Manipulating Stress Receptor Signaling to Enhance Immunosuppression and Prolong Survival of Vascularized Composite Tissue

: Vascularized composite tissue allotransplantation (VCA) can replace severely 28 damaged body parts but the unavoidable toxicity of high doses of immunosuppressive drugs, 29 such as tacrolimus, required results in significant morbidity. Here we tested whether we could 30 suppress immune activity in a mouse model of VCA by mimicking the natural immune 31 suppression generated by nervous system-induced signaling of adrenergic receptors (AR) by 32 using a safe and well-studied β-AR agonist (terbutaline). Using wild-type and β 2 -AR-knockout 33 (KO) mice, we found that increased β 2 -AR signaling results in delayed rejection in VCA 34 recipients, even with subtherapeutic doses of tacrolimus, and this was associated with changes 35 in immune contexture, expression of pro-inflammatory cytokines and chemokines, and 36 function of endothelial adhesion molecules. We propose that β-AR agonists can be used safely 37 to mimic the natural suppression of immune responses generated by adrenergic stress signaling 38 and thereby reduce the dose needed of other more toxic immunosuppressive drugs. 39 40 41


Introduction
threatening morbidities from toxic immunosuppressive drugs. 87 The nervous and immune systems have been found to interact closely in host defense 88 and stress responses 17 . Although the relationship of the hypothalamus-pituitary-adrenal (HPA) 89 axis and cortisol has been well studied 18,19 , the natural role of the sympathetic nervous system 90 in regulating immune responses is not fully understood 20,21 . Sympathetic and parasympathetic 91 nerves are found innervating immune organs and near immune cells throughout the body. 92 Extensive research now shows that neurotransmitter interactions between norepinephrine (NE) 93 and β-adrenergic receptors (AR) regulate the immune system and that dysregulation of neuro-94 immune signaling pathways contributes to the pathogenesis of immune disorders [22][23][24] . 95 Recently, we have shown that pharmacologic blockade of the β-AR with the common 96 pan-β-blocker propranolol reduces tumor growth rates by interfering with the suppressive 97 activity of adrenergic stress on adaptive immune responses. Specifically, we found that 98 blocking β-AR with propranolol increased glucose uptake and glycolysis needed for metabolic 99 reprogramming during T cell activation and anti-tumor immunity [25][26][27][28] . We also found that β2-100 AR has an important role for immune modulation of T cells and myeloid-derived suppressor 101 cells (MDSC) 29 . The strong, naturally occurring immunosuppressive potential of β-AR 102 5 signaling is consistent with our observations that adrenergic stress or addition of β-AR agonists 103 can suppress graft versus host disease (GVHD) following bone marrow transplantation 104 (BMT) [29][30][31] . 105 These data led us to investigate whether targeting the natural ability of β-AR to regulate 106 immunity in response to adrenergic stress signaling could be used to more safely promote an 107 immunosuppressive environment in transplanted grafts following VCA, thus allowing a 108 reduction of the dose of the more toxic immunosuppressant drug tacrolimus. Here, we 109 investigated the impact of targeting β2-AR using the β2-blocker terbutaline, on graft rejection 110 rate and immune contexture using wild-type and β2-AR-knockout (KO) mice. We found that 111 increased β2-AR activation results in delayed rejection responses in VCA recipients without 112 detectable toxicity and this occurred through mechanisms involving pro-inflammatory 113 cytokines and chemokine suppression as well as inhibition of endothelial adhesion molecules. 114 More importantly, we were able to extend graft survival using a subtherapeutic dose of 115 tacrolimus combined with β2-AR agonist. Together, these data reveal an immediately feasible 116 strategy which can be tested in patients receiving VCA or other types of allotransplants.

