key: cord-0294436-6hy2iocp
authors: Rosado, Mitchell R. Sanchez; Marzan-Rivera, Nicole; Watowich, Marina M.; Pantoja, Petraleigh; Pavez-Fox, Melissa A.; Siracusa, Erin R.; Cooper, Eve B.; Negron-Del Valle, Josue E.; Phillips, Daniel; Ruiz-Lambides, Angelina; Martinez, Melween I.; Montague, Michael J.; Platt, Michael L.; Higham, James P.; Brent, Lauren J. N.; Sariol, Carlos A.; Snyder-Mackler, Noah
title: Sociodemographic effects on immune cell composition in a free-ranging non-human primate
date: 2021-12-07
journal: bioRxiv
DOI: 10.1101/2021.12.06.471383
sha: ae4d5c6aebe511a3f2544054858c4cfe2d79dc41
doc_id: 294436
cord_uid: 6hy2iocp

Aging results in declines in immune function and increases in inflammation, which underlie many age-related diseases. These immunosenescent signatures are similar to those seen in individuals exposed to social adversity, who may age more rapidly than those unexposed. Yet, it is unclear how social adversity alters immunity across demographic factors - data that are essential to identify how it might increase aging-related diseases. Here, we investigated how age, sex, and social adversity predicted immune cell proportions in 250 rhesus macaques living in a semi-naturalistic colony. As macaques aged, they exhibited signatures of immunosenescence. Older individuals had signatures of diminished antibody production and adaptive immunity, with declines in CD20+ B cells, CD20+/CD3+ cell ratio, and the CD4+/CD8+ T cell ratio. At all ages, females had higher CD20+/CD3+ and CD4+/CD8+ ratios, indicative of a stronger antibody and adaptive immune response that may facilitate pathogen clearance even with increasing age. Older individuals had signatures of inflammation, with higher proportions of CD3+/CD8+ Cytotoxic T cells, CD16+/CD3- Natural Killer cells, CD3+/CD4+/CD25+ and CD3+/CD8+/CD25+ T regulatory cells, and CD14+/CD16+/HLA-DR+ intermediate monocytes, combined with lower levels of CD14+/CD16-/HLA-DR+ classical monocytes. Notably, we found an interaction between age and social adversity, where low-status individuals had higher proportions of CD3+/CD4+/CD25+ T regulatory cells for their age, compared to higher-status individuals. Together, our study identifies immune cell types that are affected by age and sex in the premier nonhuman primate model of human biology and behavior, and demonstrate a novel link between inflammatory CD4+ T regulatory cells and social adversity.

The average human lifespan has almost doubled over the past century, leading to an 56 increase in the prevalence of diseases of aging, like cardiovascular disease (1), 57 autoimmune disease (2), diabetes, arthritis, and cognitive decline (3). As individuals age, 58

