Mice
Six- to ten-week-old mixed-sex C57BL/6J, B6.Cg-Tg(TcraTcrb)425Cbn/J (OT-II), B6.129P2(C)-Ightm2Cgn/J (B1-8) and B6.129S2-IghtmICgn/J (μMT) mice were purchased from Jackson Laboratory and Charles River and subsequently bred and housed at Yale University. PROX1-CreERT2;CDH5-Dre;R26-STOP-mCherry and VEGFR3-CreERT2;R26-mTmG mice were gifts from the laboratory of J.-L.T. All procedures used in this study complied with federal guidelines and the institutional policies of the Yale School of Medicine Animal Care and Use Committee. Age- and sex-matched animals were randomly assigned to control or treatment groups in each experiment. No statistical methods were used to predetermine sample sizes. Sample sizes were empirically determined based on previously published studies and to ensure sufficient statistical power. Investigators were not blinded to experimental groups, as measurements were not subjective.
Cells
GL261–Luc cells were a gift from J. Zhou (Yale Neurosurgery) and were cultured in RPMI supplemented with 10% FBS, 1% penicillin–streptomycin and 1% sodium pyruvate. CT2A–BFP cells were a gift from T. Mathivet (Paris Centre de Recherche Cardiovasculaire). B16 cells were a gift from N. Palm (Yale Immunobiology). Cells were negative for mycoplasma contamination.
Bacteria and viruses
The HSV-1 KOS strain and HSV-2 strains 186syn−TK− and 186syn+ were gifts from D. Knipe (Harvard Medical School). These viruses were propagated and titrated on Vero cells (ATCC CCL-81) as previously described43. Vero cells were negative for mycoplasma contamination.
S. pneumoniae (ATCC 6303) was grown on 10% sheep blood agar plates (BD Biosciences) overnight at 37 °C (5% CO2). These colonies were then transferred to Todd Hewitt broth and grown overnight. The number of bacteria was enumerated using the optical density at 600 nm.
Optic nerve tissues and fluorescence immunocytochemistry
Tg(mrc1a:eGFP)44 adult male and female 1-year-old zebrafish (n = 3) were provided by B.W. Weinstein at the US National Institutes of Health (Bethesda, MD). Zebrafish were euthanized by an overdose of MS-222 and whole heads were excised and fixed in 60 mM HEPES-buffered 4% PFA (pH 7.4) overnight at room temperature. Samples were then decalcified in 10% w/v EDTA solution (pH 7.4) for 5 days at room temperature. Heads were then snap frozen in OCT and 30-μm-thick cryosections were obtained. Cryosections containing the optic nerve were stained with the primary antibodies mouse anti-zebrafish Zns2 (ZIRC, ZDB-ATB-081002-34; 1:50) and chicken anti-GFP (Aves, catalogue no. GFP-1010; 1:500) followed by the secondary antibodies AF488-conjugated goat anti-mouse IgG (Jackson Immuno 115-545-146) and Cy5-conjugated donkey anti-chicken IgY (Jackson Immuno 703-175-155) with DAPI (5 μg ml−1) for nuclear staining. Images were acquired in a Zeiss LSM.
Eyes from rabbits and pigs were collected post-mortem from animals euthanized owing to unrelated health conditions. Eyes were removed within an hour of death and immersion fixed in 10% neutral-buffered formalin. Intact eyes from two female rhesus macaques (Macaca mulatta), between 25–28 years of age and euthanized owing to unrelated health conditions were collected post-mortem. Human optic nerves and chiasmas were obtained through the Yale Pathology Tissue Services through the Tissue Procurement and Distribution service. A mini standard operating procedure was written and approved for these samples. Eyes were removed within an hour of death and immersion fixed in 10% neutral-buffered formalin. Optic nerves with meningeal sheaths were cut and processed through the iDISCO protocol below.