119
Safety and cross-reactivity with tacrolimus of selective 2-AR agonist 120 Our previous work has established a reliable and consistent platform using a pre-clinical 121 murine model of hind limb VCA 32 to investigate novel therapies to prolong transplant survival 122 The transplanted graft consisted of skin, fat, muscle, bone and blood vessels (Supplementary 123 Fig. 1a), a complex combination of tissues similar to those which are often used in VCA. 124 While we know from the literature and its clinical safety profile that terbutaline is 125 considered to be a safe drug, we wanted to be sure that there were no cardiac physiology 126 problems generated by this β2-AR agonist alone, or in combination with the drug tacrolimus, 127 6 a drug that causes significant immunosuppression and commonly used in the transplant 128 setting 13,33 . We assessed blood pressure (BP) and heart rate (HR) with a non-invasive tail cuff 129 system 34 . Normal mice were acclimated to the device for 10 days prior to collecting readings.

130
No statistical difference was detected in systolic/diastolic BP or HR between mice treated daily 131 with β2-AR agonist or vehicle over a course of 11 days ( Supplementary Fig. 1b). We also 132 measured plasma levels of tacrolimus in the mice treated with two different doses; full dose 133 (fTac, i.e., a dose known to generally maintain allografts long term 32 ), and half dose (hTac, i.e., 134 a dose that only delays graft rejection) were measured. As expected, significantly higher drug 135 concentrations were measured in the fTac injected group compared to the hTac group, but there 136 was no significant difference in tacrolimus concentrations between mice treated with 137 tacrolimus alone and mice treated with tacrolimus and β2-AR agonist combination 138 ( Supplementary Fig. 1c). In addition, no changes of body weight and physical appearance were 139 detected suggesting that it is safe to use terbutaline alone and in combination with tacrolimus 140 without drug interaction.  VCA, but over 85% of the vehicle injected recipients showed grade 3 or 4 rejection 7 days after 151 VCA. In contrast, β2-AR agonist-treated recipient mice had less severe rejection (grade 2) at 152 7 day 7, which was further improved by the addition of subtherapeutic dose of tacrolimus (hTac), 153 with some grafts surviving with grade 2 rejection at day 10 ( Fig. 1b). Although no evidence of 154 gross rejection was observed at day 5, various histologic rejection grades were detected with 155 H&E. β2-agonist treatments delayed rejection responses with/without subtherapeutic dose of 156 tacrolimus compared to the vehicle injected group (Fig. 1c).

157
To determine whether infiltrating cells originated from the donor or the recipient, we 158 took advantage of differential CD45 isoform usage of C57BL/6 recipient (CD45.1) and 159 BALB/c donor (CD45.2) mice. By day 7 post-VCA, over 90% of leukocytes within grafts with the β2-agonist treated group ( Supplementary Fig. 3e). The majority of T cells in the grafts 176 7 days after VCA were EM cells in both groups, and these values were higher than healthy 177 8 donor (pre-transplanted grafts) control specimens ( Supplementary Fig. 3f).

178
To test whether we could use a β-AR agonist to mimic natural NE and β-AR interactions 179 and reduce the dose of tacrolimus needed for immunosuppression, we tested tacrolimus at half 180 dose (hTac) in combination with either vehicle or β2-agonist. Grade 4 graft rejection was 181 observed 10 days after VCA in the hTac+vehicle group, while the hTac+β2 group showed 182 mainly only Grade 2 with partial Grade 3 rejection with intact skin histology (Fig. 2e).

183
Abundant CD4 + T and CD8 + T cell infiltration was present 7 days after VCA in the 184 hTac+vehicle group, and then the values dropped significantly 3 days later. β2-agonist 185 treatments significantly decreased the number of infiltrating CD4 + T and CD8 + T cells at day 7 186 compared to the hTac+vehicle group, but more CD4 + T cell infiltration was found in the β2-187 agonist treated group than the vehicle injected group at day 10 representing remnant immune 188 responses in the graft (Fig. 2f). There was no statistical difference in the proportion of CM and 189 EM T cell populations in grafts 7 days after VCA between two groups using hTac (Fig. 2g).

