Peptide GGGCGGSK(azide) with a free N-terminus and C-terminal amide was synthesized following standard solid phase synthesis protocols (24). All Fmoc amino acids were purchased from Chempep, Inc. Fmoc-Lys(azide)-OH was used as a building block to provide the bioorthogonal handle. Peptides were purified by reverse phase HPLC. Their identity was confirmed by LC-MS prior to maleimido-DFO coupling to the cysteine thiol. Peptides were further purified by reverse phase HPLC and their identity was confirmed again by LC-MS.

Ca²⁺-independent heptamutant sortase A derived from S. aureus (10 μM final concentration, 10x stock in 50 mM Tris, pH 7.4, 150 mM NaCl) and probe (1 mM final concentration, 50x stock) were added to VHH-X118 (200 μM final concentration) in phosphate-buffered saline (PBS). The resulting mixture was incubated at 4°C overnight. Ni-NTA (0.5 ml) was added to the reaction mixture and incubated for 20 minutes to remove sortase and unreacted VHHs. The mixture was centrifuged and the supernatant was collected and purified by size exclusion chromatography (Superdex 75- GE Life Sciences), and analyzed by SDS-PAGE and LC-MS. PEGylated VHH-X118 was generated by reacting the bioorthogonal azide group with dibenzocyclooctyne DBCO-(PEG)₂₀ overnight. The end product was analyzed by SDS-PAGE to confirm efficient coupling.

IAV was quantified by flow cytometry, using a method adapted from one previously described (25). Briefly, confluent Madin-Darby canine kidney cells were infected in triplicate with 2-fold serial dilutions of IAV WSN/33 (in DMEM, 0.2% BSA) for 1 hour. The inoculum was removed and replaced with DMEM, 0.2% (w/v) BSA for 5 hours. Cells were washed with PBS, trypsinized, and fixed with 4% formaldehyde in PBS. Cells were stained with 1 μg/ml VHH62 (anti-IAV NP)-Alexa Fluor 647 under permeabilizing conditions. Fluorescence was quantified using a BD Accuri C6 Plus. The NP-positive population was determined by comparison with an uninfected control population ( Figure S1A ). Data were processed using the FlowJo software package (TreeStar Inc). A linear regression model was applied ( Figure S1B ). The slope of the line of best fit was used to determine the percentage of infected cells, which was multiplied by the number of cells per well to yield the number of viral particles per well.

Age-matched, 6-week old, female CD45.2 C57BL/6J mice (n≥3 in each group) were purchased from the Jackson Laboratory. Mice were anesthetized with isoflurane and infected by the intranasal route with 4 x 10⁴ infectious units of IAV WSN/33 diluted in PBS. Control mice were similarly given an equal volume of PBS intranasally. Infection was tracked by monitoring daily weight loss. Mice were euthanized with CO₂ when weight loss exceeded 25% of initial body weight and/or animals displayed signs of severe distress, or if no weight was recovered 9 dpi.

IAV-infected and control mice were anesthetized using isoflurane and injected via the retro-orbital route with approximately 25 μCi of ⁸⁹Zr-VHH-X118-PEG₂₀. PET-CT procedures have been described in detail (21, 26). Briefly, approximately 24 h post-administration of ⁸⁹Zr-VHH-X118-PEG₂₀, mice were anesthetized using 2% isoflurane in O₂ at a flow rate of approximately 1 L/minute. Mice were imaged with a G8 PET-CT small-animal scanner (PerkinElmer). PET images were acquired over a 20 minute period, which was followed by approximately 2 minutes of CT acquisition. As a standard for absolute intensity, a fixed quantity of radioactivity was imaged using 5-fold serial dilutions of radioisotope in PCR strip tubes.

We processed and quantified PET images using VivoQuant software. The CT scan was used as a guide to generate 3D regions of interest (ROIs) to represent regions corresponding to the lungs and MLN. ROIs were drawn for each image corresponding to the absence of CT signal in the ribcage, surrounding the heart as a means of identifying pulmonary space. We used preset CT values to view the area corresponding to the lungs, and preset CT intensities (-1400, 400) were used to generate the lung ROIs. For the MLN region, ROIs were generated by creating a spherical region with a 10-pixel diameter centered around the point of highest PET intensity of MLN. An additional ROI was drawn in the quadriceps muscle of the hind leg of each mouse, avoiding bones and LNs, which was subtracted as background. Once all ROIs were generated, statistical information for each ROI containing mean PET signal, was exported and processed. We normalized signal intensities to background signal for each mouse by using the ratio between PET signal in the lungs or MLN to PET signal in the quadriceps muscle. The level of significance was determined using a Student’s t-test.

