Mycobacterial infection uncovers plasticity of Kupffer cells

Mycobacterial infection uncovers plasticity of Kupffer cells | 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 Mycobacterial infection uncovers plasticity of Kupffer cells View ORCID Profile Jana Neuber , View ORCID Profile Florens Lohrmann , Samuel Wald , Merve Göçer , Anne Kathrin Lösslein , David Obwegs , View ORCID Profile Vitka Gres , Torsten Goldmann , View ORCID Profile Manuel Rogg , Christoph Schell , Sebastian Preißl , Sagar , View ORCID Profile Philipp Henneke doi: https://doi.org/10.1101/2025.02.07.636999 Jana Neuber 1 Institute for Infection Prevention and Control, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany 2 Center for Chronic Immunodeficiency, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany 3 Faculty of Biology, University of Freiburg , Freiburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Jana Neuber Florens Lohrmann 1 Institute for Infection Prevention and Control, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany 4 Department for Pediatrics and Adolescent Medicine, University Medical Center, Medical Faculty, University of Freiburg , Freiburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Florens Lohrmann Samuel Wald 5 Institute of Experimental and Clinical Pharmacology and Toxicology, Faculty of Medicine, University of Freiburg , Freiburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Merve Göçer 1 Institute for Infection Prevention and Control, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany 2 Center for Chronic Immunodeficiency, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany 3 Faculty of Biology, University of Freiburg , Freiburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Anne Kathrin Lösslein 1 Institute for Infection Prevention and Control, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany 2 Center for Chronic Immunodeficiency, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany 6 Institute for Medial Microbiology and Hygiene, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site David Obwegs 3 Faculty of Biology, University of Freiburg , Freiburg, Germany 7 Department of Medicine II, University Medical Center, Medical Faculty, University of Freiburg , Freiburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Vitka Gres 1 Institute for Infection Prevention and Control, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany 2 Center for Chronic Immunodeficiency, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Vitka Gres Torsten Goldmann 8 Research Center Borstel, Leibniz Lung Center, Member of the German Center for Lung Research (DZL) , Borstel, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Manuel Rogg 9 Institute for Surgical Pathology, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Manuel Rogg Christoph Schell 9 Institute for Surgical Pathology, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Sebastian Preißl 5 Institute of Experimental and Clinical Pharmacology and Toxicology, Faculty of Medicine, University of Freiburg , Freiburg, Germany 10 CIBSS-Center for Integrative Biological Signaling Studies, University of Freiburg , Freiburg, Germany 11 Department of Pharmacology and Toxicology, Institute of Pharmaceutical Sciences, University of Graz , 8010 Graz, Austria Find this author on Google Scholar Find this author on PubMed Search for this author on this site Sagar 7 Department of Medicine II, University Medical Center, Medical Faculty, University of Freiburg , Freiburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site Philipp Henneke 1 Institute for Infection Prevention and Control, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany 2 Center for Chronic Immunodeficiency, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany 10 CIBSS-Center for Integrative Biological Signaling Studies, University of Freiburg , Freiburg, Germany 12 Institute for Immunodeficiency, Medical Center and Faculty of Medicine, University of Freiburg , Freiburg, Germany Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Philipp Henneke For correspondence: philipp.henneke{at}uniklinik-freiburg.de Abstract Full Text Info/History Metrics Supplementary material Preview PDF Abstract Bona fide Kupffer cells (KCs) are prenatally seeded and show unique functional and immunophenotypic features among tissue macrophages. They are considered as terminally differentiated, and adaptability in disease is attributed to recruited, monocyte-derived KCs. Here, we investigated the extent of KC plasticity and the impact of origin in mycobacterial infections that target macrophages and can persist for months. Fate-mapping combined with high-resolution imaging revealed the emergence of a unique, infection specific KC subset which downregulated the signature markers CLEC4F and VSIG4 (“KC low ”). KC low were derived from bona fide KCs and located exclusively to granuloma cores. In contrast, monocyte-derived macrophages were contained at the granuloma borders and contributed to this tissue reaction. ATAC and single-cell RNA sequencing identified a specific signature of KC low with high antimycobacterial activity and specialization to a hypoxic microenvironment. Despite their fundamental deviation from the classical KC phenotype, KC low showed remarkable adaptability, and were capable to return to a homeostatic-like KC state. Accordingly, mycobacterial infections unmask KCs as highly plastic cells, capable of responding to extreme environmental changes. Main Tissue-resident liver macrophages, called Kupffer cells (KCs), are prenatally seeded by either progenitors from the yolk sac or fetal liver 1 . Due to their distinct anatomical localization within the sinusoids, they are poised to capture blood-borne pathogens 2 . In addition, they play pivotal roles in the clearance of senescent erythrocytes, as well as gut-derived microbial products 3 . In homeostasis, KCs self-renew and receive negligible input from monocytes 1 . However, during liver damage, infection-mediated cell death or after experimental KC depletion, monocytes enter the liver in high numbers and differentiate into monocyte-derived KCs (MoKCs) with the acquisition of a KC specific marker set including CLEC4F, VSIG4, or TIM-4 4 , 5 . The upregulation dynamics of KC markers in MoKCs vary, e.g. expression of CLEC4F can be detected after around one week, while TIM-4 is only partially expressed after one month in experimental KC depletion 6 . The initiation of KC differentiation requires KCs to extend protrusions into the space of Disse, where they interact with liver sinusoidal endothelial cells (LSECs) and hepatic stellate cells (HSCs) 4 , 7 , 8 . Accordingly, KCs are prime examples of terminally differentiated macrophages with highest specification to their tissue niche early in life. In line with this morphological uniqueness, they are the only macrophages that express CLEC4F 6 . Like other tissue macrophages, KCs are targeted by mycobacteria, which chronically perturb local innate immunity 9 , 10 . Mycobacteria have specifically adapted to macrophages in a coevolution extending over tens of thousands of years 11 . Thus, they have developed mechanisms to not only persist in this, in principle, hostile environment 12 , 13 , but to use macrophages as a proliferative niche 14 . Infections with mycobacteria lead to the formation of granulomas, highly dynamic cellular structures consisting of specialized macrophage subsets, in particular multinucleated giant cells (MGCs), foamy macrophages, or epithelioid macrophages 15 , 16 [Losslein, AK, Henneke, P, Ann Rev Immunol (in press)]. Granulomas are thought to limit mycobacterial dissemination, but also serve as reservoirs for mycobacterial persistence and may facilitate reactivation and spread, e.g. by necrotic transformation and promotion of tissue damage 17 . The preference of mycobacteria for macrophages including their slow replication rate makes them a unique tool to investigate the role of bona fide KCs and monocytes in tissue macrophage plasticity and diversity. At the same time, the investigation of tissue-specific macrophage responses during chronic infections may provide fundamental insights into how macrophages adapt to challenges. Exploitation of the dual reporter fate-mapping mouse Clec4f Cre -tdTomato-NLS : ROSA26 EYFP enabled long-term tracing of KCs by multidimensional analysis, thereby uncovering the emergence of a unique, highly specialized, and adapted KC subset, which we denominated KC low . KC low were characterized by the loss of KC identity markers, metabolic alterations, - i.e., the upregulation of inducible nitric oxide synthase (iNOS), a key enzyme in granuloma formation and antimycobacterial defense 18 , 19 - and enhanced efferocytosis. These changes were accompanied by a distinct localization of KC low to granuloma cores with direct mycobacterial contact, while bone marrow-derived cells induced and surrounded granulomas. Yet, KC low retained unexpected plasticity and were able to regain classical immunophenotypic KC traits. Results A specific KC subset forms the core of granulomas In order to dissect KC dynamics during mycobacterial infections with high resolution, we created a dual reporter mouse by crossing the Clec4f Cre -tdTomato-NLS to a ROSA26 EYFP mouse, with tdTomato expression under control of the KC specific Clec4f promoter and tagged with a nuclear localisation signal (NLS) 20 ( Fig. 1a ). Thus, steady-state KCs appear both tdTomato and YFP. Download figure Open in new tab Figure 1: Identification of a granuloma specific Kupffer cell population (a) Experimental setup of Clec4f Cre-tdTomato-NLS :ROSA26 EYFP mice i.v. injected with 10 7 cfu of BCG-BFP or PBS for (b-k). (b) Spleen size and weight during infection (n=(17, 4, 12, 18, 10, 5), mice respectively). (c) Bacterial burden in spleen and liver during infection. nspleen=(4, 10, 18, 10, 5) and nliver=(4, 12, 18, 10, 5), in sequence. One mouse at 10 wpi was below the detection limit for the liver. (d) Immunofluorescence (IF) staining of a liver 4 wpi. Dotted lines mark granulomas. Arrowhead indicates BCG. Scale bars: 20 µm. (e, f) Flow cytometry of KC populations in Clec4f Cre-tdTomato-NLS :ROSA26 EYFP mice with high (KC high ) or low (KC low ) tdTomato expression, or KCs lacking TIM-4 during BCG-BFP (e) or WT M. avium (f) infection. nBCG= (13, 4, 7, 11, 10, 5) and n M. avium =(3, 5, 5, 5), mice respectively. (g) Flow cytometry of Ki-67 in sorted KCs of PBS injected mice or at 2 wpi with BCG-BFP (KC high and KC low ). Isotype controls of the respective KC population are depicted in white. n=3 per group. (h) Flow cytometry gating of YFP + KCs after PBS injection, 4 and 10 wpi with BCG-BFP as used in (e). Percentages of gated populations are depicted in the graph. (i) IF staining of Clec4f Cre-tdTomato-NLS :ROSA26 EYFP livers of PBS injected mice, 4 and 10 wpi. Dotted lines mark granulomas. Scale bars: 50 µm. (j) Distribution of YFP + KC subsets analysed from microscopy images as shown in (i). n=(3, 3, 4, 3, 4) in sequence. (k) Percentage of granuloma-associated KC high and KC low based on (i, j). (l) IF staining of human livers from a healthy donor and two patients infected with Mycobacterium tuberculosis ( Mtb ). Dotted lines mark granulomas. Scale bars: 50 µm. Data represent mean ± SEM with each symbol depicting one mouse are from at least 2 independent experiments. One-way ANOVA with Tukey‘s multiple comparisons test. * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001. dpi, days post infection. wpi, weeks post infection. Intravenous infection with 10 7 colony forming units (cfu) of Mycobacterium (M.) bovis Bacillus Calmette-Guerin (BCG) led to a persistent, systemic infection, associated with splenomegaly most pronounced at 4 weeks post infection (wpi) ( Fig. 1b ). Splenic and hepatic bacterial burdens peaked in the first week and then showed a slow, gradual decrease over several months, indicating a chronic infection ( Fig. 1c ). We observed a similar progression in response to M. avium (Extended Data Fig. 1a), the most common cause of nontuberculous infections in humans in many parts of the world 