Vascular Microarchitecture of Murine Colitis-Associated Lymphoid Angiogenesis.код для вставкиСкачать
THE ANATOMICAL RECORD 292:621–632 (2009) Vascular Microarchitecture of Murine Colitis-Associated Lymphoid Angiogenesis ASLIHAN TURHAN,1 MIAO LIN,1 GRACE S. LEE,1 LINO F. MIELE,1 AKIRA TSUDA,2 MORITZ A. KONERDING,3 AND STEVEN J. MENTZER1* 1 Laboratory of Adaptive and Regenerative Biology, Brigham and Women’s Hospital, Harvard Medical School, Boston, Massachusetts 2 Molecular and Integrative Physiological Sciences, Harvard School of Public Health, Boston, Massachusetts 3 Department of Anatomy, Johannes Gutenberg University, Mainz, Germany ABSTRACT In permissive tissues, such as the gut and synovium, chronic inﬂammation can result in the ectopic development of anatomic structures that resemble lymph nodes. These inﬂammation-induced structures, termed lymphoid neogenesis or tertiary lymphoid organs, may reﬂect differential stromal responsiveness to the process of lymphoid neogenesis. To investigate the structural reorganization of the microcirculation involved in colonic lymphoid neogenesis, we studied a murine model of dextran sodium sulfate (DSS)-induced colitis. Standard 2-dimensional histology demonstrated both submucosal and intramucosal lymphoid structures in DSSinduced colitis. A spatial frequency analysis of serial histologic sections suggested that most intramucosal lymphoid aggregates developed de novo. Intravital microscopy of intravascular tracers conﬁrmed that the developing intramucosal aggregates were supplied by capillaries arising from the quasi-polygonal mucosal plexus. Confocal optical sections and whole mount morphometry demonstrated capillary networks (185 46 lm diameter) involving six to ten capillaries with a luminal diameter of 6.8 1.1 lm. Microdissection and angiogenesis PCR array analysis demonstrated enhanced expression of multiple angiogenic genes including CCL2, CXCL2, CXCL5, Il-1b, MMP9, and TNF within the mucosal plexus. Intravital microscopy of tracer particle ﬂow velocities demonstrated a marked decrease in ﬂow velocity from 808 901 lm/sec within the feeding mucosal plexus to 491 155 lm/sec within the capillary structures. We conclude that the development of ectopic lymphoid tissue requires signiﬁcant structural remodeling of the stromal microcirculation. A feature of permissive tissues may be the capacity for lymphoid angioC 2009 Wiley-Liss, Inc. genesis. Anat Rec, 292:621–632, 2009. V Key words: microcirculation; genesis; colitis Abbreviations used: CFSE ¼ 5-(and-6)-carboxyﬂuorescein diacetate, succinimidyl ester; 2D ¼ 2-dimensional; 3D ¼ 3-dimensional; DiI ¼ 1,1-dioctadecyl-3,3,3,3-tetramethylindocarbocyanine perchlorate; DiO ¼ 3,30 -dioctadecyloxacarbocyanine perchlorate; DMSO ¼ dimethyl sulfoxide; FITC ¼ ﬂuoroscein isothiocyanate; H and E ¼ hematoxylin and eosin; HEV ¼ high endothelial venule; Mab ¼ monoclonal antibodies; PBS ¼ phosphate buffered saline; TLO ¼ tertiary lymphoid organs. Grant sponsor: NIH; Grant numbers: HL47078 and HL75426. C 2009 WILEY-LISS, INC. V lymphoid neogenesis; angio- *Correspondence to: Steven J. Mentzer, Room 259, Brigham and Women’s Hospital, 75 Francis Street, Boston, MA 02115. E-mail: [email protected] Received 11 September 2008; Accepted 21 January 2009 DOI 10.1002/ar.20902 Published online in Wiley InterScience (www.interscience.wiley. com). 622 TURHAN ET AL. In normal circumstances, the peripheral immune system is organized into secondary lymphoid organs such as regional lymph nodes and Peyer’s patches (Picker and Butcher, 1992). In some chronic inﬂammatory diseases, anatomic structures that resemble lymph nodes with B cell follicles and T cell zones form de novo. These inﬂammation-induced ectopic lymphoid structures have been termed lymphoid neogenesis or tertiary lymphoid organs (Ruddle, 1999; Hjelmstrom, 2001). Some tissues, including the gut, have been associated with lymphoid neogenesis in inﬂammatory disease states, whereas other tissues, such as the skin, are rarely associated with ectopic lymphoid aggregates (Aloisi and Pujol-Borrell, 2006). These tissue-speciﬁc differences suggest the importance of stromal responsiveness to lymphoid neogenesis. Attempts to investigate stromal adaptation to chronic inﬂammation have largely focused on endothelial cells. Endothelial cell acquisition of adhesive and chemoattractant properties has been proposed as a mechanism for regulating lymphoid trafﬁc to inﬂamed tissues (von Andrian and Mempel, 2003). High endothelial venules (HEV), endothelial cells characterized by a cuboidal morphology, and a distinctive molecular phenotype (e.g., PNAdþ and CCL21þ), are a variable ﬁnding in chronically inﬂamed tissues (Armengol et al., 2001; Weninger et al., 2003; Barone et al., 2005; Manzo et al., 2005). Much of the variability in endothelial adhesive function and chemokine expression may reﬂect the underlying microvascular architecture. Stromal reorganization, including the adaptive structural change in the microvasculature, may provide an explanation for variability in both endothelial phenotype and tissue responsiveness. To investigate the structural reorganization of the microcirculation involved in lymphoid neogenesis, we studied a murine model of dextran sodium sulfate (DSS)-induced colitis. Both submucosal and intramucosal lymphoid aggregates were identiﬁed in the mouse colon. A frequency analysis suggested that most intramucosal lymphoid aggregates developed de novo. The developing intramucosal lymphoid aggregates were supplied by capillaries arising from the quasi-polygonal mucosal plexus. The time course of aggregate development and the gene expression within the