Mice were euthanized by asphyxiation with CO₂ and spleens were extracted by dissection. Spleens were homogenized into single-cell suspensions. Red blood cells were removed by hypotonic lysis. T cells were isolated from these splenocytes using the Dynabeads^® Untouched™ Mouse T Cells Kit following the suppliers’ recommended procedures. The lungs of day 9-infected CD45.2 C57BL/6J mice were extracted by dissection, and T cells were isolated using the Dynabeads^® Untouched™ Mouse T Cells Kit with additional incubations with anti-CD326 and anti-CD31 antibodies to deplete epithelial and endothelial cells, respectively.

Aside from virus quantification assays, all data were acquired on a Fortessa instrument (BD Biosciences) and analyzed using FlowJo software Cells obtained from the lungs and MLN of CD45.2 C57BL/6J mice were used for flow cytometry. Cells were resuspended in PBS (137 mM NaCl, 2.7 mM KCl, 10 mM Na₂HPO₄, 1.8 mM KH₂PO₄) with 2% (v/v) fetal bovine serum and passed through 40-μm cell strainers to obtain single-cell suspensions prior to antibody staining (30 minutes at 4°C). All antibodies were obtained from BioLegend (San Diego, CA). Student’s t-test was applied for statistical analysis.

96-well ELISpot plates (BD ELISPOT Mouse IFNγ ELISPOT Set, BD Biosciences, San Jose, CA) were coated with an IFNγ capture antibody (BD Biosciences) in PBS overnight at 4°C, followed by incubation with complete RPMI-1640 medium for 2 hours at room temperature (RT). Single cell suspensions from selected organs of IAV-infected and uninfected mice were prepared. Red blood cells were removed by hypotonic lysis. Quadruplicate ELISpot wells containing mononuclear cells, were supplemented with the IAV NP peptide (366-374; ASNENMETM) (2 μg/ml), to serve as a H2-D^b-restricted epitope from the Influenza A/PR/8/34 nucleoprotein (27, 28). As a control, medium without added IAV peptide was used. ELISpot plates were incubated at 37°C for 18 hours, washed and incubated with a biotinylated IFNγ detection antibody (BD Biosciences) for 2 hours, followed by incubation with a streptavidin-horse radish peroxidase (HRP) conjugate (BD Biosciences) for 1 hour at RT. ELISpot plates were developed with 3-amino-9-ethyl-carbazole substrate (BD ELISPOT AEC Substrate Set) and dried. Spots were counted using an ImmunoSpot Analyzer.

Splenocytes from CD45.2 C57BL/6J mice were cultured in plates pre-coated with anti-CD3 (5 µg/ml) and anti-CD28 (1 µg/ml) antibodies in complete RPMI-1640 medium supplemented with 250 ng/ml of mouse IL-2 produced in house. Following activation for 24 hours, cells were transferred to fresh dishes without anti-CD3 and anti-CD28 in complete RPMI-1640 supplemented with 250 ng/ml of mouse IL-2, followed by two additional days of culture. Cells were washed three times with PBS and cell numbers were determined prior to transfer experiments.

Cells were labeled with non-PEGylated ⁸⁹Zr-VHH-X118 in PBS at 4°C for 20 minutes with constant agitation. Cells were then washed three times with PBS to remove unbound ⁸⁹Zr-VHH-X118. Labeled cells (6 x 10⁶) were then transferred into the retro-orbital plexus of each CD45.1 C57BL/6J mouse. Images were acquired as previously described 1 and 24 h post-T cell transfer.

T cells purified from lungs of day 9-infected CD45.2 BL/6J mice were stained with FITC anti-mouse CD8 and Alexa700 anti-mouse CD45.2 antibodies (BioLegend). The CD8⁺CD45.2⁺ T cell population was sorted on a BD FACSAria III sorter.

