as chemoattractants for other monocytes, neutrophils, NK cells, and T-cells 43 . ACMono-CXCL8 had high expression of markers for Module M5 (adjusted p = 1.60×10 -142 ), which was enriched for Gene Ontology terms related to vesicle secretion ( e.g., Proton-Transporting V-type ATPase (GO:0033180), positive regulation of exosomal secretion (GO:1903543), and multivesicular body secretion (GO:0032585) and with the interleukin-1-mediated signaling pathway (GO:0070498). Similarly, ACMono-CXCL8 had elevated expression of Module M9 (adjusted p = 6.07×10 -101 ), which consists primarily of genes associated with vesicle production. This cluster also highly expressed markers of the “IL1B” MoMac Cluster 15 (adjusted p = 2.30×10 -141 ), as well as the NF-κB transcription factor inhibitor NFKBIA, which regulates the pro-inflammatory activities of NF-κB downstream of TLR4. Collectively, these enrichments suggest that ACMono-CXCL8 may have M1-like activity. ACMono-CXCL5, the smallest monocyte population, expressed activated monocyte ligands such as CXCL5 and CXCL3 that are known for their roles as chemoattractants for myeloid cells. In common with ACMono-CXCL8, ACMono-CXCL5 had elevated expression Module M9 (adjusted p = 2.85×10 -6 ) , and MoMac Cluster 15 (adjusted p = 2.95×10 -12 ). Macrophages (Mac) represented 5.24% of cells in the monocyte lineage. In addition to genes that are largely expressed in macrophages such as GPNMB, PLD3 and CTSD , these cells were marked by high expression of pro-inflammatory genes such as HIF1A-AS3 44 FTL 45 , and MDM2 46 . FTL expression in macrophages has been associated with exposure to oxidized low-density lipoprotein 47 . Consistent with this, we found high expression of several genes associated with lipid-associated macrophages (e.g., CSTB , CTSD , GPNMB , LGALS3 , PLD3 , and PLIN2 48 ), which are also induced in macrophages in a murine model of stroke 49 . Interestingly, we also detected high expression for markers of MoMac cluster 3 (adjusted p = 1.50×10 -26 ), a TREM2 + , M2-like population. Overall, these annotations demonstrate the presence of macrophages in the CSF after IVH and provide evidence for heterogeneous reactive states. The final sub-population (4.21%) expressed markers of dendritic cells (DC). We annotated these cells as DC based on shared markers with MoMac cluster 14 (adjusted p = 5.54×10 -22 ), described by the MoMac authors as dendritic cell contamination. We note, however, that dendritic cells share several markers with intermediate monocytes, making a definitive annotation difficult. Some cells in the cluster expressed intermediate monocyte markers such as CIITA , suggesting a mixed population ( Fig. S5D ) . Notably, this population was also distinguished by elevated expression of MHC class II markers such as HLA-DRA, HLA-DRB1, HLA-DQA1 and HLA-DQA2 , suggesting that these cells may have the ability to interact with and activate CD4⁺ T cells. Lymphocyte diversity in the intraventricular CSF The 1,990 nuclei identified in our data as lymphocytes expressed canonical pan-lymphocyte and T cell markers, such as PDE3B , FOXP1 , MBNL1 , ANK3 , BCL2 , FYN , SKAP1 and CD247 . A small population of these cells expressed B-cell markers ( CD19 , CD20 , CD21 ), consistent with observations for intraparenchymal hematoma and CSF in subarachnoid hemorrhage by flow cytometry 15 , 50 . De novo clustering of lymphocytes did not clearly define known T- or B-cell subtypes, most likely because of the relatively small number of cells representing each subtype. Therefore, to better characterize lymphocyte populations, we performed a label transfer analysis by integrating our lymphocytes with those from a large PBMC reference atlas and projected our data onto the UMAP from the reference data to identify shared marker genes and confirm annotations 51 ( Fig. 5 ). This analysis suggested that most of the lymphocytes we detected were CD4 + subtypes, especially central memory T-cells (28.5%) and CD4 + naïve T-cells (41.6%). CD8 + effector memory T-cells represented an additional 14.7% of lymphocytes. Double-negative CD4 - CD8 - T-cells represented 10.4% of all lymphocytes, suggesting an enrichment of activate T-cell states. Consistent with this, further sub-clustering revealed a population expressing pro-inflammatory markers such as S100A8 , S100A9 , CSF3R , IFITM2 , and CXCR2 , a marker of γδ T-cells 52 . Intercellular communication networks in the intraventricular CSF Chemokine ligands and their receptors were markers for several sub-populations of neutrophils, monocytes, and lymphocytes. For instance, the chemokine CXCL8 was expressed almost exclusively by ACMono-CXCL8, while its receptor, CXCR2, was expressed primarily by neutrophils. The specific expression of these ligand-receptor pairs provides the potential for signaling between populations. To more rigorously investigate this putative cell-cell communication, we made use of Immune Dictionary 53 and the CellChat 54 , 55 package in R. The Immune Dictionary provides single-cell transcriptomic profiles of 17 immune cell types in response to treatment with each of 86 cytokines in mice. We reasoned that similarities between cells in our data set and those responsive to specific cytokines in the Immune Dictionary imply common stimuli. We observed expression of IL-1α-induced genes in all three neutrophil populations ( Fig. 6A , Table S8 ). In the IFN-activated neutrophil population, expression signatures of interferon responses (Type I and Type II) and IL-1 signaling (IL-1α, IL-1β, IL-15, IL-18, and IL-36α) were evident. While IFN-activated neutrophils were IL-1 responsive, they were not IL-1 productive, as they were negatively associated with co-expression module N10, a hub gene of which was IL1B . This is consistent with observations that Type I interferon can suppress IL-1 production 31 . In contrast, monocytes generally showed a greater expression of genes that are suppressed by IFN-α1 and IFN-γ stimulation ( Fig. S8 A-D ). Monocytes also expressed IL-1α- and IL-1β induced genes at high levels. Interestingly, PROK2 monocytes expressed higher levels of genes that were both induced and suppressed by IL-1α and IL-1β, suggesting a complex remodeling of the IL-1-modulated transcriptome. ACMono-VCAN, ACMono-CXCL5, and DC cells expressed a signature observed in monocytes treated with IL-33 ( Fig. 6B ), which is an IL-1 family cytokine that is released in response to cell stress or damage. Macrophages shared with monocytes an IFN-α1 and IFN-γ suppressed expression profile, but also highly expressed genes that were suppressed by IL-1α. Download figure Open in new tab Figure 6. Intracellular communication in ventricular CSF. A. Neutrophils in CSF scored by genes upregulated in Neutrophils from the Immune Dictionary upon IFNa1 exposure. B. Monocytes, dendritic cells and macrophages in CSF scored by genes upregulated in Monocytes from the Immune Dictionary upon IL33 exposure. C. Circos plot showing strength of outgoing and incoming signals across cell states in the vCSF, based on CellChat analysis. D. Heatmap of outgoing and incoming signal patterns for cytokine classes across all the cell states in the vCSF, based on CellChat analysis. Download figure Open in new tab Figure 7. Flow cytometry of CSF after IVH. Aggregate data from n=3 samples from n=2 patients is shown, for a total of 72,725 