mucosa suggest that the structural changes were the result of inﬂammationinduced sprouting angiogenesis. METHODS changed to water for the remainder of the experimental period. Clinical Colitis Score Using RFID tagging of each mouse (AVID, www.avidmicrochip.com), body weights and clinical scores were recorded daily. A modiﬁcation of a previously described method (Waidmann et al., 2002), the colitis score incorporated posture (0 normal; 1 abnormal), activity level (0 normal; 1 abnormal), rufﬂed fur (0 absent; 1 present), rectal prolapse (0 absent; 1 present), feces (0 normal; 2 liquid; 4 bloody), weight loss (0 10%; 2 ¼ 10%–20%; 4 20%). A score of less than 3 was considered minimal or no colitis, 3–5 was moderate colitis, and a score greater than 5 was severe colitis. Microvideo Endoscopy As previously described (Ravnic et al., 2007a), the microvideo endoscopy was performed using a multi-purpose rigid endoscope (KSVEA Rigid; 64018 BSA) (Karl Storz, Germany) with a 2.7 mm diameter and 18 cm length. The rigid optical system included a 30 degree wide angle forward oblique telescope. The KSVEA rigid endoscope used a 175 Watt Xenon light source. The analog video images were digitized for archiving and analysis. Stereologic Sampling Small mucosal tissue blocks (10 15 mm) were freshly excised from DSS-treated mice, embedded in OCT compound (Miles Labs, Elkhart, IN), and prepared for cryosectioning. Vertical cryosections were prepared in 7–10 lm thickness slides, stained with hematoxylin and eosin (H and E), and evaluated for lymphoid aggregates. Preliminary microscopic evaluation of each block was performed to ensure an acceptable tissue preparation. In adequately prepared specimens, the number of lymphoid aggregates was assessed based on H and E staining (Carlsen et al., 2002). Multiple sections were obtained from two parallel regions a minimum of 750 lm apart; more sections were obtained in regions of apparent conﬂuence so that discrete aggregates could be judged. The histology sections were evaluated with a 500 lm 750 lm grid projection by at least two independent observers. The mean aggregate number of these two regions was recorded for each time point. Mice C57B/6 mice (Jackson Laboratory, Bar Harbor, ME), 25–33 g, were used in all experiments. The care of the animals was consistent with guidelines of the American Association for Accreditation of Laboratory Animal Care (Bethesda, MD). Dextran Sulfate Administration In C57/B6 mice, the DSS (TdB Consultancy AB, Uppsala, Sweden) model of colitis was similar to that described previously (Okayasu et al., 1990). Brieﬂy, DSS was freshly prepared and added daily to the drinking water at a ﬁnal concentration of 3%. The mice were assessed daily for clinical signs and total body weight. The DSS treatment was continued for 6 days then Antibodies Immunohistochemistry was performed with commercially available primary antibodies used at a 1:50 concentration. The anti-CD4 (GK1.5, Abcam, Cambridge, UK), anti-CD19 (1D3, BD Pharmigen), anti-CD11b (M1/ 70, BD Pharmigen, San Jose, CA), F4/80 (CI:A3-1, BD Pharmigen) antibodies were used with a goat anti-rat biotinylated second antibody and developed with neutralite avidin-texas red conjugate (Southern Biotechnology, Birmingham, AL). The biotinylated anti-CD31 (MEC7.46, Cell Sciences, Canton, MA) was developed with the neutralite avidin-texas red conjugate (Southern Biotechnology). The Flk-1/KDR/VEGFR2 antibody (ThermoFisher Scientiﬁc) and the anti-CD20 antibody (EP459Y, Abcam) were detected with Qdot 525 goat MURINE COLITIS-ASSOCIATED LYMPHOID ANGIOGENESIS F(ab’)2 anti-rabbit IgG conjugate (Invitrogen, Eugene, OR). Immunoﬂuorescence Staining Cryostat sections were obtained from colon specimens were treated with O.C.T. compound and snap frozen. After warming the slide to 27 C, the sections were ﬁxed for 10 min (2% paraformaldehyde and PBS at pH 7.43). The slides were washed with buffer (PBS, 5% sheep serum, 0.1% azide, 1 mM MgCl2, 1 mM CaCl2) and blocked with 20% sheep serum, 20% goat serum, 0.1% azide in PBS. The slides were treated with monoclonal antibodies (Mab) at 10–20 lg/mL. The slides were incubated for 1 hr at 27 C and washed twice. The detection antibody was added and incubated for 20 min at 27 C. The slides were washed twice and examined by ﬂuorescence microscopy. Optical System The exteriorized tissue was imaged using a Nikon Eclipse TE2000 inverted epiﬂuorescence microscope using Nikon objectives of 10, 20, and 40 linear magniﬁcation with inﬁnity correction. An X-Cite (Exfo; Vanier, Canada) 120 watt metal halide light source and a liquid light guide was used to illuminate the tissue samples. Excitation and emission ﬁlters (Chroma, Rockingham, VT) in separate LEP motorized ﬁlter wheels were controlled by a MAC5000 controller (Ludl, Hawthorne, NY) and MetaMorph software 7.5 (MDS Analytical Technologies, Downingtown, PA). The CFSE tracer (ex 480 nm, em 520 nm) was imaged with 25 nm band pass ﬁlters (Omega, Brattleboro, VT). The intravital videomicroscopy 14-bit ﬂuorescent images were digitally recorded on a C9100-02 camera (Hamamatsu, Japan). The C9100-02 camera has a hermetic vacuum-sealed aircooled head and on-chip electron gain multiplication (2000). Images with 1000 1000 pixel resolution were routinely obtained at 50 fps; frame rates exceeding 50 fps were obtained with binning and subarrays. The images were recorded in image stacks comprising 30 sec–10 min video sequences. 