Myeloid-derived suppressor cell and regulatory T cell frequencies in canine myasthenia gravis: A pilot study

Abstract

Myasthenia gravis (MG) is a T cell-dependent, B cell-mediated autoimmune disease. Little is known about its cellular pathogenesis in dogs. This study provides the first preliminary assessment of the frequency of myeloid-derived suppressor cells (MDSCs) and regulatory T cells (Tregs) in the peripheral blood of dogs with seropositive generalized MG. No alteration in frequency of either MDSCs or Tregs in dogs with MG was observed when compared to those in either seronegative dogs with diagnoses other than MG, or healthy dogs. A longitudinal study in three dogs with MG revealed no correlation between the relative numbers of either population and the clinical course of disease. Neither the frequency of MDSCs nor of Tregs showed a correlation with anti-AChR antibody titer in dogs with MG. These findings suggest that aberrations in the frequency of either immunosuppressive population do not occur in MG, but they need to be validated in large-scale prospective studies.

Myasthenia gravis (MG) is an autoimmune disorder mainly caused by an antibody-mediated autoimmune response to muscle acetylcholine receptors (AChRs), which impairs neuromuscular transmission leading to muscle weakness (Ludwig et al., 2017). Although T cells are not directly involved in damaging the neuromuscular junction, T helper cells have an essential pathogenic role by permitting and facilitating the synthesis of high-affinity autoantibodies (Wang et al., 2004). In contrast, regulatory T cells (Tregs), a specialized subset of CD4⁺ T cells expressing forkhead box P3 (FoxP3), may play a protective role in MG. While studies of the frequency of circulating Tregs in human myasthenic patients show controversial results, their functional defects have been consistently reported (Balandina et al., 2005; Fattorossi et al., 2005; Li et al., 2008; Luther et al., 2009; Thiruppathi et al., 2012; Gradolatto et al., 2014; Alahgholi-Hajibehzad et al., 2015; Kohler et al., 2017). Passive transfer of polyclonal CD4⁺CD25⁺ Tregs suppresses experimental autoimmune MG (EAMG) in rats (Aricha et al., 2008; Nessi et al., 2010; Aricha et al., 2015).

Myeloid-derived suppressor cells (MDSCs) are a heterogeneous population of immature myeloid cells with immunosuppressive functions, whose differentiation is driven by inflammatory signals characteristic of a variety of cancers and autoimmune diseases (Boros et al., 2016; Gabrilovich, 2017). Despite suppressing T cell responses in vitro, MDSCs have shown conflicting roles in autoimmunity. While some studies suggest a pro-inflammatory role of MDSCs in rheumatoid arthritis (Zhang et al., 2015; Guo et al., 2016), others reported that MDSCs mediate immunosuppression in experimental autoimmune arthritis (Fujii et al., 2013; Park et al., 2018). Adoptive transfer of MDSCs reduced weakness in a murine EAMG model (Li et al., 2014), but few studies have examined their involvement in MG.

Spontaneous canine MG replicates the human disease in many regards, including its clinical presentation (Shelton et al., 1997), the presence of pathogenic anti-AChR antibodies (Shelton et al., 1988), a favorable response to anticholinesterases (Khorzad et al., 2011), comorbidity with thymomas (Marx et al., 2015), and association with genes located within the major histocompatibility complex (Wolf et al., 2017). Despite increasing knowledge of this disease, we still know little about the cellular pathomechanisms of canine MG.

To evaluate the frequency of MDSCs and Tregs in canine MG, peripheral blood samples were collected from dogs with seropositive MG or healthy dogs recruited at the School of Veterinary Medicine of the University of Pennsylvania, following informed consent granted by the owners of the dogs. This study was approved by the University of Pennsylvania’s Institutional Animal Care and Use Committee (Protocol number 806573; Approval date, 7 December 2018). An anti-AChR titer > 0.6 nmol/L was diagnostic for MG. Nine dogs diagnosed with seropositive generalized MG, four disease control (DC) dogs with clinical signs consistent with MG but a negative anti-AChR titer, and six healthy control (HC) dogs were recruited (Supplementary Table 1). Of the nine myasthenic dogs, six had clinical signs of MG and a positive anti-AChR titer, designated as symptomatic MG; the remaining three had a positive anti-AChR titer at the time of initial diagnosis, but negative titers at the time of sampling and no, or minimal, clinical signs.

Peripheral blood mononuclear cells (PBMCs) were isolated from the blood by density gradient centrifugation using Histopaque-1077, before labelling with antibodies for the identification of MDSCs (Supplementary Table 2; Goulart et al., 2019). For Treg analysis, freshly isolated PBMCs were stained with a fixable viability dye, before labelling with antibodies specific for extracellular epitopes (Supplementary Table 2) and then fixation, permeabilization, and labelling with anti-FoxP3. Data were acquired by a FACSCantoII flow cytometer, and analyzed using FlowJo (version 10, Tree Star). The gating strategy for MDSCs and Tregs followed previously published protocols with minor modifications (Pinheiro et al., 2011; Goulart et al., 2019; Wu et al., 2019; Fig. 1A).

Fig. 1.