Mouse optic nerve sheaths were dissected after mice were anaesthetized and transcardially perfused with cold PBS and 4% PFA (Sigma-Aldrich) sequentially. Mice were first decapitated, and the skull was exposed by cutting off the skin and scalp. One midline cut down the skull and two transverse cuts on both sides of the skull were made. Tweezers were used to peel and remove both halves of the skull to expose the brain. The brain was lifted from the posterior, and vasculature connecting the brain to other parts of the head was cut until the brain could be lifted enough to expose the optic tract. Brain tissue was removed after cutting just anteriorly to the optic chiasma. The extracranial optic canal was unroofed by removing parts of the skull above the eyes to expose the intracanalicular and canal segments of the optic nerve. Tissue surrounding the optic nerve was carefully dissected to free the nerve, and the eye was cut off at the optic nerve head. After removing the optic nerve, the optic nerve sheath was cut along the length of the nerve and removed for staining and whole-mount imaging.
The PROX1-CreERT2;CDH5-Dre;R26-STOP-mCherry and VEGFR3-CreERT2; R26-mTmG mice were injected through the i.p. route with 100 μl tamoxifen (10 mg ml−1; Sigma-Aldrich, T5648) for 5 consecutive days and optic nerve sheaths were collected 2 days later as above. Then optic nerve sheaths were fixed in 1% PFA for 1 h and immediately processed in a blocking solution (10% normal donkey serum, 1% bovine serum albumin, 0.3% PBS–Triton X-100) for 1 h at room temperature. For detection of lymphatic vessels, samples were incubated with primary antibodies overnight at 4 °C, and then washed five times at room temperature in PBS with 0.5% Triton X-100, before incubation with fluorescence-conjugated secondary antibodies diluted in PBS with 5% normal donkey serum. Lymphatic vessel images were acquired using a Leica confocal microscope (Stellaris 8). The following antibodies were used: goat anti-mouse VEGFR3 (R&D, No. AF743, 1:400), rat anti-mouse LYVE1 (R&D, No. MAB2125, 1:400), Syrian hamster anti-mouse podoplanin (BioLegend,127402, 1:500), rabbit anti-Prox1 (Angiobio, 11-002 P, 1:200), Armenian hamster anti-mouse CD31 (Gene Tex, 2H8, 1:1,000). The primary antibodies were detected with appropriate AF405-, AF488-, AF555- and AF647-conjugated secondary antibodies (Thermo Fisher, 1:500) after 2 h of incubation at room temperature. ProLong Gold Antifade Mountant (Invitrogen, P36930) was used for mounting the sections.
Antibodies for flow cytometry
Anti-CD3 (145-2C11, APC, 152306, 1:200), anti-CD4 (RM4-5, PerCP, 100538, 1:400; RM4-5, BV605, 100548, 1:400), anti-CD8α (53-6.7, BV605, 100744, 1:400; 53-6.7, BV785, 100750, 1:400), anti-CD11b (M1/70, BV711, 101242, 1:500), anti-CD19 (6D5, APC–Cy7, 115530, 1:400), anti-IA and IE (M5/114.15.2, AF488, 107616, 1:800), anti-CD44 (IM7, AF700, 103026, 1:200; BV421, 103040, 1:200), anti-CD45 (30-F11, APC–Cy7, 103116, 1:200), anti-CD45.1 (A20, BV785, 110743, 1:300), anti-CD45.2 (104, Pacific Blue, 109820, 1:200), anti-CD64 (X54-5/7.1, PE, 139304, 1:300), anti-CD95 (Jo2, PE–Cy7, 557653, 1:400), anti-B220 (Ra3-6B2, AF700, 103232, 1:400), anti-GL7 (GL7, fluorescein isothiocyanate (FITC), 144603, 1:400), anti-NK1.1 (PK136, APC–Cy7, 108724, 1:400), anti-TCRβ (H57-597, APC–Cy7, 109220,1:200) were purchased from BD Biosciences or BioLegend. Anti-Igλ light chain, (JC5-1, FITC, 130-098-415, 1:400) was purchased from Miltenyi Biotec.
Isolation of mononuclear cells
For brain tissues, tissues were collected and incubated in a digestion cocktail containing 1 mg ml−1 collagenase D (Roche) and 30 μg ml−1 DNase I (Sigma-Aldrich) in RPMI at 37 °C for 45 min. Tissues were pipetted to break tissue down and filtered through a 70-μm filter. Then, cells were mixed in 3 ml of 25% Percoll (Sigma-Aldrich) solution and centrifuged at 580g for 15 min without brake. The Percoll layer was removed, and cell pellets were treated with 0.5 ml ACK buffer and spun for 5 min at 500g. Then, the cell pellets were resuspended in FACS buffer (PBS + 2% FBS + 1 mM EDTA) for staining.