21 . Both infection models caused formation of liver granulomas as early as 2 wpi, which decreased over time in number and size (Extended Data Fig 1b, c). Additionally, we identified mycobacteria residing in granulomas, which contained abundant iNOS-expressing macrophages, indicating an active and mycobacteria-specific inflammatory response ( Fig. 1d ). Flow cytometry analysis uncovered KC diversification after infection with either BCG or M. avium . The PBS control group contained only tdTomato and YFP double positive bona fide KCs, which additionally expressed the resident tissue macrophage marker TIM-4 6 , and were termed KC high . Interestingly, KC high increased steeply until 2 wpi, before returning to homeostatic level ( Fig. 1e, f ). This was in line with initial proliferation of KC high ( Fig. 1g ), which likely served to control the infection until monocyte influx compensated for an increased macrophage need. Notably, early in infection (2-4 wpi), a second YFP + KC subset emerged, defined by low to abrogated tdTomato expression (KC low ). In addition, TIM-4 - KCs were identified from 4 wpi onwards and peaked at 10 wpi ( Fig. 1e, f, h ). Microscopic analysis confirmed the appearance of KC low , predominantly at 4 wpi ( Fig. 1i, j ). Yet, in contrast to the flow cytometry data, KC low persisted until late infection time points (10- 12 wpi), when granuloma numbers were substantially decreased ( Fig. 1i, j ). In view of the relatively low number of cells, it seems probable, that the microscopy, which enabled focusing on granulomas, captured KC low better than flow cytometry. On the other hand, the late microscopic emergence of TIM-4 - KCs accurately reflected the flow cytometry data. Clec4f reporter downregulation specifically occurred for KCs in granuloma cores, while KCs surrounding granulomas remained KC high ( Fig. 1i ). For KC high , the peak association to granulomas was at 4 wpi (∼32 % of total KC high ), while up to ∼90 % of KC low resided within granulomas at this infection stage ( Fig. 1k ). The relative association of KC low with granulomas decreased until 12 wpi, but remained higher than for KC high ( Fig. 1k ). In other words, KC low rather than KC high stood out as granuloma macrophages. The maintenance of the KC phenotype and its acquisition by MoKCs respectively have been attributed to the interaction with HSCs via bone morphogenetic proteins and DLL4-Notch signaling with LSECs 4 , 7 , 8 . Therefore, we explored by microscopy whether the phenotypical changes of KC low were caused by a loss of heterocellular interactions. We found a marked increase of HSCs around granulomas revealed by staining for desmin 22 (Extended Data Fig. 2a). Moreover, KC low resided partly in the immediate vicinity of HSCs and, albeit at a lower extend, LSECs, even within granulomas (Extended Data Fig. 2b, c). These observations argued against loss of heterocellular contact as the central cause for CLEC4F downregulation in infection, but rather pointed at infection-induced macrophage reprogramming. Next, we investigated whether KC low analogues in mycobacterial infections exist in humans. Notably, liver biopsies from patients with Mycobacterium tuberculosis ( Mtb ) infection or other mycobacterial species are rare 23 , as Mtb infections are usually diagnosed by lung-related clinical and microbiological findings, and the puncturing of the diseased liver poses a considerable risk. Yet, we identified samples from two patients with confirmed Mtb infections and liver granuloma formation (Extended Data Fig. 1d). Staining for VSIG4 and CD163 identified KCs as double positive 24 , 25 , both in healthy and diseased liver tissue outside granulomas. However, inside granulomas, macrophages expressed only CD163 ( Fig. 1l ). The lack of the prototypic KC marker VSIG4 in granuloma-associated macrophages suggested the existence of KC low analogous cells in humans with Mtb infections. Taken together, mycobacterial infections led to the emergence of a unique KC subset - KC low - that lost key immunophenotypic markers and distinctly localized to granuloma cores with close mycobacterial contact. KC low are derived from resident KCs In order to explore the KC low origin, we combined the intravenous BGC infection model with chimeric mice generated by transplantation of Clec4f Cre -tdTomato-NLS : ROSA26 EYFP bone marrow (BM) into CD45.1 recipients, or vice versa ( Fig. 2a ). To prevent KC damage, livers were shielded during irradiation at the expense of a mixed chimerism (Extended Data Fig. 3a). The blood chimerism of circulating monocytes was used to calculate adjusted percentages of donor-derived cells in the liver. We found donor-derived TIM-4 + KCs to increase late in infection ( Fig. 2b ). Moreover, granuloma cores were mainly formed by recipient KCs at 4 wpi, while donor cells predominantly surrounded granulomas ( Fig. 2c, d and Extended Data Fig. 3b). Most of the donor cells lacked CLEC4F or TIM-4 expression, but expressed MHC-II (Extended Data Fig. 3c). In contrast, splenic granulomas contained a significant proportion of donor- derived cells, which were also the predominant cells expressing iNOS (Extended Data Fig. 3d). At 10 wpi, both resident and donor-derived cells localized to hepatic granuloma cores ( Fig. 2c, d ). Accordingly, mycobacterial granuloma macrophages in the liver were spatially and dynamically organized related to their origin, appearing to result from organ specific cues. Download figure Open in new tab Figure 2: Origin of the KC subsets during infection (a) Experimental setup for (b-e). Livers were shielded from radiation and bone marrow subsequently transplanted. After 8 weeks, mice were i.v. injected with WT BCG or PBS. Clec4f Cre-tdTomato-NLS :ROSA26 EYFP were transplanted with CD45.1 bone marrow and vice versa. (b) Chimerism of TIM-4 + KCs, adjusted to blood monocyte chimerism. nleft=(3, 4, 5) and nright=(5, 5, 5, 6, 5) in sequence. One-way ANOVA with Dunnett‘s multiple comparisons test compared to the PBS condition. (c) Immunofluorescence staining (IF) of a CD45.1 liver transplanted with Clec4f Cre-tdTomato- NLS :ROSA26 EYFP bone marrow at 4 and 10 wpi. Staining against CD45.2 (donor, green) is in the same channel as the tdTomato signal. Dotted lines mark granulomas. Scale bar: 50 µm. (d) Relative mean fluorescence intensity (MFI) for CD45.2 (bone marrow (BM)-derived) or CD45.1 (resident) calculated as a ratio of liver granuloma cores and granuloma borders based on microscopic images shown in (c). Analysis was performed for 4 and 10 wpi. One dot represents averaged values for one mouse. n=4 per group. Two-tailed unpaired t test. (e) Flow cytometry analysis of KC high , KC low , and TIM-4 - KCs from transplanted mice in (b). (f, g) Flow cytometry analysis of tdTomato expression in Ly6G + granulocytes and Ly6C high monocytes (f) or in TIM-4 + VSIG4 + KCs and TIM-4 - VSIG4 + KCs (g) in livers of Ms4a3 Cre :ROSA26 tdTomato mice after BCG-BFP infection. n=(5, 5, 3), mice respectively. Two mice in the PBS condition and 4 wpi were i.v. infected via the retro-orbital route. (h) IF of a Ms4a3 Cre :ROSA26 tdTomato liver 4 wpi. Arrowheads mark exemplary tdTomato + CLEC4F + cells. Scale bar: 50 µm. (i) Spleen weight and hepatic bacterial burden comparing WT (nspleen weight=(7, 7, 9, 7), ncfu=(7, 7, 7), mice respectively) and Ccr2 -I- (nspleen weight=(4, 6, 5, 5), ncfu=(7, 6, 6) in sequence) mice. Infections were performed with WT BCG. One mouse at 10 wpi was below the detection limit for hepatic bacterial burden. (j) Flow cytometry of KC high and KC low in Clec4f Cre-tdTomato-NLS :ROSA26 EYFP (data from Fig. 1e ) and Clec4f Cre-tdTomato-NLS :ROSA26 EYFP :Ccr2 -I- (n=(7, 4, 8, 5) in sequence) mice. (k) IF 2 wpi of WT and Ccr2 -I- mouse livers. Scale bars: 100 µm. (l) Quantification of the liver area occupied by granulomas in microscopy images shown in (k), defined as CD68 + clusters with a minimal size of 350 µm 2 , as well as their average size. One dot represents a mouse, n WT =(5, 6, 4) and nCcr2-/-=(5, 5, 5) respectively. Infections were performed with WT BCG. Data represent mean ± SEM with each symbol depicting one mouse and are derived from at least 2 independent experiments. One-way ANOVA (e-g) or two-way ANOVA (i, j, l) with Tukey‘s multiple comparisons test, unless otherwise stated. * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001. Furthermore, we noted that donor-derived KC high slowly increased during prolonged infection, while donor-derived KC low remained rare ( Fig. 2e ). In contrast, increased recipient-derived KC low at 4 wpi indicated that KC low derived from bona fide KC without substantial contribution of BM-derived cells. Additionally, and in accordance with the literature, TIM-4 - KCs were mainly donor-derived 6 , 26 . The increase of TIM-4 - recipient KCs likely reflected the contribution of remaining recipient BM cells due to the mixed chimerism ( Fig. 2e ). To reinforce these findings, we analyzed Ms4a3 Cre :ROSA26 tdTomato fate-mapping mice, in which granulocyte-monocyte progenitor (GMP)-derived cells, and thus the vast majority of monocytes, are indefinitely labelled with tdTomato 27 . As expected, a high percentage of granulocytes and Ly6C high monocytes in the liver were tdTomato-positive ( Fig. 2f ). In line with our expectations, tdTomato + TIM-4 - KCs rapidly increased in number during infection, while tdTomato + TIM-4 + KCs displayed a slow but progressive increase ( Fig. 2g ). Additionally, and in full alignment with the model that mycobacterial liver granuloma cores are composed of resident KCs, microscopy of Ms4a3 Cre :ROSA26 tdTomato livers revealed a lack of tdTomato + , and thus monocyte-derived cells, in granuloma centers at 4 wpi ( Fig. 2h ). Of note, some tdTomato + cell agglomerates were visible, yet, since they were negative for TIM-4, they represented immature granulomas. Moreover, we identified occasional tdTomato + cells, which coexpressed CLEC4F and in part TIM-4, localized outside of granulomas, indicating a transitional stage towards KC high ( Fig. 2h ). Subsequently, we explored the role of monocytes during infections, given the distinct spatial distribution of BM-derived macrophages at granuloma margins 4 wpi. Accordingly, we infected Ccr2 -I- mice, which are deficient in circulating Ly6C high monocytes (Extended Data Fig. 3e) 28 , and found a reduced splenomegaly and increased splenic bacterial loads, while hepatic bacterial counts were comparable to WT mice early in infection ( Fig. 2i and Extended Data Fig. 3f). Moreover, livers of Ccr2 -I- mice were devoid of Ly6C high monocytes during infection and showed lower granulocyte numbers as compared to WT mice at 2 and 4 wpi (Extended Data Fig. 3e, g). Comparison of the KC subsets within the Clec4f reporter line crossed onto a Ccr2 -I- background revealed reduced KC high and KC low numbers at 2 wpi within the knockout mice ( Fig. 2j ). Yet, KC high were markedly increased 4 wpi in Ccr2 -I- Clec4f reporter mice, making it tempting to speculate that this increase in KC high compensated for the lack of monocytes. Furthermore, at 10 wpi Ccr2 -I- Clec4f reporter mice had fewer TIM-4 - KCs (Extended Data Fig. 3h), in line with their BM origin. Additionally, Ccr2 -I- mice displayed significantly smaller and fewer liver granulomas ( Fig. 2k, l ), indicating the role of monocytes in granuloma propagation and explaining the observed reduction of KC low in Ccr2 -I- Clec4f reporter mice, especially at 2 wpi. In summary, we have demonstrated that KC low are a progeny of bona fide KCs. The role of monocytes in granuloma formation appeared to be rather indirect: they did not directly contribute to granulomas, but influenced granuloma size and counts, without substantially affecting bacterial clearance. Thus, bona fide KCs appeared to be the main granuloma cells, as well as key effectors for bacterial control. KC low lose their Kupffer cell identity To further explore the identity and function of KC low , we performed bulk ATAC sequencing (seq) (gating strategy in Extended Data Fig. 4a) and bulk RNAseq on the three identified KC populations at 4 wpi, and of KCs after