events. A. Gating strategy for identifying the singlet, live, CD45 + cell population (53,424 events) with CD3 and CD14 used to distinguish neutrophils (orange), monocytes (blue), and lymphocytes (green). Back gating onto the SSC-A vs. CD45 plot shows separation of these populations based on the characteristic distribution of these cells types by side scatter. B. Neutrophil population is further gated by CD10 and CD62L. Each quadrant is subsequently gated by either CD16b, CXCR1, or CXCR2. C. Monocyte population is further gated by CD62L and CD163. Each quadrant is subsequently gated by CXCR1. We further delineated intercellular communication networks using CellChat, which uses the expression patterns of receptor-ligand pairs to identify cell types with the potential for cell-cell communication. Doing so, we found that the expression of ELR-motif CXC chemokines linked ACMono-CXCL5 and ACMono-CXCL8 to chemokine receptor CXCR1- and CXCR2-expressing neutrophils. This pathway is particularly important for neutrophil recruitment and activation, suggesting a role for macrophages in facilitating this function. In addition, ACMono-CXCL8 contribute CC chemokines that tie them to IFN-activated neutrophils specifically. CellChat also revealed interactions between subtypes of neutrophils, mediated by ADGRE and SELPLG/SELL. Other interactions are shown in Fig. 6C and 6D . Integrating observations from Immune Dictionary and CellChat suggested complex behaviors not evident from either data set alone. For example, from Immune Dictionary we observed that ACMono-CXCL5 shared common gene expression features with macrophages stimulated by the IL-1 family cytokine IL-33. CellChat indicated that ACMono-CXCL5 also contribute IL-1, suggesting an IL-1 signaling positive feedback loop in this macrophage subtype. Development of a flow cytometry panel for neutrophil and monocyte diversity in the ventricular CSF Our snRNA-seq data suggest that specific immune cell subtypes are enriched in the CSF of IVH patients compared to their prevalence in other contexts. Monitoring the abundance and dynamics of these immune cell subtypes could potentially reveal biomarkers that predict clinical outcomes. However, snRNA-seq is not a practical method to track these populations in clinical settings. Leveraging our snRNA-seq data, we sought to develop a panel of cell surface markers to distinguish neutrophil and monocyte subtypes by flow cytometry. We therefore developed a flow cytometry panel to delineate functional states of neutrophils and monocytes, the two most common cell types observed in our snRNA-seq data. To leverage the insights of our snRNA-seq data, we derived this panel empirically from our data set, as opposed to relying on existing cytometry-based classifications. After identifying cluster of differentiation (CD) genes enriched in specific clusters, we used FluoroFinder (Broomfield, CO) to select a subset of markers for which reliable antibodies with compatible fluorophores existed. Using this panel, we performed flow cytometry analysis of CSF samples on three samples from two participants with IVH not included in the snRNA-seq analysis. Both participants had experienced thalamic ICH with IVH, one as a result of hypertension and other as a result of a ruptured arteriovenous malformation. As expected, CD45+ immune cells could be distinguished as granulocyte, monocyte, and lymphocyte populations using CD3 and CD14 gating, which were further confirmed by forward and side scatter profiles ( Figure 6A ). Based on neutrophil clustering from our snRNA-seq data, we hypothesized that three canonical neutrophil surface markers (CD10, CD62L, and CD16b) could help resolve distinct neutrophil subsets 56 . The dominant population (>90% of neutrophils) was CD10 + CD62L + CD16b high , consistent with mature circulating neutrophils. CD10 presented a particular interpretive challenge: while transcriptional data, including our own and that from Wigerblad et al. 16 , suggest CD10 as a marker of nascent neutrophils, it is better established as a surface protein associated with mature neutrophils. Consistent with this, we observed a small population of CD10 - CD62L + CD16b high neutrophils (<2%) consistent with a nascent phenotype, but no CD10 + CD62L + CD16b high , as otherwise might have been predicted directly from our transcriptomic data ( Figure 6B ) . From our transcriptomic data, we would infer that quiescent neutrophils are CD10 - CD62L - CD16b low . However, we again anticipated that the protein-level expression of CD10 would differ for these cells, and that quiescent cells should be CD10 + CD62L - CD16b low . Though both populations are small, the CD10 + CD62L - CD16b low population is much more common in our flow cytometry data than is the CD10 - CD62L - CD16b low population ( Figure 6B ). A distinct subset of a CD10 - CD62L - population is CD10 - CD62L - CD16b high . Cross-referencing with our snRNA-seq data, this population is most consistent with interferon-activated neutrophils, though the markers shown here do not directly test for that state. Additionally, we identified by flow cytometry a CD10 + CD62L - CD16b + subset not resolved in our snRNA-seq clustering analysis but reported in the literature as T cell–suppressive neutrophils 57 . Similarly, we hypothesized that monocyte subtypes could be distinguished using the markers CD163, CD62L, CXCR1, CXCR2, CD50, and CD87. We found that of these, CD163, CD62L, and CXCR1 were the only markers with discriminating power in our flow cytometry data set ( Figure 6C ). We note that we detected nascent, quiescent, and interferon-activated populations — which together dominate the snRNA-seq neutrophil profile -- at very low levels in our flow cytometry data. It is unclear whether this arises from greater sensitivity of snRNA-seq; misclassification by snRNA-seq due to mismatch between gene and protein expression; limitations of the flow cytometry panel; or patient-specific differences in neutrophil subtype. Taken together, however, our flow cytometry results indicate that it may be possible to translate leukocyte functional states inferred from single-nucleus transcriptomics into clinically-feasible flow cytometry panels. Discussion In this study, we generated a single-nucleus transcriptomic atlas of the ventricular CSF in IVH. We leveraged nuclear extraction as a strategy to enrich for cell types — specifically neutrophils — that are often challenging to capture in single-cell experiments because of their fragility. Our results revealed a diversity of cell types, including distinct subpopulations of neutrophils, monocytes, and lymphocytes. Notably, we captured a large and diverse population of neutrophils, including a population of interferon-activated neutrophils that has not been described in the central nervous system previously. As single-nucleus RNA sequencing is not feasible for clinical diagnostic purposes, we validated potential flow cytometry markers for cellular sub-clusters we identified. This demonstrates the feasibility of translating biomarkers derived from snRNA-seq data into clinically practical flow cytometry markers. A strength of our study that distinguishes it from previous studies is the detection of a large population of neutrophils in the CSF. Neutrophils are expected to be among the most common leukocytes present in the CSF during the