All animal protocols were conducted in accordance with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All animals were maintained according to the guidelines of the Animal Resources Children’s Hospital. These studies were approved by the Boston Children’s Hospital Institutional Animal Care and Use Committee (protocol #16-12-3328). All infections and PET imaging procedures were performed under isoflurane anesthesia and all efforts were made to minimize suffering.

In order to follow CD8⁺ T cells noninvasively over the course of IAV infection, we used a VHH specific for murine CD8α, VHH-X118 (21). We produced two versions of ⁸⁹Zr-labeled VHH-X118, both made via modification of the VHH’s C-terminal sortase recognition motif (LPETG). This modification allows the use of a peptide containing an N-terminal GGGC sequence equipped with a radiometal chelator to serve as a nucleophile in a sortase A-mediated transpeptidation reaction (29). By incorporation of maleimido-desferrioxamine (DFO), we installed ⁸⁹Zr for PET imaging. We also generated a PEGylated version of ⁸⁹Zr-labeled VHH-X118 by using an azide-substituted lysine in the course of synthesis of the nucleophile to incorporate polyethylene glycol (PEG) moieties for in vivo injections, effectively reducing non-specific retention in the kidneys ( Figure 1 ) (21). The agent used for ex vivo labeling of CD8⁺ T cells prior to use in transfer experiments, as described below, did not include a PEG moiety.

To establish the whole-body distribution of CD8⁺ T cells during IAV infection, we acquired CD8⁺ PET images using ⁸⁹Zr-labeled VHH-X118, and in parallel, tracked weight loss after intranasal inoculation of CD45.2 C57BL/6J mice (n≥3 for each group) with a sub-lethal dose (4 x 10⁴ infectious units) of influenza A/WSN/33 virus (H1N1). In uninfected mice, CD8⁺ T cells were distributed in the cervical, axillary and brachial (lung draining), mediastinal (lung draining), popliteal, renal, iliac, and inguinal lymph nodes, as well as in the spleen, consistent with previous observations (21). PET signals in the organs of elimination (kidneys, liver and bladder) and the site of injection (retro-orbital plexus) are non-specific and common occurrences when using nanobodies as imaging agents (21, 30, 31). During the first week of IAV infection, mice experienced weight loss, paralleled by increases in PET signal in the mediastinal lymph node (MLN) and lungs ( Figure 2 ). Attribution of PET signals to particular anatomical structures was confirmed by imaging dissected organs, including MLN, lungs, heart, and thymus ( Figure S2 ). At 4 days post-infection (dpi), mice showed a striking increase in CD8⁺ T cells in the area corresponding to the MLN and in the draining lymph nodes, as inferred from the PET signal. At the peak of infection (6 dpi) as determined by weight loss, a diffuse pattern of CD8 signal was present in the lungs, likely caused by CD8⁺ T cell migration from secondary lymphoid organs towards particular foci of infection. This signal became more concentrated and localized in specific regions of the lungs over the course of the infection, starting at 9 dpi. The PET signal decreased around 18 dpi, when mice regained weight in the recovery phase, and finally disappeared from the lungs by 21 dpi.

In order to determine whether the PET signals observed in the lungs and MLN of mice over the course of infection constituted measurable significant increases, we quantified the PET signal from the animals in Figure 2 using VivoQuant software. Specifically, we generated 3D regions of interest (ROIs) representing the lungs ( Figure 3A ; cyan), MLN ( Figure 3B ; cyan), and quadriceps muscle (control; green) to quantify the PET signal within the volume of the ROI. Mean intensity was calculated by dividing the PET signal by the volume of the ROI. This value was further normalized within each mouse to the intensity of the quadriceps muscle, in order to account for possible variations in multiple experiments and across mice. CD8 signal intensity in the lungs increased during infection and remained elevated from days 6 through 12 of IAV infection. Lung signal decreased after day 12, as mice recovered from IAV infection. These results correlate with morbidity, as mice reached the peak of their weight loss in this interval ( Figure 2 , inset graphs). Similarly, we observed a trend towards an increase in CD8 signal intensity in the MLN beginning at 4 dpi, which reached significance (P<0.05) between days 9 and 15 of the infection, before eventually decreasing during the recovery phase of the infection ( Figures 3C, D ). An increase in CD8 signal intensity should be detectable in the MLN of infected mice prior to its appearance in the lungs, as T cells are activated in lymph nodes prior to migrating to the lungs in response to chemokines (32).