PBS injection. The principle component analysis (PCA) of the ATACseq data, as well as analysis of the significantly differential accessible regions (DARs) revealed distinct genome-wide chromatin accessibility profiles for each subtype ( Fig. 3a and Extended Data Fig. 4b). The DARs of KC low were particularly close to genes involved in cellular response to low oxygen, suggesting reduced oxygen availability within granulomas. Additionally, enriched gene ontology (GO) terms of genes close to DAR of KC low included actin filament and fibronectin binding (Extended Data Fig. 4c), indicating their contribution to granuloma organization. Surprisingly, enriched GO terms of genes close to DAR of TIM-4 - KCs were related to bacterial response ( Fig. 3b ), even though they appeared late in infection and were only loosely associated with granulomas. Download figure Open in new tab Figure 3: KC low lose identity markers and adapt to their new microenvironmental niche (a) Principle component analysis (PCA) plot from ATAC sequencing (seq) of sorted KC high , KC low , and TIM-4 - KCs at 4 wpi, as well as KC high from PBS treated mice. n = 5 for KC high and n = 4 for the other populations from two independent infection experiments. (b) Top 10 enriched GO-terms of biological processes for KC low vs. all other populations (rest) (left) and TIM-4 - KCs vs. rest (right) within the ATACseq data. (c) Integrative Genomics Viewer of ATACseq displaying chromatin accessibility for indicated genes representative for one sample per population. Significantly differentially accessible regions (DARs) are depicted below the tracks in blue for downregulated and red for upregulated accessibility comparing one population vs. all other populations. (d) Heatmap of Kupffer cell signature genes in bulk RNAseq data for sorted KC high , KC low , and TIM-4 - KCs at 4 wpi, as well as for KC high from PBS treated mice. n = 3 per group. (e) Heatmap of differentially expressed genes of the GO term “cholesterol metabolic process” (GO:0008203) between PBS and KC low in the bulk RNAseq data. (f) Heatmap of differentially expressed genes of the GO term “response to hypoxia” (GO:0001666) in the bulk RNAseq data. Chromatin accessibility in the gene locus for Clec4f was decreased in KC low and a complete absence of accessibility was observed in the Timd4 locus for TIM-4 - KCs ( Fig. 3c ). Additionally, chromatin accessibility in KC low was high for Hif1a and substantially increased in the Nos2 gene ( Fig. 3c ). In contrast, key transcription factors of KCs 4 , 29 remained largely unchanged (Extended Data Fig. 4d). Bulk RNA sequencing uncovered distinct expression profiles of the predefined populations (Extended Data Fig. 4e). KC low exhibited drastically reduced expression of many KC signature genes, most notably for ld3 , Clec4f , Cd207 , and Vsig4 ( Fig. 3d ). TIM-4 - KCs did not yet upregulate these genes, indicating their early transitional stage towards the KC phenotype ( Fig. 3d ). It is worth mentioning that, although TIM-4 - KCs exhibited reduced Clec4f expression compared to KCs from PBS mice, the Clec4f promotor activity was sufficient to drive Cre/Lox recombination, as TIM-4 - KCs were identified as YFP + . GO terms for KC low indicated metabolic rewiring and increased chemokine receptor binding compared to KC high , suggesting differences in microenvironmental cues (Extended Data Fig. 4f). Mycobacteria can directly and profoundly affect the macrophage lipid metabolism 30 , 31 and the accumulation of cholesterol is essential for the transformation of macrophages towards multinucleated giant cells 32 . While our mouse model did not yield MGCs, KC low were in direct contact with mycobacteria, which might explain the altered gene expression pattern of the cholesterol metabolism ( Fig. 3e ). In humans, the necrotic and hypoxic transformation of granulomas facilitates mycobacterial spread 33 . While WT mice did not show visible granuloma necrosis, in line with the literature 34 , 35 , KC low still upregulated many genes associated to hypoxia ( Fig. 3f ), in accordance with GO terms resulting from ATACseq ( Fig. 3b ). It seems likely that oxygen availability in granuloma macrophages is altered due to vascular reorganization 36 , as indicated by increased Vegfa expression ( Fig. 3b and Extended Data Fig. 4h). Overall, KC low displayed a fundamentally altered transcriptional profile compared to KC high with more than 1500 differentially expressed genes (Extended Data Fig. 4g). In particular, KC low upregulated inflammatory genes like Nos2 , ltgax , and Saa3 (Extended Data Fig. 4i), as well as genes involved in tissue remodelling [ Vegfa and Mmp12 (Extended Data Fig. 4h)], arguing for a pivotal role in granuloma formation and granuloma macrophage function. In summary, KC low and bona fide KC high are distinct liver specific macrophage species with a unique immunophenotype and function as indicated by their transcriptional profiles and epigenetic traits, although they are of similar origin. Accordingly, our findings indicated previously underappreciated plasticity of KCs. ldentification of infection-specific KC clusters We performed single cell RNA sequencing (scRNAseq) to precisely resolve the heterogeneity of KCs in infection on sorted KC high , KC low , TIM-4 - KCs, and CD11b high or TIM-4 + cells, which contained mainly monocytes, from mice at 4 and 10 wpi, as well as corresponding populations from PBS injected control mice (sorting strategy in Extended Data Fig. 5a). Due to low cell numbers, the KC subsets were separately isolated and enriched. After removal of contaminating cells, we identified 17 distinct clusters ( Fig. 4a ). Based on the cell classification used for sorting, we found monocytes clustered together and separately from the KC subsets (Extended Data Fig. 5b). Download figure Open in new tab Figure 4: Single cell sequencing of KCs and monocytes during infection (a) Uniform manifold approximation and projection (UMAP) of sorted cells from Clec4f Cre- tdTomato-NLS :ROSA26 EYFP mice after PBS injection or 4 and 10 weeks after BCG-BFP infection used for scRNA sequencing. In total 48,500 cells. (b) Cell cluster definition and average expression of marker genes depicted as dot plot. (c) UMAP for the aggregated gene score “efferocytosis”. (d) Heatmap of the, up to, top 50 differentially expressed genes comparing 4 wpi vs. PBS in cluster 1, calculated as pseudobulk data. (e) Top 10 GO terms in cluster 1 upregulated 4 wpi compared to the PBS control, calculated for the pseudobulk dataset. Values depicted as 100 were calculated with fold enrichment > 100. (f) Heatmap of the number of cell-cell interactions determined with CellChat exclusively for cells 4 wpi. The x-axis depicts the receivers, while the y-axis shows the clusters sending signals. Cluster 16 had no interactions and is therefore not depicted. (g) Heatmap of signalling patterns in decreased order according to the relative strength, illustrated for outgoing and incoming patterns for cells at 4 wpi. (h) Chord diagram for the CCL pathway with the interactions between the different clusters at 4 wpi, as well as the top receptor-ligand interactions contributing to the pathway. (i) Reclustered UMAP, subsetted on the KC low cluster 7 from (a). In total 3183 cells. (j) Feature plots depicting the expression of the above labelled genes. (k) Dot plot for marker genes defining the different subclusters from (i). Next, we calculated differentially expressed genes for each cluster (Extended Data Fig. 5c), thus distinguishing six KC high (1, 4, 5, 6, 8, and 15), one KC low (7), two TIM-4 - KCs (3 and 12), and six monocyte (0, 2, 9, 10, 11, and 14) clusters ( Fig. 4b ). Cluster 16 represented liver capsular macrophages and cluster 13 a small KC subset, defined by high Ccl5 expression, sorted among KC low and TIM-4 - KCs ( Fig. 4b and Extended Data Fig. 5b). The reported contribution of CCL5 as a potent chemoattractant for lymphocytes in early granuloma formation 37 indicated an important role of cluster 13 in this context, further enhanced by cluster 7 and 9. In accordance with our bulk RNAseq data and immunofluorescence staining, Nos2 was highly expressed in cluster 7 (KC low ). Additionally, monocyte cluster 9 showed increased expression of Nos2 ( Fig. 4b and Extended Data Fig. 5d). Ccr2 expression was restricted to monocyte populations, while Timd4 and Clec4f , as well as the transcriptions factors Nr1h3 and Spic , characterized KC populations (Extended Data Fig. 5d). Cluster 9 potentially had a modest contribution to bacterial clearance, given the slightly elevated bacterial burden observed in Ccr2 -I- mice ( Fig. 2i ) lacking monocytes. Association of the clusters with the treatment modalities and time points revealed that mainly clusters 0, 3, 6, 7, 9, and 13 were enriched in infected mice, particularly at 4 wpi (Extended Data Fig. 5e, f). KC high clusters 1 and 4 were primarily present in the PBS condition, indicating transcriptional changes in nearly all KCs during infection, in accordance with the bulk RNAseq data. However, these alterations seemed to be reversible, since the cluster distribution converged with the PBS condition at 10 wpi (Extended Data Fig. 5e, f). The comparison of GO terms specific for KC low (cluster 7) with those in steady state KC high (cluster 0), infection-associated KC high (cluster 6), or Nos2 -expressing monocytes (cluster 9), uncovered an enrichment for endocytosis, and phagocytosis (Extended Data Fig. 5g). Additionally, KC low upregulated genes associated with fatty acid synthesis and type II interferon response (Extended Data Fig. 5g), which mediates important roles in mycobacterial defense, e.g. via the induction of nitric oxide 38 . Given that GO terms of cluster 7 were linked with phagocytosis and the recognized role of KCs in efferocytosis 39 , which has been shown to contribute to mycobacterial killing 40 and granuloma macrophage specialization 41 , we assessed the efferocytic gene expression profile of KC low . The aggregated gene score for “efferocytosis” revealed an increased expression of associated genes in clusters 6, 7, and 13 ( Fig. 4c ). To further investigate the differences of KC high in infection, we performed pseudobulk analysis of cluster 1, comparing the PBS condition with cells isolated at 4 wpi. KC high of infected mice, exhibited significant transcriptional changes ( Fig. 4d ), and were associated with cytolysis and type II interferon-mediated signaling ( Fig. 4e ). Additionally, we found an upregulated metabolic activity, including fructokinase and hexokinase ( Fig. 4e ). The metabolic breakdown of sugars is an important liver function 42 , and these findings point towards increased energy consumption in KCs, required to control infection. The CellChat algorithm revealed that most interactions at 4 wpi originated from cluster 3, 7, 8, and 12, with cluster 7 notably receiving most signals ( Fig. 4f, g ). The top signalling patterns were associated with antigen processing and presentation (APP, MHC-II), adhesion (VCAM, ICAM), and chemokines (CCL) ( Fig. 4g ). Outgoing CCL signalling was present in all clusters, except for cluster 16, but most prominent in cluster 7, while infection-associated monocytes (cluster 0) received the largest input ( Fig. 4g, h ). This was in accordance with active monocyte recruitment to the site of infection. Among specific receptor-ligand pairs, Ccl5 and Ccr5 formed the top interaction, followed by Ccl6 and Ccr2 ( Fig. 4h ). Subclustering of KC low (cluster 7) to obtain a higher resolution identified six distinct subclusters ( Fig. 4i and Extended Data Fig. 5h). Feature plots of Timd4 and Nos2 revealed discrete expression patterns, with cluster 4 exhibiting the most inflammatory profile, while expression of Timd4 was lower than in other clusters ( Fig. 4j, k ). Additionally, differential expression of Cd5l and Lgals3 divided the clusters ( Fig. 4j, k ). Cd5l , encoding for “apoptosis inhibitor of macrophages”, has been shown to be upregulated during M. avium infection, leading to increased apoptosis resistance in macrophages and may therefore facilitate