Immunity ArticleAdrenergic Nerves GovernCircadian Leukocyte Recruitment to TissuesChristoph Scheiermann,1,3 Yuya Kunisaki,1,3 Daniel Lucas,1 Andrew Chow,1,2 Jung-Eun Jang,1,2 Dachuan Zhang,1Daigo Hashimoto,2 Miriam Merad,2 and Paul S. Frenette1,2,*1Ruth L. and David S. Gottesman Institute for Stem Cell Biology and Regenerative Medicine, Albert Einstein College of Medicine, New York,NY 10461, USA2Department of Medicine, Mount Sinai School of Medicine, New York, NY 10029, USA3These authors contributed equally to this work*Correspondence: email@example.com http://dx.doi.org/10.1016/j.immuni.2012.05.021SUMMARY 2011; Springer, 1994; Vestweber and Blanks, 1999; Wagner and Frenette, 2008). This sequential multistep process is regulatedThe multistep sequence leading to leukocyte migra- by signals in situ from adhesion receptors and by soluble factorstion is thought to be locally regulated at the inﬂam- (e.g., cytokines and chemoattractants), thereby enabling endo-matory site. Here, we show that broad systemic thelial cells to serve as gatekeepers at the interface of bloodprograms involving long-range signals from the and tissues.sympathetic nervous system (SNS) delivered by Although leukocyte migration in inﬂammatory scenarios has been studied intensely, the regulation of leukocyte trafﬁckingadrenergic nerves regulate rhythmic recruitment of under homeostasis is less understood. Steady-state migrationleukocytes in tissues. Constitutive leukocyte adhe- of hematopoietic stem cells (HSCs) and lymphocytes in lym-sion and migration in murine bone marrow (BM) and phoid and nonlymphoid tissues has been described as part ofskeletal-muscle microvasculature ﬂuctuated with normal immunosurveillance that maximizes encounters withcircadian peak values at night. Migratory oscillations, potential pathogens (Massberg et al., 2007; Sigmundsdottiraltered by experimental jet lag, were implemented and Butcher, 2008; von Andrian and Mackay, 2000). It hasby perivascular SNS ﬁbers acting on b-adrenorecep- been assumed that similar surveillance mechanisms exist fortors expressed on nonhematopoietic cells and lead- myeloid cells whose migration to tissues exposed to the externaling to tissue-speciﬁc, differential circadian oscilla- environment (e.g., skin and gut) keeps pathogens at bay. Consti-tions in the expression of endothelial cell adhesion tutive, low-level expression of endothelial adhesion moleculesmolecules and chemokines. We showed that these probably regulates myeloid cell trafﬁcking, because mice lacking major adhesion pathways are susceptible to spontaneousrhythms have physiological consequences through bacterial infections (Bullard et al., 1996; Forlow et al., 2002; Fren-alteration of hematopoietic cell recruitment and over- ette et al., 1996). Given that leukocytes play key roles in re-all survival in models of septic shock, sickle cell vaso- generative processes, one would predict that the organism alsoocclusion, and BM transplantation. These data pro- possesses broad ‘‘housekeeping’’ programs that maintain thevide unique insights in the leukocyte adhesion integrity of all tissues, irrespective of probabilities of infection.cascade and the potential for time-based therapeu- Circadian rhythms regulate several vital biological processestics for transplantation and inﬂammatory diseases. through internal molecular clocks (Dibner et al., 2010; Green et al., 2008). Blood leukocyte numbers have long been known to exhibit circadian oscillations (Haus and Smolensky, 1999),INTRODUCTION and more recent studies have revealed that the release of HSCs and progenitor cells (HSPCs) from the BM follows similarLeukocyte recruitment is critical for combating pathogens in the ´ rhythms (Lucas et al., 2008; Mendez-Ferrer et al., 2008). Interest-periphery as well as for bone marrow (BM) repopulation after ingly, speciﬁc circadian times have been linked with the onset oftransplantation. Much progress has been made over the past acute diseases, notably in the cardiovascular system (Mullertwo decades in our understanding of the major molecular mech- et al., 1985; Willich et al., 1987). Emerging data, in turn, indicateanisms involved in leukocyte recruitment in response to an that chronic perturbations of circadian rhythms promote vas-inﬂammatory challenge. Leukocytes initially tether and roll on cular diseases (Anea et al., 2009; Brown et al., 2009). Althoughendothelial cell P- and E-selectins, allowing signals from