first several days after hemorrhage 58 . Neutrophil depletion has been shown to be protective in a range of hemorrhagic brain disease models 59 – 61 . Yet previous sequencing studies in these diseases have selected against neutrophils in sample processing. Sansing and colleagues reported single-cell RNA sequencing data from the intraparenchymal hematoma of a patient who underwent clot evacuation after an intracerebral hemorrhage as part of the MISTIE III phase III clinical trial 62 and have recently extended this work to a larger group of subjects undergoing clot evacuation 63 . Their study provides a rich time-dependent atlas of leukocytes in the intracerebral hematoma, but only a small population of neutrophils was detected because they subjected samples to a specific granulocyte depletion. Similarly, the two reported single-cell studies of CSF in subarachnoid hemorrhage 64 , 65 specifically selected against neutrophils, one profiling PBMCs and the other profiling CD8 + CD161 + cells. Until recently, neutrophils were thought to display a relatively limited functional repertoire, owing to their relatively low transcriptional activity and short lifespans. Single-cell RNA sequencing has revealed that, in fact, neutrophils are highly diverse cell populations 16 , 66 , 67 , 68 . This expanded understanding of the diversity of neutrophils prompted us to investigate this cell population more comprehensively in IVH. Neutrophils may be particularly challenging to capture in single-cell studies. They are often short-lived and highly sensitive, rendering them difficult to capture with methods that have long or complex workflows. This raises the question of why we were able to identify a large population of neutrophils when others have not. We analyzed CSF instead of parenchymal hematoma, and these may have different cell populations. However, numerous neutrophils can be isolated from parenchymal hematoma by flow cytometry 69 . It is likely that differences in our sample preparation account for the preservation of neutrophils. The use of single-nucleus instead of single-cell RNA sequencing may improve the identification of neutrophils, as all cells undergo lysis, reducing the bias against fragile neutrophils. While we cannot fully exclude the possibility that the nuclei we identified as deriving from neutrophils actually arise from other cell types, the expression of canonical neutrophil markers in these cells supports their assignment. A strength of single-nucleus RNA sequencing over traditional cytometric approaches is the ability to distinguish subtle cellular phenotypes and activation states. The advent of these methods has revealed that neutrophils, previously thought to be a relatively homogenous cell type, in actuality exhibit a variety of states 16 . In this case, we identified three major neutrophil functional states, though we anticipate that with larger sample sizes and improved isolation techniques, it is likely we would be able to classify these states even more specifically. We have shown that these are consistent with classifications of neutrophils derived from peripheral blood in independent data sets 16 , 19 . One particularly interesting neutrophil state is the IFN-activated neutrophil. To our knowledge, this subtype of neutrophils has not been reported in the human central nervous system previously, though it has been detected in rodents 70 . Interferons have been known to stimulate neutrophils for nearly half a century 71 . It was recently reported that interferon-activated neutrophils play an important role in tumor control 17 , 72 and the response to SARS-CoV-2 infection 18 . Knockout of the Type I interferon receptor results in increased neutrophil accumulation in infected tissue 73 , 74 , while interferon-gamma decreases neutrophil accumulation in such tissue 75 . Our analysis based on Immune Dictionary suggests that Type I and II responses may be particularly relevant to IFN-activated neutrophils in IVH. A study of soluble factors in the CSF after IVH identified time-dependent increases in the Type II interferon IFN-γ that were augmented in patients treated with intraventricular tissue plasminogen activator; no subjects in our study underwent such treatment 76 . Similarly, IFN-γ levels in the CSF were associated with white matter damage in premature infants with posthemorrhagic hydrocephalus 77 . Little is known about the role of Type I interferons in human IVH, though a nonsense single-nucleotide polymorphism in the gene encoding IFN-ε is associated with the development of ICH 78 . In a rat germinal matrix hemorrhage model, treatment with Rh-IFN-α improved neurological function 79 . Our findings suggest that the Type I interferon response in general, and interferon-activated neutrophils in specific, merit further investigation in the pathophysiology of intraventricular hemorrhage. This represents a potentially targetable pathway for modulating neutrophils for future therapies. Our data also reflect a diversity of monocytes/macrophages that occupy a range of functional states. These states corresponded to strong differential expression of genes encoding certain secreted proteins, which may give some indication to their role in IVH and other acute brain injuries. The distinction we observed between macrophages expressing lipid metabolism genes and monocytes expressing cytokine-stimulated genes was previously noted for tumor-associated cells 32 . We further classified monocytes into four activated states defined by their expression of genes encoding specific secretory proteins. VCAN ACMono and PROK2 ACMono exhibited transcriptomic profiles that could be described broadly as pro-inflammatory and interferon-responsive, gene co-expression network analysis revealed markedly different profiles for these cellular states. Increased deposition of versican (encoded by VCAN ), a protein that is involved in myelination, has been noted in the brains of rabbit pups subjected to IVH 80 . Overexpression of PROK2 has been shown to reduce neurodegeneration and functional deficits in a mouse model of TBI 81 . Both the VCAN ACMono and PROK2 ACMono express interferon-response genes, again supporting an outsize role of interferon signaling in the immune response to IVH. Further insight can be gleaned receptor-ligand interactions, such as CXCL8 and CXCR2, which we have characterized more extensively using CellChat. These results indicate cross-talk within subtypes of individual cell types (e.g., PROK2 with CXCL8 ACMono) or between cell types (e.g., PROK2 ACMono with IFN-activated neutrophils). Major roles for the ELR-motif CXC chemokines, CC chemokines, IL-1 family cytokines, and interferons were observed. We also observed monocyte/macrophage phenotypes that have been associated with disease. For example, both snRNA-seq and flow cytometry revealed a substantial number of CD163 + monocytes/macrophages, a population that has been associated with disease severity in aneurysmal subarachnoid hemorrhage 82 . The distribution of lymphocytes we observed in the CSF is intriguing and merits further investigation. The majority of cells were CD4 + T cells. It has been reported that the majority of T cells in the CSF under normal conditions exhibit a central memory phenotype 83 . This was one of the major subtypes we observed in IVH, though not the most common. These cells