mTORC1 activation in presumed classical monocytes: observed correlation with human size variation and neuropsychiatric disease

Introduction

Although formal biochemical analysis of aging pathways is a relatively novel academic discipline, medical modulation of nutrient sensing pathways in humans has been widespread since the approval of lithium in the United States in 1970. Lithium, a glycogen synthase kinase inhibitor, has proven efficacy not only in bipolar disorders [1] but also more recently in Alzheimer’s disease prophylaxis [2], and is known to extend Drosophila lifespan [3]. The combination of lithium with drugs that inhibit the mammalian target of rapamycin complex 1 (mTORC1) (by rapamycin) and mitogen-activated protein kinase (by trametinib) pathways, can, in Drosophila, increase lithium’s 11% lifespan extension to 48% [4].

The inexpensive generic drugs ketamine and rapamycin, that are efficacious in both chronic pain and depression [5, 6], modulate different nutrient sensing pathways. These medications have proximal impacts on the innate nitrogen sensing pathway in every cell in the body via mTORC1, although potentially in opposing directions. Ketamine blocks an extracellular glutamate sensor, the N-methyl-D-aspartate (NMDA) receptor, which is preferentially activated during high activity in nerve cells. The resulting decrease in voltage-gated calcium flux and mTORC2 downregulation may stabilize small G protein Rheb and activate mTORC1 [7, 8]. In contrast, rapamycin mimics amino acid restriction and subsequently downregulates glycolytic-dependent growth and inflammatory processes, resulting in mTORC1 inhibition [4].

Optimal population use of these inexpensive medications in patients requires identification of biomarkers which specify the dose requirements for each drug in each individual. Absent aggressive innovation in diagnostic technology, it is likely the prophylactic effects of lithium, ketamine, or rapamycin will be introduced into the community too late to protect hundreds of millions of individuals who will ultimately die from Alzheimer’s disease or widespread behavioral consequences of depression. An additional barrier is minimal commercial financial interest to address “the last mile problem,” educating the individual patient to take generic stigmatized medications lifelong with potential immediate side effect versus uncertain long-term benefit.

Treatment-resistant psychiatric patients, refractory to standard Food and Drug Administration (FDA) approved medications for depression, in contrast, actively demand novel treatments and diagnostics. Although high dimensional tissue assays, both genomic and metabolomic, are recently commercially available, in our experience, practical utility is limited by cost, convenience, excessive dimensionality of multiple findings, and uncertain applicability of observed peripheral somatic tissue pathophysiology to brain pathophysiology.

In response to patients’ unmet needs, we observed that recent developments in Western blot technology and conceptual advances in biochemical pathways derived from yeast models potentially allow for cost-effective estimation of mTORC1 activity in the individual patient.

We hypothesized that biochemical activity in activated monocytes is a cost-effective estimate of baseline human mTORC1 gain of function. The cost ($45) and convenience of the acquisition of blood cells is trivial. Although blood is composed of a heterogeneous group of cells, from a biochemical perspective most cells in the blood have minimal mTORC1 activity [9]. The exception to this generalization may be monocytes; in mice models phosphorylation of p70 ribosomal S6 kinase (p70S6K), a downstream target of mTORC1, was found to be uniquely elevated in monocytes relative to neutrophil or lymphocyte subsets [9]. In physiological settings, Ly6hi “classical monocytes” with elevated Ki67+activity transition into terminally differentiated Ly6low tissue-resident monocytes during exposure to colony stimulating factor 1 (CSF-1) from damaged portions of the vascular endothelium [10, 11]. Pathological overgrowth of these transitioning cells may be observed in numerous varied disease phenotypes: systemic juvenile idiopathic arthritis [9], lithium-responsive bipolar disorder [1], and rapamycin treatment in amyotrophic lateral sclerosis [12]. Unfortunately, blood samples may be less than optimal as markedly higher expression of PI3K-AKT, glycolytic genes, and Hif1a are observed in wound “M1-like” macrophages at day 2 post injury compared with blood monocytes [13].

The objective of this study was to determine whether presumed macrophage mTORC1 activity, as measured by p70S6K phosphorylation, has implications in terms of identifying undiscovered patient phenotypes or indications for specific medications.

Results

Discussion

Our data suggest that human variability of mTORC1 gain of function observed during the differentiation of stem-like monocytes into vascular tissue-resident macrophages correlates with physical size, subsets of neuropsychiatric disease, and clinical ketamine or rapamycin response.

The major strength of this study is the absence of commercial interests. Funding for this study ultimately derived from patients with treatment-refractory disease with a potentially life-threatening course. All medications of interest studied were generic, historically underutilized, and relevant to their probable benefit absent commercial marketing. No professional conflicts of interest regarding patents, pending grant applications, or publication prominence required for professional advancement were present. In the absence of commercial conflicts and subsequent required regulatory delays, the latency between project design and successful clinical application was short, six months. This latency is entirely consistent with the intent of recent regulatory developments in Federal and Washington state Right-to-Try laws [15].

