Interleukin-17 drives sex-dependent weight loss and changes in feeding behaviour 1 during Trypanosoma brucei infection

28 Previous work has demonstrated that Trypanosoma brucei occupy several adipose tissue 29 depots, including the subcutaneous adipose tissue in mice and humans, and due to its 30 proximity to the skin, it is proposed to be an important for transmission. Here, we demonstrate 31 that parasites in the inguinal white adipose tissue (iWAT) niche induce sexually dimorphic 32 responses. During infection, male mice experience reduced adipose tissue mass, altered 33 tissue function, and changes in feeding behaviour, whereas females do not. This tissue 34 impairment correlates with an accumulation of T H 17 T cells in the iWAT. Genetic ablation of 35 IL-17A/F abolishes infection-associated weight loss and alters feeding behaviour, limiting 36 tissue wasting in male mice. Importantly, we detected a significant elevation in serum IL-17A 37 in sleeping sickness patients, indicating that IL-17A/F signalling is also conserved in humans. 38 We propose a model whereby IL-17A/F acts locally in adipocytes via engagement with its 39 cognate receptor leading to lipolysis and tissue wasting, and/or systemically, via signalling in 40 the hypothalamus to modulate feeding behaviour. Together, our findings suggest key sex- 41 dependent roles for IL-17A/F in regulating adipose tissue and energy balance, as well as a 42 coordinator of brain-adipose tissue communication during sleeping sickness, opening new 43 directions to understand energy balance during infection. 44 45 46 47 48 49 50 influences adipose tissue wasting and feeding behaviour. These results provide novel insights into the role of IL-17A/F as a regulator of adipose tissue structure and function and feeding behaviour during infection. Furthermore, these findings support the utility of T. brucei into response of during how changes in cellular metabolism influence systemic metabolism. Furthermore, this highlights the sex-dependent effects of IL-17A/F on the adipose tissue responses to Future work is needed to determine the pathways underpinning these sex as well as the potential sources of IL-17A/F during infection, which dimorphic responses in


Introduction
In both humans and livestock, trypanosome infections are known to cause weight loss, 141 and this has been recapitulated in male mouse models of infection (Trindade et al., 2016).

142
However, direct comparisons have not been made between male and female mice to assess 143 whether infection induces weight loss in a sexually dimorphic manner. To test this, we infected 144 age-matched male and female C57BL/6 mice for a period of 25 days. We first wanted to 145 determine that mice were successfully infected and whether there were differences between 146 the levels of circulating parasites between sexes. Parasitaemia measurements followed a 147 characteristic pattern, with no significant differences between sexes ( Figure 1A). There were 148 also no significant differences in the clinical scores of the mice ( Figure 1B)

154
Weight loss may be explained as a consequence of adipose tissue wasting (Dahlman 155 et al., 2010), as a consequence of changes in feeding behaviour (Aviello et al., 2021), or both.

156
To understand this in more detail, we measured gross food intake as a proxy for feeding inguinal white adipose tissue (iWAT), which is analogous to the subcutaneous white adipose tissue beneath the skin in humans, as this constitutes an important parasite niche for disease 172 transmission, especially in asymptomatic carriers (Capewell et al., 2016). In a seminal study, energy metabolism. However, using pathway enrichment analysis, we found extensive 226 downregulation of genes related to mitochondrial carbon (including Idh1, Sdha, Mdh1, and 227 subunits of NADH:ubiquinone oxidoreductase and ATPase) and lipid metabolism and storage 228 (including Pnpla2,Cpt2,Plin1,Plin4,and Plin5) Table 2).

233
Next, we sought to understand which gene pathways are discriminatory between male 234 and female mice. Pathway enrichment analysis of genes exclusively upregulated in females 235 highlighted genes related to folate metabolism (including Mocs1, Mocos, and Dhfr), and 236 protein processing in the endoplasmic reticulum (including Ero1,Pdia4,and Pdia6) 237 (Supplementary Table 2). In contrast, several immune-related pathways were exclusively 238 upregulated in males during infection, including those related to T cell signalling and T helper 239 (TH) cell differentiation: TH1, TH2, and TH17 cells (Supplementary Table 2). Upregulated TH1-240 and TH2-related transcripts included Nfatc1, Cd4, Cd3e, Runx3, and Gata3, suggesting that 241 TH1 cells may be a significant contributor to interferon gamma (IFNg) production during 242 infection in the adipose tissue. Additionally, we also detected a significant upregulation of 243 genes typically associated with differentiation of TH17 effector cells, including Irf4, Cd4, Cd3e, 244 Il21r, and Il6ra ( Figure 3D). Taken together, we conclude that the iWAT of male mice 245 experiences a robust pro-inflammatory response potentially mediated by TH1 and TH17 246 effector T cells, whereas the iWAT of infected female mice display transcriptional signatures 247 associated with metabolism and protein synthesis.