When analysing lymphocytes, an LN or a spleen was put in a 60 mm × 15 mm Petri dish containing 2 ml FACS buffer and was ground between two frosted microscope slides. When analysing DCs, an LN or a spleen was digested as above. Cell suspensions were filtered through a 70-μm filter and spun for 5 min at 500g. Then, the cell pellets were resuspended in FACS buffer for staining.
Flow cytometry
Preparation of single-cell suspensions from spleens, LNs and brains is described above. Nonspecific binding was blocked using an Fc receptor-blocking solution (TruStain FcX, BioLegend, 101320, 1:200) for 10 min at 4 °C before immunostaining. Subsequently, the cells were stained with corresponding antibodies for 30 min at 4 °C. Then, cells were washed to remove excess antibodies and resuspended in FACS buffer. Samples were run on an Attune NxT flow cytometer and then analysed using FlowJo software (10.8.1, Tree Star).
Enzyme-linked immunosorbent assay
CSF and serum were collected from mice as previously described9. The serum and CSF were then diluted with PBS containing 0.1% BSA in a 1:1 ratio. Plates (96-well) were coated with 100 μl of heat-inactivated or PFA-inactivated purified HSV-2 (104 to 105 plaque-forming units equivalent per 100 μl) for virus-specific immunoglobulin measurement or goat anti-mouse immunoglobulin (SouthernBiotech, 1010-01, 1:1,000) and then incubated overnight at 4 °C. These plates were then washed with PBS–Tween 20 and blocked for 2 h with 5% FBS in PBS. Samples were then plated in the wells and incubated for at least 4 h at room temperature. After being washed in PBS–Tween 20, HRP-conjugated anti-mouse immunoglobulin antibodies (SouthernBiotech, 1010-05, 1:5,000) were added in the wells for 1 h, followed by washing and addition of TMB solution (eBioscience). Reactions were stopped with 1 N H2SO4, and absorbance was measured at 450 nm. The total antibody titres were defined by using an immunoglobulin standard (C57BL/6 mouse immunoglobulin panel; SouthernBiotech).
Western blot
rAAV-RFP-infected retinas or control retinas were digested with cocktail containing 1 mg ml−1 collagenase D (Roche) and 30 μg ml−1 DNase I (Sigma-Aldrich) in RPMI at 37 °C for 45 min. Tissues were pipetted to break tissue down and filtered through a 70-μm filter. Then, cells were lysed in RIPA buffer and boiled for 5 min with sample buffer. Western blotting was carried out in a similar manner to that previously reported9. In brief, 15% gels were used and run at 10 mA per gel for 30 min and 40 mA per gel until ladder separation. Wet transfer was carried out at 120 mA per gel for 90 min on ice. RFP-Tag rabbit polyclonal antibodies (OriGene Technologies, catalogue no. AP09229PU-N) were used at a concentration of 1:1,000 and incubated overnight in a cold room. After being washed, HRP-conjugated anti-rabbit IgG secondary antibodies (Thermo Fisher, G-21234) were used at a concentration of 1:5,000 at room temperature for 2 h and imaged using the ChemiDoc MP imaging system (Bio-Rad).
ELISpot assay
Mice were injected with rAAV-RFP through the IVT or AC route. Their dCLNs, sCLNs and retinas were collected 10 days later. Single-cell suspensions were prepared and co-cultured with splenocytes at a ratio of 1:5 with the presence of rAAV-RFP virus peptides (SNYNKSVNV and NGRDSLVNPGPAMAS). rAAV-specific immune responses were quantified using an ELISpot assay (mouse IFNγ ELISpot Kit; R&D, catalogue no. EL485), following the manufacturer’s instructions for the assay.
Flank tumour inoculation and treatment
Mice were anaesthetized using a mixture of ketamine (50 mg kg−1) and xylazine (5 mg kg−1), and the flank was shaved and disinfected. A 1-ml syringe with a 30-G needle was used to deliver 100 μl of 500,000 B16 cells subcutaneously. Then, mice were treated with irradiated B16 cells (250,000 cells) through s.c., i.c., AC or IVT administration routes (day 7) along with anti-PD1 (RMP1-14) antibodies (days 7, 9 and 11) through the i.p. route, and their survival was monitored.