chemo- the mechanisms are still undeﬁned, numerous studies havekines and endothelial receptors to activate leukocyte integrins demonstrated strong associations between high leukocytefor binding to intercellular cell adhesion molecule-1 (ICAM-1) counts and various ischemic vascular diseases (Coller, 2005;and vascular cell adhesion molecule-1 (VCAM-1). These high- Margolis et al., 2005). Here, we tested the hypothesis that circa-afﬁnity interactions lead to leukocyte arrest on endothelial cells dian-controlled neural signals inﬂuence leukocyte behavior andand, subsequently, diapedesis toward an inﬂammatory site or the inﬂammatory response. We show that leukocyte recruitmentfor engraftment in the BM (Butcher, 1991; Ley et al., 2007; Muller, to tissues under homeostasis was not a continuous process but290 Immunity 37, 290–301, August 24, 2012 ª2012 Elsevier Inc.
ImmunityLeukocyte Recruitment Regulated by NervesA B C D Figure 1. Circadian Oscillations of Hemato- 0.4 ** ** poietic Cell Recruitment to the Bone 10.0 ** * 10.0 Marrow Donor Gr1+ Mac1+ cellsDonor BM cells (x104) 8 30 Blood WBC counts (A) Time course of homed adoptively transferred Donor LSK cells 0.3 CFU-C homing 7.5 7.5 6 BM cells (red) and corresponding WBC numbers in 20 blood (3 103/ml, black). Light and dark cycles are 5.0 0.2 5.0 indicated by white and black bars, respectively. 4 n = 4–15. 0.1 10 2.5 (B) Quantiﬁcation of homing of donor LSK cells 2.5 2 *** (3 103) 3 hr after 5 3 106 BM cells were transferred to recipients. n = 8–9. 0.0 0 0.0 0 0.0 ZT5 ZT13 ZT5 ZT13 ZT5 ZT13 (C and D) Circadian oscillations in homing of ZT 1 5 9 13 17 21 1 5 CFU-C (in %) (C) and neutrophils (3 104) (D). n = E F G ZT5 H 9–14. 35 * 25 *** *** (E and F) Quantiﬁcations of absolute numbers of 3 Rolling Flux Fraction (%) Rolling leukocytes/min 30 rolling leukocytes in BM sinusoids (E) and the (/100 m2 vessel area) 20 Adherent BM cells 25 rolling ﬂux fraction (F). n = 33–43 vessels from 8–9 2 mice per group. 20 15 (G and H) In vivo images (G) and quantiﬁcation (H) 15 ZT13 of ﬂuorescently labeled adherent BM cells (green) 10 10 1 after adoptive transfer. BM sinusoids were identi- 5 ﬁed with R6G (red). n = 60 areas from 7 mice per 5 group. 0 0 0 *p < 0.05, **p < 0.01, ***p < 0.001. The scale ZT5 ZT13 ZT5 ZT13 ZT5 ZT13 bar represents 50 mm. See also Figure S1 and Table S1.rather exhibited circadian oscillations, and that these rhythms, dian time has a signiﬁcant impact on the rhythmic recruitment oforchestrated by the molecular clock via adrenergic nerves, can different hematopoietic cell populations to the BM.impact the outcome of disease. Rhythmic Leukocyte Recruitment to Skeletal MuscleRESULTS We next investigated whether circadian oscillations also dictated leukocyte recruitment to peripheral tissues using the cremasterHematopoietic Cell Recruitment in the BM Operates muscle as a model to study leukocyte-endothelial cell interac-under Circadian Control tions in real time. Initial investigations by whole-mount immu-Total circulating leukocyte counts oscillate in murine blood, noﬂuorescence staining of unstimulated muscle tissues forpeaking $5 hr after the onset of light (zeitgeber time, ZT5) and the pan-leukocyte marker CD45 and myeloid antigens F4/80 orexhibiting a trough at ZT13 (Figure 1A), conﬁrming previous Gr-1 revealed numerous extravascular CD45+ cells along-reports (Haus and Smolensky, 1999). These rhythms are ob- side postcapillary venules identiﬁed by expression of platelet- ´served for HSCs (Mendez-Ferrer et al., 2008) as well as all major endothelial cell adhesion molecule-1 (PECAM-1), vascular-leukocyte subsets (Figure S1A available online), but not for eryth- endothelial cadherin (VE-Cadherin), and ephrin-receptor B4rocytes or platelets (data not shown). (EphB4) (Figures 2A and S2A). Interestingly, extravascular leuko- To test whether the recruitment of leukocytes from the blood cyte numbers exhibited circadian ﬂuctuations, with CD45+F480+to tissues also exhibited circadian preferences, we initially inves- macrophages representing the predominant leukocyte popula-tigated hematopoietic cell recruitment to the BM. When we tion in muscle (81% of CD45+ cells versus 2% for neutrophils,assessed the recruitment potential of adoptively transferred data not shown), peaking at ZT13 along postcapillary venules,BM cells throughout the day, we observed signiﬁcant circadian where leukocyte inﬁltration mainly occurs (Figures 2B, 2C,oscillations