have been purported to enter the CSF via the choroid plexus 84 . Limitations of our study include a small sample size and insufficient material for genotyping, both of which reduce our ability to delineate differences inflammatory profiles that are associated with disease severity. These samples were convenience samples that were obtained when specific clinical events (EVD placement, EVD removal, or diagnostic testing) prompted CSF sampling, increasing the heterogeneity of the samples and potentially creating sampling bias. In addition, low cell counts from individual subjects further limit the generalizability of our results. Nevertheless, our study is one of very few single-cell or single-nucleus transcriptomic studies of cerebrospinal fluid for hemorrhagic brain disease, and substantially increases the data available in the literature. While we hypothesize that the single-nucleus approach has allowed us to capture difficult-to-detect neutrophil states, this approach does have drawbacks. Nuclei contain substantially fewer transcripts than do whole cells, potentially diminishing our ability to identify transcriptionally regulated pathways relevant to disease biology. Nevertheless, even small nuclear transcriptomes have proven reliable for cell type classification in our study. While single-cell transcriptomic methods are unlikely to be translatable directly to clinical diagnostics, they may generate candidates for the development of flow cytometry panels. We provide a proof-of-concept for this here, but with too few subjects to demonstrate the utility of these markers for diagnostic or prognostic purposes. It would be valuable to extend our approaches to elucidate time-dependent changes in blood and CSF leukocytes across a larger cohort of patients in ICH, SAH, IVH, and TBI. Previous studies report divergence in leukocytes sampled from the blood and central nervous system over time 15 , 62 . The distinct immune profile in the CSF and brain after IVH may yield insights into the risk for common complications, such as persistent hydrocephalus. Such information could help to inform the timing of surgical interventions and reveal strategies to guide treatment of post-hemorrhagic hydrocephalus. Taken together, our results provide the most comprehensive cellular atlas of the CSF in IVH to date. This atlas reveals a rich diversity of cellular populations, including myeloid states not previously described in the human central nervous system, including interferon-responsive neutrophils. We have further demonstrated that transcriptional markers from these populations can in many cases be validated at a protein level with flow cytometry. We anticipate that these results will set the stage for larger studies that elucidate the relevance of specific leukocyte subtypes to disease severity, progression, and outcome. Eventually, this will yield new diagnostic tools and therapeutic targets for intraventricular hemorrhage, a common, catastrophic disease with few effective treatments. Materials and Methods Sex as a biological variable As we collected convenience samples and batched nuclei for sequencing (see below), we were unable to interrogate sex as a biological variable specifically; N=6 samples were from male subjects and N=1 sample was from a female subject. For the period of enrollment, a recruitment flow chart is shown in Figure S9 . Statistics Statistical methods used in this study are described in the sections pertaining to each specific assay or analytical tool. Study approval Subjects (N=7; ICH=6, SAH=1) are a subset of those included in a prospective cohort study of patients aged 18-88 with potentially life-threatening neurologic illness ( NCT04189471 ). The University of Maryland, Baltimore, institutional review board (IRB) approved this study, entitled “Improving Outcomes for Patients with Life-Threatening Neurologic Illnesses,” approval: number HP-00056063, date: August 1, 2013. Written informed consent was obtained from subjects (if they had decision-making capacity) or their legally authorized representatives. Study procedures were performed in accordance with the ethical standards of the responsible committee on human experimentation and with the Helsinki Declaration of 1975. Detailed clinical characteristics of study participants were recorded in a secure database, including Cytospin cytology and computed tomography (CT) scans obtained from the participants as part of the standard of care. Sample preparation and single-nucleus RNA sequencing Of the CSF aspirated from the proximal port of the EVD during routine clinical practice, excess sample was obtained and transported to the laboratory for immediate processing. Obtained CSF was collected on EDTA (purple top vacutainers), with a processing time <1 hour from the time of collection. RNAse and Proteinase/phosphatase were inhibited by adding 1uM DEPC (Sigma D5758) and 1XHALT (Invitrogen 1861281). The cellular component of CSF was obtained by spinning the sample at 350g for 10 min. The pellet was resuspended, and red blood cells were depleted using EasySep magnetic beads (StemCell Technologies cat18170), and Miltenyi microcolumns (Miltenyi cat130-042-701). Samples were then washed and stored at - 80C. Half of the samples were stored as pellets and half in cryopreservation media (Cryostor, StemCell Technologies cat 100-1061). All the samples were stored at -80°C for <3 months. For pelleted samples, nuclei extraction was performed directly from the pellets. For samples stored in cryopreservation media, nuclei isolation was performed on the cellular fraction after washing with PBS+1% BSA at 350g/5 minutes at 4C. Nuclei isolation was performed using a detergent-free nuclei isolation kit, Invent (cat IN-024), according to the manufacturer’s instructions. Briefly, nuclei were resuspended in buffer A, then vortexed for 20s at level 9 on Genie2 Vortex, then placed on the spin column at 16000g/30s. The pass-through was washed with buffer B and counted on a MoxiGoII (Orflo) using 1uM Propidium Iodide (Sigma P-4864). We loaded two pools, one pool containing samples stored as pellets and other with cellular fraction cryoprotected. A total of 30K isolated nuclei per pool in PBS+1% BSA were loaded in a 10X Chromium X. Sequencing libraries were prepared by the Maryland Genomics core facility at the Institute for Genome Sciences using Chromium Single Cell 3’ Kit v3.1 reagents, according to manufacturer’s instructions. Libraries were sequenced on a NovaSeq6000 sequencer (Illumina) to an average depth of at least 25,000 raw reads per cell. Flow cytometry CSF samples were acquired from the EVD buretrol chamber. Samples were stabilized in EDTA tubes during transport from the intensive care unit to the laboratory and kept at room temperature. Samples were then spun at 300g for 5 min, at 4°C. Pellets were then resuspended in 100 µL of staining buffer (phosphate buffered saline with 1% bovine serum albumin and 1% EDTA - A9418 and 324506, Sigma Aldrich), and incubated for ten minutes on ice with Fc Block (ThermoFisher). Pellets were resuspended in 50 µL of staining buffer mixed with antibodies and incubated for an additional 15 minutes in dark. A 13-antibody panel was used for this experiment. Antibodies included mouse anti-human anti-CD45 (363-0459-41, ThermoFisher), mouse anti-human anti-CD3 (344825, BioLegend), mouse anti-human anti-CD14 (367109, BioLegend), human anti-human anti-CD16b (130-123-643, Miltenyi Biotec), mouse anti-human anti-CD163 (333611, BioLegend), human anti-human anti-CD50 (130-115-309, Miltenyi Biotec), mouse anti-human anti-CD183 (353721, BioLegend), goat anti-human anti-CD121a (FAB269F-025, R&D Systems), mouse anti-human anti-CD62L (567989, BD Biosciences), mouse anti-human anti-CD10 (312241, BioLegend), human anti-human anti-CD181 (130-115-950, Miltenyi Biotec), human anti-human anti-CD182 (130-131-233, Miltenyi Biotec), mouse anti-human anti-CD87 (749063, BD Biosciences). 