3-Dimensional Tissue Mounts The structure of the colon microcirculation was characterized by ﬂuorescent vessel painting (Ravnic et al., 2005). After systemic heparinization, the aorta was cannulated and perfused with 15 mL of 37 C phosphate buffered saline (PBS) followed by perfusion with a buffered 2.5% glutaraldehyde solution (Sigma). The systemic circulation was perfused with 1,1-dioctadecyl-3,3,3,3-tetramethylindocarbocyanine perchlorate (DiI) or 3,30 dioctadecyloxacarbocyanine perchlorate (DiO) (10–25 mL) as described previously (Ravnic et al., 2005). Immediately following tracer infusion, the organs were harvested, prepared in a 25 C PBS bath, and ﬁxed overnight between glass slides in 4% formalin. After a brief rinse with distilled water, the specimens were stained with DAPI (Vector, Burlingame, CA) and permanently mounted with Vectashield mounting medium (Vector). The ﬂuorescently labeled microvessels were imaged using a Nikon Eclipse TE2000 inverted epiﬂuorescence microscope. Structured illumination confocal 623 microscopy was performed with an Optigrid system (Qioptiq, Rochester, NY) (Lee et al., 2008). The Optigrid uses a one-dimensional optical grid in the form of a Ronchi grating mounted on a piezo-electrically driven actuator. The pattern is moved perpendicular to the grid lines three times producing three separate images that are digitally recombined using a proprietary software algorithm (Volocity 4.4; Improvision, Natick, MA). Mucosal Microdissection After euthanasia, subtotal colectomy (ascending to transverse) was performed. The lumen was ﬂushed and opened along the mesenteric border (McDonald and Newberry, 2007). The mucosa was copiously irrigated with cold PBS (4 C) until all debris was removed as determined by stereomicroscopy. The colon wall was immobilized on a standard microscope slide and the mucosa, superﬁcial to the lamina propria, was removed using gentle dissection with a second microscope slide. Limited dissection of the intact superﬁcial (50–100 lm thick) mucosa was conﬁrmed by light microscopy. Tissue Processing Standard RNA isolation procedures were used, including separate laboratory space for tissue harvesting, RNA isolation, and PCR processing. Pipets and consumables were regularly treated by UV irradiation; work surfaces were routinely cleaned with DNA Exitus Plus (Applichem, Cheshire, CT), standard disinfectants and UV treatment. Routine wipe tests of work areas were performed to screen for nucleic acid contamination. RNA Isolation Total RNA was isolated using Qiagen RNeasy Midi Kit (Qiagen, Valencia, CA). Brieﬂy, the fresh tissue was triturated using a 20-G needle until uniformly homogeneous. The tissue lysate was centrifuged at 3000g for 10 min and the supernatant (lysate) was removed by pipetting. An equal volume of 70% ethanol was added to lysate and gently mixed. The sample was placed in an RNeasy midi column, centrifuged for 5 min at 3000g and the ﬂow-trough was discarded. After additional RPE buffer was added to the column, the tube was again centrifuged for 5 min at 3000g to dry the RNeasy silica-gel membrane. The RNeasy column was transferred to a collection tube and elution was performed using RNase-free water and centrifugation for 3 min at 3000g. Generally, a second elution step was not performed. Genomic DNA contamination was eliminated by RNase-Free DNase Set (Qiagen). Brieﬂy, 1–2 lg of potentially contaminated RNA was treated with DNase buffer, RNase inhibitor, and DNase I. In all RNA isolations, the total RNA quality was assessed by using an Agilent 2100 Bioanalyzer (Agilent Technologies, Palo Alto, CA). RNA integrity numbers (RIN) (Schroeder et al., 2006) of the RNA samples were uniformally greater than 7.3 (mean, 8.5; range, 7.3–9.8). Angiogenesis PCR Arrays First Strand cDNA Synthesis used RT2 First Strand Kit from SuperArray Bioscience Corporation. Mouse Angiogenesis RT2 Proﬁler PCR Array and RT2 Real- 624 TURHAN ET AL. Timer SyBR Green/ROX PCR Mix were purchased from SuperArray Bioscience Corporation (Frederick, MD). Angiogenesis Genes The angiogenesis genes examined in our study include the following (abbreviation, gene name): Angiopoietin 1 (Angpt1, 1110046O21Rik/Ang-1), Angiopoietin 2 (Angpt2, Agpt2/Ang- 2), Chemokine (C-C motif) ligand 2 (CCL2, AI323594/HC11), Collagen, type IV, alpha 3 (Col4a3, [a]3(IV)/alpha3(IV)), Colony Stimulating Factor 3 (granulocyte) (CSF3, Csfg/G-CSF), Chemokine (C-X-C motif) ligand 1 (CXCL1, Fsp/Gro1)), Chemokine (C-X-C motif) ligand 2 (CXCL2, CINC-2a/GROb), Chemokine (C-X-C motif) ligand 5 (CXCL5, AMCF-II/ENA-78), Fibroblast growth factor 1 (Fgf1, Dffrx/Fam), Fibroblast growth factor 2 (Fgf2, Fgf-2/Fgfb), Fibroblast growth factor 6 (Fgf6, Fgf-6), Fibroblast growth factor receptor 3 (Fgfr3, Fgfr-3/ HBGFR), Heart and neural crest derivatives transcript 2 (Hand2, AI225906/AI661148), Interferon gamma (Ifng, IFN-g/IFN-gamma), Interleukin 1 beta (Il1b, IL-1beta/Il1b), Interleukin 6 (Il6, Il-6), Leptin (Lep, ob/obese), Matrix metallopeptidase 9 (MMP9, AW743869/B), T-box 4 (Tbx4, 3930401C23), Transforming growth factor alpha (Tgfa, wa-1/wa1), Transforming growth factor, beta 1 (Tgfb1, TGF-beta1/TGFbeta1), Transforming growth factor, beta 2 (Tgfb2, BB105277/Tgf-beta2), Transforming growth factor, beta 3 (Tgfb3, Tgfb-3), Tumor necrosis factor (TNF, DIF/TNF-alpha), Thymidine phosphorylase (Tymp, 2900072D10Rik/Ecgf1), Vascular endothelial growth factor A(Vegfa, VEGF- A/VEGF120), Vascular endothelial growth factor B (Vegfb, VEGF-B/Vrf), Vascular endothelial growth factor C (Vegfc, AW228853/VEGF-C). Quantitative PCR Real-time PCR was performed with SYBR green qPCR master mixes that include a chemically-modiﬁed hot start Taq DNA polymerase (SABioscience). PCR was performed on ABI 7300 Real-Time PCR System (Applied Biosystems). For all reactions, the thermal cycling conditions were an initial 50 C for 2 min