mycobacterial growth 43 . On the other hand, knockout mice for Lgals3 , encoding for galectin-3, were found to be associated with heightened susceptibility to Mtb infection 44 , 45 . This divergent expression pattern supported a model where macrophages at the infection site were functionally distinct: Cluster 2 and 4, as Nos2 and Hif1a expressing clusters, were dedicated to controlling infection, while cluster 0 and 1, with a transcription profile reminiscent of KC high , appeared permissive for mycobacterial persistence. KC low are able to revert to KC high Since KC low were characterized by a profound loss of KC identity while being polarized for essential antibacterial functions, we wondered about their fate after infection resolution when both granulomas and KC low disappeared ( Fig. 1e and Fig. 2k ). This disappearance could be due to cell death, or a return to the classical KC high phenotype, rendering previous KC low undistinguishable from persisting KC high or MoKCs. RNA velocity trajectory analysis of all macrophage clusters based on the scRNAseq data identified connections at 4 wpi between cluster 7 (KC low ) towards cluster 8 (infection-associated KC high ). From there, additional connections existed at 10 wpi to cluster 1, suggesting that KC low may revert to KC high -like cells in a progressive manner ( Fig. 5a ). Furthermore, KC high cluster 5 converged towards cluster 1 in all three conditions, suggesting it as a transitional stage of KCs ( Fig. 5a ). In line with our previous assumption, TIM-4 - KCs (cluster 3, 12) showed connections towards KC high (cluster 6) ( Fig. 5a ). Overall, the data indicated cluster 1 as a final trajectory of KCs during infection resolution and in the steady state (Extended Data Fig. 6a). Download figure Open in new tab Figure 5: KC low show remarkable plasticity (a) Trajectory analysis of the macrophage clusters contained in the scRNAseq dataset split by condition (PBS, 4 wpi, and 10 wpi), calculated based on scvelo using the dynamic model. (b) Flow cytometry of TIM-4, tdTomato, and VSIG4 expression, gated on live YFP + cells, of cultured KC high (n=6), KC low (n=6), and TIM-4 - KCs (n=5 or n=4 on D5) isolated from Clec4f Cre-tdTomato-NLS :ROSA26 EYFP livers 4 wpi with BCG-BFP. Statistical significance was calculated between a KC subset and KC high at the same day (asterisks, * P < 0.0001 and stated p-values) and for each KC subset during the cultivation time (shown as hashtags compared to D1, # P ≤ 0.05, ## P ≤ 0.01) using a Mixed-effects model with Tukey’s multiple comparison test. (c) Confocal imaging of cultured KC high and KC low from (c). Depicted are two representative pictures (top and bottom) per condition and cultivation day. Bright fields are overlaid with the YFP (yellow, left) or tdTomato (red, right) reporter signal. Scale bar: 10 µm. (d) Relative gene expression of Mrc1 and Hif1a normalized to Hprt during the cultivation of KC high (n=4-6), KC low (n=3-6), and Tim-4 - KCs (n=3-4). For Hif1a two values of TIM-4 - KCs at D5 are depicted as 0, as they were non-detectable. Mixed-effects model with Tukey’s multiple comparison test with * P ≤ 0.05. Each symbol depicts one mouse. (e) Schematic representation depicting the appearance and localization of the identified KC subsets during the course of infection. Data in (b-d) represent mean ± SEM and is derived from 2-3 independent experiments. In order to corroborate these findings, we set out to analyze the KC fate in vitro . However, KCs are notoriously difficult to manipulate and analyze ex vivo since they downregulate key markers in conventional 2D cultures 46 or in precision-cut liver sections 47 . Thus, we developed a novel system enabling in vitro cultivation of functional and viable KCs with preservation of KC identity parameters for several days. Our improved cultivation method involved plating sorted KCs from untreated Clec4f Cre -tdTomato-NLS : ROSA26 EYFP mice on Geltrex TM coated wells (Extended Data Fig. 6b-d). To assess KC functionality, we analyzed their ability to avidly phagocytose BCG-BFP. Cultured KCs remained viable over 5 days with the majority of KCs continuing to express TIM- 4 (∼89 %) and maintaining Clec4f reporter expression (∼69 %) (Extended Data Fig. 6e, f). In contrast, both markers substantially decreased after infection with high BCG concentrations (Extended Data Fig. 6f). Moreover, VSIG4 expression was maintained with a transient decrease at D3 for the high BCG concentration (Extended Data Fig. 6f). The downregulation of both tdTomato and VSIG4 expression hinted towards the emergence of a KC low analog in vitro in the presence of mycobacteria. KCs phagocytosed BCG over 5 days in culture (Extended Fig. 6g, h), thus demonstrating the preservation of energy consuming and therefore demanding functions of KCs in Geltrex TM . Next, we sorted and cultured the predefined KC subsets extracted from infected mice. We found maintained TIM-4 expression in KC low , while KC high exhibited a marginal decrease until day 5 ( Fig. 5b ). Notably, KC low regained both tdTomato and VSIG4 expression over time, ultimately reaching levels comparable to KC high ( Fig. 5b, c ). These findings suggested an unprecedented phenotypic plasticity of KCs in the absence of structural liver cells. Of note, TIM-4 - KCs did not increase TIM-4 expression in culture. Instead, ∼80 % of TIM-4 - KCs lost tdTomato expression already after one day in culture ( Fig. 5b ). On the transcriptional level, Mrc1 expression, as a key hallmark across liver macrophages, was maintained over the cultivation period. Additionally, and consistent with the transformation of KC low to a KC high -like phenotype, we observed for them a slight downregulation of Hif1a over time ( Fig. 5d ). Overall, it seems important to note that the described in vitro KC culture of infected mice yielded in only a relatively small number of cells. Accordingly, although we embarked on adoptive transfers of KC low we did not succeed in recovering sufficient numbers of transferred cells. Taken together, we find that KC possess a remarkable plasticity, leading to far reaching identity loss and maximal spatial adaptations during chronic infection, yet reversion to classical bona fide KC features after infection clearance ( Fig. 5e ). Discussion Bona fide KCs are extraordinary macrophages due to their prenatal origin, their localization in a unique microenvironment within liver sinusoids, and their frequent exposure to foreign material, e.g., to microbial components drained from the intestine. The established filtering capacity of KCs, which form the body’s largest macrophage population, manifests in their efficient phagocytic capacity, visible already minutes after intravenous bacterial injection 10 , 48 . The impact of the microanatomical and functional KC niche is reflected by a highly specific transcriptional profile, leading to the expression of the C-type lectin receptor CLEC4F. The dependency of KCs on the tissue environment is supported by their rapid loss in liver injuries and infections 5 , 49 – 51 , as well as the pronounced difficulty of maintaining KCs in culture 46 , 47 . The disappearance of bona fide KCs is linked with an opening of the niche and replacement by monocyte-derived macrophages, which display high plasticity during differentiation and are primed to rapidly acquire an inflammatory phenotype 5 , 49 , 52 , 53 . Here we employed an infection model with M. bovis BCG to revisit the concept of labor division between KCs, which perform homeostatic functions, and MoKCs, which occupy newly formed niches due to tissue alterations and help to contain the spread of infection. The chronic mycobacterial infection, uncovered substantial adaptability of bona fide KCs, culminating in the emergence of the KC low subset, characterized by the loss of lineage surface markers. Cues from their new microenvironment in granuloma cores, with a high density of mycobacteria, imposed this transformation process on KCs in both mice and humans. Downregulation of CLEC4F in liver fibrosis or infection with Leishmania infantum have been described before 54 , 55 . Tissue reorganization and communication breakdown of KCs with HSCs and LSECs were suggested by Peiseler et al. to cause diminished expression of the lineage determining transcription factor Nr1h3 55 . However, we found KC low to undergo their identity shift despite retaining partial close contact with HSCs and LSECs, arguing against loss of heterocellular interactions as the critical factor for the differentiation trajectory described here. It seems worth noting that Nr1h3 expression was decreased in KC low compared to KC high , yet, the expression levels were highly elevated compared to monocyte populations infiltrating the liver. Furthermore, ATAC sequencing revealed no significant changes in the chromatin accessibility of lineage determining KC transcription factors. One of the inflammatory genes upregulated on KC low was ltgax (encoding for CD11c), a marker often used to denominated dendritic cells (DCs) 56 . DCs were reported to be recruited to maturing granulomas 16 and to phagocytose mycobacteria 57 , 58 . With the help of a CD11c-eYFP reporter mouse, Harding et al. found YFP + cells, defined as DCs, as the predominant granuloma cells, able of seeding new liver granulomas 58 . In contrast, our fate-mapping investigations strongly suggest, that CD11c in granuloma cores identifies transformed bona fide KCs (KC low ). The misidentification of the cells is further promoted by the upregulation of the antigen processing and presentation machinery within KC low . Yet, it remains to be investigated if KC low are able to traffic between sites and seed granulomas, as Harding et al. proposed for the YFP + cells 58 . In addition, KC low exhibited some similarities with lipid-associated macrophage-like KCs (LAM- like KCs) observed in non-infectious liver injury 59 . Both cell subsets were derived from bona fide KCs, show efferocytic activity, and upregulated lipid-associated genes, Mmp12, and ltgax 59 . Yet, KC low are clearly distinguished from LAM-like KCs by extensive loss of both CLEC4F and VSIG4 expression, and by being maintained by differentiation of KC high rather than self-proliferation. Next to KC low , we also found a temporal emergence of TIM-4 - KCs, peaking late in infection. Data generated in Ms4a3 Cre :ROSA26 tdTomato mice indicated TIM-4 - KCs to be BM-derived, confirming expectations based on prior reports concerning TIM-4 upregulation on MoKCs 6 . The late appearance of this subset in infection is in accordance with the time consuming process of complete KC maturation. Corroborating this notion, we observed the highest recruitment of monocytes at 4 wpi, at which time few TIM-4 - KCs were present. However, only 70-80 % of TIM-4 - KCs in Ms4a3 Cre :ROSA26 tdTomato mice expressed tdTomato, suggesting that either a part of the population was derived from monocyte-dendritic cell progenitors (MDPs) and not GMPs, or that bona fide KCs downregulated TIM-4, thereby contributing to TIM-4 - KCs. The presence of TIM-4 - KCs in Clec4f Cre -tdTomato-NLS : ROSA26 EYFP : Ccr2 -I- mice argues for a downregulation of TIM-4 by a subset of KC high . Interestingly, TIM-4 - KCs showed an epigenetic profile that resembled highly inflammatory cells, although this was not reflected on the transcriptional level, likely representing an epigenetic priming originating from BM progenitors. This prevention to develop the full inflammatory phenotype was potentially caused by insufficient contact with BCG, as TIM-4 - KCs had only a slight contribution to granulomas. Previous studies have identified transcriptional changes of BM progenitors in the presence of BCG, leading to an enhanced antimycobacterial capacity of monocytes, supporting our hypothesis of primed but “frustrated” TIM-4 - KCs 60 . Combined with the Ccr2 -I- mouse data, it seems likely that monocytes and their progeny, such as TIM-4 - KCs, are not essential for the initial bacterial clearance in our model. Yet, the presence of monocytes was required for granuloma induction as their lack caused a decrease in granuloma