Adoptive transfer
To directly analyse the immune response (antigen-specific T and B cells) in dCLNs and sCLNs after IVT or AC injection, we transferred OT-II and B1-8 cells and analysed their response after different immunization routes.
To analyse antigen-specific B cell response, we followed a previously reported method45. In brief, CD45.2 C57BL/6 recipient mice (6–8 weeks of age) were primed by i.p. immunization with 50 mg of OVA (Sigma, A5503) precipitated in alum at a 2:1 ratio in PBS. Two weeks later, resting B cells were isolated from CD45.1.2 B1-8 mice with a mouse B cell isolation kit (Stemcell, 19854). Then, the B cells were labelled with CellTrace Violet Cell Proliferation Kit (Thermo Fisher, C34557). A total of 5 million cells were transferred intravenously into recipient mice. Eight hours later, the mice were immunized with 10 μg NP20–OVA (Biosearchtech, N-5051) through IVT or AC injection. Igλ+ B1-8hi cell proliferation and germinal centre formation in dCLNs and sCLNs were analysed at day 7.
For antigen-specific T cell response, OVA-specific CD4+ T cells were isolated from CD45.1 OT-II mice with mouse CD4+ T Cell Isolation Kit (Stemcell, 19852). Then, the CD4+ T cells were labelled using the CFSE Cell Proliferation Kit (Thermo Fisher, C34554). A total of 5 million cells were transferred intravenously into CD45.2 recipient mice. Eighteen hours later, the mice were immunized with 2 μl 50 μg OVA (Sigma, A5503) plus 1 μg poly(I:C) (Invivogen, tlrl-picw) through IVT or AC injection. A 150 μg quantity of FTY720 (Sigma, SML0700) was i.p. injected to inhibit the circulation of primed T cells 24 h after immunization46. OVA-specific CD4+ T cell proliferation in ingLNs, dCLNs and sCLNs was analysed 72 h after immunization.
IVT, AC, i.c. and ICM injection
Mice aged 6–10 weeks old were anaesthetized through i.p. injection of a mixture of ketamine (50 mg kg−1) and xylazine (5 mg kg−1). For IVT or AC injection, the eye was dilated with tropicamide ophthalmic solution. For IVT injection, a 30-G needle was used to puncture a hole at 1 mm posterior to the corneoscleral junction. A blunt-ended Hamilton syringe with 1 μl of dye, HSV-1, HSV-2, S. pneumoniae or irradiated tumour cells was inserted into the vitreous humour about 1–2 mm deep and administered at a rate of 1 μl min−1. For AC injection, the hole was punctured close to the corneoscleral junction and a blunt-ended Hamilton syringe was inserted into the AC about 1 to 2 mm deep. After IVT or AC injection, petrolatum ophthalmic ointment was applied on the eye to prevent cataract formation. The method for i.c. injection was similar to that for tumour inoculation, but 3 μl of HSV-1, HSV-2, S. pneumoniae or irradiated tumour cells was administered. ICM injection was carried out as previously described9. The mice were kept on heating pads and continuously monitored until recovery after the injection.
Virus and bacteria immunization and challenge
Mice were anaesthetized with a mixture of ketamine (50 mg kg−1) and xylazine (5 mg kg−1). For HSV immunization, 106 plaque-forming units of heat-inactivated HSV-1 or HSV-2 were AC, IVT, i.p. or i.c. injected. Thirty days later, these mice were challenged through the i.c. route with 105 HSV-1 or HSV-2 and their survival was monitored. For some experiments, CD4+ T cell-depleting antibodies (BioXCell, GK1.5, No. BE0003-1, 200 μg for 3 days) were injected before rechallenge. Similar experiments were carried out with S. pneumoniae; the immunization dose was 104 of heat-inactivated bacteria and the challenge dose was 104 of live bacteria.