in the homing of total hematopoietic cells to the S2B, and S2C), but not along arterioles or capillaries (Fig-BM, whose pattern ran in antiphase with that of blood, exhibiting ure S2D). Staining for Ki67 revealed very few positive leukocytesa peak at ZT13 and a trough in the daytime (ZT5) (Figure 1A). with no detectable circadian rhythm (data not shown), suggest-Circadian recruitment was observed for LineageÀSca-1+c-kit+ ing that the observed leukocyte increase at night was not due to(LSK) cells (Figure 1B), colony-forming units in culture (CFU-C, proliferation.Figure 1C), and neutrophils (Figure 1D). We next quantiﬁed leukocyte recruitment in real time in the To investigate the sequence of events in vivo, we visualized exteriorized cremaster muscle through MFIM after injection ofthe interactions between circulating leukocytes and BM sinu- very low doses of ﬂuorescently conjugated antibodies for identi-soidal endothelium using multichannel ﬂuorescence intravital ﬁcation of leukocyte populations in vivo (Figure 2D) (Chiang et al.,microscopy (MFIM) of the calvarial BM as previously described 2007). Under surgically induced trauma conditions, the numbers(Chiang et al., 2007; Mazo et al., 1998). The number of endoge- of adherent neutrophils and monocytes signiﬁcantly increasednous rolling leukocytes in BM sinusoids was increased at ZT13 at night, whereas lymphocyte numbers remained unchangedcompared to ZT5 (Figures 1E and 1F), as was the number of (Figures 2E–2G), and overall leukocyte rolling was not affectedadherent cells after adoptive transfer (Figures 1G and 1H and (data not shown). No circadian differences in hemodynamicFigures S1B and S1C). These ﬁndings suggested that the circa- parameters such as blood ﬂow or wall shear rates were observed Immunity 37, 290–301, August 24, 2012 ª2012 Elsevier Inc. 291
Immunity Leukocyte Recruitment Regulated by NervesA B *** Figure 2. Circadian Oscillations of Leuko- leukocytes (/50x100 m2) leukocytes (/50x100 m2) 8 cyte Recruitment to Skeletal Muscle Extravasated CD45+ ZT5 (A–C) Ex vivo images (A) and quantiﬁcations 6 (B and C) of extravasated CD45+ and F4/80+ leukocytes situated around postcapillary venules 4 as analyzed by whole-mount immunoﬂuorescence staining of cremaster-muscle tissues. n = 6. 2 (D–G) Adherent leukocyte subsets as determined by multichannel ﬂuorescence intravital micros- CD45 F4/80 merged+ PECAM-1 0 C ZT 5 ZT 13 copy. In vivo images showing antibody-stained ZT13 adherent leukocytes, (D); adherent neutrophils, (E) ** Extravasated F4/80+ 6 and n in (D); monocytes, (F) and arrows, m in (D); and lymphocytes, (G). n = 142–148 vessels quan- 4 tiﬁed from 7 mice per group. *p < 0.05, **p < 0.01, ***p < 0.001. Scale bars represent 50 mm (A) and 10 mm (D). See also Fig- 2 ure S2 and Table S1. 0 ZT5 ZT13D CD45 Gr1 F4/80 merged+BF in sorted cremasteric endothelial cells; m peak and trough values overlapped with m m n those of leukocyte recruitment (Fig- ure 3A). By contrast, and consistent with n the circadian differences observed in adhesion, but not rolling, in the muscle (Figures 2E and 2F and data not shown), Adherent CD45+Gr1-F4/80+ cellsE F G Adherent CD45+Gr1+F4/80- cells Adherent CD45+Gr1-F4/80- cells 0.8 ** 0.7 * 0.6 there was no difference in the expression (/100 m vessel segment) (/100 m vessel segment) (/100 m vessel segment) 0.6 0.5 of endothelial selectins (Sele and Selp), 0.6 0.5 0.4 Vcam1, Icam2, or Cd44 (Figures 3A and 0.4 S3E). Immunoﬂuorescence staining of 0.4 0.3 0.3 frozen tissue sections corroborated these 0.2 data at the protein level (Figures 3B and 0.2 0.2 0.1 S3F). We conﬁrmed circadian ﬂuctua- 0.1 tions of ICAM-1 on the endothelial cell 0.0 0.0 0.0 surface in vivo by quantifying the num- ZT5 ZT13 ZT5 ZT13 ZT5 ZT13 bers of intravenously (i.v.) injected ad- herent green or red ﬂuorescent micro-at either time point in BM, nor in skeletal muscle (Table S1). Thus, spheres coated with anti-ICAM-1 or isotype control antibodieseven when an inﬂammatory response is triggered, the experi- using MFIM (Figure 3C, left bars). Oscillations were also seenmental time can inﬂuence the recruitment of myeloid cells in in hematopoietic Icam1À/À radiation chimeras, excluding thetrauma-stimulated tissues. Together, these data argue for anal- potential contribution from ICAM-1-bearing blood leukocytesogous mechanisms orchestrating circadian recruitment to BM (Figure 3C, right bars). We also investigated the circadianand