number and size consistent with reports from aerosol Mtb lung infections 61 . Monocytes and their precursors appear to have important roles in specific mycobacterial infection situations, in particular when the antimycobacterial capacity of resident tissue macrophages is overwhelmed, e.g., in soft tissue infections with BCG 53 , or as effector cells giving rise to specialized granuloma cells, e.g., MGCs 32 . Given our finding that mainly resident macrophages, i.e., KCs, were present in liver granuloma cores at peak infection and upregulated inflammatory markers, while the reverse was observed in the spleen, it is likely that the mycobacterial response is tissue specific. The unique reaction of liver is supported by the rarity of human hepatic mycobacterial infections, while other tissues are more commonly target of extrapulmonary Mtb infections 62 , 63 . The return to homeostatic conditions in our model was coupled with the disappearance of KC low , which may indicate their inevitable cell death, caused by exhaustion after bacterial clearance and a terminal differentiation state suggested by the loss of lineage-determining markers. However, our newly established in vitro culture uncovered a before unknown plasticity of KCs, which retained, at least in part, the ability to regain KC signature properties. In other words, becoming KC low , which was associated with relocalization into granulomas and therefore the adaptation to a new microenvironment, was not a one-way road. Extraction from this inflammatory context enabled the reversion to a KC high -like phenotype. Moreover, maintenance of the KC phenotype is linked to constant provision of transforming growth factor-b (TGF-β) by LSECs 8 . However, in our hands, KC marker expression was preserved in pure KC high culture, strongly suggesting that sustained interactions with LSECs or HSCs are not a prerequisite for the KC identity preservation over shorter times. As TGF-β can also be produced by KCs 64 – 66 , autocrine signalling might have compensated for the heterocellular interaction. In contrast, TIM-4 - KCs failed to upregulate TIM-4 and lost their Clec4f reporter expression rapidly under the same conditions, indicating that they were not sufficiently primed towards the KC phenotype. In summary, our data argues for an unprecedented plasticity of a priori liver-resident KCs that are both an integral part of the longitudinal diversification of the tissue macrophage landscape with integration of monocyte progeny, and the mycobacterial control, leading to the restoration of homeostasis. Methods Mice C57BL/6J mice were purchased from Jackson Laboratories (USA) or Charles River Laboratories (Germany). Ccr2 -I- (B6.129S4- Ccr2 tm1lfc /J), CD45.1 (B6.SJL- Ptprc a Pepc b /BoyJ or C57BL/6J- Ptprc em6Lutzy /J), Clec4f Cre -tdTomato-NLS (C57BL/6J- Clec4f em1(cre)Glass I J), and Ms4a3 Cre mice (C57BL/6J-Ms4a3 em2(cre)Fgnx /J) were purchased from Jackson Laboratories (USA). For fate mapping, animals were either crossed to ROSA26 EYFP (B6.129X1- Gt(ROSA)26Sor tm1(EYFP)Cos I J) or ROSA26 tdTomato (B6.Cg-Gt(ROSA)26Sor tm9(CAG-tdTomato)Hze /J) mice, which were also purchased from Jackson Laboratories (USA). Mice were bred in the CEMT animal facility in Freiburg, Germany, and housed under specific pathogen-free conditions with food and water ad libitum . Day and night cycles were set to 12h. At the start of experiments mice were typically between 6-9 weeks of age and both sexes were used. All animal experiments were approved by the Federal Ministry for Nature, Environment and Consumer’s protection of the state of Baden-WOrttemberg (proposal numbers: G19/171, G20/157, G22/076). Human samples Liver biopsies were collected from individuals with diagnosed Mycobacterium tuberculosis infections and were provided by the BioMaterialBank Nord, Borstel, Germany. Two samples contained liver granulomas and were used for the analysis. Diagnosis was based on culture, staining and/ or PCR. Healthy liver tissue was archived at the Institute of Surgical Pathology, Freiburg, Germany (E 324/09-121068). All experimental procedures were approved by an ethics committee at the Universitat zu LObeck (proposal number: 2024-535). Bacterial culture M. bovis BCG was a kind gift of Prof. Dr. Dirk Wagner. BCG-BFP was generated by electroporation (2500 V, 25 µF, 1000 Ohm) of 100 µl competent BCG with 1 µg DNA of a purified BFP plasmid (Addgene plasmid # 30177; http://n2t.net/addgene:30177 ; RRID:Addgene_30177) 67 . Bacteria were grown in liquid culture of 7H9 broth with 10 % OADC and 0.5 % glycerol for 24 h. Afterwards, 50 µg/ml hygromycin were added for selection. Cultures were grown up to an OD600 of 1, pelleted and aliquoted in 7H9 broth with 10 % OADC and 15 % glycerol. Aliquots were stored at -80 °C till usage. M. avium strain 104, as well as a fluorescently labelled M. avium RFP were a kind gift of Prof. Dr. Trude Flo. 20 µg/ml kanamycin were added for selection of M. avium RFP. Infection Bacterial stocks were thawed and pulled through a 27 G syringe several times to obtain a single cell solution. Additionally, stocks were thrice sonicated for 30 s followed each round by vigorous vortexing. Mice were injected i.v. into the tail vein - if not otherwise indicated - with 10 7 cfu M. bovis BCG or M. avium in 100 µl, or received 100 µl PBS as a control injection. BCG-BFP was used for the infections unless otherwise stated. Bacterial burden in organs To determine colony forming units (cfu), livers and spleens were weighed and smashed through a 70 µm filter within 500 µl PBS. Serial dilutions of the cell suspension were plated on 7H10 agar plates and incubated at 37 °C and 5 % CO2 for up to four weeks. Bacterial counts below the detection limit were depicted as half the detection limit, and used as such for statistical analysis. Irradiation and transplantation For transplantations, bone marrow was isolated from CD45.1 mice or Clec4f Cre -tdTom- NLS: ROSA26EYFP mice which were sex and age matched to the recipient. To this end, the tibia and femur from the donor mice were removed, cleaned from tissue and flushed with PBS through a 70 µm cell strainer. Cells were pelleted by centrifugation (314 g, 7 min, 4 °C). Recipient mice were anesthetized with a ketamine-xylazine mix and livers shielded with a 2 cm wide and 1 mm thick led plate. Irradiation was performed with 9 Gray. Afterwards, 5*10 6 isolated bone marrow cells were i.v. injected. After eight weeks, mice were injected either with PBS or BCG i.v. as described above. To assess the liver chimerism, donor liver macrophages were normalized to the blood monocyte chimerism on the day of analysis. Data points were excluded in case the blood chimerism was below 14 %. In addition, two mice 4 wpi were excluded due to unusually low bacterial burdens. Tissue preparation Liver cell isolation Mice were perfused with cold PBS, livers harvested, weighed and cut into small pieces. Livers were then digested for 45 min in a horizontal shaker at 37 °C within PBS containing 10 % FBS, 10 mg/ml Collagenase IV (Worthington), and 1 mg/ml DNase I (SigmaAldrich). The cell suspension was filtered through a 70 µm cell strainer, washed with 10 % FBS in PBS and centrifuged at 300 g, 5 min, 4 °C. The supernatant was discarded and the washing step repeated. Afterwards, red blood cell lysis was performed with 1x RBC lysis buffer (ThermoFisher) according to the manufacturer’s protocol. Cells were incubated with CD16/32 for 10 min on ice, diluted with 1 % FBS, 2 mM EDTA in PBS (FACS buffer), and stained for 30 min at 4 °C for flow cytometry. Cells were acquired on a Gallios (Beckmann), LSR Fortessa (BD Bioscience) flow cytometer or used for cell sorting on a MoFlo Astrios EQ (Beckman Coulter) or Aria Fusion (BD) device. Additional gating strategies can be found in Supplemental Figure 1. Samples used for in vitro cultures, Ki-67 staining, for ATAC or single-cell RNA sequencing, were additionally processed by a 33 % percoll (Cytiva) after the digestion step. Samples were centrifuged for 12 min at 693 g without a break and the acceleration was set to 3. The supernatant was discarded and the pellet washed with FACS buffer. Afterwards, RBC lysis was performed and cells treated with CD16/32, before they were subjected to the antibody staining. KC populations 2 wpi and from PBS control mice were sorted for later Ki-67 staining. Cells were subjected to fixable viability dye eFluor 450 (Invitrogen). Afterwards, they were fixed according to the manufacturer’s manual using the Foxp3/Transcription Factor Staining Buffer Set (Invitrogen), followed by staining for Ki-67 and subsequent acquisition on a LSR Fortessa (BD Bioscience) flow cytometer. Blood cell isolation Blood was drawn from the eye of mice and collected in heparin tubes. Up to 100 µl blood were subjected twice to red blood cell lysis with the 1x RBC lysis buffer (ThermoFisher) according to the manufacture’s protocol. Cell suspensions were then incubated with CD16/32 for 10 min on ice and stained for flow cytometry. Gating strategies can be found in Supplemental Figure 1. Kupffer cell culture KCs from untreated Clec4f Cre -tdTomato-NLS : ROSA26 EYFP mice, or KC low , KC high , and TIM-4 - KCs from mice 4 wpi were isolated and sorted as described above and used for culture experiments. 96-well U-bottom plates (Falcon) were coated with 75 µl of 1:200 diluted Geltrex TM Reduced Growth Factor Basement Membrane Matrix (ThermoFisher) within Dulbecco’s Modified Eagle Medium (DMEM) containing GlutaMAX (ThermoFisher). The matrix was hardened for 2 h at 37 °C and 5 % CO2 and afterwards washed with 100 µl DMEM. Per well 20.000 cells within 200 µl medium (DMEM with 10 % FBS and 0.5 % ciprofloxacin) were plated and supplemented with 20 ng/ml M-CSF (PeproTech). Steady-state KCs were additionally treated with BCG-BFP at an MOI of 1 or 10, or left untreated. The cells were incubated at 37 °C and 5 % CO2 for up to 5 days and harvested for analysis by incubating with accutase (SigmaAldrich) for 10 min at 37 °C, followed by two times washing with FACS buffer. The cell suspension was incubated with CD16/32 for 10 min on ice, and stained with VSIG4 PE-Cy7, TIM-4 AF647, and CD45 PerCP-Cy5.5 for 30 min at 4 °C. Afterwards, steady-state KCs were stained with fixable viability dye eFluor 780 (ThermoFisher) for 20 min before acquisition on a LSR Fortessa (BD Bioscience). KCs isolated from mice after infection were stained shortly before the acquisition with DAPI (BioLegend) instead. Confocal images of KC cultures grown on 8 well Ibidi slides with 40.000 cells per well were taken on a LSM880 with a 20x (N.A. 0.8) objective. Additionally, for transcriptomic assessment, the cells were resuspended in RNA lysis buffer (QIAGEN RNeasy Micro kit TM ) containing 1 % β-mercaptoethanol after harvesting with accutase. RNA was isolated according to the manufacturer’s instruction of the QIAGEN RNeasy Plus Micro kit TM . qPCR Isolated RNA was reverse transcribed to cDNA using the SuperScript TM IV VILO TM Master Mix (Thermo Fisher) according to the manufacturer’s instruction. For quantitative PCR 5 µl of Absolute qPCR SYBR Green Mix (Thermo Fisher) were added to 0.05 µl of each primer, 2.9 µl water, and 2 µl cDNA. Samples were run on 384 well plates using a Roche Lightcycler TM . Relative gene expression was normalized to the expression of hypoxanthine-guanine phosphoribosyltransferase 1 ( Hprt1) . Primer sequences are listed in table 1 below. View this table: View inline View popup Download powerpoint Table 1: Primer sequences Histological analysis Livers and spleens were removed and fixed in 4 % PFA at 4 °C for 4 h. The samples were then transferred to 20 % sucrose till they sunk to the bottom of the tube and flash frozen with TissueTek (Sakura) in liquid nitrogen. Fixed organs were cut on a cryotome at 8 or 30 µm thickness at -20 °C. Sections were permeabilized with wash buffer (PBS containing 1 % BSA, 0.1 % Triton-X100) at 4 °C for 5-8 h. Afterwards, blocking with CD16/32 was performed for 30 min at 4 °C. The tissue slices were then incubated with antibodies and Hoechst 33342 at 4 °C overnight. Samples were washed