Imaging and quantification of tracer transport
For quantification of fluorescence intensity in the eye or LNs, dextran conjugated to either FITC (40 kDa and 70 kDa) or tetramethylrhodamine (40 kDa and 70 kDa) was injected into the AC or vitreous humour through AC or IVT injection, respectively. To analyse the kinetics of dye drainage from the eye, the eye was collected at serial time points and was homogenized in 150 μl PBS using the bead beating method (Lysing Matrix D, 116913500, MP Biomedicals). Then, homogenized tissue was centrifuged at 10,000g for 10 min and 100 μl supernatant was collect into a 96-well plate and fluorescence intensity was read with emission and excitation wavelengths of 494 nm and 514 nm or 555 nm and 585 nm. To measure the dye drainage into LNs, sCLNs and dCLNs were collected 12 h after dye injection. The fluorescence intensity was measured as above.
For measuring nanoparticle draining in the serum, we used a previously published protocol17. In brief, serum from mice was isolated and placed on microscope slides to allow small-volume high-sensitivity detection of tracer transport in the blood.
For IVIS imaging of tracer transport, the eye, sCLNs and dCLNs were collected after dye injection through either an AC or IVT administration route. They were imaged using the IVIS Spectrum In Vivo Imaging System (PerkinElmer).
For IVT tracer transport in vivo, 0.2 μl AF647–OVA (2 mg ml−1, Thermo Fisher, O34784) was IVT injected into the vitreous humour. A 100 μl volume of sodium fluorescein (1 mg ml−1, Santa Cruz, sc-206026) was injected into blood to label blood vessels. The kinetics of AF647–OVA drainage was tracked with the Phoenix MICRON IV imaging microscope.
For analysing the co-localization of optic nerve sheath lymphatics with IVT tracer, 1 μg anti-mouse LYVE1 antibody (R&D, MAB2125) was IVT injected into the vitreous humour. Optic nerve sheaths were collected 2 h later and stained as above.
For tracking the kinetics of dye draining in the eye, 1 μl AF647–OVA (AF647–OVA (2 mg ml−1; Thermo Fisher, O34784) was IVT injected into the vitreous humour. The eyes were enucleated at the indicated time points and processed in a manner similar to that previously reported47. In brief, eyes were fixed in Hartman’s fixation buffer, and three windows on the eye were created using the previously described window technique47. Serial 10-μm sections were obtained using a cryostat (Leica CM190) following dehydration of the tissue in a sucrose gradient up to 30% sucrose. The sections were mounted with ProLong Gold Antifade Mountant with DAPI (Invitrogen, P36931) and imaged with a Leica confocal microscope (Stellaris 8).
CSF collection
For CSF collection, mice were anaesthetized through i.p. injection of a mixture of ketamine (50 mg kg−1) and xylazine (5 mg kg−1). The dorsal neck was shaved and sterilized. A 1-cm incision was made at the base of the skull, and the dorsal neck muscles were separated using forceps to expose the cisterna magna. A custom-pulled micropipette (0.75/1 1brl GF; Stoelting) was used to penetrate the dura to collect CSF.
Brain tumour inoculation and IVIS imaging
Tumour inoculation was carried out as previously described with slight modifications9. Mice were anaesthetized through i.p. injection using a mixture of ketamine (50 mg kg−1) and xylazine (5 mg kg−1). Mouse heads were shaved and scalps were sterilized. A midline scalp incision was made and a burr hole was drilled 2 mm lateral to the sagittal suture and 0.5 mm posterior to the bregma with a 25-G needle. Then, mice were placed in a stereotaxic frame. A 10-μl Hamilton syringe loaded with 3 μl GL261–Luc cells (105 cells) was inserted into the burr hole at a depth of 4 mm from the surface of the brain and left to equilibrate for 1 min before infusion. A micro-infusion pump (World Precision) was used to infuse at 1 μl min−1. The syringe was left in place for another minute after the infusion was finished. The skin was stapled and cleaned. Following intramuscular administration of an analgesic (meloxicam and buprenorphine, 1 mg kg−1), mice were placed in a heated cage until full recovery. We tracked tumour size weekly thorough IVIS imaging. Mice were anaesthetized using isoflurane and injected through the i.p. route with d-luciferin potassium salt bioluminescent substrate (PerkinElmer, 122799, 200 μl, 30 mg ml−1). After 10 min, mice were imaged using the IVIS Spectrum In Vivo Imaging System (PerkinElmer).
Preparation of tissue for analysis of CT2A–BFP tumour antigen drainage
i.c. CT2A–BFP tumours were analysed 14 days after injection. Tumours, meninges, IngLNs, dCLNs and sCLNs were collected. Mononuclear cells were isolated and stained.