skeletal muscle across a wide range of hematopoietic line- expression levels of chemokines known to be involved in myeloidages and differentiation stages. cell trafﬁcking in cremasteric endothelial cells via qPCR. We detected a clear circadian rhythm for Ccl2 (Figure 3D), but notOscillations of Adhesion Molecules and Chemokines for other chemokines tested, including Ccl3, Ccl5, Cxcl1,Mediate Circadian Leukocyte Recruitment Cxcl2, and Cx3cl1 (Figure S3G). In contrast to the muscle, BMThe circadian differences in leukocyte rolling and/or adhesion in endothelial cells exhibited circadian differences in P-selectin,the BM and skeletal muscle suggest substantial diurnal changes E-selectin, and VCAM-1, but not in ICAM-1 at either the RNAin the ability of endothelial cells to recruit leukocytes. We used or the protein level (Figures 3E and S3H).transgenic mice expressing green ﬂuorescent protein (GFP) To investigate the functional relevance of ICAM-1 and CCL2under the Tie-2 promoter (Tie2-Gfp) to isolate endothelial cells ﬂuctuations in skeletal muscle, we examined the numbers of(Motoike et al., 2000). Whole-mount staining of cremaster- extravasated leukocytes in Icam1À/À mice and in mice deﬁcientmuscle and BM tissues from Tie2-Gfp mice, as well as ﬂow in the CCL2 receptor (Ccr2À/À) via whole-mount ex vivo immuno-cytometry of collagenase IV-digested tissues using antibodies ﬂuorescence. Consistent with our expression results, no circa-directed against PECAM-1 and VE-Cadherin, conﬁrmed endo- dian rhythm was apparent, demonstrating the critical require-thelial GFP expression (Figures S3A–S3D). We therefore sorted ment of these molecules in this activity (Figure 3F).the CD45ÀVE-Cadherin+PECAM-1+GFP+ cell subset for quanti- To assess the functional relevance of ﬂuctuations in adhesion-tative PCR (qPCR) analyses (Figures S3C and S3D). Interest- molecule expression in BM, we examined the role of P- andingly, we observed signiﬁcant ﬂuctuations in Icam1 expression E-selectins. MFIM of BM microvessels of SeleÀ/ÀSelpÀ/À mice292 Immunity 37, 290–301, August 24, 2012 ª2012 Elsevier Inc.
ImmunityLeukocyte Recruitment Regulated by Nerves A Selp Sele Vcam1 Icam1 ment to the BM was independent of ICAM-1, because Icam1À/À 1.5 1.5 1.5 2 *** mice exhibited no alterations in circadian oscillations in the BM Skeletal muscle ECs relative to wild-type (WT) animals (Figure S3I). mRNA (fold) 1.5 1 1 1 Taken together, these data demonstrate that tissue-speciﬁc 1 expression of key promigratory factors ﬂuctuates within the 0.5 0.5 0.5 endothelial cells of these tissues. Of importance, the differences 0.5 in their molecular signatures correlate with the oscillations 0 0 0 0 observed in their respective functions, i.e., adhesion in the cre- ZT 1 13 ZT 1 13 ZT 1 13 ZT 1 13 master muscle (ICAM-1, CCL2) and both rolling and adhesion ICAM-1 B C * D Ccl2 in the BM (endothelial selectins, VCAM-1). isotype beads/cremaster * Adherent anti-ICAM-1/ 2500 8 3.5 * 3 * mRNA (fold) 2000 6 2.5 Requirements of Local Adrenergic Nerves 1500 Most prior studies have focused on the relevance of humoral MFI 2 4 1000 1.5 factors in synchronizing circadian rhythms (Dimitrov et al., 2 1 2009; Haus, 2007). We have recently described a key role for 500 0.5 0 0 local innervation in the circadian release of HSCs from the BM 0 ZT 5 13 ZT 5 13 5 13 ZT 1 13 ´ (Mendez-Ferrer et al., 2008). To assess the function of local inner- WT hematop. Icam1-/- vation in hematopoietic cell recruitment to tissues, we adopted E two surgical approaches to denervate the cremaster muscle Selp Sele Vcam1 Icam1 2 and BM. For the cremaster muscle, we denervated mice unilater- 2.5 * 2.5 *** 3 **Bone marrow ECs 2 ally through microsection of the genitofemoral nerve (GFNx) 2 1.5 mRNA (fold) 2 (Lucio et al., 2001) that innervates this tissue (Zempoalteca 1.5 1.5 1 et al., 2002). For the BM, we sympathectomized mice through 1 1 unilateral surgical ablation of the superior cervical ganglion 1 0.5 0.5 0.5 (SCGx) (Alito et al., 1987), which resulted in ptosis on the dener- 0 0 0 0 vated side (Figure S4A), as seen in Horner’s syndrome (Walton ZT 1 13 ZT 1 13 ZT 1 13 ZT 1 13 and Buono, 2003). Denervation was conﬁrmed via whole-mount F G immunoﬂuorescence staining for the sympathetic-nerve marker *** *** Extravasated leukocytes 8 tyrosine hydroxylase (TH) (Zhou et al., 1995), which revealed (/100 µm2 vessel area) 8 Adherent BM cells marked reductions in the number of TH+ nerve ﬁbers in the neu- (/50x100 m2) 6 6 rectomized sides of the tissues compared to the sham-operated 4 4 contralateral sides (Figures 4A and 4B). Thus, these approaches allowed us to investigate the role of local innervation in nerve- 2 2 intact and denervated tissues of the same mouse. 