100 µL of staining buffer were added, cells were mixed and then spun at 300g for five minutes at 4° C. The supernatant was removed, cells resuspended in 200 µL of staining buffer, and 2 µL of 7-amino-actinomycin D (7-AAD) (420404, BioLegend) were added before flow acquisition. Flow acquisition was performed on a Cytek Aurora 4 Spectral Cytometer (Cytek Biosciences, Freemont, CA, USA) (lasers: 355nm, 405nm, 488nm, 640nm) at the University of Maryland Greenebaum Comprehensive Cancer Center Flow Cytometry Shared Service. At least 5,000 events were acquired and analyzed with FCS Express V7 software (De Novo Software, Pasadena, CA, USA). snRNA-seq cell clustering and annotation Initial data processing was performed with CellRanger (7.2.0), yielding a cell x gene matrix for each pooled sample. Seurat 51 v4.0.5 was used for batch integration and clustering, using the following parameters. We selected high-quality cells with at least 300 unique molecules and a maximum of 20 percent mitochondrial counts. SCT integration was performed with 5000 variable genes and 20 anchors. Louvain clustering was performed using 20 principal components. An optimal clustering resolution was determined by using the mrtree clustering optimization method 85 , which was given resolutions from 0 to 5, with 3.5 (35 clusters) being optimal for the entire population. Cell clusters were annotated to established cell types based on canonical markers from immune cells in other contexts 86 – 88 . Subsequently, the most abundant major cell types — neutrophils and monocytes — were sub-clustered, following the same procedures starting from the SCT integration step. Cellular populations derived from sub-clustering were compared to neutrophil and monocyte data sets that have been described by scRNA-seq in other contexts 16 , 19 , 32 . Cells in our snRNA-seq dataset were scored using the combined expression of marker genes from comparison datasets with the AddModuleScore() function in Seurat. Genes from reference datasets were selected with a log 2 FC > 0.585 (greater than 1.5 fold change) and adjusted p-value < 0.05. We also considered cell annotations derived from label transfer using the Seurat FindTransferAnchors and TransferData functions with 30 principal components and otherwise default parameters. Label transfer was performed separately for monocytes and lymphocytes using level 2 annotations from scRNA-seq of human peripheral blood mononuclear cells 51 . Gene co-expression networks in neutrophils and monocytes Gene co-expression modules were identified through k-means clustering of imputed read counts, as previously described 89 . First, we performed a zero-preserving imputation with ALRA 90 to reduce dropouts and improve correlation structure. Next, we performed k-means clustering (parameters). Modules with tightly correlated centroids were merged, and genes with insufficient correlation to the centroids of the merged modules were dropped. These steps were performed using the cells from Pool 2, then projected onto the cells from Pool 1. This strategy reduces the potential for batch effects and provided an independent test set to evaluate the robustness of the modules. Gene Ontology annotations for the modules were obtained using the enrichR package in R from the GO Molecular Function 2023, GO Cellular Component 2023 and GO Biological Process 2023 databases. Cell-cell signaling networks Cytokine signaling activity among cell types was predicted by two methods: We used the CellChat 54 (v 2.2.0) package in R to predict the incoming and outgoing signals and the associated pathways for each cell type. Cell populations with fewer than 50 cells were excluded due to insufficient data. Incoming) cytokine responses were further evaluated by scoring each cell (for the activity of signatures defined by the set of genes up- or down-regulated (FDR < 0.05) in immune cell populations after exposure to each of 86 cytokines 53 . As the Immune Dictionary was generated in mice, our analysis used human genes with one-to-one orthology. Scores were calculated with Seurat AddModuleScore() for each cytokine / cell type condition in which at least ten differentially expressed genes were detected. Data availability We have made our expression data available on the Neuroscience Multi-Omic Analytics (NeMO) platform 91 ( https://nemoanalytics.org/p?s=c4d7a449 ) to facilitate secondary data visualization and analysis. Sequencing data will be available on dbGaP (pending). Code generated for this study is available at https://github.com/Ciryam-Lab/vCSF-Hemorrhage.git . Figure Legends Download figure Open in new tab Supplementary Figure S1. UMAPs of ventricular CSF colored by quality metrics. UMAPs of snRNA-seq of ventricular CSF (11,191 nuclei from seven study participants) colored by A. Preservation method. B. RNA Counts. C. Percentage Mitochondrial Counts. Download figure Open in new tab Supplementary Figure S2. UMAPs of the neutrophil subset colored by quality metrics. UMAPs of snRNA-seq of the neutrophil subset colored by A. Preservation method. B. RNA Counts. C. Percentage Mitochondrial Counts. D. Dot plot showing marker genes for neutrophil subtypes. Download figure Open in new tab Supplementary Figure S3. Annotation of neutrophil subtypes. Violin plots of neutrophil subtypes scored by marker genes from cell states found in blood during sepsis response 19 . Download figure Open in new tab Supplementary Figure S4. Gene co-expression Modules. 12 neutrophil gene co-expression modules differentially expressed across neutrophil subtypes. Download figure Open in new tab Supplementary Figure S5. UMAPs of the monocyte subset colored by quality metrics. UMAPs of snRNA-seq of the monocyte subset colored by A. Preservation method. B. RNA Counts. C. Percentage Mitochondrial Counts. D. Dot plot showing marker genes for monocyte subtypes. Download figure Open in new tab Supplementary Figure S6. Annotation of monocyte subtypes. Violin plots of monocyte subtypes scored by marker genes from all the cell states found in the MoMac dataset 32 . Download figure Open in new tab Supplementary Figure S7. Gene co-expression Modules. 