three times with wash buffer, after which secondary antibodies were added for 1 h at room temperature. After additional washing, samples were mounted with ProLong TM Diamond Antifade Mountant (Invitrogen) or with SlowFade TM Diamond Antifade Mountant (Invitrogen) if used for 3D reconstruction. Sections were acquired on a Zeiss LSM710, LSM880, or LSM980 confocal microscope with a 20x objective (N.A. 0.8). Analysis was performed with ZEN (Zeiss, version 2012) and Fiji (version 1.53t). For 3D reconstructions, performed in IMARIS (Oxford Instruments Group, version 10.1.1), confocal images were taken on a Zeiss LSM880 with a 63x (N.A. 1.4, oil) objective. Microscopic analysis of KC subsets Livers of Clec4f Cre -tdTomato-NLS : ROSA26 EYFP mice were stained with TIM-4 AF647 and Hoechst 33342. Granulomas were outlined, and YFP + cells with Hoechst 33342 nuclei counted, both outside and within granulomas. Based on the expression of TIM-4 and tdTomato cells were assigned to the KC subsets. At least 9 pictures per mouse were evaluated. Assessment of granuloma counts in the liver For granuloma assessments, slides were stained with CD68 AF647, F4/80 AF488, and Hoechst 33342. In addition, sections 10 wpi were quenched with TrueBlack (Biotium) according to the manufacturer’s protocol before mounting. For each mouse three 8 µm thick sections, with at least 120 µm distance between them, were acquired as a tile scan with 5 % overlap on a LSM880 with a 20x (N.A. 0.8) objective. The pictures were stitched using the Hoechst 33342 signal in ZEN and analyzed with Fiji. The area of the liver was determined and a threshold for the CD68 signal was set. Afterwards, cell clusters were automatically determined with the “analyze particle” function and the settings for a minimal size of 350 µm 2 , as well as the inclusion of holes. The three sections were averaged for each mouse to represent a single value. Microscopic analysis of cell origin in granulomas For the assessment of cell origin in granulomas after BCG infection, sections from CD45.1 irradiated mice transplanted with CD45.2 ( Clec4f Cre -tdTomato-NLS : ROSA26 EYFP ) bone marrow and liver shielding were stained with CD45.1 APC, CD45.2 PE and Hoechst 33342. Single snap shots on an LSM880 with a 20x objective were taken in a blinded fashion by only using the Hoechst 33342 channel during acquisition. Analysis was performed in ZEN with the profile function. To be unbiased, the channels for CD45.1 and CD45.2 were set to the same color and two intersecting lines per granuloma were drawn to determine the MFI for the length of these lines. The first and last 15 µm of the lines were defined as “border”, while the rest was considered “inner” granuloma. The two measurements were averaged, and to obtain the ratio between inner and border granuloma, the values were divided by each other and represented as the relative MFI. MFI values of 15 granulomas were averaged for one mouse 4 wpi, while at least 10 granulomas per mouse were assessed for 10 wpi. Microscopic analysis of iNOS expression in the spleen Spleen sections from CD45.1 irradiated mice transplanted with CD45.2 ( Clec4f Cre -tdTomato- NLS : ROSA26 EYFP ) bone marrow and liver shielding were stained with CD45.1 e450, CD45.2 PE, CD68 AF647 and iNOS AF488. 10 pictures per mouse were taken on an LSM880 with a 20x objective and analyzed in Fiji. Areas of iNOS, CD45.2, and CD45.1 were determined by setting a fixed threshold for each channel. Then the overlap between iNOS and CD45.2 or iNOS and CD45.1 were evaluated with the “AND” function. Data was depicted as percentage of CD45.1 or CD45.2 overlapping iNOS area compared to the total iNOS area. Multicycle immunofluorescence staining For multicycle IF stainings, slides were blocked for 1 h at room temperature with wash buffer and then stained for 3 h at room temperature. The antibodies per cycle can be found in table 2 . Every staining round Hoechst 33342 was included to have a reference for merging the pictures later on. After staining, samples were washed three times with wash buffer and mounted with SlowFade TM Diamond Antifade Mountant (Thermo Fisher). View this table: View inline View popup Download powerpoint Table 2: Antibodies used for multicycle IF Sections were acquired on a Zeiss LSM980 confocal microscope with a 20x objective (N.A. 0.8). Afterwards, slides were demounted in PBS and destained for 20 min at room temperature with 0.002 % SDS, 20 mM NaOH in H2O under rocking and light exposure. Sections were then four times washed with H2O, mounted and acquired on the microscope as a reference for the background noise. Slides, were then subjected to a new round of staining and destaining. Pictures were analyzed with ZEN. Histological preparation of human samples Paraffin embedded patient liver samples were cut with 2 µm thickness, dried and de- paraffinized. Antigen retrieval was performed in citrate buffer for 5 min. Afterwards, the tissue was blocked in 5 % BSA in PBS for 30 min and then incubated with the primary antibodies against CD163 and VSIG4 over night at 4 °C. Antibodies were diluted in Antibody Diluent with background reducing components (Agilent Dako). Sections were washed with 1:10 diluted Dako Wash Buffer 10x (Agilent Dako) and then secondary antibodies, wheat germ agglutinin (WGA), and Hoechst 33342 were added for 45 min. Sections were washed again and mounted with ProLong TM Gold Antifade (Invitrogen) mounting medium. Additionally, serial sections were used for a hematoxylin and eosin (H&E) staining. Images were acquired on a Zeiss LSM880 confocal microscope or an Axioscan 7 (Zeiss) with a 20x objective. Analysis was performed with ZEN. Bulk RNA sequencing For bulk RNA sequencing KCs from three PBS injected mice and KC low , KC high , and TIM-4 - KCs of three mice 4 wpi were directly sorted into RNA lysis buffer containing 1 % β- mercaptoethanol. RNA was isolated with the ExtractMe Total RNA Micro Spin Kit (Blirt) according to the manufacturer’s protocol. Sequencing was performed by CeGaT (TObingen, Germany) using the NovaSeq 6000 (Illumina) system with 2x 100 bp read length with 50 million clusters/sample. The company provided trimmed reads (done with STAR version 2.7.3) which were aligned to the reference genome mm10. The data was analyzed using DESeq2 (version 1.40.2) in R (version 4.3.2) 68 . Genes with row sums below 2 were excluded. The PCA plot was produced by the plotPCA function in R. All heatmaps depict calculated Z values from normalized reads. Volcano plots were prepared in Prism TM 10 (GraphPad) by depicting log2 fold changes and the adjusted p values (padj), for which the Benjamini-Hochberg adjustment was used. Genes with a padj 2 or < -2. GO terms for differentially expressed genes with padj < 0.05, were determined with PANTHER (version 19.0, https://www.pantherdb.org/ ). Assay for transposase-accessible chromatin (ATAC) sequencing ATACseq was performed on liver macrophages from five mice 4 wpi, and four PBS control mice. The KC low and TIM-4 - population from one infected mouse were excluded due to a technical issue. Sorted KCs were frozen in RPMI containing 10 % DMSO with 10 % FBS. After thawing at 37°C, samples were centrifuged (500 g, 5 min, 4 °C) and resuspended in 250 µl OMNI buffer (10 mM Tris-HCl (pH 7.5), 10 mM NaCl, 3 mM MgCl2, 0.1 % IGEPAL-CA630 (SigmaAldrich), 0.1 % Tween-20, 0.01 % Digitonin (Promega) in DMSO). After 5 min of permeabilization, nuclei were centrifuged (500 g, 5 min, 4 °C) and the supernatant discarded. 180 µl tagmentation buffer (33 mM Tris-acetate, 66 mM K-acetate (SigmaAldrich), 11 mM Mg- acetate, 16 % N,N -Dimethylformamide (EMD Millipore)) was added without disturbing the pellet. After centrifugation (500 g, 5 min, 4 °C) the nuclei were resuspended in 15 µl cold tagmentation buffer, counted, and the concentration adjusted. 1 µl Tn5 Illumina Tagment DNA Enzyme (Illumina) was added to 20 µl nuclei suspension and incubated for 60 min with 500 rpm at 37 °C. Tagmented DNA was purified using the MinElute Reaction Cleanup Kit (Qiagen). DNA fragment amplification was performed by addition of 25 µl NEBNext High-Fidelity 2x PCR MasterMix (New England Biolabs) to 10 µl tagmented DNA, 0.7 µl 24 UDI for Tagmented libraries – Set I (Diagenode) and 14.3 µl Molecular Biology Grade Water (Corning) using 8 PCR-cycles. Amplified DNA was purified with the MinElute Reaction Cleanup Kit (Qiagen). Very small and large fragments were removed using a double-sided bead clean-up with a right side clean-up ratio of 0.55X and a left side clean-up ratio of 1.5X (AMPure XP beads, Beckmann Coulter). Fragment size distribution of final libraries was assessed with a TapeStation (Agilent). Libraries were sequenced on a NextSeq 1000 (Illumina). Processing of the sequencing data was performed with Galaxy Europe ( https://usegalaxy.eu ). First, adapters were trimmed with Cutadapt (version 4.0) and the sequences aligned to the reference genome mm10 with Bowtie2 (version 2.4.5). Properly paired reads with MAPQ ≥ 30 were kept for analysis. Mitochondrial reads and reads falling into blacklisted genomic regions for mm10 defined by ENCODE 69 were excluded. Duplicates were removed with Picard’s MarkDuplicates (version 2.18.2.3) tool with a lenient validation stringency. Peaks were called using MACS2 (version 2.2.7.1). Downstream analysis was performed in R (version 4.3.2) with DESeq2 (version 1.40.2) and batch correction, adjusting the function with “design = ∼ batch + group”. Genes with row sums below 10 were excluded. Chromatin accessible regions were depicted with the Integrative Genomics Viewer 70 (version 2.16.2). To assess the associated genomic regions and generate gene ontology terms, the online tool GREAT (version 4.0.4, https://great.stanford.edu/public/html/ ) was used with the species assembly mm10 for genes with padj 0.5. The settings were “basal plus extension” with 5 kbp upstream, 1 kbp downstream, and a maximal extension of 1000 kbp. To contrast a single group vs. all others and derive gene ontology terms, DESeq2 was rerun with the new definition of groups. Volcano plots were prepared in Prism TM 10 (GraphPad) by depicting log2 fold changes and the adjusted p values (padj), for which the Benjamini-Hochberg adjustment was used. Genes with a padj 2 or < -2. Peak visualization was created using the Integrative Genomics Viewer. Significantly differentially accessible regions were calculated with DESeq2 for a single group vs. all others and depicted in blue for significantly downregulated and in red for significantly upregulated accessible regions. Single-cell RNA sequencing For the single-cell RNA sequencing (scRNAseq), liver cells of three mice per group (PBS, 4 or 10 weeks after infection) were prepared as described above. During the antibody staining step, each sample was additionally incubated by a specific TotalSeq TM hashtag antibody (BioLegend), contained in table 3 , to allow for multiplexing. View this table: View inline View popup Download powerpoint Table 3: Hashtag antibodies per sample Cells were sorted according to the strategy in Extended Figure 5A and afterwards 30.000-100.000 cells of the same population from different mice were pooled. For the TIM-4 - KC population 4 wpi one mouse was excluded due to low cell numbers. ScRNAseq was performed using the Chromium Next GEM Single Cell 5’ Kit v2, and libraries were constructed with the Library Construction Kit by 10X Genomics. Samples were sequenced on a NovaSeq6000 (Illumina). Fastq files were aligned with the CellRanger software (7.2.0) with prebuilt mouse mm10 reference and downstream analysis was carried out in R (version 4.3.2) with the Seurat package (version 5.0.1) 71 . For demultiplexing, HTODemux with a positive quantile threshold of 0.99 was employed and only singlets were kept for further processing. Cells with more than 5 % mitochondrial transcripts, and less than 200 or more than 5000 detectable genes were excluded. The datasets from the different samples were subsequently merged, and subjected to the NormalizeData(), FindVariableFeatures(), ScaleData(), and RunPCA() algorithm with default parameters. For correction of