This study has several limitations. First, this study involved only female patients and cannot be extrapolated to males. So, a study in a male cohort is required given elevated bimodal expression of synaptic disease in males (overgrowth in autism vs. undergrowth in schizophrenia) [16, 17], decreased relative risk of hyperinflammatory manifestations (somatic autoimmune disease and central nervous system mood disorders) [18], and the highly variable differences in response to life-extending interventions between sexes, such as reduced IGF-1 or mTOR signaling, observed in mice [19]. Second, to confirm the proposed tentative model, isolation of monocyte subpopulations from the blood using cell sorting is needed. Third, non-linear distributions of mTORC1 activity require further investigations using alternative readouts e.g., 4E-binding protein 1 (4EBP1) or AKT subtypes, which might allow for a more robust biochemical gradient when estimating medication selection and dose. Fourth, the results of multiple non-preregistered statistical comparisons within a restricted clinical sample require confirmation in a study with a larger number of patients before applying the observed effect sizes to broader populations. Fifth, absent randomized assignment of patients to treatment and control groups, it cannot be ruled out that there is a confounding selection bias for ketamine-maintained patients to be more predisposed to trying novel therapies. A larger, randomized trial will be required to confirm our results.

The strong association between a fixed physical structure (i.e., head size and pupil distance) and dynamic macrophage differentiation is consistent with known mTORC1 effects on macrocephaly [20]. This finding provides face validity to the concept that systemic mTORC1 gain of function can be estimated easily, albeit incompletely, with simple physical measurements and crude blood analysis. Age, exercise, and diet-related hypertrophic lipodystrophy manifest as increased waist circumference and BMI and may rely on different proximal upstream activators of mTORC1 activated later in development than in infancy and adolescence.

The miscellaneous associations between multiple neuropsychiatric syndromes (anxiety and autism) and mTORC1 activity in blood monocytes direct attention to the structural role of tissue-resident macrophages in the brain, the microglia. In the brain, are physical tension and psychological tension the same? The brain is a closed space due to the inflexibility of the skull. All learning is a zero-sum game. Therefore, an augmented synaptic trace associated with a new experience requires the corresponding phagocytosis of a less preferred synapse. The extracellular matrix, in turn, must be digested to avoid “tension” as different synaptic “eigenvalues” embedded in brain activity “eigenvector” grow in size with time and distort the morphology of connecting tissue or present risk of local excitotoxicity [21]. This conceptual framework suggests that different densities of the extracellular matrix and embedded synaptic structure may manifest as dimensional behavioral traits in humans.

The association of mTORC1 activity with ketamine response is consistent with the literature. Reductions in serum TNF-α were found to correlate with clinical antidepressant response after intravenous ketamine treatment; these changes were observed within 40 minutes [22]. Recent work in invasive vs. non-invasive ovarian cancer revealed that high levels of tumor-derived N-acetyl aspartate competitively inhibited the NMDA receptor on M1 tissue-resident macrophages and induced M2 polarization [23]. The resultant biochemical profile within the proliferative MKI67+ M2 macrophage (GLS1 upregulation, SLCA5/SLCA7 upregulation) strongly corresponded to the biochemical phenotype observed in the whole blood, inferred monocyte fraction, in human lithium responders [1]. These converging observations suggested the possibility that the M2 homeostatic microglia population (CD163+) in the brain of depressed patients is of inadequate size to allow for adequate synaptic pruning or trophic support [8]. The upregulation of BCL-2 in lithium responders [1] could potentially suggest protection against premature senescence of the long-lived microglia stem cell population, which migrates into the brain from the yolk sac at week 4 of embryonic development.

The association of mTORC1 activity with rapamycin response is consistent with the literature, specifically rapamycin’s efficacy in treating tuberous sclerosis [24]. Rapamycin also has documented efficacy in a host of polygenetic autoimmune disorders, specifically in the prototype disorder lupus [25] and the case-specific disorder ulcerative colitis [26]. Its proximal mechanism of action may relate to biasing differentiation of naive T-cells towards the FoxP3+ T-regulatory and CD8+CD45R0+ memory subtypes dependent on fatty acid oxidation and reducing interleukin (IL)-4 and IL-17 production by CD3+CD4-CD8- double negative T-cells.

The observed disjunction between relative phosphorylated p70S6K and absolute p70S6K abundance correlations with clinical phenotypes suggested careful attention to the different physiological drivers of metabolic reprogramming in macrophages. The total amount of available p70S6K and its active phosphorylated form may only partially control macrophage activity. Note that phosphoinositide 3-kinase (PI3K)-γ inhibition alone promoted the “M1-phenotype (IL-12 excretion)” relative to the “M2-phenotype (IL-10 excretion)” in the context of lipopolysaccharide (LPS) stimulation without affecting p70S6K activity [27]. Even within the “M2-phenotype” in cultured microglial cells, PI3K/TORC1 independent (IL-4) and dependent (ATP/purine) extracellular receptor activation pathways appeared to be present [28].

In conclusion, recent conceptual, technological, and regulatory developments may allow for the rapid development of a simple blood assay supporting both widespread and yet targeted use of generic medications known to extend both healthspan and lifespan. This blood assay attempts to capture a snapshot of “stem-like” monocytes that will transform into tissue-resident macrophages. The resulting image of innate mTORC1 and mTORC2 activation may provide insights into the management of neuropsychiatric disease. The current study provides a proof of concept that requires replication in larger cohorts and using isolated monocyte subpopulations.

Materials and Methods

Supplementary Materials

Supplementary Figures

Supplementary Tables

Supplementary Data