191
In contrast to observations within grafts by flow cytometry (Fig. 2d, g), β2-AR agonist 192 treatment significantly decreased the representation of CD4 + and CD8 + memory (CM and EM) 193 T cell populations in the recipient's systemic blood compartment (spleen) at day 5 (i.e., before 194 the emergence of signs of rejection; Fig. 3a), however, by day 7 (i.e., once gross signs of 195 rejection are apparent) these differences were lost (Fig. 3b). This finding was more pronounced 196 in the EM T cell population in mice treated with hTac. Additional decreases in the CD4 + T and 197 CD8 + T EM populations were found with β2-agonist treatment at day 7 (Fig. 3c), and these 198 differences are lost by day 10 (Fig. 3d). β2-agonist decreased the compositions of Th1 199 population significantly compared to the vehicle group at day 5 and 7 without tacrolimus (Fig. 200 3e) and at day 7 and 10 with hTac (Fig. 3f). Interestingly, higher Treg (CD4 + CD25 + Foxp3 + ) 201 levels in the graft (Fig. 2c) and the body (spleen; Fig. 3e) did not predict a better prognosis for suppression of memory T cell populations were not detected in the blood (Fig. 4b), and there 224 was no survival benefit of grafts given β2-agonist following cessation of tacrolimus (Fig. 4c).

225
However, with a scenario using hTac with either vehicle or β2-agonist (Fig. 4d), the dose of 226 tacrolimus was reduced by half 14 days after a full dose of tacrolimus daily treatment (fTac to 227 hTac). Even though no statistical difference on CD4 + and CD8 + memory T cell populations 228 was detected between two groups in the blood (Fig. 4e), there was a significant graft survival 229 benefit in the β2-agonist group compared to the vehicle group (Fig. 4f).  donor were examined. WT donors were exposed to 8 Gy (gray) irradiation the day before VCA 252 to eradicate BM cells, and then grafts including a femur was transplanted to β2-AR KO 253 recipients followed by vehicle or β2-agonist injections. Both groups showed mild xerosis of 254 grafts with intact skin anatomy (Fig. 6a), and distinct effects of β2-agonist treatment were lost 255 in the CD4 + and CD8 + memory T cell (Fig. 6b), Th1 and Treg (Fig. 6c) populations 7 days after 256 VCA. However, significantly lower numbers of CD4 + and CD8 + T cell infiltrations were found 257 in the β2-agonist treated graft than the vehicle group (Fig. 6d). Further, β2-AR KO donor was 258 irradiated, and then WT BMT was performed to generate a chimeric model including β2-AR 259 KO stromal cells and WT BM, and vice versa (Fig. 7a). Grafts composed of β2-AR KO stromal  Fig. 5c, d). Here, we tested a strategy that mimics the natural ability of nerves and adrenergic stress 295 signaling to suppress T cell mediated immune responses to prolong survival of a complex tissue 296 allograft.
Importantly has never been tested previously in either preclinical or clinical settings.

310
In non-β2-AR agonist treated animals, CD4 + T and CD8 + T cells were increased

375
One problem with our approach is that terbutaline is a relatively short acting β2-agonist.

376
It is likely -that the effects of β2-AR modulation with longer acting agonists (e.g., bambuterol, 377 salmeterol) could be superior to terbutaline. Another limitation is that our study is focused upon water. After removing slides from the Bond they were dehydrated, cleared, and cover-slipped. Chimeras were generated between BALB/c WT and β2-AR KO mice as donors.

478
Recipient mice were lethally irradiated with 8.0 Gy of total body irradiation (Cesium 137 source).

479
One day after irradiation, BM was reconstituted with the intravenous injection via a tail vein 480 of 10 × 10 6 donor cells. Reconstituted mice were used 8 weeks after BMT.   Table S1. Visual and histologic grading system for assessment of rejection after VCA