Antigen and G-Protein Coupled Receptor signaling differentially control CD8 T cell motility immediately before and after virus clearance in a primary infection

In mice, experimental influenza virus infection stimulates CD8 T cell infiltration of the airways. Virus is cleared by day 9, and between days 8 and 9 there is an abrupt change in CD8 T cell motility behavior transitioning from low velocity and high confinement on day 8, to high velocity with continued confinement on day 9. We hypothesized that it is loss of virus and/or antigen signals in the context of high chemokine levels that drives the T cells into a rapid surveillance mode. Virus infection induces chemokine production, which may change when the virus is cleared. We therefore sought to examine this period of rapid changes to the T cell environment in the tissue and seek evidence on the roles of peptide-MHC and chemokine receptor interactions. Experiments were performed to block G protein coupled receptor (GPCR) signaling with Pertussis toxin (Ptx). Ptx treatment generally reduced cell velocities and mildly increased confinement, except on day 8 when velocity increased and confinement was relieved, suggesting chemokine mediated arrest. Blocking specific peptide-MHC with monoclonal antibody unexpectedly decreased velocities on days 7 through 9, suggesting TCR/peptide-MHC interactions promote cell mobility in the tissue. Together, these results suggest the T cells are engaged with antigen bearing and chemokine producing cells that affect motility in ways that vary with the day after infection. The increase in velocities on day 9 were reversed by addition of specific peptide, consistent with the idea that antigen signals become limiting on day 9 compared to earlier time points. Thus, antigen and chemokine signals act to alternately promote and restrict CD8 T cell motility until the point of virus clearance, suggesting the switch in motility behavior on day 9 may be due to a combination of limiting antigen in the presence of high chemokine signals as the virus is cleared.


132
133 Surgical set-up: The hair is removed with a shaver from one hind leg, thigh to groin, exposing the 134 skin for the MouseOX sensor. The hair is removed from the thoracic area with scissors and/or a 135 shaver. The animal is placed in a supine position, on a warming blanket, once a surgical plane 136 of anesthesia is determined by lack of both pedal and palpebral reflex. The coat is opened from 137 below the chin to the top of the ribcage. The salivary glands are separated to reveal the muscles 138 covering the trachea. Using round forceps, the muscles are separated to expose the trachea.
139 The forceps are then inserted beneath the trachea to lift and separate it from the muscle and 140 mouse body. A small flexible plastic support is placed in the space created by the forceps to 141 permanently hold the trachea above and separated from the muscle, surrounding tissue and coat.

142
143 The animal is moved to the previously warmed stage. A small incision is made between the 144 cartilage rings below the larynx. The steel cannula is inserted into the opening in the trachea until 7 146 The cannula is secured on the stage with a support that holds it in position, so it is correctly aligned 147 with the trachea. It is held in place with 2 screws that prevent it from moving in transport, or during 148 the attachment of the respirator. The forepaws are secured to the stage with surgical tape to 149 maintain position of the mouse body on the stage. A few drops of saline are placed on the 150 exposed tracheal tissue to prevent drying.

151
152 Maintaining blood O2 saturation and temperature: The cannula is quickly attached to a Harvard 153 Inspira Advanced Safety Ventilator, and both 100% O2 and 0.5% isofluorane flow is started, 154 according to mouse weight. The MouseOX Plus thigh sensor is attached to the exposed thigh 155 and monitoring is begun immediately. Oxygenation levels are maintained at 95% and heart rate 156 ranges between 250 and 600 beats per minute. The rectal body-temperature is continuously 157 monitored and maintained using a small animal temperature controller that is connected to a rectal 158 probe and a feedback-regulated rodent heating pad. After achieving stable physiology, and 159 verification of lack of both pedal and palpebral reflex, the pancuronium bromide (0.4 mg/kg) is 160 administered, based on body weight. The saline covering the exposed trachea is blotted away 161 and replaced with 0.05% agarose to seal the exposed area. Once the agarose is solidified, a 162 support ring is placed over the imaging area and covered with a piece of plastic wrap.
163 Approximately 5 ml of water is pooled over the imaging area to submerse the objective.  213 If the T cells are responding to chemokine signals that direct them towards the infected epithelium 214 on days 6-8, then we would predict that interference with these signals would result in delayed 215 migration to the airway surfaces. We therefore developed a computational approach to measure 216 the depth of the T cells in the trachea relative to the airway surface. The measurements were 217 conducted from days 6-9 of infection. In the control mice, T cell abundance closest to the airways 218 peaked on day 7 and actually decreased on days 8 and 9 ( Figure 2). Consistent with the   14 249 Re-introduction of antigen: We reasoned that the changes on day 9 were the result of a 250 significant reduction in signals received from ABC as the virus is cleared [9]. To address this, we 251 re-introduced antigen in the form of soluble SIINFEKL peptides on days 8 and 9. The introduction 252 of peptides had no effect on motility parameters on day 8 when virus and, presumably, ABC are 253 still abundant. However, on day 9, the introduction of peptides resulted in significantly decreased 254 velocities and displacement, increased numbers of cells arrested, and heightened confinement 255 compared to vehicle controls ( Figure 4A-C). This is consistent with the results of pMHC blocking 256 on day 8, and strongly suggests that the balance of antigen signals in the presence of continued 257 chemokine signaling result marked changes in velocity and confinement that occur between day 258 8 and day 9. This is consistent with the interpretation that the still highly activated T cells switch  16 280 narrow in cross section, and there is dense extracellular matrix underlying the epithelial surface), 281 high CXCL9 and CXCL10 chemokine concentrations at the epithelium, which can contribute to 282 arrest, and presumably engagement with antigen bearing cells in the form of either APC or 283 infected targets. This is discussed further below. 284 285 On day 9 when virus is cleared, the CD8 T cells normally exhibit an increase in cell velocities, and 286 yet remain confined. Imaging of this behavior shows rapid extension and retraction of T cell 287 protrusions, but little distance traveled (compare Supplemental Movies S1 and S2). We have 288 interpreted this collective behavior as rapid surveillance for residual antigen and virus by the 289 highly activated effector T cells. Treatment with Ptx at this time point significantly decreased all 290 the motility parameters, suggesting a requirement for GPCR signals for optimal surveillance of 291 the tissue as the infection resolves. How chemokines and chemokine receptors affect changes to 292 T cell motility on day 9 is unresolved. The chemokines CXCL9 and CXCL10 are expressed 293 constitutively by airway epithelium, and increase during infection [25,26]. In addition to potentially 294 providing chemotactic and chemokinetic signals, they may be needed for integrin activation [27], 295 each of which could contribute to the cell behavior we observe, though we have not formally tested 296 these mechanisms in our model.

298
The experiments blocking MHC with specific monoclonal antibodies showed significant effects on 299 CD8 T cell motility. However, the effects were not entirely as expected. If we take the hypothesis 300 that engaging antigen bearing cells will cause cells to arrest, at least briefly, as has been observed 301 in other T cell motility studies [28], then we would predict that blocking antigen signals would 302 partially or completely reverse cell arrest. On day 6 there were no differences in the motility 303 parameters between pMHC mAb treated mice and isotype control mice, likely because few T cells 304 had found their targets. But on days 7-9, velocities were consistently reduced with pMHC mAb 305 treatment, with greatest magnitude of the effect on day 8, a time when we expect maximal