250
Based on these observations, we hypothesised that TH17 cells are important for the 251 adipose immune response to infection. To quantify the different populations of CD4 + T cells,

252
including TH17 cells, present in the iWAT during chronic T. brucei infection, we utilised mass 253 cytometry by time of flight (CyTOF), enabling us to gain as much information as possible from 254 the wasted adipose tissue. Previous studies, exploring the gWAT described an expansion of B cells, macrophages, granulocytes, and T cells (Machado et al., 2021). Using a broad panel,

266
Regarding the T cell effector population, we observed an increase in the proportion of 267 CD3ε + TCRβ + CD4 + T cells in the iWAT of infected mice ( Figure 4F) and, in particular, we 268 identified an expansion of T effector (Teff) cells (CD44 + and CD69 + ) ( Figure 4G). Whilst the 269 frequency of CD4 + T cells increased in both males and females, the iWAT of infected females   Figure 5B) Il17af -/mice over the course of infection. We observed that the first 293 peak of parasitaemia was similar between wildtype and Il17af -/male mice, but at late 294 timepoints the Il17af -/males displayed consistent higher levels of parasitaemia than their 295 wildtype counterparts, indicating that IL-17A/F is essential to control infection during the 296 chronic stage of the infection in male mice. In contrast, although not significant, the first peak 297 of parasitaemia in Il17af -/females was lower than in wild type mice, and the second peak of 298 parasitaemia was delayed. Deletion of Il17a/f was also associated with an earlier onset and 299 increased severity of clinical symptoms in both males ( Figure 5C) and females ( Figure 5D).

300
Il17af -/male and female mice started to exhibit clinical symptoms (piloerection and hunching) 301 from 3 and 7 dpi, respectively, whereas wild type mice started to experience these symptoms

308
Our data so far indicate that chronic T. brucei infection leads to an accumulation of 309 TH17 effector cells in the iWAT, as well as a significant increase of Il17a receptor expression, 310 indicating both local IL-17 production and a potential increase in IL-17A receptor signalling.

311
Therefore, we next asked whether IL-17A/F insufficiency leads to weight loss during infection.

312
As previously in this study, we monitored the weight of both male and female mice during the 313 course of infection. Unlike wild type male mice, infected Il17af knockout (Il17af -/-) males 314 maintained their weight until 19 dpi, after which they started to gain significantly more weight 315 than infected wild type males ( Figure 6A). In contrast, infected female Il17af -/mice started to 316 gain weight in a similar pattern to wild type females, although similarly to males, knockout 317 females gained more weight that their wild type counterparts ( Figure 6B). Some of this weight 318 gain was related to increased splenomegaly in Il17af -/females, compared with their wild type 319 counterparts (Supplementary Figure 2A).

320
It was previously observed that when it is administered to wild type naïve mice, IL-17A 321 suppresses food intake (Nogueira et al., 2020). Therefore, to understand whether IL-17A 322 influences feeding behaviour and drives weight loss during infection, gross food intake was 323 measured for naïve and infected Il17af -/mice. Until 9 dpi, wild type and Il17af -/male mice 324 reduced their food intake at the same rate ( Figure 6C). However, after 11dpi, Il17af -/mice 325 started to increase their food intake above that of the wild types. Indeed, food intake for 326 infected Il17af -/mice was higher between 15 and 23 dpi compared with at the onset of 327 infection. Unlike male mice, feeding behaviour was indistinguishable between infected female 328 Il17af -/and wild type mice ( Figure 6D). Together, this may suggest that IL-17A/F regulate 329 bodyweight by potentially suppressing the appetite of male mice during infection, but not 330 female mice. Further supporting this hypothesis, we found that naïve l17af -/ male mice have Figure 2B). Upon infection, Il17af -/knockout male mice also retained more of their iWAT mass 334 compared with their wild type counterparts, suggesting that they experienced less wasting. In ( Figure 6F), supporting the assertion that the effects of IL-17A/F are sex-dependent.

338
To understand whether deletion of Il17af -/impacts iWAT lipid content, we performed 339 H&E staining on the iWAT ( Figure 6G) and measured lipid droplet size in naïve and infected 340 male ( Figure 6H) and female ( Figure 6I) mice and compared these with wild type animals.

341
The iWAT adipocyte size was indistinguishable between genotypes in naïve animals.

342
However, upon infection, Il17af -/males retained a higher frequency of larger lipid droplets than 343 wild type males ( Figure 6H). When comparing this to female mice, we found that deletion of that IL-17A/F signalling drives lipid usage in adipocytes in males, but not in females, during 346 infection. We reasoned that the changes in weight, feeding behaviour, and lipid usage may in general, and neurons controlling appetite and feeding behaviour in particular, are primed to 356 sense IL-17 via Il17ra. In conjunction with increased levels of circulating IL-17A, we propose adipose tissue wasting in the subcutaneous adipose tissue depot, and whether this is 364 mediated by components of the immune system. Throughout this study, we uncovered 365 evidence demonstrating that male mice are more susceptible to trypanosomiasis-induced 366 weight loss and adipose tissue wasting than female mice, and that IL-17A/F plays a key role 367 in this sexually dimorphic pathology. Furthermore, we found that IL-17A/F suppresses food 368 intake in males but not in females, potentially suggesting a dual effect of this cytokine; either 369 locally in adipocytes, or in the hypothalamus to promote changes in feeding behaviour. Indeed, 370 changes in feeding behaviour due to altered hypothalamic signalling may play a key role in 371 the pathology that we observe during infection and may be an indirect consequence of 372 perturbations within the central nervous system. Notably, we also observed elevation of 373 circulating IL-17A in HAT patients, suggesting that this cytokine plays a role in the human 374 immune response to T. brucei infection. It is, therefore, tempting to speculate that IL-17A could 375 also play a role in mediating the weight loss experienced by patients infected with T. brucei.          Table 2. Generated data was converted to parasite copy number using a standard 695 curve.