and 95 C for 10 min followed by 40 cycles of denaturation at 95 C for 15 sec and simultaneous annealing and extension at 60 C for 1 min. via the tail vein with 150 lL of the prepared solution. The CFSE and FITC-dextran tracers (ex 480 nm, em 520 nm) were imaged with 25 nm band pass ﬁlters (Omega). Multi-Frame Particle Tracking Tracking of the green and infra-red particles was performed on digitally recorded and distance calibrated multi-image ‘‘stacks’’ (Ravnic et al., 2006). The image stacks produced a sequential time history of velocity and direction as the acquired images were time stamped based on the 100 mHz system bus clock of the Xeon processor (Intel, Santa Clara, CA). The movement of individual particles was tracked using the MetaMorph (MDS Analytical Technologies) object tracking applications. The intensity centroids of the particles were identiﬁed and their displacements tracked through planes in the source image stack. For displacement reference, the algorithm used the location of the particle at its ﬁrst position in the stack. Each particle was imaged as a high contrast ﬂuorescent disk and its position was determined with sub pixel accuracy. The image of the particle was tracked using a cross correlation centroid-ﬁnding algorithm to determine the best match of the particle/ cell position in successive images. With routine distance calibration, the overlay of the image stack provided a quantitative assessment of the particle/cell path. From the XY coordinates, velocity, mean displacement, and mean vector length were calculated. Time-Series Flow Visualization The stream acquired images were stacked to create a time-series of 500 or 1000 consecutive frames. The stacks were systematically analyzed to ensure the absence of motion artifact. The stack ‘‘maximum’’ operation selected the highest intensity value for each pixel location throughout the time-series. The resultant image produced a time-series reconstruction of particle locations during the time interval of the image stack. Nanoparticles Statistical Analysis The nanoparticles were developed by Molecular Probes (Invitrogen, Eugene, OR) for intravascular particle tracking (Ravnic et al., 2007b). Characteristics of the particles included superior ﬂuorescence intensity, small size (500 nm), and low surface charge content (6.2 lEq/ g). The nanoparticles used in this study were green (ex 488; em 510) and infra-red (ex 655 nm; em 710 nm). Gene expression was calculated using the comparative cycle threshold (Ct) method (Livak and Schmittgen, 2001). Although the data was monitored for nonideal efﬁciencies, comparable ampliﬁcation of the target genes and reference genes was assumed. Every effort to optimize the reaction efﬁciency was made. Validation assays using serial dilutions of the target and reference genes were not routinely performed. The DSS-induced colitis and control data were plotted as a scattergram and a linear regression was calculated with 95% prediction bands after the data was imported into Origin 8.0 (OriginLab, North Hampton, MA). Linear regression was uniformly P < 0.0001. In nanoparticle velocity analyses, the unpaired Student’s t-test for samples of unequal variances was used to calculate statistical signiﬁcance. The data was expressed as mean one standard deviation. The signiﬁcance level for the sample distribution was deﬁned as P < 0.01. Plasma-Marker Fluorescence Labeling A 5-(and-6)-carboxyﬂuorescein diacetate, succinimidyl ester (CFSE) (Invitrogen, Eugene, OR) labeling solution was prepared in dimethyl sulfoxide (DMSO) as described (Becker et al., 2004; Ravnic et al., 2006). The freshly prepared CFSE (400 lL) was injected into the tail vein of an anesthetized mouse. In some mice, a 10% 250,000 kD ﬂuorescein isothiocyanate (FITC)-dextran (Sigma) solution in normal saline was prepared. Mice were injected MURINE COLITIS-ASSOCIATED LYMPHOID ANGIOGENESIS Fig. 1. DSS-induced colitis in adult mice. After DSS exposure for 5 days, the mice demonstrated persistent inﬂammatory changes in the colonic mucosa. Microendoscopy of the mouse descending colon 5 to 60 days after the initial DSS exposure showed gross mucosal changes, including erythema and ulceration, reminiscent of human colitis (A, control; B, DSS colitis). (C–F) Hematoxylin and eosin histology 625 of submucosal and intramucosal mononuclear aggregates (white ovals) in mice 28–30 days after the onset of DSS colitis. The mononuclear inﬁltration in many regions of the colon was nearly transmural (C and D). Other areas demonstrated intramucosal aggregates superﬁcial to the lamina propria (E and F)(Bar A,B ¼ 200 lm; C,D ¼ 160 lm). 626 TURHAN ET AL. RESULTS Colon Lymphoid Neogenesis Consistent with previous reports (Neurath et al., 2000), the mice in this study (N ¼ 92 mice) initially developed weight loss and clinical signs of colitis: the mean weight dropped to 76% 6% of baseline on day 7 and gradually recovered to 100% 4% of baseline weight on day 28. Similarly, the clinical colitis scores, including rufﬂed fur, inactivity, and diarrhea, peaked on day 8 (score 10 4) and returned to baseline on day 19 (score 0.3 2). Despite the clinical improvement over the ﬁrst 2–3 weeks, microendoscopy demonstrated ongoing colonic inﬂammation (Fig. 1). Serial histologic sections demonstrated submucosal aggregates that frequently involved both the submucosa and mucosal crypts; these aggregates appeared to span the lamina propria (Fig. 1C,D). Relatively smaller intramucosal aggregates were identiﬁed within the mucosa; that is, superﬁcial to the lamina propria (Fig. 1E,F). Lymphoid Membrane Markers Fig. 2. Histologic studies of lymphoid aggregates in both acute