0 0 Whereas brightﬁeld intravital microscopy (BIM) studies of the ZT 5 13 5 13 5 13 ZT 5 13 5 13 WT Icam1-/- Ccr2-/- WT Sele-/-Selp-/- cremasteric microvasculature revealed no circadian oscillations in leukocyte rolling (Figure 4C), leukocyte adhesion (Figure 4D)Figure 3. Oscillations of Promigratory Factors Mediate Circadian and extravasation (Figure 4E) were signiﬁcantly elevated at nightLeukocyte Recruitment in contralateral muscle tissues, with no oscillations observed in(A) qPCR analysis of sorted cremasteric endothelial cells (ECs) for P-selectin the denervated side. In addition, GFNx completely abolished(Selp), E-selectin (Sele), Vcam1, and Icam1. n = 12–16. the nightly increase in ICAM-1 expression (Figure S4B), suggest-(B) Quantiﬁcation of ICAM-1 protein expression in muscle by confocal ing that the nighttime surge in ICAM-1 is controlled locally byimmunoﬂuorescence imaging of frozen sections. n = 9. MFI, mean ﬂuores- adrenergic sympathetic-nerve ﬁbers. Adoptively transferredcence intensity. BM cells in SCGx animals exhibited circadian ﬂuctuations in(C) MFIM quantiﬁcation of speciﬁc (anti-ICAM-1-coated) versus nonspeciﬁc(immunoglobulin G-coated) ﬂuorescent microsphere adhesion in the cre- the numbers of adherent cells in the nerve-intact side, whereasmasteric microvasculature in WT and hematopoietic Icam1À/À/WT BM the denervated side did not show any oscillations (Figure 4F).chimeras. n = 5–6. No circadian differences were observed in hemodynamic(D) qPCR analysis of sorted cremasteric endothelial cells for Ccl2. n = 6. parameters, tissue weight, or vascular density in neurectomized(E) qPCR of sorted BM endothelial cells (ECs). n = 6–7. animals (Figures S4C–S4E and Table S2). Together, these(F) Numbers of extravasated CD45+ leukocytes as analyzed by whole-mount studies demonstrate the importance of local innervation forimmunoﬂuorescence staining of cremaster-muscle tissues in WT control,Icam1À/À, and Ccr2À/À animals. n = 6–7. circadian oscillations in hematopoietic cell recruitment to both(G) Quantiﬁcation of adherent ﬂuorescently labeled cells to BM sinusoids in WT skeletal muscle and BM.and SelpÀ/ÀSeleÀ/À mice after adoptive transfer. n = 31–75 areas from 4–8 Signals from the SNS are transmitted from the brain to periph-mice per group. eral tissues by norepinephrine through adrenoreceptors (Elen-*p < 0.05, **p < 0.01, ***p < 0.001. See also Figure S3. kov et al., 2000). To examine which adrenergic receptors were important for the circadian regulation of hematopoietic cell traf-revealed that, in these mice, circadian oscillations in leukocyte ﬁcking, we adoptively transferred ﬂuorescently labeled WTrecruitment were ablated (Figure 3G). In stark contrast, and in BM cells into b2- or b3-adrenoreceptor (Adrb2 or Adrb3)-agreement with the expression data, circadian leukocyte recruit- deﬁcient recipients. We found that circadian oscillations in cell Immunity 37, 290–301, August 24, 2012 ª2012 Elsevier Inc. 293
Immunity Leukocyte Recruitment Regulated by Nerves Cremaster muscle Bone marrow Figure 4. Requirements of Local Adrenergic NervesA Control GFNx B Control SCGx (A–F) Unilateral surgical denervation of the geni- tofemoral nerve (GFNx) or the superior cervical ganglion (SCGx). (A and B) Images of TH+ nerve ﬁbers (red) and associated vessels (PECAM-1, blue) in cremaster muscle (A) or calvarium (B) 4 weeks after GFNx or SCGx of the contralateral control and operated TH PECAM-1 sides. (/50x100 m2 vessel segment)C D E (C–E) BIM microscopy of exteriorized cremaster- * (/100 m vessel segment) Extravasated leukocytes Rolling leukocytes/min 120 1.2 2.0 * muscle tissues quantifying rolling (C), adhesion (D), Adherent leukocytes 100 1.0 and transmigration (E) after GFNx. n = 28–40 1.5 vessels from 3–4 mice per group. 