11 monocyte modules differentially expressed across monocyte subtypes. Download figure Open in new tab Supplementary Figure S8 Cytokine Response Scores. Heatmap showing of CSF A. neutrophils, B., monocytes, C. macrophages, and D. dendritic cells showing mean log fold changes of scores based on differential gene expression in response to specific cytokines derived from Immune Dictionary. All interleukin and interferon cytokines eliciting a response in that cell type in Immune Dictionary are plotted. Supplementary Table S1. Marker genes for all cell types. Supplementary Table S2. Marker genes for neutrophil subtypes. Supplementary Table S3. Neutrophil subtype marker genes from prior studies and enrichment scores. Supplementary Table S4. Neutrophil gene co-expression module membership and functional enrichment. Supplementary Table S5. Marker genes for monocyte subtypes. Supplementary Table S6. Monocyte subtype marker genes from prior studies and enrichment scores. Supplementary Table S7. Monocyte gene co-expression module membership and functional enrichment. Supplementary Table S8. Cytokine activity scores using Immune Dictionary signatures and CellChat. Supplementary Table S9. Flow chart of subject recruitment. Download figure Open in new tab Acknowledgments P.C. received funding from the Henry M. Jackson Foundation for the Advancement of Military Medicine, Neurocritical Care Society, and Passano Foundation. P.C. and S.A.A. received funding from the Brain Aneurysm Foundation. S.M. was supported by NIH R01 MH052716. The authors thank Flow Cytometry Shared Service of the University of Maryland Marlene and Stewart Greenebaum Comprehensive Cancer Center. This publication was supported by funds through the Maryland Department of Health’s Cigarette Restitution Fund Program – CH-649-CRF and the National Cancer Institute - Cancer Center Support Grant (CCSG) - P30CA134274. Graphical abstract and Figure 1 schematic created in BioRender ( https://BioRender.com/q4n1xt9 ). Footnotes Significant revisions to text and figures. References 1. ↵ Gaberel , T. , Magheru , C. & Emery , E . Management of non-traumatic intraventricular hemorrhage . Neurosurg Rev 35 , 485 – 494 ; discussion 494-495 ( 2012 ). OpenUrl CrossRef PubMed 2. ↵ Ziai , W. C. et al. Occurrence and impact of intracranial pressure elevation during treatment of severe intraventricular hemorrhage . Crit Care Med 40 , 1601 – 1608 ( 2012 ). OpenUrl CrossRef PubMed 3. ↵ Bhattathiri , P. S. , Gregson , B. , Prasad , K. S. M. , Mendelow , A. D. , & STICH Investigators . Intraventricular hemorrhage and hydrocephalus after spontaneous intracerebral hemorrhage: results from the STICH trial . Acta Neurochir Suppl 96 , 65 – 68 ( 2006 ). OpenUrl CrossRef PubMed Web of Science 4. ↵ Shah , V. A. et al. One-Year Outcome Trajectories and Factors Associated with Functional Recovery Among Survivors of Intracerebral and Intraventricular Hemorrhage With Initial Severe Disability . JAMA Neurol 79 , 856 – 868 ( 2022 ). OpenUrl PubMed 5. ↵ Lovasik , B. P. et al. The Efect of External Ventricular Drain Use in Intracerebral Hemorrhage . World Neurosurg 94 , 309 – 318 ( 2016 ). OpenUrl CrossRef PubMed 6. ↵ Herrick , D. B. et al. Determinants of external ventricular drain placement and associated outcomes in patients with spontaneous intraventricular hemorrhage . Neurocrit Care 21 , 426 – 434 ( 2014 ). OpenUrl CrossRef PubMed 7. ↵ Hanley , D. F. et al. Thrombolytic removal of intraventricular haemorrhage in treatment of severe stroke: results of the randomised, multicentre, multiregion, placebo-controlled CLEAR III trial . Lancet 389 , 603 – 611 ( 2017 ). OpenUrl CrossRef PubMed 8. ↵ Akinduro , O. O. et al. Predictors of Ventriculoperitoneal shunting following Subarachnoid Hemorrhage treated with External Ventricular Drainage . Neurocrit Care 32 , 755 – 764 ( 2020 ). OpenUrl CrossRef PubMed 9. ↵ Sadegh , C. et al. Choroid plexus-targeted NKCC1 overexpression to treat post-hemorrhagic hydrocephalus . Neuron 111 , 1591 – 1608.e4 ( 2023 ). OpenUrl CrossRef PubMed 10. ↵ Karimy , J. K. et al. Inflammation-dependent cerebrospinal fluid hypersecretion by the choroid plexus epithelium in posthemorrhagic hydrocephalus . Nat Med 23 , 997 – 1003 ( 2017 ). OpenUrl CrossRef PubMed 11. ↵ Garton , T. et al. Intraventricular Hemorrhage: the Role of Blood Components in Secondary Injury and Hydrocephalus . Transl Stroke Res 7 , 447 – 451 ( 2016 ). OpenUrl CrossRef PubMed 12. Wang , Y. et al. Irisin ameliorates neuroinflammation and neuronal apoptosis through integrin αVβ5/AMPK signaling pathway after intracerebral hemorrhage in mice . J Neuroinflammation 19 , 82 ( 2022 ). OpenUrl CrossRef PubMed 13. Shen , J. et al. Neurogenesis after primary intracerebral hemorrhage in adult human brain . J Cereb Blood Flow Metab 28 , 1460 – 1468 ( 2008 ). OpenUrl CrossRef PubMed 14. ↵ Xiong , X.-Y. & Yang , Q.-W . Rethinking the roles of inflammation in the intracerebral hemorrhage . Transl Stroke Res 6 , 339 – 341 ( 2015 ). OpenUrl CrossRef PubMed 15. ↵ Moraes , L. et al. Immune cells subpopulations in cerebrospinal fluid and peripheral blood of patients with Aneurysmal Subarachnoid Hemorrhage . Springerplus 4 , 195 ( 2015 ). OpenUrl CrossRef PubMed 16. ↵ Wigerblad , G. et al. Single-Cell Analysis Reveals the Range of Transcriptional States of Circulating Human Neutrophils . J Immunol 209 , 772 – 782 ( 2022 ). OpenUrl Abstract / FREE Full Text 17. ↵ Gungabeesoon , J. et al. A neutrophil response linked to tumor control in immunotherapy . Cell 186 , 1448 – 1464.e20 ( 2023 ). OpenUrl CrossRef PubMed 18. ↵ Sinha , S. et al. Dexamethasone modulates immature neutrophils and interferon programming in severe COVID-19 . Nat Med 28 , 201 – 211 ( 2022 ). OpenUrl CrossRef PubMed 19. ↵ Kwok , A. J. et al. Neutrophils and emergency granulopoiesis drive immune suppression and an extreme response endotype during sepsis . Nat Immunol 24 , 767 – 779 ( 2023 ). OpenUrl CrossRef PubMed 20. ↵ Zhang , X. , Tang , X. , Liu , K. , Hamblin , M. H. & Yin , K.-J . Long Noncoding RNA Malat1 Regulates Cerebrovascular Pathologies in Ischemic Stroke . J Neurosci 37 , 1797 – 1806 ( 2017 ). OpenUrl Abstract / FREE Full Text 21. Pan , Y. et al. Novel Insights into the Emerging Role of Neat1 and Its Efects Downstream in the Regulation of Inflammation . J Inflamm Res 15 , 557 – 571 ( 2022 ). OpenUrl CrossRef PubMed 22. Agarwal , S. et al. The long non-coding RNA LUCAT1 is a negative feedback regulator of interferon responses in humans . Nat Commun 11 , 6348 ( 2020 ). OpenUrl PubMed 23. ↵ Vierbuchen , T. et al. The lncRNA LUCAT1 is elevated in inflammatory disease and restrains inflammation by regulating the splicing and stability of NR4A2 . Proceedings of the National Academy of Sciences 120 , e2213715120 ( 2023 ). OpenUrl PubMed 24. ↵ Xu , X. et al. GPCR-mediated PLCβγ/PKCβ/PKD signaling pathway regulates the cofilin phosphatase slingshot 2 in neutrophil chemotaxis . Mol Biol Cell 26 , 874 – 886 ( 2015 ). OpenUrl Abstract / FREE Full Text 25. Kunisaki , Y. et al. DOCK2 is a Rac activator that regulates motility and polarity during neutrophil chemotaxis . J Cell Biol 174 , 647 – 652 ( 2006 ). OpenUrl Abstract / FREE Full Text 26. ↵ Strieter , R. M. et al. Cytokine-induced neutrophil-derived interleukin-8 . Am J Pathol 141 , 397 – 407 ( 1992 ). OpenUrl PubMed Web of Science 27. ↵ Abdollahpour , H. et al. The phenotype of human STK4 deficiency . Blood 119 , 3450 – 3457 ( 2012 ). OpenUrl Abstract / FREE Full Text 28. Hu , T. et al. Single-cell analysis defines highly specific leukemia-induced neutrophils and links MMP8 expression to recruitment of tumor associated neutrophils during FGFR1 driven leukemogenesis . Exp Hematol Oncol 13 , 49 ( 2024 ). OpenUrl PubMed 29. ↵ Hoogendijk , A. J. et al. Kinase activity is impaired in neutrophils of sepsis patients . Haematologica 104 , e233 – e235 ( 2019 ). OpenUrl FREE Full Text 30. ↵ Goss , K. et al. Single-cell RNA-sequencing of circulating eosinophils from asthma patients reveals an inflammatory