and chronic DSS-induced colitis. Random and systematic sections were obtained from DSS colitis mice at various time points after the induction of colitis (refer Methods). Histologic sections demonstrated an increasing frequency of submucosal (N ¼ 39 mice; closed circles) and intramucosal (N ¼ 27 mice; open circles) mononuclear aggregates. Linear curve ﬁt for the submucosal (solid line) and intramucosal aggregates (dashed line) are shown with 95% conﬁdence bands (dotted lines). Fig. 3. Immunohistochemical phenotyping of the submucosal and intramucosal aggregates. Immunophenotyping demonstrated T-cell (CD4) and B-cell (CD19) expression within both the submucosal and intramucosal aggregates (day 30 shown). The B-cell prevalence, as documented by CD19 and CD20 expression, increased in frequency between 30 and 60 days and was associated with an increase in the Spatial frequency analysis of the mononuclear aggregates within the submucosa demonstrated an increase in size and prevalence during the 60 day study period (Fig. 2, close circles; R ¼ 0.79; F ¼ 61.8; P < 0.0001). Similarly, the smaller intramucosal aggregates also increased in size and prevalence (Fig. 2, open circles; R ¼ 0.62; F ¼ 16.4; P < 0.0001). As expected, immunophenotyping demonstrated T-cell (CD4) and B-cell (CD19) expression within the submucosal aggregates (Fig. 3A–C). The prevalence of B cells within the aggregates, as demonstrated by anti-CD19 and anti-CD20 staining, increased in frequency between 30 and 60 days. The increase in B-cell frequency, demonstrated by immunoﬂuorescence staining, was associated with an increasing frequency of lymphoid follicles. Although most intramucosal and development of lymphoid follicles (not shown). Intramucosal aggregates occasionally demonstrated a notable density of monocytes with a prominent CD11b and F4/80 expression (Bar ¼ 100 lm) (day 30 shown). Other intramucosal nodules demonstrated more balanced distribution of lymphocytes and monocytes. MURINE COLITIS-ASSOCIATED LYMPHOID ANGIOGENESIS 627 Fig. 4. Fluorescent vessel painting of the mucosal plexus microcirculation 30 days after the onset of DSS-induced colitis. The colitis mice had the microcirculation ﬂushed, ﬁxed, and labeled with the intravascular lipophilic tracer (red ¼ DiI). The colon was then prepared as a whole mount and examined by wide ﬁeld (A–D) and structured illumination confocal microscopy. Intramucosal structures, consistent with the distribution of intramucosal aggregates, were visible on lower power (A,C). Higher magniﬁcation (B,D) demonstrated a small vessel network in the plane of, and contiguous with, the mucosal plexus (B,D Bar ¼ 80 lm). submucosal aggregates were phenotypically similar, 6%– 10% of the intramucosal mononuclear aggregates demonstrated a predominance of monocytoid markers (CD11b and F4/80) (Fig. 3D–F). vascular microarchitecture, the mucosal plexus was examined using intravital videomicroscopy and ﬂuorescent vessel painting. In control mice, the mucosal plexus—a quasi-polygonal network of vessels surrounding the mucosal crypts—was a continuous plexus without a specialized vascular supply to lymphoid tissue. In contrast, 21 to 60 days after the onset of inﬂammation, intravital microscopy studies of the mucosal plexus demonstrated distinctive microcirculatory structures composed of a small network of capillaries. The structures Mucosal Plexus Angiogenesis The development of intramucosal lymphoid tissue suggested the possibility of structural changes in the mucosal microcirculation. To investigate changes in the 628 TURHAN ET AL. Fig. 5. Angiogenic gene expression in chemically-induced colitis (A) 7 days, (B) 14 days, (C) 31 days, and (D) 65 days after the onset of DSS exposure. In the SuperArray assay, angiogenic inﬂammatory mediators including CCL2, CXCL2, CXCL5, MMP9, IL-1b, and TNF are shown as solid squares (n). The angiogenic factors Angpt1, Angpt2, Vegfa, Vegfb, and Vegfc are shown as solid triangles (s). The remainder of the SuperArray angiogenesis PCR array genes are shown as open circles (O). A linear regression was performed on the data from each time point: (A) 0.975, (B) 0.857, (C) 0.916, and (D) 0.993 (P < 0.0001); 95% prediction bands are shown. were 185 46 lm (N ¼ 12) in diameter and appeared to be contiguous with the mucosal plexus (Fig. 4). Morphometry based on ﬂuorescent vessel painting and 3D tissue mounts of the capillaries indicated a microvessel diameter of 6.8 1.1 lm (N ¼ 6). Consistent with the histologic analysis, the rarity of similar structures in control mice suggested that the capillary structures developed de novo. To explore the gene expression potentially involved in sprouting angiogenesis, mRNA was isolated from microdissected mucosal plexus in DSSinduced colitis and control mice. The expression of genes implicated in angiogenesis was explored using the angiogenesis pathway PCR arrays at 4 timepoints after the induction of DSS colitis: 7, 14, 31, and 65 days (Fig. 5). The expression of CXCL2, Il-1b, CXCL5, CCL2, TNF, and MMP9 peaked at 14 days after the onset of DSSinduced colitis (Fig. 5B). In this bulk RNA, angiogenic factors with a less notable inﬂammatory association, such as Angpt1, Angpt2, Vegfa, Vegfb, and Vegfc, were not signiﬁcantly elevated relative to controls (Fig. 6). Intravital Microscopy of the Lymphoid Aggregate The functional implications of the lymphoid angiogenesis was investigated by ﬂuorescent intravital videomicroscopy. In the chronic phase, 30–60 days after the onset of chemically-induced colitis, intravenously injected ﬂuorescent nanoparticles were tracked through the mucosal capillary structures (Fig. 7A,B). Nanoparticle ﬂow