Parabiosis
Parabiosis was carried out as previously described with slight modifications3. Naive or IVT-immunized C57BL/6 mice of similar age and weight were anaesthetized with a mixture of ketamine (50 mg kg−1) and xylazine (5 mg kg−1). After shaving the corresponding lateral aspects of each mouse, the skin was cleaned and sterilized with an alcohol prep pad and Betadine surgical scrub. Matching skin incisions were made from just above the knee upwards to the olecranon, and two mice were sutured together with Ethicon 5-0 coated Vicryl absorbable sutures. Then, the skin was stapled and Neosporin + Pain Relief Ointment was applied on the incisions. During the surgery, mice were kept on heating pads and continuously monitored until recovery.
Ligation of sCLNs or dCLNs
For ligation of LNs, mice were anaesthetized with a mixture of ketamine (50 mg kg−1) and xylazine (5 mg kg−1) and the rostral neck was shaved and sterilized. A 2-cm incision was made, and the sCLNs and dCLNs were sequentially exposed using forceps. Their afferent lymph vessels were cauterized or kept intact on the basis of the experiment conditions. Then, the incision was closed with a 5-0 Vicryl suture, and mice were subjected to the same postoperative procedures as above.
rAAV transduction and imaging
Wild-type mice were IVT injected with rAAV (3 × 1011 viral genomes) with PBS, VEGFC (1 μg) or sVEGFR3 (1 μg). Then, 1 month later, these mice were rechallenged with rAAV-RFP (3 × 1011 viral genomes). Eyes were either imaged on the Phoenix Micron IV or collected 1 month after rAAV-RFP transduction and fixed with 1% PFA overnight at 4 °C. The retina whole mount was carefully dissected and imaged with a Leica confocal microscope (Stellaris 8).
Three-dimensional imaging of solvent-cleared organs
iDISCO was carried out as previously described (http://www.idisco.info)22. The following antibodies were used: goat anti-mouse VEGFR3 (R&D, No. AF743,1:400), rat anti-mouse LYVE1 (R&D, MAB2125,1:400), rabbit anti-mouse LYVE1 (AngioBio, No. 11-034,1:200), mouse anti-human VEGFR3 (Santa Cruz Biotechnology, SC-28297, 1:200), rabbit anti-human LYVE1 (Angio-Proteomie, 102-PA50S, 1:200), goat anti-mouse IgG–AF647 (Invitrogen, A21235, 1:500), donkey anti-goat IgG–AF647 (Invitrogen, A21447, 1:500), goat anti-rabbit IgG–AF555 (Invitrogen, A21428, 1:500). Subsequently, the transparent optic nerves with optic nerve sheaths were imaged using a Leica confocal microscope (Stellaris 8). Three-dimensional rendering was completed using Imaris 8 software (Oxford Instruments).
Spatial transcriptomics
The 10X Visium Spatial Gene Expression for FFPE slide (PN-1000185) and associated protocols (CG000483) from 10X Genomics were used. Mouse eyeballs were fixed47, processed and sectioned48 as previously described. Transverse sections of 5 μm in thickness of eyeball were cut using a microtome (RM2255, Leica Biosystems) and carefully placed within the fiducial frame on the Visium slide, and then sections were air dried at room temperature overnight and stored in a desiccant container before spatial transcriptomics experiment. The FFPE sections were baked, stained with haematoxylin–eosin and then imaged using a Keyence bz-x800 all-in-one fluorescence microscope. Then, cell permeabilization and library preparation was carried out following the Visium Spatial Gene Expression FFPE User Guide using the supplied reagents (10X Genomics). The generated libraries were sequenced and analysed using Space Ranger (version 2.1.0), and data were analysed using Seurat 4.9.9.9040.
Image processing and analysis
Quantitative analysis of rAAV-infected cells was carried out using either FIJI or ImageJ image-processing software (NIH or Bethesda).
Statistical analysis
All statistical analyses were carried out using GraphPad Prism software. Data were analysed with a two-tailed unpaired Student’s t-test or a one-way ANOVA with multiple comparisons testing (Dunnett) with Prism software. Statistical significance is defined as *P < 0.05, **P < 0.01 and ***P < 0.001.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