80 0.8 60 0.6 (F–H) Quantiﬁcation of adherent ﬂuorescently 1.0 labeled cells to BM sinusoids after SCGx (F) or in 40 0.4 0.5 Adrb2À/À (G) or Adrb3À/À (H) mice after adoptive 20 0.2 transfer. n = 28–45 vessels from 5–6 mice per 0 0.0 0.0 group. ZT 5 13 5 13 ZT 5 13 5 13 ZT 5 13 5 13 Control GFNx Control GFNx *p < 0.05, ***p < 0.001. Scale bars represent Control GFNx 100 mm. See also Figure S4 and Table S2. F G H *** *** 4 ***(/100 µm2 vessel area) 4 8 Adherent BM cells 3 6 3 diated recipients (Figure 5C). After isopro- 2 4 2 terenol treatment, the number of recruited 1 donor BM cells and progenitor cells was 1 2 signiﬁcantly increased compared to that 0 0 ZT5 ZT13 ZT5 ZT13 0 ZT5 ZT13 ZT5 ZT13 of the control (Figures 5D and 5E), ZT5 ZT13 ZT5 ZT13 Control SCGx WT Adrb2-/- WT Adrb3-/- whereas this effect was abrogated in SelpÀ/ÀSeleÀ/À or anti-VCAM-1-treated animals (Figures 5E and 5F). Furthermore,recruitment to the BM were signiﬁcantly reduced in both recip- survival was markedly improved in the isoproterenol-treatedient strains (Figures 4G and 4H). In addition, circadian differ- animals when limiting numbers of BM cells were injected forences in the expression of BM endothelial cell adhesion mole- transplantation (Figure S5B). To determine whether this effectcules were ablated in these mice (Figures S4F and S4G). was dependent on Adrb2 or Adrb3, we treated mice with speciﬁcSimilarly, no circadian oscillations were observed in the numbers agonists (clenbuterol or BRL37344, respectively). We found thatof extravascular leukocytes in the cremaster muscle of Adrb2À/À only BRL37344 exhibited a similar effect to that of isoproterenolor Adrb3À/À animals at steady state (Figures S4H and S4I), and (Figures 5G and S5A and data not shown), indicating thatBM chimeras for both strains revealed the critical role of non- b3-adrenoreceptor signaling is sufﬁcient to promote hematopoi-hematopoietic b2- and b3-adrenoreceptors in circadian leuko- etic recruitment to the BM. BRL37344 treatment also enhancedcyte recruitment to this tissue (Figures S4J–S4L). Furthermore, the homing of long-term repopulating HSCs (LT-HSCs) (Fig-the nighttime surge in ICAM-1 expression was averted when ure 5H). Because we performed BM transplantation 24 hr aftermice were treated with the b3-adrenoreceptor-speciﬁc antago- treatment with isoproterenol or BRL37344, the enhanced recruit-nist SR59230A (Figure S4M). These data clearly indicate that ment was probably due to an effect on the stroma rather than onthe endothelial oscillations in adhesion-molecule expression the infused BM cells. We found no differences in CXCL12 levelsrequire local delivery of adrenaline and signaling through between groups at the time of BM transplantation (Figures S5Cb-adrenoreceptors. and S5D). These data suggested that transplantation performed at night or after pharmacological treatment with b3 agonistsBiological Relevance of Circadian BM Recruitment could potentially reduce the number of HSPCs needed for suc-We next examined whether the circadian regulation of hemato- cessful long-term engraftment.poietic cell recruitment to the BM could be exploited in thesetting of BM transplantation. The injection of lethally irradiated Circadian Rhythms in Leukocyte Recruitment Aremice with limiting numbers of BM cells (25,000 cells) led to Entrained by Photic Cuesmarked differences in survival. Whereas most animals trans- Circadian rhythms are entrained by light in that photic inputplanted in the morning succumbed to death due to hematopoi- is interpreted by suprachiasmatic nuclei, the fundamentaletic failure (Figure S5A), all animals survived the procedure timekeeper of the mammalian clock (Klein et al., 1991; Ralphwhen transplanted at ZT13 (Figures 5A and 5B). To test whether et al., 1990). We thus tested whether alterations in photic inputmimicking an enhanced adrenergic tone at night could emulate by induction of an experimental jet lag would be sufﬁcient tothis phenotype, we treated mice with the pan-b-adrenoreceptor modify rhythms in leukocyte recruitment. Advancing the lightagonist isoproterenol, which signiﬁcantly upregulated expres- regime by 12 hr after the light phase (Figure 6A) completely abol-sion of P- and E-selectins and VCAM-1, but not ICAM-1, in irra- ished the circadian rhythms in leukocyte recruitment to skeletal294 Immunity 37, 290–301, August 24, 2012 ª2012 Elsevier Inc.