demonstrated that the particles passed directly from the mucosal plexus into the capillary structures, conﬁrming both structural and functional continuity. Frequently, the particles exited the capillary structures and passed into deeper collecting veins. The ﬂow through the mucosal plexus structures was notable for a MURINE COLITIS-ASSOCIATED LYMPHOID ANGIOGENESIS 629 Fig. 6. Angiogenic gene expression in chemically-induced colitis after the onset of DSS exposure. Microdissected colon mucosa was obtained from colitis and control mice at 7, 14, 31, and 65 days after the onset of colitis. Gene expression was determined by real-time qPCR. Gene expression is presented relative to control values. signiﬁcant decrease in ﬂow velocity when compared to the feeding vessels within the mucosal plexus (Fig. 7C– E). The analysis of intravital microscopy recordings (N ¼ 6 mice) showed that particles passing into these de novo vessels demonstrated a mean velocity of 491 155 lm/ sec. In contrast, the feeding vessels of the mucosal plexus demonstrated a mean velocity of 808 901 lm/ sec (Fig. 7F; P < 0.01). DISCUSSION In this report, we studied the microvascular adaptations associated with prolonged inﬂammation in DSSinduced murine colitis. Although both submucosal and intramucosal lymphoid aggregates were identiﬁed, the development of lymphoid neogenesis within the superﬁcial mucosa was associated with structural reorganization of the microcirculation. A frequency analysis suggested that most intramucosal lymphoid aggregates developed de novo. The intramucosal aggregates were supplied by capillaries arising from the quasi-polygonal mucosal plexus. The expression of genes associated with both inﬂammation and angiogenesis suggested that the structural changes were the result of inﬂammationinduced sprouting angiogenesis. The structural changes observed in this study highlight the importance of the mucosal stroma in sustaining a peripheral immune response. Previous work has focused on the plasticity of vascular endothelial cells in adapting to peripheral inﬂammation. Endothelial cells can undergo dramatic inﬂammation-induced changes in morphology—from ﬂat conduit lining cells to cuboidal HEV cells (Freemont, 1988; Sasaki et al., 1994; Peng, 1996). On a molecular level, the stromal-endothelial interactions in ulcerative colitis and rheumatoid arthritis can stimulate the expression of PNAd, CCL21, and CXCL13 proteins (Takemura et al., 2001; Salomonsson et al., 2003; Carlsen et al., 2004; Manzo et al., 2005). Consistent with other studies of inﬂammation (Weninger et al., 2003), PNAd expression in our model was 630 TURHAN ET AL. Fig. 7. Intravital microscopy of DSS-induced intramucosal aggregates after intravenous injection of ﬂuorescent plasma markers and infra-red nanoparticle tracers. (A) A plasma-marker angiogram of an intramucosal capillary structure in DSS colitis. A large submucosal vein draining some of the capillaries is seen (arrow). (B) The same mouse with digital recombination of 500 consecutive infra-red images obtained at 20 msec intervals (Bar ¼ 100 lm). Fluorescence reﬂects the infra-red nanoparticles within the capillary structure during the 500 image time series. (C–E) Instantaneous velocities of three nanoparticles as they were tracked through intramucosal structures (N ¼ 3 mice); arrows delineate the anatomic extent of the capillary structures. (F) Combined data of 600 particles tracked in six mice; the mean velocity of particles in vessels feeding the structure and passing through the structure are shown. The data was plotted with the box deﬁning the 25th and 75th velocity precentiles with the whiskers deﬁning the ﬁfth and 95th percentile. The median value was plotted as a square. inconsistent (not shown) suggesting that the variability in HEV morphology and PNAd expression may reﬂect different stages of ectopic lymphoid aggregate development. Regardless, the signiﬁcant structural remodeling—including the apparent sprouting growth of a complex arrangement of capillaries—suggests that stromal adaptations play an important role in the pathophysiology of prolonged DSS-induced colitis. Our exploratory analysis of mRNA expression within the mucosal plexus demonstrated several mediators previously associated with angiogenesis. Three members of the CXC chemokine family, known to promote angiogenesis (Strieter et al., 2005), were expressed at high levels during the peak of the inﬂammation. mRNA From the CXCL1, CXCL2, and CXCL5 genes were expressed at levels 90-to 8000-fold greater than controls. These CXC family chemokines signal through the CXCR2 receptor— a receptor that has been implicated in vivo in models of corneal neovascularization (Addison et al., 2000) and wound repair (Devalaraja et al., 2000). The association of MMP9 with tissue remodeling (Page-McCaw et al., 2007) suggests a functional role for extracellular proteases in the remodeling necessary for the development of both lymphoid aggregates and lymphoid angiogenesis. Furthermore, the inﬂammatory mediators IL-1b and TNF also have been implicated in angiogenesis (Maruotti et al., 2006). In contrast, the expression of angiogenic factors such as Angpt1 and Vegfa were not elevated relative to controls. This ﬁnding may reﬂect the bulk sampling of mRNA. Although the mucosal plexus was microdissected from the remainder of the colon wall, the samples included bulk mRNA from the perivascular inﬂammatory cells as well as the mucosal plexus vessels. Discrete spatial sampling, enabled by laser capture MURINE COLITIS-ASSOCIATED LYMPHOID ANGIOGENESIS microdissection, may be necessary to elucidate the participation of these