there is a disruption in the homeostatic balance between innate and adaptive immunity 59 linked to both increases in age-related disease and susceptibility to infection. This 60 imbalance is reflective of two age-related changes, namely increased inflammation males have a dampened age-related increase in inflammatory cells (7). Although these 77 studies have provided some insight into sex differences in immunological aging, there 78 is still a lack of data demonstrating how these differences affect individuals across their 79 lifespan and the specific cell types that contribute to these differences. 80 81 Heterogeneity in aging can also be attributed to life experiences, such as exposure to 82 social adversity, which can influence the onset and progression of disease and, ultimately, 83 mortality (8). Social adversity has been linked to accelerated aging as measured using 84 biomarkers like epigenetic age and telomere attrition (9, 10). There are also broad 85 similarities between the effect of age and social adversity on peripheral immune function 86 (11). For instance, early life adversity in humans has been linked to increases in 87 proinflammatory T cells (12) -a characteristic usually seen with increasing age. However, 88 it is still unclear if and how social adversity alters immune function across the life course 89 -data that are essential to identifying how social adversity might lead to increased aging-90 related disease and death, or, conversely, how social advantage can help protect an 91 individual from the effects of aging. Taken together, there is mounting evidence that social which individuals can access resources (15), which is a feature also commonly seen in 108 human populations. Like humans, social status in macaques affects health and survival 109 Here, we characterized the immunological effects of aging and quantified whether social 113 adversity recapitulates or interacts with these effects. We studied a free-ranging 114 population of rhesus macaques living on the island of Cayo Santiago, Puerto Rico where 115 we could simultaneously measure the effects of aging, sex, and social adversity in a semi-116 naturalistic social setting with minimal human intervention. Using objective measures of 117 social adversity, we were able to avoid some of the biases that sometimes affect studies 118 of humans that rely on self-reporting and questionnaires (18, 19) . We identified which 119 immune cell proportions exhibit age-associated changes, how these proportions are 120 affected by sex and social adversity, and if/how the social adversity-induced changes 121 recapitulate the immunological changes seen during aging. October 2020 -February 2021) into 6ml K2 EDTA tubes (Beckton, Dickson and Company, 143 cat #367899). We sampled a total of 250 animals (115 females; 135 males) spanning 144 their natural lifespan (mean age = 11.05 years, range 0-28 years; Figure 1A) . (27). To control for group size variation, dominance rank was calculated as the 208 percentage of same-sex group peers that a subject outranked; a score above 80% 209 corresponded to high-ranking animals, scores from 50-79% to medium ranking animals 210 and scores below 49% to low ranking animals (28). Overall, we were able to quantify 211 To evaluate each cell type at a more granular level, we employed the same linear mixed-226 effects to model the proportion of each cell type and certain cell-type ratios (e.g., 227 CD4+/CD8+) as a function of age, sex, and sample period. We also included an 228 interaction between age and sex to identify sex-dependent age-related associations. For 229 the subset of samples where social status information was available (n = 140), we also 230 modeled cell proportions as a function of social status (rank), age, sex, and sample 231 period, again with individual ID as a random effect. We also included an interaction 232 between rank and age as well as for rank and sex to test if changes in the relationship 233 between cell type proportion and social status were dependent on either of these 234

Results 237

Age was the strongest predictor of immune cell composition in our dataset, and was dataset was roughly balanced between males and females and captured the entire 264 natural lifespan of macaques in this population. We calculated social status by assigning 265 dominance ranks to 140 animals using observational data collected in the year before 266 each sample was collected. Animals were assigned to one of three dominance ranks: 267 high (n = 31), medium (n = 42) and low (n = 67). The social status dataset is a subset of 268 the original age dataset because behavioral data were not available for all study animals. Figure 1) , and when 281 paired with an age-associated increase in cytotoxic CD8+ T cells (βCD8 = 0.60 ± 0.09, p = 282 1.4 x 10 -10 , Figure 2C) , there was a strong and significant age-associated decline in the 283 CD4+/CD8+ ratio (βCD4:CD8 = -0.06 ± 0.008, p = 1.9 x 10 -13 , Figure 2D) . 284

We then examined less common, but extremely immunologically important, regulatory T 286 cell populations (CD25+) that are involved in immunity suppression and maintenance of 287 self-tolerance (29). Proportions of both CD4+ and CD8+ T regulatory cells increased 288 significantly with age (CD3+CD4+CD25+: β = 0.11 ± 0.02, p = 2.4 x 10 -9 , Figure 2E ; 289 CD3+CD8+CD25+: β = 0.01 ± 0.005, p = 5.8 x 10 -3 , Figure 2F ). Innate immune system 290 cells also showed significant changes with age, with two monocyte populations presenting 291

weak, but differing age-related changes in abundance. increasing age (β = 0.31 ± 0.13, p = 0.02, Figure 3B ). The proportion of CD16+CD3-NK 308