batch effects, the data was integrated with harmony 72 . FindNeighbours and FindClusters were applied with the top 30 harmony dimensions (determined by the ElbowPlot() function) and a resolution of 0.6. To depict the data RunUMAP() was calculated for the harmony reduction and plotted with DimPlot. Based on the clustering and gene expression patterns, contaminating clusters that did not contain macrophages or monocytes were excluded. The total remaining 48,500 cells were reclustered with the functions FindNeighbours and FindClusters, using the top 30 harmony dimensions and a resolution of 0.7. In addition, cluster 7 was further subsetted and reclustered using the top 30 harmony dimensions and a resolution of 0.3. For the aggregated counts of the term “efferocytosis”, the function AddModuleScore() with the genes in Table 4 was used, identified with Mouse Genome Informatics ( https://www.informatics.jax.org/ ) and present in the scRNAseq data. View this table: View inline View popup Download powerpoint Table 4: Genes used for the term efferocytosis Differentially expressed genes were determined with the FindAllMarkers() function. Pathway enrichment analysis was performed with DEenrichRPlot including a maximum of 2000 genes comparing cluster 7 to either cluster 0, 6 or 9. Otherwise the standard settings were used and the enrich.database was set to GO_Biological_Process_2023. Cell communication was determined with the CellChat package (version 2.1.1), separated for the controls, 4 wpi, and 10 wpi samples. The functions identifyOverExpressedGenes(), identifyOverExpressedInteractions(), and computeCommunProb() were employed, flowed by a filtering using filterCommunication() with a minimum of 10 cells. computeCommunProbPathway() was calculated before using aggregateNet() and depiction of the data in a heatmap or chord diagram. For pseudobulk analysis, AggregateExpression was calculated for PBS and 4 wpi samples of cluster 1. FindMarkers with DESeq2 was used to determine differentially expressed genes. GO terms for biological process and molecular function were determined with PANTHER (version 19.0, https://www.pantherdb.org/ ) for genes with padj < 0.05. Heatmap was plotted with the 50 most upregulated and downregulated genes with padj < 0.05. RNA velocity was calculated using the scVelo (0.3.2) and scanpy (1.9.8) packages in Python (3.12.2) 73 . In a first step, the data was filtered and normalized, genes that did not have at least 20 counts for the spliced and unspliced layer were excluded. The top 2000 variable features were picked and further processed with the function sc.pp.neighbors() and scv.pp.moments(). For the trajectory analysis the dynamical model was calculated with scv.tl.recover_dynamics()and scv.tl.velocity for the macrophage clusters (clusters 1, 3, 4, 5, 6, 7, 8, 12, 13, 15) individually in each condition. To avoid confounding effects by the different batches, mouse 1 in the PBS condition was excluded. Schematics and figures Schematics were created in BioRender.com (Henneke, P. (2025): https://BioRender.com/y64d004 and https://BioRender.com/c51o935 ) and figures were created using both Prism TM 10 (GraphPad) and Inkscape (version 1.3.2). Statistical analysis Statistical analysis was performed in Prism TM 10 (GraphPad). The usage of the statistical test is indicated in the figure legends. Differences were considered as significant if p-values were ≤ 0.05. ns, not significant, * P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001, **** P ≤ 0.0001, unless otherwise indicated. Data is depicted as mean ± SEM unless otherwise specified. Antibodies View this table: View inline View popup Table 5: Anti-mouse antibodies used in this study View this table: View inline View popup Download powerpoint Table 6: Anti-human antibodies used in this study Contributions PH and JN conceptualized the study and wrote the manuscript. JN performed most of the experiments and prepared the figures. FL helped with the analysis and provided critical intellectual input. SW performed the ATAC sequencing under supervision of SP and helped with the analysis. MG helped with the establishment of the Kupffer cell culture. AKL provided critical intellectual input. DO helped with the trajectory analysis under the supervision of S. S performed the scRNA sequencing and helped with its analysis. VG performed retro-orbital infections. TG provided human samples of patients infected with Mtb and performed the H&E staining. MR and CS established the immunofluorescence staining of the human livers and CS confirmed granuloma presence. All co-authors have read and edited the manuscript. Acknowledgements We are very grateful for expert technical assistance of Katja Grawe, Anita Imm, Reem Alsumati, and Adriana Greco. We are indebted to the Lighthouse Core Facility at the Center for Chronic Immunodeficiency of the Medical center at the University of Freiburg, as well as the Center for Experimental Models and Transgenic Services in Freiburg for their instrumental technical assistance in flow cytometry, microscopy, and mouse work. The Lighthouse Core Facility is funded in part by the Medical Faculty, University of Freiburg (2023/A2-Fol; 2023/B3- Fol) and the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) (450392965). Additionally, we like to thank the BioMaterialBank Nord for providing human samples. The BioMaterialBank Nord is supported by the German Center for Lung Research. The BioMaterialBank Nord is member of popgen 2.0 network (P2N). FL and AKL were recipients of the Clinician scientist program IMM-PACT stipend, funded by the DFG (413517907). CS has received funding through the SFB1453 (431984000), SCHE 2092/4-1 (RP9, CP2, CP3) (241702976 and 438496892), SFB1160 (project-ID 256073931), and the Heisenberg program (501370692). Funding to SP was provided by the DFG (491676693 TRR 359 - PILOT). Sagar is supported by the Department of Medicine II, Freiburg University Medical Center, Faculty of Medicine, University of Freiburg and by the DFG (491676693 TRR 359 - PILOT). PH has received funding by the DFG (283781347, 254895677, 446317895-SPP EXIT, 259373024 - TRR 167 NeuroMac, and 491676693 TRR 359 - PILOT). References ↵ Ginhoux , F. & Guilliams , M . Tissue-Resident Macrophage Ontogeny and Homeostasis . lmmunity 44 , 439 – 449 , doi: 10.1016/j.immuni.2016.02.024 ( 2016 ). OpenUrl CrossRef PubMed ↵ Li , W. , Chang , N. & Li , L. Heterogeneity and Function of Kupffer Cells in Liver Injury . Front lmmunol 13 , 940867 , doi: 10.3389/fimmu.2022.940867 ( 2022 ). OpenUrl CrossRef ↵ Bennett , H. , Troutman , T. D. , Sakai , M. & Glass , C. K . Epigenetic Regulation of Kupffer Cell Function in Health and Disease . Front lmmunol 11 , 609618 , doi: 10.3389/fimmu.2020.609618 ( 2020 ). OpenUrl CrossRef ↵ Bonnardel, J. et al. Stellate Cells, Hepatocytes , and Endothelial Cells Imprint the Kupffer Cell Identity on Monocytes Colonizing the Liver Macrophage Niche . lmmunity 51 , 638 – 654 e639 , doi: 10.1016/j.immuni.2019.08.017 ( 2019 ). OpenUrl CrossRef PubMed ↵ Li , W. , Yang , Y. , Yang , L. , Chang , N. & Li , L . Monocyte-derived Kupffer cells dominate in the Kupffer cell pool during liver injury . Cell Rep 42 , 113164 , doi: 10.1016/j.celrep.2023.113164 ( 2023 ). OpenUrl CrossRef PubMed ↵ Scott, C. L. et al. Bone marrow-derived monocytes give rise to self-renewing and fully differentiated Kupffer cells . Nat Commun 7 , 10321 , doi: 10.1038/ncomms10321 ( 2016 ). OpenUrl CrossRef PubMed ↵ Zhao, D. et al. ALK1 signaling is required for the homeostasis of Kupffer cells and prevention of bacterial infection . J Clin lnvest 132 , doi: 10.1172/JCI150489 ( 2022 ). OpenUrl CrossRef ↵ Sakai, M. et al. Liver-Derived Signals Sequentially Reprogram Myeloid Enhancers to Initiate and Maintain Kupffer Cell Identity . lmmunity 51 , 655 – 670 e658 , doi: 10.1016/j.immuni.2019.09.002 ( 2019 ). OpenUrl CrossRef PubMed ↵ Cambier , C. J. , Falkow , S. & Ramakrishnan , L . Host evasion and exploitation schemes of Mycobacterium tuberculosis . Cell 159 , 1497 – 1509 , doi: 10.1016/j.cell.2014.11.024 ( 2014 ). OpenUrl CrossRef PubMed Web of Science ↵ Egen, J. G. et al. Macrophage and T cell dynamics during the development and disintegration of mycobacterial granulomas . lmmunity 28 , 271 – 284 , doi: 10.1016/j.immuni.2007.12.010 ( 2008 ). OpenUrl CrossRef PubMed Web of Science ↵ Comas, I. et al. Out-of-Africa migration and Neolithic coexpansion of Mycobacterium tuberculosis with modern humans . Nat Genet 45 , 1176 – 1182 , doi: 10.1038/ng.2744 ( 2013 ). OpenUrl CrossRef PubMed ↵ Awuh , J. A. & Flo , T. H . Author Correction: Molecular basis of mycobacterial survival in macrophages . Cell Mol Life Sci 75 , 161 , doi: 10.1007/s00018-017-2683-x ( 2018 ). OpenUrl CrossRef ↵ Chandra , P. , Grigsby , S. J. & Philips , J. A . Immune evasion and provocation by Mycobacterium tuberculosis . Nat Rev Microbiol 20 , 750 – 766 , doi: 10.1038/s41579-022-00763-4 ( 2022 ). OpenUrl CrossRef PubMed ↵ Cohen, S. B. et al. Alveolar Macrophages Provide an Early Mycobacterium tuberculosis Niche and Initiate Dissemination . Cell Host Microbe 24 , 439 – 446 e434 , doi: 10.1016/j.chom.2018.08.001 ( 2018 ). OpenUrl CrossRef PubMed ↵ Pagan , A. J. & Ramakrishnan, L. The Formation and Function of Granulomas . Annu Rev lmmunol 36 , 639 – 665 , doi: 10.1146/annurev-immunol-032712-100022 ( 2018 ). OpenUrl CrossRef PubMed ↵ Ramakrishnan , L . Revisiting the role of the granuloma in tuberculosis . Nat Rev lmmunol 12 , 352 – 366 , doi: 10.1038/nri3211 ( 2012 ). OpenUrl CrossRef PubMed ↵ Cronan , M. R . In the Thick of It: Formation of the Tuberculous Granuloma and Its Effects on Host and Therapeutic Responses . Front lmmunol 13 , 820134 , doi: 10.3389/fimmu.2022.820134 ( 2022 ). OpenUrl CrossRef PubMed ↵ Ehlers, S. et al. NOS2-derived nitric oxide regulates the size, quantity and quality of granuloma formation in Mycobacterium avium-infected mice without affecting bacterial loads . lmmunology 98 , 313 – 323 , doi: 10.1046/j.1365-2567.1999.00875.x ( 1999 ). OpenUrl CrossRef PubMed Web of Science ↵ Gharun, K. et al. Mycobacteria exploit nitric oxide-induced transformation of macrophages into permissive giant cells . EMBO Rep 18 , 2144 – 2159 , doi: 10.15252/embr.201744121 ( 2017 ). OpenUrl Abstract / FREE Full Text ↵ Yang, C. Y. et al. CLEC4F is an inducible C-type lectin in F4/80-positive cells and is involved in alpha-galactosylceramide presentation in liver . PLoS One 8 , e65070 , doi: 10.1371/journal.pone.0065070 ( 2013 ). OpenUrl CrossRef PubMed ↵ To , K. , Cao , R. , Yegiazaryan , A. , Owens , J. & Venketaraman , V . General Overview of Nontuberculous Mycobacteria Opportunistic Pathogens: Mycobacterium avium and Mycobacterium abscessus . J Clin Med 9 , doi: 10.3390/jcm9082541 ( 2020 ). OpenUrl CrossRef PubMed ↵ Niki, T. et al. Comparison of glial fibrillary acidic protein and desmin staining in normal and CCl4-induced fibrotic rat livers . Hepatology 23 , 1538 – 1545 , doi: 10.1002/hep.510230634 ( 1996 ). OpenUrl CrossRef PubMed Web of Science ↵ Niu, Q. et al. Characterization of pathological features and immune microenvironment in hepatic tuberculosis and pulmonary tuberculosis . Front Cell lnfect Microbiol 14 , 1418225 , doi: 10.3389/fcimb.2024.1418225 ( 2024 ). OpenUrl CrossRef ↵ Roca Suarez , A. A. et al. Protocol for isolating CD163(+) Kupffer cells from human liver resections . STAR Protoc 5 , 103359 , doi: 10.1016/j.xpro.2024.103359 ( 2024 ). OpenUrl CrossRef ↵ Tran, S. et al. Impaired Kupffer Cell Self-Renewal Alters the Liver Response to Lipid Overload during Non-alcoholic Steatohepatitis . lmmunity 53 , 627 – 640 e625 , doi: 10.1016/j.immuni.2020.06.003 ( 2020 ). OpenUrl CrossRef PubMed ↵ Beattie, L. et al. Bone marrow-derived and resident liver macrophages display unique transcriptomic signatures but similar biological functions . J Hepatol 65 , 758 – 768 , doi: 10.1016/j.jhep.2016.05.037 ( 2016 ). OpenUrl CrossRef PubMed ↵ Liu, Z. et al. Fate Mapping via Ms4a3-Expression History Traces Monocyte-Derived Cells . Cell 178 , 1509 – 1525 e1519 , doi: 10.1016/j.cell.2019.08.009 ( 2019 ). OpenUrl CrossRef PubMed ↵ Nascimento, M. et al. Ly6Chi