factors. Gut-associated lymphoid tissue has been separated into effector sites, which consist of lymphocytes scattered throughout the superﬁcial mucosal tissue and the induction sites present in organized lymphoid tissues (Mowat, 2003; Spahn and Kucharzik, 2004). The inductive sites include Peyer’s patches, mesenteric lymph nodes, and isolated lymphoid follicles. The contemporary understanding of immunologic function is that antigen presentation and the generation of antigen-speciﬁc effector cells occurs in inductive tissues, and that effector cells migrate into superﬁcial mucosal tissues. Our observation of progressive organization of the superﬁcial mucosal compartment suggests that this initial functional distinction may evolve during the subacute phase of inﬂammation leading to the presence of both inductive and effector elements with the chronically inﬂamed colonic mucosa. This functional evolution within mucosal tissue is suggested by human studies of secondary and ectopic lymphoid tissue (Manzo et al., 2007). Distorted crypt architecture, intramucosal inﬂammatory cells, and lymphoid aggregates were features present in 79% of ongoing inﬂammatory bowel disease and were highly predictive of chronic colitis (Surawicz and Belic, 1984). The development of inductive sites may also provide an explanation for the spatial distribution of the intramucosal aggregates. The sporadic distribution of intramucosal lymphoid aggregates does not reﬂect any vascular microarchitectural feature that might predispose to capillary sprouting and lymphoid angiogenesis. Rather than a structural predisposition to lymphoidassociated angiogenesis, we suspect that the spatial distribution of lymphoid aggregates may reﬂect the location of antigen presenting cells, such as dendritic cells, in the initiation of lymphoid neogenesis (Carragher et al., 2008). The importance of antigen presenting cells in lymphoid neogenesis may also help explain the occasional concentration of monocytoid cells within the intramucosal aggregates. An advantage of our study was the use of intravascular tracers to demonstrate structural continuity between the capillary structures and the mucosal plexus. Because the particles were inert and charge-neutral, they could be tracked through the microcirculation without the concern of unanticipated biomolecular interactions with vascular lining cells. The particles provided a useful measure of both ﬂow velocity and network ﬂow ﬁelds within the sprouting capillary structures. An interesting observation was the diminished ﬂow velocity within the capillary structures; velocities were sufﬁciently diminished to be within the physiologic range of rolling velocities in secondary lymphoid tissue (Stein et al., 1999). Thus, even if some of the typical secondary lymphoid organ receptor-ligand interactions were not present, the microhemodynamic conditions were nonetheless suitable for lymphoid adhesion and transmigration (Li et al., 2001). Finally, an assessment of capillary structure and intravacular tracer velocity permits an assessment of blood ﬂow within the mucosal aggregates. Assuming an average of 10 capillaries with cylindrical geometry, and a cardiac output of 20 mL/min (Janssen et al., 2002), the perfusion of an average capillary structure would be 0.0015% of cardiac output; that is, 10% of the perfu- 631 sion of a popliteal lymph node in the sheep (0.014% cardiac output) or a comparable node in the rabbit (0.011%) (Hay and Hobbs, 1977). This ﬁnding indicates that even small and developing lymphoid structures possess the potential for a substantial exposure to circulating lymphocytes and a signiﬁcant contribution to the peripheral immune response. LITERATURE CITED Addison CL, Daniel TO, Burdick MD, Liu H, Ehlert JE, Xue YY, Buechi L, Walz A, Richmond A, Strieter RM. 2000. The CXC chemokine receptor 2, CXCR2, is the putative receptor for ELRþ CXC chemokine-induced angiogenic activity. J Immunol 165: 5269–5277. Aloisi F, Pujol-Borrell R. 2006. Lymphoid neogenesis in chronic inﬂammatory diseases. Nat Rev Immunol 6:205–217. Armengol MP, Juan M, Lucas-Martin A, Fernandez-Figueras MT, Jaraquemada D, Gallart T, Pujol-Borrell R. 2001. Thyroid autoimmune disease: demonstration of thyroid antigen-speciﬁc B cells and recombination-activating gene expression in chemokinecontaining active intrathyroidal germinal centers. Am J Pathol 159:861–873. Barone F, Bombardieri M, Manzo A, Blades MC, Morgan PR, Challacombe SJ, Valesini G, Pitzalis C. 2005. Association of CXCL13 and CCL21 expression with the progressive organization of lymphoid-like structures in Sjogren’s syndrome. Arthritis Rheum 52:1773–1784. Becker HM, Chen M, Hay JB, Cybulsky MI. 2004. Tracking of leukocyte recruitment into tissues of mice by in situ labeling of blood cells with the ﬂuorescent dye CFDA SE. J Immunol Methods 286:69–78. Carlsen HS, Baekkevold ES, Johansen FE, Haraldsen G, Brandtzaeg P. 2002. B cell attracting chemokine 1 (CXCL13) and its receptor CXCR5 are expressed in normal and aberrant gut associated lymphoid tissue. Gut 51:364–371. Carlsen HS, Baekkevold ES, Morton HC, Haraldsen G, Brandtzaeg P. 2004. Monocyte-like and mature macrophages produce CXCL13 (B cell-attracting chemokine 1) in inﬂammatory lesions with lymphoid neogenesis. Blood 104:3021–3027. Carragher DM, Rangel-Moreno J, Randall TD. 2008. Ectopic lymphoid tissues and local immunity. Semin Immunol 20:26–42. Devalaraja RM, Nanney LB, Du J, Qian Q, Yu Y, Devalaraja MN, Richmond A. 2000. Delayed wound healing in CXCR2 knockout mice. J Invest Dermatol 115:234–244. Freemont AJ. 1988. HEV-like vessels in lymphoid and non-lymphoid tissues. Int J Tissue React 10:85–88. Hay