ImmunityLeukocyte Recruitment Regulated by Nerves A C P-selectin E-selectin VCAM-1 ICAM-1 Figure 5. b-Adrenergic Stimulation 1.5 2 3.5 1.25 Enhances Hematopoietic Progenitor Cell 100 * *** 3 *** Reconstitution after TransplantationPercent survival 80 1 (A) Survival curves after circadian BM trans- ** 2.5 MFI (fold) 1 plantation (BMT) with limiting numbers of BM cells 60 2 0.75 1 (2.5 3 104) into lethally irradiated recipients. n = 10. 40 1.5 0.50 0.5 (B) Representative micrographs of hematoxylin ZT5 1 20 0.25 and eosin (H&E)-stained sections 30 days after ZT13 0.5 circadian BMT. 0 0 0 0 0 0 5 10 15 20 25 30 PBS Iso PBS Iso PBS Iso PBS Iso (C) Flow-cytometric analysis of endothelial cell Time after BMT (days) adhesion molecule expression after PBS or isoproterenol (iso) treatment in recipients. n = 5. B ZT5 D E * F 300 *** * 8 MFI, mean ﬂuorescence intensity. ** Donor BM cells (x104) 800 Donor CFU-C (%) Donor CFU-C (%) (D) Quantiﬁcation of donor BM cells that homed to 6 600 the BM of recipients treated with PBS or iso. n = 10. 200 (E and F) Quantiﬁcation of donor CFU-C that 4 400 homed to the BM of recipients treated with PBS or ZT13 100 iso in WT or SelpÀ/ÀSeleÀ/À mice (E) or after 2 200 treatment with an isotype (isot.) or anti-VCAM-1 blocking antibody (F). n = 4–9. 0 0 0 PBS Iso PBS Iso PBS Iso PBS Iso Iso (G) Survival curves after transplantation with WT Sele-/-Selp-/- Ab isot. VCAM-1 limiting numbers of BM cells (2.5 3 104) into lethally irradiated recipients pretreated with PBS G H 40 100 or BRL37344. n = 18–20. PBS * % of CD45.1 cells (H) Effect of BRL37344 on homing of LT-HSCs. Percent survival 80 30 BRL37344 *** * n = 6. 60 *p < 0.05, **p < 0.01, ***p < 0.001. The scale bar 20 * * represents 100 mm. See also Figure S5. 40 PBS 10 20 BRL37344 0 0 0 5 10 15 20 25 30 0 4 8 12 16 Time after BMT (days) circadian oscillations (Figure 6K, left Time after BMT (weeks) bars). We also observed induction of endothelial cell adhesion moleculemuscle (Figure 6B) and the oscillations in endothelial ICAM-1 expression (P- and E-selectin, VCAM-1, and Cd44) as well asexpression (Figure 6C). Experimental jet lag also eliminated the induction of chemokines (Ccl2, Ccl5, Cxcl1, and Cxcl2) afterincreased homing to the BM at night (Figure 6D). To assess TNF-a stimulation (Figures S6A–S6H), but no rhythm was detect-further the role of the molecular clock in the circadian ﬂuctua- able. Icam2, Ccl3, and Cx3cl1 levels were not altered aftertions of leukocytes, we evaluated the role of BMAL-1, a key tran- inﬂammation and did not exhibit oscillations (Figures S6I–S6K).scription factor required to entrain circadian oscillations (Levi When we induced jet lag and treated mice with TNF-a usingand Schibler, 2007). We found that circadian oscillations of the same protocol, circadian rhythms in white blood cell (WBC)leukocyte numbers in both blood and tissues were completely counts were inhibited (Figure 6I, right bars) and neutrophil inﬁltra-abolished in Bmal1À/À animals kept in constant darkness for tion was almost completely abolished (Figures 6H, right panelthree weeks, in contrast to the normal oscillations observed in and 6J, right bars), whereas the expression of endothelial cellheterozygous and WT littermates (Figures 6E–6G). These results adhesion molecules was dramatically reduced compared tounderscore the importance of light as an environmental cue controls kept under a normal light regime (Figure 6K, right barsrequired for synchronizing the clock machinery that entrains and Figures S6A–S6C). These experiments further illustrate therhythms in leukocyte recruitment to tissues. signiﬁcance of circadian time and light in leukocyte recruitment under both steady-state and inﬂammatory conditions.Circadian Time Inﬂuences Leukocyte Recruitmentin Inﬂammation Inﬂammatory Leukocyte Recruitment Oscillates inTo assess whether circadian leukocyte recruitment could alter Sickle Cell Diseasethe inﬂammatory response, we injected mice with tumor We next aimed to evaluate the relevance of circadian leukocytenecrosis factor alpha (TNF-a) and analyzed cremaster-muscle recruitment in inﬂammation in models of sickle cell diseasetissues 8 hr later, either at ZT5 or ZT13. Whereas leukocyte (SCD) and septic shock, pathologies in which neutrophil recruit-rhythms in blood were signiﬁcantly blunted after stimulation ment has been shown to play a critical role (Frenette and Atweh,(Figure 6I, left bars), neutrophil inﬁltration showed a dramatic 2007; Hewett et al., 1992; Thomas et al., 1992). SCD mice chal-increase compared to PBS-treated animals (Figures 6H, left lenged with a TNF-a-induced model of vaso-occlusion (Fig-panel and 6J, left bars). Circadian rhythms in leukocyte recruit- ure S7A) exhibited signiﬁcantly increased leukocyte recruitmentment remained intact in inﬂammation, with ZT13 showing signif- to the cremaster muscle when the experiment was performed aticantly higher neutrophil inﬁltration than ZT5 (Figures 6H, left and night compared to during the daytime (Figures S7B and S7C). In6J, left). In addition, ICAM-1 expression in endothelial cells was addition, adherent leukocytes showed marked elevations insigniﬁcantly increased after TNF-a stimulation and exhibited Mac-1-integrin activation at night (Figure S7D), a key molecule Immunity 37, 290–301, August 24, 2012 ª2012 Elsevier Inc. 295