To characterize the population of infiltrating T cells at 9 dpi and uninfected mice, we performed flow cytometry on cells isolated from the lungs and MLN. The ratio of infiltrating CD4⁺ versus CD8⁺ T cells was determined by gating on the CD3⁺CD45⁺ population ( Figure 4A ). The number of infiltrating lymphocytes in the lungs of day 9-infected mice was significantly increased ( Figure 4C ), in agreement with the observed increase in PET signal. To better identify the T cell subset present in the infiltrating population and to establish the percentage of CD44⁺CD62L⁺ [central memory T cells (T_CM)] and CD44⁺CD62L^- [effector T cells (T_EFF/T_EM)] cells, we gated on either the CD8⁺ or CD4⁺ T cell population ( Figure 4B ). The majority of infiltrating CD8⁺ and CD4⁺ T cells in the lungs were CD44⁺CD62L^- (T_EFF/T_EM). No significant increase in CD8⁺CD44⁺CD62L⁺(T_CM) was detected in the lungs of the infected animals ( Figure 4C ).

We next confirmed that IAV-specific T cell responses could be detected in the lungs, MLN, and spleen by using a peptide corresponding to the immunodominant H-2D^b-restricted IAV NP epitope, ASNENMETM, in an IFNγ ELISpot assay. As expected, we saw an increase in the number of IAV-specific CD8⁺ T cells in the lungs, MLN, and spleen in infected mice ( Figure 4D ). We detected low levels of IFNγ-producing cells in the media-only controls in the lungs from infected animals, presumably due to the persistence of antigen and/or previously activated IFNγ-producing T cells.

Transfer of virus-specific CD8⁺ T cells can clear IAV in a B cell-deficient mouse model (33). In such settings, the tissue distribution of donor CD8⁺ T cells remains unclear. Does the inflamed environment of the lungs provide signals that direct and retain T cells? Do T cells instructed to deal with a pulmonary insult display lung homing properties, independent of whether the lungs are infected or not? In any case, little is known about the in vivo distribution of T cells immediately after transfer into recipients. We approached these questions by transfer of ⁸⁹Zr anti-CD8α-labeled CD8⁺ T cells to trace their distribution in the recipients at 1 hour and 24 hours post-transfer, using whole-body immuno-PET imaging. T cells were harvested and purified from the lungs of IAV-infected mice 9 dpi, with additional negative selection steps to remove epithelial and endothelial cells, followed by cytofluorimetry to assess purity of the T cell population to be transferred. We likewise harvested splenocytes from uninfected control mice and purified CD8⁺ T cells from them. These naïve T cells were then activated on plate-bound anti-CD3 and anti-CD28, in medium supplemented with IL-2 (34). CD8⁺ T cells from day 9-infected mice or activated CD8⁺ T cells from control mice were labeled in suspension with ⁸⁹Zr-VHH-X118. Labeled cells were then transferred into either IAV-infected (4 dpi) or uninfected control mice via retro-orbital injection. PET images were acquired at 1 hour and 24 hours post-transfer ( Figure 5A ).

Upon transfer into day 4-infected recipient mice, ex vivo activated, control CD8⁺ T cells initially localized to the lungs and spleen ( Figure 5B , top panels). After 24 hours, the signal corresponding to CD8⁺ T cells decreased in the lungs. Activated CD8⁺ T cells generated from control splenocytes transferred into naïve mice localized mostly to the spleen and remained there after 24 hours. CD8⁺ T cells from day 9-infected mice initially migrated to the lungs and spleen when transferred into day 4-infected and uninfected recipients, imaged 1 hour post-transfer ( Figure 5B , bottom panels). After 24 hours, a signal corresponding to these CD8⁺ T cells remained detectable in the lungs of day 4-infected mice and appeared to increase in intensity in the spleen and MLN. Uninfected mice also retained a signal corresponding to CD8⁺ T cells from infected mice in their lungs 24 hours post-transfer, but to a lesser extent, probably due to the lack of an inflammatory environment. CD8⁺ T cells harvested from day 9-infected mice did not initially populate the spleens of uninfected mice in large numbers, nor did they appear to migrate there over the 24 hour observation period.