non-antigen specific -also increased significantly with age (β = 0.35 ± 0.10, p = 3.2 x 10 -310 4 , Figure 3C ). Together, these changes indicate an age-related decline in adaptive 311 immunity paired with an increase in inflammation profile by innate immune cells, 312 potentially disrupting a "healthy" homeostatic immune system. Intriguingly, we also identified social status-dependent age-related changes in T 368 regulatory cells. Together these changes likely shape immune responses to future 369 pathogenic challenges as well as the development of inflammatory-related diseases. We also identified important sex-differences in cells linked to adaptive immune cell 419

populations. Females had more CD4+ T cells across the life-course compared to males, 420 which recapitulates changes seen in humans (43). Furthermore, we found sex specific 421 differences in the CD4+/CD8+ T cell ratio, with males having a lower ratio across the life-422 course than females. CD4+ T cells aid in the affinity maturation process that results in the 423 production of antibodies in the lymph nodes (44). Thus, their higher abundance in females 424 across the life-course, which is also reflected in a lower CD4+/CD8+ ratio for males, may 425 be linked to a better and more effective antibody response and thus adaptive immune 426 protection in females against invading pathogens (45). 427 428 A strength of our study was the ability to quantify measures of social adversity (i.e., 429 dominance rank), and to test if and how social status altered age-related immune 430 changes. Contrary to our expectations, we found no main effect of social status on the 431 abundance of any immune cell population when controlling for age and sex. This may be 432 due to the fact that we only had behavioral data, and thus social status measures, for a subset of adults in our dataset (n = 140). Nevertheless, we did detect a strong interaction between age and status on the proportion of the inflammatory CD4+ T regulatory cells. 435

individuals on the bottom of the social hierarchy, which typically experience the most 437 social adversity. Thus, older individuals exhibited the strongest increase in these 438 inflammatory cells if they were also low social status. This finding points to the age-439 dependence of some effects of social adversity. In other words, as individuals get older, 440 the adversity associated with low status disproportionally increases inflammation, which 441 may accelerate the onset of aging-related diseases and early mortality, which might help 442 to explain why low status is associated with shorter lifespan in this population (46, 47) . 443

This also provides a novel addition to previous work that established links between social 444 adversity and inflammation in humans and rhesus macaques (48, 49) . 445

In conclusion, our results demonstrate that, at the level of circulating immune cell 447

proportions, macaques and humans exhibit very similar age-related changes indicating 448 both immunosenescence and inflammaging. Sex differences in these patterns indicate 449 that the immune system has evolved similarly in different species and that responses are 450 conserved between them, with females typically presenting a better and more effective 451 adaptive immune response. Furthermore, we report that low social status is associated 452 with a stronger increase in inflammation with aging. Although we found age-related 453 changes in both adaptive and innate immune cells, we did not measure specific adaptive 454 immune cells, such as the effector and memory subsets of B cells and T cells, which can 455 change with age. In future studies, it will be important to measure other innate immune 456 cell types, such as dendritic cells and granulocytes, because these cell types are key to 457 the antigen presentation process and adaptive immune response development. While we 458

found an interaction between age and social status, we were limited by our sample size \ 

Rhesus macaques as a tractable physiological 535 model of human ageing

Social determinants of health and survival in 539 humans and other animals

Early life adversity and telomere length: a meta-analysis

Sociality predicts individual variation in 546 the immunity of free-ranging rhesus macaques

Common 549 method biases in behavioral research: a critical review of the literature and 11

Human monocyte subsets and phenotypes in major chronic 627 inflammatory diseases

From monocytes to M1/M2 macrophages: 630 phenotypical vs. functional differentiation

Chronic kidney disease severity is associated with 634 selective expansion of a distinctive intermediate monocyte subpopulation. Frontiers 635 in immunology

Sex differences in the blood transcriptome identify robust 638 changes in immune cell proportions with aging and influenza infection

Social stress up-regulates inflammatory gene expression in the 656 leukocyte transcriptome via β-adrenergic induction of myelopoiesis

early social stress promotes Inflammation and Disease Risk in Rhesus Monkeys