signature . iScience 28 , 112609 ( 2025 ). OpenUrl PubMed 31. ↵ Guarda , G. et al. Type I interferon inhibits interleukin-1 production and inflammasome activation . Immunity 34 , 213 – 223 ( 2011 ). OpenUrl CrossRef PubMed Web of Science 32. ↵ Mulder , K. et al. Cross-tissue single-cell landscape of human monocytes and macrophages in health and disease . Immunity 54 , 1883 – 1900.e5 ( 2021 ). OpenUrl CrossRef PubMed 33. ↵ Masuda , A. et al. Versican is upregulated in circulating monocytes in patients with systemic sclerosis and amplifies a CCL2-mediated pathogenic loop . Arthritis Res Ther 15 , R74 ( 2013 ). OpenUrl CrossRef PubMed 34. ↵ Wight , T. N. , Kang , I. & Merrilees , M. J . Versican and the control of inflammation . Matrix Biol 35 , 152 – 161 ( 2014 ). OpenUrl CrossRef PubMed 35. ↵ Chang , M. Y. et al. Versican is produced by Trif- and type I interferon-dependent signaling in macrophages and contributes to fine control of innate immunity in lungs . Am J Physiol Lung Cell Mol Physiol 313 , L1069 – L1086 ( 2017 ). OpenUrl CrossRef PubMed 36. ↵ Wong , K. L. et al. Gene expression profiling reveals the defining features of the classical, intermediate, and nonclassical human monocyte subsets . Blood 118 , e16 – e31 ( 2011 ). OpenUrl Abstract / FREE Full Text 37. ↵ LeCouter , J. , Zlot , C. , Tejada , M. , Peale , F. & Ferrara , N . Bv8 and endocrine gland-derived vascular endothelial growth factor stimulate hematopoiesis and hematopoietic cell mobilization . Proc Natl Acad Sci U S A 101 , 16813 – 16818 ( 2004 ). OpenUrl Abstract / FREE Full Text 38. ↵ Martucci , C. et al. Bv8, the amphibian homologue of the mammalian prokineticins, induces a proinflammatory phenotype of mouse macrophages . Br J Pharmacol 147 , 225 – 234 ( 2006 ). OpenUrl CrossRef PubMed Web of Science 39. ↵ Shojaei , F. et al. Bv8 regulates myeloid-cell-dependent tumour angiogenesis . Nature 450 , 825 – 831 ( 2007 ). OpenUrl CrossRef PubMed Web of Science 40. ↵ Severini , C. et al. Bv8/prokineticin 2 is involved in Aβ-induced neurotoxicity . Sci Rep 5 , 15301 ( 2015 ). OpenUrl PubMed 41. ↵ Ng , K. L. et al. Dependence of olfactory bulb neurogenesis on prokineticin 2 signaling . Science 308 , 1923 – 1927 ( 2005 ). OpenUrl Abstract / FREE Full Text 42. ↵ Mundim , M. V. et al. A new function for Prokineticin 2: Recruitment of SVZ-derived neuroblasts to the injured cortex in a mouse model of traumatic brain injury . Mol Cell Neurosci 94 , 1 – 10 ( 2019 ). OpenUrl PubMed 43. ↵ Zlotnik , A. & Yoshie , O . The chemokine superfamily revisited . Immunity 36 , 705 – 716 ( 2012 ). OpenUrl CrossRef PubMed Web of Science 44. ↵ Liu , C. et al. Chronic Stress Stimulates Protumor Macrophage Polarization to Propel Lung Cancer Progression . Cancer Res 85 , 2429 – 2447 ( 2025 ). OpenUrl PubMed 45. ↵ Fan , Y. et al. The efect of anti-inflammatory properties of ferritin light chain on lipopolysaccharide-induced inflammatory response in murine macrophages . Biochim Biophys Acta 1843 , 2775 – 2783 ( 2014 ). OpenUrl CrossRef PubMed 46. ↵ Han , N.-R. , Moon , P.-D. , Kim , H.-M. & Jeong , H.-J . TSLP Exacerbates Septic Inflammation via Murine Double Minute 2 (MDM2) Signaling Pathway . J Clin Med 8 , 1350 ( 2019 ). OpenUrl PubMed 47. ↵ Marques , L. , Negre-Salvayre , A. , Costa , L. & Canonne-Hergaux , F . Iron gene expression profile in atherogenic Mox macrophages . Biochim Biophys Acta 1862 , 1137 – 1146 ( 2016 ). OpenUrl CrossRef 48. ↵ Xu , Y. et al. Identification of conserved and tissue-restricted transcriptional profiles for lipid associated macrophages . Commun Biol 8 , 953 ( 2025 ). OpenUrl PubMed 49. ↵ Beuker , C. et al. Stroke induces disease-specific myeloid cells in the brain parenchyma and pia . Nat Commun 13 , 945 ( 2022 ). OpenUrl CrossRef PubMed 50. ↵ Jiang , C. et al. Immune changes in peripheral blood and hematoma of patients with intracerebral hemorrhage . FASEB J 34 , 2774 – 2791 ( 2020 ). OpenUrl CrossRef PubMed 51. ↵ Hao , Y. et al. Integrated analysis of multimodal single-cell data . Cell 184 , 3573 – 3587.e29 ( 2021 ). OpenUrl CrossRef PubMed 52. ↵ Glatzel , A. et al. Patterns of chemokine receptor expression on peripheral blood gamma delta T lymphocytes: strong expression of CCR5 is a selective feature of V delta 2/V gamma 9 gamma delta T cells . J Immunol 168 , 4920 – 4929 ( 2002 ). OpenUrl Abstract / FREE Full Text 53. ↵ Cui , A. et al. Dictionary of immune responses to cytokines at single-cell resolution . Nature 625 , 377 – 384 ( 2024 ). OpenUrl CrossRef PubMed 54. ↵ Jin , S. et al. Inference and analysis of cell-cell communication using CellChat . Nat Commun 12 , 1088 ( 2021 ). OpenUrl CrossRef PubMed 55. ↵ Jin , S. , Plikus , M. V. & Nie , Q . CellChat for systematic analysis of cell–cell communication from single-cell transcriptomics . Nat Protoc 20 , 180 – 219 ( 2025 ). OpenUrl CrossRef PubMed 56. ↵ Grieshaber-Bouyer , R. & Nigrovic , P. A . Neutrophil Heterogeneity as Therapeutic Opportunity in Immune-Mediated Disease . Front Immunol 10 , 346 ( 2019 ). OpenUrl CrossRef PubMed 57. ↵ Kamp , V. M. et al. Human suppressive neutrophils CD16bright/CD62Ldim exhibit decreased adhesion . J Leukoc Biol 92 , 1011 – 1020 ( 2012 ). OpenUrl CrossRef PubMed Web of Science 58. ↵ Gusdon , A. M. et al. CSF and serum inflammatory response and association with outcomes in spontaneous intracerebral hemorrhage with intraventricular extension: an analysis of the CLEAR-III Trial . J Neuroinflammation 18 , 179 ( 2021 ). OpenUrl PubMed 59. ↵ Moxon-Emre , I. & Schlichter , L. C . Neutrophil depletion reduces blood-brain barrier breakdown, axon injury, and inflammation after intracerebral hemorrhage . J Neuropathol Exp Neurol 70 , 218 – 235 ( 2011 ). OpenUrl CrossRef PubMed 60. Provencio , J. J. et al. Neutrophil depletion after subarachnoid hemorrhage improves memory via NMDA receptors . Brain Behav Immun 54 , 233 – 242 ( 2016 ). OpenUrl CrossRef PubMed 61. ↵ Kenne , E. , Erlandsson , A. , Lindbom , L. , Hillered , L. & Clausen , F . Neutrophil depletion reduces edema formation and tissue loss following traumatic brain injury in mice . J Neuroinflammation 9 , 17 ( 2012 ). OpenUrl CrossRef PubMed 62. ↵ Goods , B. A. et al. Leukocyte dynamics after intracerebral hemorrhage in a living patient reveal rapid adaptations to tissue milieu . JCI Insight 6 , e145857 , 145857 ( 2021 ). OpenUrl 63. ↵ Kawamura , Y. et al. Single-Cell Temporal Atlas of Myeloid Cells in the Live Haemorrhagic Brain . 2024.12.24.630187 Preprint at doi: 10.1101/2024.12.24.630187 ( 2024 ). OpenUrl Abstract / FREE Full Text 64. ↵ Meng , N. et al. Single-cell transcriptomic landscape reveals the role of intermediate monocytes in aneurysmal subarachnoid hemorrhage . Front Cell Dev Biol 12 , 1401573 ( 2024 ). OpenUrl PubMed 65. ↵ Sanchez , S. et al. Modulation of the Immunological Milieu in Acute Aneurysmal Subarachnoid Hemorrhage: The Potential Role of Monocytes Through CXCL10 Secretion . Transl Stroke Res 16 , 88 – 95 ( 2025 ). OpenUrl CrossRef PubMed 66. ↵ Xie , X. et al. Single-cell transcriptome profiling reveals neutrophil heterogeneity in homeostasis and infection . Nat Immunol 21 , 1119 – 1133 ( 2020 ). OpenUrl CrossRef PubMed 67. ↵ Haruna , N.