Immunity Leukocyte Recruitment Regulated by Nerves Figure 6. Circadian Rhythms in Leukocyte Recruitment Are Entrained by Photic Cues (A) Light regime under normal and jet-lag condi- tions as entrained using a light cycler. (B) Quantiﬁcation of extravasated CD45+ leuko- cytes in the cremaster muscle as analyzed by whole-mount immunoﬂuorescence staining in mice under normal-light and jet-lag conditions. n = 3–6. (C) Quantiﬁcation of speciﬁc versus nonspeciﬁc sphere adhesion in hematopoietic Icam1À/À/WT BM chimeras under normal-light and jet-lag con- ditions. n = 5–6. (D) Quantiﬁcation of adherent ﬂuorescently labeled cells to BM sinusoids in mice under normal-light and jet-lag conditions. n = 46–61 areas quantiﬁed from 4–5 mice per group. (E–G) Circadian oscillations in Bmal1À/À mice and control littermates after 3 weeks of constant darkness. (E) Blood leukocyte counts (3 103/ml). n = 4–8. (F) Numbers of extravasated CD45+ leukocytes as analyzed by whole-mount immunoﬂuorescence staining of cremaster-muscle tissues. n = 24–30 vessels from 4 mice per group. (G) Numbers of adherent ﬂuorescently labeled cells to BM sinusoids after adoptive transfer. n = 22–31 areas quantiﬁed from 3 mice per group. (H–J) Representative images (PBS in insets) (H), blood leukocyte counts (3 103/ml) (I), and quanti- ﬁcation of TNFa-induced neutrophil inﬁltration in sections of the cremaster muscle (J) under normal-light and jet-lag conditions. n = 3–9. (K) Quantiﬁcation of ICAM-1 protein expression in frozen sections harvested from the same mice. MFI, mean ﬂuorescence intensity. CT, circadian time. *p < 0.05, **p < 0.01, ***p < 0.001. The scale bar represents 50 mm. See also Figure S6. studies using BM harvested from Vav1- Cre;Rosa26-lsl-Luc mice expressing luci- ferase selectively in hematopoietic cells. Consistent with our studies in musclemediating heterotypic interactions between activated adherent and BM, we observed strong recruitment of adoptively trans-leukocytes and circulating red blood cells (RBCs) (Hidalgo ferred cells to the liver that had a circadian rhythm peaking atet al., 2009). This translated to a $2-fold increase in WBC-RBC night (Figure 7A). These rhythms coincided with oscillationsinteractions (Figure S7E) and a signiﬁcant decrease of the in endothelial ICAM-1 and VCAM-1 expression (Figure 7B),mean venular-blood-ﬂow rates (Figure S7F). Ultimately, the over- whereas expression of endothelial selectins was unchangedall survival of SCD mice was signiﬁcantly reduced at night. (data not shown). To test further the relevance of these ﬂuctua-Lethality was due to disseminated intravascular coagulation, tions, we challenged mice with a lethal dose of endotoxin (lipo-as evidenced by microthrombi in the lungs as well as scattered polysaccharide [LPS]) in the morning or at night. Consistentfoci of liver necrosis (Figures S7G–S7I and data not shown). with our results with TNF-a, both adhesion-molecule expressionImportantly, we did not observe circadian differences in TNF- (Figure 7C) and leukocyte recruitment exhibited strong oscilla-receptor expression at the investigated time points (Figure S7J). tions, with neutrophil inﬁltration into liver peaking when LPSThese data indicate that circadian leukocyte recruitment can was administered at night (Figures 7D–7F). In contrast, diurnalexacerbate sickle cell vaso-occlusion and compromise survival. variations in neutrophil recruitment to this tissue were ablated in Icam1À/À animals (Figures 7D–7F). Similarly to TNF-receptorCircadian Time Governs Leukocyte Recruitment and expression, we also did not observe circadian rhythms of TLR4Survival in Septic Shock (the principal receptor for LPS) at these time points (data notTo gain further insight into the relevance of circadian leukocyte shown). In line with prior studies (Feigin et al., 1969; Halbergrecruitment to vital organs, we performed adoptive transfer et al., 1960), survival was greatly reduced when mice were296 Immunity 37, 290–301, August 24, 2012 ª2012 Elsevier Inc.