monocyte recruitment is responsible for Th2 associated host-protective macrophage accumulation in liver inflammation due to schistosomiasis . PLoS Pathog 10 , e1004282 , doi: 10.1371/journal.ppat.1004282 ( 2014 ). OpenUrl CrossRef PubMed ↵ Mass, E. et al. Specification of tissue-resident macrophages during organogenesis . Science 353 , doi: 10.1126/science.aaf4238 ( 2016 ). OpenUrl Abstract / FREE Full Text ↵ Chen, Y. et al. Mycobacterium tuberculosis/Mycobacterium bovis triggered different variations in lipid composition of Bovine Alveolar Macrophages . Sci Rep 12 , 13115 , doi: 10.1038/s41598-022-17531-2 ( 2022 ). OpenUrl CrossRef ↵ Gago , G. , Diacovich , L. & Gramajo , H . Lipid metabolism and its implication in mycobacteria-host interaction . Curr Opin Microbiol 41 , 36 – 42 , doi: 10.1016/j.mib.2017.11.020 ( 2018 ). OpenUrl CrossRef PubMed ↵ Losslein, A. K. et al. Monocyte progenitors give rise to multinucleated giant cells . Nat Commun 12 , 2027 , doi: 10.1038/s41467-021-22103-5 ( 2021 ). OpenUrl CrossRef ↵ Remot , A. , Doz , E. & Winter , N . Neutrophils and Close Relatives in the Hypoxic Environment of the Tuberculous Granuloma: New Avenues for Host-Directed Therapies? Front lmmunol 10 , 417 , doi: 10.3389/fimmu.2019.00417 ( 2019 ). OpenUrl CrossRef PubMed ↵ Via, L. E. et al. Tuberculous granulomas are hypoxic in guinea pigs, rabbits, and nonhuman primates. lnfect lmmun 76 , 2333 – 2340 , doi: 10.1128/IAI.01515-07 ( 2008 ). OpenUrl Abstract / FREE Full Text ↵ Domingo-Gonzalez, R. et al. Interleukin-17 limits hypoxia-inducible factor 1alpha and development of hypoxic granulomas during tuberculosis . JCl lnsight 2 , doi: 10.1172/jci.insight.92973 ( 2017 ). OpenUrl CrossRef ↵ Russell , D. G. , Cardona , P. J. , Kim , M. J. , Allain , S. & Altare , F . Foamy macrophages and the progression of the human tuberculosis granuloma . Nat lmmunol 10 , 943 – 948 , doi: 10.1038/ni.1781 ( 2009 ). OpenUrl CrossRef PubMed Web of Science ↵ Vesosky , B. , Rottinghaus , E. K. , Stromberg , P. , Turner , J. & Beamer , G . CCL5 participates in early protection against Mycobacterium tuberculosis . J Leukoc Biol 87 , 1153 – 1165 , doi: 10.1189/jlb.1109742 ( 2010 ). OpenUrl CrossRef PubMed Web of Science ↵ Shanmuganathan, G. et al. Role of Interferons in Mycobacterium tuberculosis Infection . Clin Pract 12 , 788 – 796 , doi: 10.3390/clinpract12050082 ( 2022 ). OpenUrl CrossRef PubMed ↵ Horst , A. K. , Tiegs , G. & Diehl , L . Contribution of Macrophage Efferocytosis to Liver Homeostasis and Disease . Front lmmunol 10 , 2670 , doi: 10.3389/fimmu.2019.02670 ( 2019 ). OpenUrl CrossRef PubMed ↵ Martin, C. J. et al. Efferocytosis is an innate antibacterial mechanism . Cell Host Microbe 12 , 289 – 300 , doi: 10.1016/j.chom.2012.06.010 ( 2012 ). OpenUrl CrossRef PubMed Web of Science ↵ Gharun, K. et al. Mycobacteria exploit nitric oxide-induced transformation of macrophages into permissive giant cells . EMBO Rep 19 , doi: 10.15252/embr.201847190 ( 2018 ). OpenUrl FREE Full Text ↵ Han , H. S. , Kang , G. , Kim , J. S. , Choi , B. H. & Koo , S. H . Regulation of glucose metabolism from a liver-centric perspective . Exp Mol Med 48 , e218 , doi: 10.1038/emm.2015.122 ( 2016 ). OpenUrl CrossRef PubMed ↵ Kajiwara, C. et al. Apoptosis Inhibitor of Macrophages Contributes to the Chronicity of Mycobacterium avium Infection by Promoting Foamy Macrophage Formation . J lmmunol 210 , 431 – 441 , doi: 10.4049/jimmunol.2200306 ( 2023 ). OpenUrl CrossRef PubMed ↵ Jia, J. et al. Galectin-3 Coordinates a Cellular System for Lysosomal Repair and Removal . Dev Cell 52 , 69 – 87 e68 , doi: 10.1016/j.devcel.2019.10.025 ( 2020 ). OpenUrl CrossRef PubMed ↵ Chauhan, S. et al. TRIMs and Galectins Globally Cooperate and TRIM16 and Galectin-3 Co-direct Autophagy in Endomembrane Damage Homeostasis . Dev Cell 39 , 13 – 27 , doi: 10.1016/j.devcel.2016.08.003 ( 2016 ). OpenUrl CrossRef PubMed ↵ Aktories, P. et al. An improved organotypic cell culture system to study tissue-resident macrophages ex vivo . Cell Rep Methods 2 , 100260 , doi: 10.1016/j.crmeth.2022.100260 ( 2022 ). OpenUrl CrossRef ↵ Li, X. et al. A conserved pathway of transdifferentiation in murine Kupffer cells . Eur J lmmunol 51 , 2452 – 2463 , doi: 10.1002/eji.202049124 ( 2021 ). OpenUrl CrossRef PubMed ↵ Broadley, S. P. et al. Dual-Track Clearance of Circulating Bacteria Balances Rapid Restoration of Blood Sterility with Induction of Adaptive Immunity . Cell Host Microbe 20 , 36 – 48 , doi: 10.1016/j.chom.2016.05.023 ( 2016 ). OpenUrl CrossRef PubMed ↵ Bleriot, C. et al. Liver-resident macrophage necroptosis orchestrates type 1 microbicidal inflammation and type-2-mediated tissue repair during bacterial infection . lmmunity 42 , 145 – 158 , doi: 10.1016/j.immuni.2014.12.020 ( 2015 ). OpenUrl CrossRef PubMed Lai, S. M. et al. Organ-Specific Fate, Recruitment , and Refilling Dynamics of Tissue- Resident Macrophages during Blood-Stage Malaria . Cell Rep 25 , 3099 – 3109 e3093 , doi: 10.1016/j.celrep.2018.11.059 ( 2018 ). OpenUrl CrossRef PubMed ↵ Hirako, I. C. et al. Uptake of Plasmodium chabaudi hemozoin drives Kupffer cell death and fuels superinfections . Sci Rep 12 , 19805 , doi: 10.1038/s41598-022-23858-7 ( 2022 ). OpenUrl CrossRef ↵ Huang , L. , Nazarova , E. V. , Tan , S. , Liu , Y. & Russell , D. G . Growth of Mycobacterium tuberculosis in vivo segregates with host macrophage metabolism and ontogeny . J Exp Med 215 , 1135 – 1152 , doi: 10.1084/jem.20172020 ( 2018 ). OpenUrl Abstract / FREE Full Text ↵ Lohrmann, F. et al. Tissue imprinting defines functional mosaic of dermal macrophages . bioRxiv , 2025.2001.2009.631670 , doi: 10.1101/2025.01.09.631670 ( 2025 ). OpenUrl Abstract / FREE Full Text ↵ Pessenda, G. et al. Kupffer cell and recruited macrophage heterogeneity orchestrate granuloma maturation and hepatic immunity in visceral leishmaniasis . bioRxiv , doi: 10.1101/2024.07.09.602717 ( 2024 ). OpenUrl Abstract / FREE Full Text ↵ Peiseler, M. et al. Kupffer cell-like syncytia replenish resident macrophage function in the fibrotic liver . Science 381 , eabq5202 , doi: 10.1126/science.abq5202 ( 2023 ). OpenUrl CrossRef PubMed ↵ Merad , M. , Sathe , P. , Helft , J. , Miller , J. & Mortha , A . The dendritic cell lineage: ontogeny and function of dendritic cells and their subsets in the steady state and the inflamed setting . Annu Rev lmmunol 31 , 563 – 604 , doi: 10.1146/annurev-immunol-020711-074950 ( 2013 ). OpenUrl CrossRef PubMed Web of Science ↵ Kim , H. & Shin , S. J . Pathological and protective roles of dendritic cells in Mycobacterium tuberculosis infection: Interaction between host immune responses and pathogen evasion . Front Cell lnfect Microbiol 12 , 891878 , doi: 10.3389/fcimb.2022.891878 ( 2022 ). OpenUrl CrossRef ↵ Harding , J. S. , Rayasam , A. , Schreiber , H. A. , Fabry , Z. & Sandor , M . Mycobacterium-Infected Dendritic Cells Disseminate Granulomatous Inflammation . Sci Rep 5 , 15248 , doi: 10.1038/srep15248 ( 2015 ). OpenUrl CrossRef PubMed ↵ De Ponti, F. F. et al. Spatially restricted and ontogenically distinct hepatic macrophages are required for tissue repair . lmmunity , doi: 10.1016/j.immuni.2025.01.002 ( 2025 ). OpenUrl CrossRef ↵ Kaufmann, E. et al. BCG Educates Hematopoietic Stem Cells to Generate Protective Innate Immunity against Tuberculosis . Cell 172 , 176 – 190 e119 , doi: 10.1016/j.cell.2017.12.031 ( 2018 ). OpenUrl CrossRef PubMed ↵ Scott , H. M. & Flynn , J. L. Mycobacterium tuberculosis in chemokine receptor 2- deficient mice: influence of dose on disease progression . lnfect lmmun 70 , 5946- 5954 , doi: 10.1128/IAI.70.11.5946-5954.2002 ( 2002 ). OpenUrl Abstract / FREE Full Text ↵ Rachwal, N. et al. Pathogen and host determinants of extrapulmonary tuberculosis among 1035 patients in Frankfurt am Main, Germany, 2008-2023 . Clin Microbiol lnfect , doi: 10.1016/j.cmi.2024.11.009 ( 2024 ). OpenUrl CrossRef ↵ Hickey , A. J. , Gounder , L. , Moosa , M. Y. & Drain , P. K . A systematic review of hepatic tuberculosis with considerations in human immunodeficiency virus co-infection . BMC lnfect Dis 15 , 209 , doi: 10.1186/s12879-015-0944-6 ( 2015 ). OpenUrl CrossRef ↵ Bissell , D. M. , Wang , S. S. , Jarnagin , W. R. & Roll , F. J . Cell-specific expression of transforming growth factor-beta in rat liver . Evidence for autocrine regulation of hepatocyte proliferation. J Clin lnvest 96 , 447 – 455 , doi: 10.1172/JCI118055 ( 1995 ). OpenUrl CrossRef PubMed Web of Science Roth , S. , Gong , W. & Gressner , A. M . Expression of different isoforms of TGF-beta and the latent TGF-beta binding protein (LTBP) by rat Kupffer cells . J Hepatol 29 , 915 – 922 , doi: 10.1016/s0168-8278(98)80119-0 ( 1998 ). OpenUrl CrossRef PubMed Web of Science ↵ Wen , Y. , Lambrecht , J. , Ju , C. & Tacke , F . Hepatic macrophages in liver homeostasis and diseases-diversity, plasticity and therapeutic opportunities . Cell Mol lmmunol 18 , 45 – 56 , doi: 10.1038/s41423-020-00558-8 ( 2021 ). OpenUrl CrossRef PubMed ↵ Takaki , K. , Davis , J. M. , Winglee , K. & Ramakrishnan , L . Evaluation of the pathogenesis and treatment of Mycobacterium marinum infection in zebrafish . Nat Protoc 8 , 1114 – 1124 , doi: 10.1038/nprot.2013.068 ( 2013 ). OpenUrl CrossRef PubMed ↵ Love , M. I. , Huber , W. & Anders , S . Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2 . Genome Biol 15 , 550 , doi: 10.1186/s13059-014-0550-8 ( 2014 ). OpenUrl CrossRef PubMed ↵ Amemiya , H. M. , Kundaje , A. & Boyle , A. P . The ENCODE Blacklist: Identification of Problematic Regions of the Genome . Sci Rep 9 , 9354 , doi: 10.1038/s41598-019-45839-z ( 2019 ). OpenUrl CrossRef PubMed ↵ Robinson, J. T. et al. Integrative genomics viewer . Nat Biotechnol 29 , 24 – 26 , doi: 10.1038/nbt.1754 ( 2011 ). OpenUrl CrossRef PubMed Web of Science ↵ Hao, Y. et al. Dictionary learning for integrative, multimodal and scalable single-cell analysis . Nat Biotechnol 42 , 293 – 304 , doi: 10.1038/s41587-023-01767-y ( 2024 ). OpenUrl CrossRef PubMed ↵ Korsunsky, I. et al. Fast, sensitive and accurate integration of single-cell data with Harmony . Nat Methods 16 , 1289 – 1296 , doi: 10.1038/s41592-019-0619-0 ( 2019 ). OpenUrl CrossRef PubMed ↵ Bergen , V. , Lange , M. , Peidli , S. , Wolf , F. A. & Theis , F. J . Generalizing RNA velocity to transient cell states through dynamical modeling . Nat Biotechnol 38 , 1408 – 1414 , doi: 10.1038/s41587-020-0591-3 ( 2020 ). OpenUrl CrossRef PubMed View the discussion thread. Back to top Previous Next Posted February 08, 2025. Download PDF Supplementary Material 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 Mycobacterial infection uncovers plasticity of Kupffer cells 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 Mycobacterial infection uncovers plasticity of Kupffer cells Jana Neuber , Florens Lohrmann , Samuel Wald , Merve Göçer , Anne Kathrin Lösslein , David Obwegs , Vitka Gres , Torsten Goldmann , Manuel Rogg , Christoph Schell , Sebastian Preißl , Sagar , Philipp Henneke bioRxiv 2025.02.07.636999; doi: https://doi.org/10.1101/2025.02.07.636999 Share This Article: Copy Citation Tools Mycobacterial infection uncovers plasticity of Kupffer cells Jana Neuber , Florens Lohrmann , Samuel Wald , Merve Göçer , Anne Kathrin Lösslein , David Obwegs , Vitka Gres , Torsten Goldmann , Manuel Rogg , Christoph Schell , Sebastian Preißl , Sagar , Philipp Henneke bioRxiv 2025.02.07.636999; doi: https://doi.org/10.1101/2025.02.07.636999 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 Immunology Subject Areas All Articles Animal Behavior and Cognition (7642) Biochemistry (17715) Bioengineering (13907) Bioinformatics (42003) Biophysics (21470) Cancer Biology (18624) Cell Biology (25533) Clinical Trials (138) Developmental Biology (13390) Ecology (19935) Epidemiology (2067) Evolutionary Biology (24356) Genetics (15617) Genomics (22529) Immunology (17753) Microbiology (40432) Molecular Biology (17200) Neuroscience (88681) Paleontology (667) Pathology (2840) Pharmacology and Toxicology (4828) Physiology (7653) Plant Biology (15161) Scientific Communication and Education (2046) Synthetic Biology (4304) Systems Biology (9826) Zoology (2271)