-F. et al. scRNA-seq profiling of human granulocytes reveals expansion of developmentally flexible neutrophil precursors with mixed neutrophil and eosinophil properties in asthma . J Leukoc Biol 116 , 1184 – 1197 ( 2024 ). OpenUrl CrossRef PubMed 68. ↵ Xiao , H. et al. High-throughput sequencing unravels the cell heterogeneity of cerebrospinal fluid in the bacterial meningitis of children . Front Immunol 13 , 872832 ( 2022 ). OpenUrl CrossRef PubMed 69. ↵ Askenase , M. H. , et al. Longitudinal transcriptomics define the stages of myeloid activation in the living human brain after intracerebral hemorrhage . Sci Immunol 6 , eabd6279 ( 2021 ). OpenUrl Abstract / FREE Full Text 70. ↵ Xu , H. et al. The choroid plexus synergizes with immune cells during neuroinflammation . Cell 187 , 4946 – 4963.e17 ( 2024 ). OpenUrl CrossRef PubMed 71. ↵ Pak , C. B. , Imanishi , J. , Kishida , T. & Matsuo , A . Increased nitroblue tetrazolium dye reduction by human neutrophils stimulated by interferon in vitro . Microbiol Immunol 24 , 717 – 723 ( 1980 ). OpenUrl CrossRef PubMed 72. ↵ Benguigui , M. et al. Interferon-stimulated neutrophils as a predictor of immunotherapy response . Cancer Cell 42 , 253 – 265.e12 ( 2024 ). OpenUrl CrossRef PubMed 73. ↵ Chen , X. et al. Expression of costimulatory TNFR2 induces resistance of CD4+FoxP3-conventional T cells to suppression by CD4+FoxP3+ regulatory T cells . J Immunol 185 , 174 – 182 ( 2010 ). OpenUrl Abstract / FREE Full Text 74. ↵ Seo , S.-U. et al. Type I Interferon Signaling Regulates Ly6Chi Monocytes and Neutrophils during Acute Viral Pneumonia in Mice . PLOS Pathogens 7 , e1001304 ( 2011 ). OpenUrl CrossRef PubMed 75. ↵ Nandi , B. & Behar , S. M . Regulation of neutrophils by interferon-γ limits lung inflammation during tuberculosis infection . J Exp Med 208 , 2251 – 2262 ( 2011 ). OpenUrl Abstract / FREE Full Text 76. ↵ Kramer , A. H. et al. Intraventricular Fibrinolysis with Tissue Plasminogen Activator is Associated with Transient Cerebrospinal Fluid Inflammation: A Randomized Controlled Trial . J Cereb Blood Flow Metab 35 , 1241 – 1248 ( 2015 ). OpenUrl CrossRef PubMed 77. ↵ Schmitz , T. et al. Interleukin-1β, Interleukin-18, and Interferon-γ Expression in the Cerebrospinal Fluid of Premature Infants with Posthemorrhagic Hydrocephalus—Markers of White Matter Damage? Pediatr Res 61 , 722 – 726 ( 2007 ). OpenUrl CrossRef PubMed Web of Science 78. ↵ Ekkert , A. , Šliachtenko , A. , Utkus , A. & Jatužis , D . Intracerebral Hemorrhage Genetics . Genes (Basel) 13 , 1250 ( 2022 ). OpenUrl PubMed 79. ↵ Li , P. et al. rh-IFN-α Attenuates Neuroinflammation and Improves Neurological Function by Inhibiting NF-κB through JAK1-STAT1/TRAF3 Pathway in an Experimental GMH Rat Model . Brain Behav Immun 79 , 174 – 185 ( 2019 ). OpenUrl PubMed 80. ↵ Vinukonda , G. et al. Intraventricular hemorrhage induces deposition of proteoglycans in premature rabbits, but their in vivo degradation with chondroitinase does not restore myelination, ventricle size and neurological recovery . Exp Neurol 247 , 630 – 644 ( 2013 ). OpenUrl PubMed 81. ↵ Bao , Z. et al. Prokineticin-2 prevents neuronal cell deaths in a model of traumatic brain injury . Nat Commun 12 , 4220 ( 2021 ). OpenUrl CrossRef PubMed 82. ↵ Thomas , A. J. , Ogilvy , C. S. , Griessenauer , C. J. & Hanafy , K. A . Macrophage CD163 expression in cerebrospinal fluid: association with subarachnoid hemorrhage outcome . J Neurosurg 131 , 47 – 53 ( 2019 ). OpenUrl CrossRef PubMed 83. ↵ de Graaf , M. T. et al. Central memory CD4+ T cells dominate the normal cerebrospinal fluid . Cytometry B Clin Cytom 80 , 43 – 50 ( 2011 ). OpenUrl CrossRef PubMed 84. ↵ Kivisäkk , P. et al. Human cerebrospinal fluid central memory CD4+ T cells: evidence for traficking through choroid plexus and meninges via P-selectin . Proc Natl Acad Sci U S A 100 , 8389 – 8394 ( 2003 ). OpenUrl Abstract / FREE Full Text 85. ↵ Peng , M. et al. Cell type hierarchy reconstruction via reconciliation of multi-resolution cluster tree . Nucleic Acids Res 49 , e91 ( 2021 ). OpenUrl CrossRef PubMed 86. ↵ Uhlen , M. et al. A genome-wide transcriptomic analysis of protein-coding genes in human blood cells . Science 366 , eaax9198 ( 2019 ). OpenUrl Abstract / FREE Full Text 87. Karlsson , M. et al. A single-cell type transcriptomics map of human tissues . Sci Adv 7 , eabh2169 ( 2021 ). OpenUrl FREE Full Text 88. ↵ The Human Protein Atlas . https://www.proteinatlas.org/ . 89. ↵ Wildermuth , E. et al. A single-cell genomic atlas for the efects of chronic ethanol exposure in the mouse dorsal striatum . Mol Psychiatry https://doi.org/10.1038/s41380-025-03014-z ( 2025 ) doi: 10.1038/s41380-025-03014-z . OpenUrl CrossRef 90. ↵ Linderman , G. C. et al. Zero-preserving imputation of single-cell RNA-seq data . Nat Commun 13 , 192 ( 2022 ). OpenUrl CrossRef PubMed 91. ↵ Orvis , J. et al. gEAR: Gene Expression Analysis Resource portal for community-driven, multi-omic data exploration . Nat Methods 18 , 843 – 844 ( 2021 ). OpenUrl CrossRef PubMed View the discussion thread. Back to top Previous Next Posted February 10, 2026. Download PDF Supplementary Material Data/Code Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following The cellular diversity of human cerebrospinal fluid following intraventricular hemorrhage revealed by single-nucleus RNA sequencing Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share The cellular diversity of human cerebrospinal fluid following intraventricular hemorrhage revealed by single-nucleus RNA sequencing Sonia Malaiya , Riccardo Serra , Marcia Cortes-Gutierrez , Bradley E. Wilhelmy , Emily Jusuf , Mayur Somalinga , Dave Peprah , Hirschel Nambiar , Kevin T. Kim , Jordan R. Saadon , Parantap D. Patel , Steven K. Yarmoska , Maureen Rakovec , John Kim , Catherine Lei , Shreyas Panchagnula , Reina Ambrocio , Meagan Cherney , Nicholas A. Morris , Natarajan Ayithan , Xiaoxuan Fan , Volodymyr Gerzanich , J. Marc Simard , Neeraj Badjatia , Gary Schwartzbauer , Gunjan Y. Parikh , Seth Ament , Prajwal Ciryam bioRxiv 2025.06.05.657834; doi: https://doi.org/10.1101/2025.06.05.657834 Share This Article: Copy Citation Tools The cellular diversity of human cerebrospinal fluid following intraventricular hemorrhage revealed by single-nucleus RNA sequencing Sonia Malaiya , Riccardo Serra , Marcia Cortes-Gutierrez , Bradley E. Wilhelmy , Emily Jusuf , Mayur Somalinga , Dave Peprah , Hirschel Nambiar , Kevin T. Kim , Jordan R. Saadon , Parantap D. Patel , Steven K. Yarmoska , Maureen Rakovec , John Kim , Catherine Lei , Shreyas Panchagnula , Reina Ambrocio , Meagan Cherney , Nicholas A. Morris , Natarajan Ayithan , Xiaoxuan Fan , Volodymyr Gerzanich , J. Marc Simard , Neeraj Badjatia , Gary Schwartzbauer , Gunjan Y. Parikh , Seth Ament , Prajwal Ciryam bioRxiv 2025.06.05.657834; doi: https://doi.org/10.1101/2025.06.05.657834 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Genomics Subject Areas All Articles Animal Behavior and Cognition (7637) Biochemistry (17705) Bioengineering (13899) Bioinformatics (41970) Biophysics (21463) Cancer Biology (18605) Cell Biology (25526) Clinical Trials (138) Developmental Biology (13385) Ecology (19911) Epidemiology (2067) Evolutionary Biology (24329) Genetics (15615) Genomics (22514) Immunology (17743) Microbiology (40424) Molecular Biology (17194) Neuroscience (88650) Paleontology (667) Pathology (2835) Pharmacology and Toxicology (4827) Physiology (7648) Plant Biology (15160) Scientific Communication and Education (2046) Synthetic Biology (4302) Systems Biology (9825) Zoology (2271)