JB, Hobbs BB. 1977. The ﬂow of blood to lymph nodes and its relation to lymphocyte trafﬁc and the immune response. J Exp Med 145:31–44. Hjelmstrom P. 2001. Lymphoid neogenesis: de novo formation of lymphoid tissue in chronic inﬂammation through expression of homing chemokines. J Leukoc Biol 69:331–339. Janssen B, Debets J, Leenders P, Smits J. 2002. Chronic measurement of cardiac output in conscious mice. Am J Physiol Regul Integr Comp Physiol 282:R928–R935. Lee GS, Miele L, Turhan A, Lin M, Hanidziar D, Konerding MA, Mentzer SJ. 2009. Spatial calibration of structured illumination ﬂuorescence microscopy using capillary tissue phantoms. Microsc Res Tech 72:85–92. Li X, Su M, West CA, He C, Swanson SJ, Secomb TW, Mentzer SJ. 2001. Effect of shear stress on efferent lymph-derived lymphocytes in contact with activated endothelial monolayers. In Vitro Cell Dev Biol 37:599–605. Livak KJ, Schmittgen TD. 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25:402–408. Manzo A, Bugatti S, Caporali R, Prevo R, Jackson DG, Uguccioni M, Buckley CD, Montecucco C, Pitzalis C. 2007. CCL21 expression pattern of human secondary lymphoid organ stroma is 632 TURHAN ET AL. conserved in inﬂammatory lesions with lymphoid neogenesis. Am J Pathol 171:1549–1562. Manzo A, Paoletti S, Carulli M, Blades MC, Barone F, Yanni G, Fitzgerald O, Bresnihan B, Caporali R, Montecucco C, Uguccioni M, Pitzalis C. 2005. Systematic microanatomical analysis of CXCL13 and CCL21 in situ production and progressive lymphoid organization in rheumatoid synovitis. Eur J Immunol 35:1347–1359. Maruotti N, Cantatore FP, Crivellato E, Vacca A, Ribatti D. 2006. Angiogenesis in rheumatoid arthritis. Histol Histopathol 21:557–566. McDonald KG, Newberry RD. 2007. Whole-mount techniques to evaluate subepithelial cellular populations in the adult mouse intestine. Biotechniques 43:50–56. Mowat AM. 2003. Anatomical basis of tolerance and immunity to intestinal antigens. Nat Rev Immunol 3:331–341. Neurath M, Fuss I, Strober W. 2000. TNBS-colitis. Int Rev Immunol 19:51–62. Okayasu I, Hatakeyama S, Yamada M, Ohkusa T, Inagaki Y, Nakaya R. 1990. A novel method in the induction of reliable experimental acute and chronic ulcerative colitis in mice. Gastroenterology 98:694–702. Page-McCaw A, Ewald AJ, Werb Z. 2007. Matrix metalloproteinases and the regulation of tissue remodelling. Nat Rev Mol Cell Biol 8:221–233. Peng FS. 1996. High endothelial venules of the rat lymph node. A review and a question: is their activity antigen speciﬁc? Anat Rec 245:593–620. Picker LJ, Butcher EC. 1992. Physiological and molecular mechanisms of lymphocyte homing. Annu Rev Immunol 10:561–591. Ravnic DJ, Jiang X, Wolloscheck T, Pratt JP, Huss H, Mentzer SJ, Konerding MA. 2005. Vessel painting of the microcirculation using ﬂuorescent lipophilic tracers. Microvasc Res 70:90–96. Ravnic DJ, Konerding MA, Huss HT, Wolloscheck T, Pratt JP, Mentzer SJ. 2007a. Murine microvideo endoscopy of the colonic microcirculation. J Surg Res 142:97–103. Ravnic DJ, Tsuda A, Turhan A, Zhang Y-Z, Pratt JP, Huss HT, Mentzer SJ. 2006. Multi-frame particle tracking in intravital imaging: deﬁning lagrangian coordinates in the microcirculation. BioTechniques 41:597–601. Ravnic DJ, Zhang Y-Z, Turhan A, Tsuda A, Pratt JP, Huss HT, Mentzer SJ. 2007b. Biological and optical properties of ﬂuorescent nanoparticles developed for intravascular imaging. Microsc Res Tech 70:776–781. Ruddle NH. 1999. Lymphoid neo-organogenesis: lymphotoxin’s role in inﬂammation and development. Immunol Res 19:119–125. Salomonsson S, Jonsson MV, Skarstein K, Brokstad KA, Hjelmstrom P, Wahren-Herlenius M, Jonsson R. 2003. Cellular basis of ectopic germinal center formation and autoantibody production in the target organ of patients with Sjogren’s syndrome. Arthritis Rheum 48:3187–3201. Sasaki K, Pabst R, Rothkotter HJ. 1994. The unique ultrastructure of high-endothelial venules in inguinal lymph nodes of the pig. Cell Tissue Res 276:85–90. Schroeder A, Mueller O, Stocker S, Salowsky R, Leiber M, Gassmann M, Lightfoot S, Menzel W, Granzow M, Ragg T. 2006. The RIN: an RNA integrity number for assigning integrity values to RNA measurements. BMC Mol Biol 7:3. Spahn TW, Kucharzik T. 2004. Modulating the intestinal immune system: the role of lymphotoxin and GALT organs. Gut 53:456–465. Stein JV, Cheng G, Stockton BM, Fors BP, Butcher EC, von Andrian UH. 1999. L-selectin-mediated leukocyte adhesion in vivo: microvillous distribution determines tethering efﬁciency, but not rolling velocity. J Exp Med 189:37–50. Strieter RM, Burdick MD, Gomperts BN, Belperio JA, Keane MP. 2005. CXC chemokines in angiogenesis. Cytokine Growth Factor Rev 16:593–609. Surawicz CM, Belic L. 1984. Rectal biopsy helps to distinguish acute self-limited colitis from idiopathic inﬂammatory bowel disease. Gastroenterology 86:104–113. Takemura S, Braun A, Crowson C, Kurtin PJ, Coﬁeld RH, O’Fallon WM, Goronzy JJ, Weyand CM. 2001. Lymphoid neogenesis in rheumatoid synovitis. J Immunol 167:1072–1080. von Andrian UH, Mempel TR. 2003. Homing and cellular trafﬁc in lymph nodes. Nat Rev Immunol 3:867–878. Waidmann M, Allemand Y, Lehmann J, di Genaro S, Bucheler N, Hamann A, Autenrieth IB. 2002. Microﬂora reactive IL-10 producing regulatory T cells are present in the colon of IL-2 deﬁcient mice but lack efﬁcacious inhibition of IFN-gamma and TNF-alpha production. Gut 50:170–179. Weninger W, Carlsen HS, Goodarzi M, Moazed F, Crowley MA, Baekkevold ES, Cavanagh LL, von Andrian UH. 2003. Naive T cell recruitment to nonlymphoid tissues: a role for endotheliumexpressed CC chemokine ligand 21 in autoimmune disease and lymphoid neogenesis. J Immunol 170:4638–4648.