The frequencies of myeloid-derived suppressor cells and regulatory T cells in dogs with myasthenia gravis. (A) Gating strategy used to identify myeloid-derived suppressor cells (MDSCs; top) and regulatory T cells (Tregs; bottom) in the peripheral blood of dogs. Cells were first gated on size, followed by the exclusion of doublets and dead cells to identify viable cells for further analysis (data not shown). Two subsets of MDSCs, and Tregs, were gated using previously published protocols, as follows: polymorphonuclear (PMN)-MDSCs, MHC II^−/lowCD11b⁺CD14⁻CADO48A⁺; monocytic (M)-MDSCs, MHC II^−/lowCD11b⁺CD14⁺CADO48A⁻; and Tregs, CD5⁺CD4⁺FoxP3⁺. Fluorescence minus one and isotype controls were used to establish regions and quadrants. (B) Comparison of the respective frequencies (%) of PMN-MDSCs, M-MDSCs (relative to total peripheral blood mononuclear cells), and Tregs (frequency of FoxP3⁺ cells among CD5⁺CD4⁺ T cells) between symptomatic myasthenic dogs (MG; n=6), and the disease control (DC; n=4) and healthy control (HC; n=6) groups. No significant difference was found between groups (PMN-MDSCs, P=0.608; M-MDSCs, P=0.395; Tregs, P=0.579; Kruskal-Wallis test). Red symbols represent dogs that were receiving treatment at the time of sampling. Only samples of the first time point with the highest anti-acetylcholine receptor (AChR) titer were included. (C) Time course of the respective frequencies of PMN-MDSCs, M-MDSCs, and Tregs in three symptomatic myasthenic dogs (numbers 12, 14, and 15) during the course of follow-up. Red symbols represent dogs that were receiving treatment at the time of sampling. Abbreviations SP and CR refer to the clinical status, ‘symptomatic’ and ‘clinical remission’, respectively. Days (d) of sampling are shown in paratheses. Dotted lines represent the mean value of Treg frequency (6.803%) of the six HC dogs. (D) Correlation analysis between anti-AChR antibody titers and the respective frequencies of PMN-MDSCs, M-MDSCs, and Tregs in all dogs with MG (n=9). Only samples of the first time point with the highest anti-AChR titer were included. No significant correlation was detected.

The frequencies of polymorphonuclear (PMN)-MDSCs, monocytic (M)-MDSCs, and Tregs in symptomatic myasthenic dogs (n=6) did not differ significantly from those in the DC (n=4) or HC (n=6) groups (Fig. 1B). Follow-up was available for three myasthenic dogs (Supplementary Table 3). The frequencies of neither the MDSC subsets nor Tregs correlated with the clinical course of MG in this small cohort (Fig. 1C). However, the frequency of Tregs in dogs 14 and 15 at the last clinical time point increased to a level higher than the mean frequency of six HC dogs, while dog 12 remained symptomatic with Treg frequency below the healthy level at all times (Fig. 1C). Furthermore, there was no correlation between cell frequencies and anti-AChR titers of the nine dogs with MG (Fig. 1D).

This study provides the first insight into the roles of MDSCs and Tregs in canine MG. Despite growing interest in MDSCs and their therapeutic potential in autoimmune diseases, limited information is available on their role in MG pathogenesis. Increased MDSC frequencies have been detected in several human autoimmune diseases (Jiao et al., 2013; Whitfield-Larry et al., 2014; Wu et al., 2016). We found no difference in the frequency of MDSCs between myasthenic and control dogs. Moreover, the frequency of MDSCs did not correlate with the clinical course of MG in three canine patients. These results suggest that changes in frequency of circulating MDSCs do not play a major role in the pathogenesis of MG in dogs. Future work will need to address whether functional defects in these cells may be incriminated in the pathogenesis of MG, as demonstrated in other autoimmune diseases such as type 1 diabetes mellitus (Whitfield-Larry et al., 2014). Future work will also need to assess the frequency and function of other myeloid regulatory subtypes in MG, such as regulatory dendritic cells (Gordon et al., 2014; Schmidt et al., 2012), macrophages (Fleming and Mosser, 2011; Manjili et al., 2014), neutrophils (Hao et al., 2013; Zemans, 2018), and eosinophils (Lingblom et al., 2017; Müller et al., 2018).

In line with most studies of human MG, our results revealed no differences between canine patients and controls in the frequency of circulating Tregs, raising the question of functional impairment in this disease. Given that functional defects of these cells have been observed in human myasthenic patients (Balandina et al., 2005; Luther et al., 2009; Thiruppathi et al., 2012; Gradolatto et al., 2014), the functional integrity of these cells in canine MG needs to be assessed in future studies. Immunosuppressive therapy using corticosteroids increased both the number and function of Tregs in two studies of human MG (Fattorossi et al., 2005; Luther et al., 2009). In our cohort, only two dogs received prednisone, preventing a meaningful comparison. However, both dogs 14 and 15 had Treg frequencies return to a level above healthy Tregs at their last clinical time point, in contrast to dog 12 with no clinical response and a Treg frequency lower than healthy dogs at all times. This raises the intriguing possibility that clinical improvement of myasthenic dogs may be associated with increased Treg frequency, but such predictions remain speculative at best in the absence of relevant follow-up studies. In addition, healthy Treg number and function are dependent on circadian rhythms across a 24-h period in humans (Bollinger et al., 2009). Blood samples of all 19 dogs in this study were collected between 9.00 a.m. and 1.00 p.m. Future work may further stratify samples according to collection times to assess whether circadian rhythm alters the number and function of MDSCs and Tregs in myasthenic and healthy dogs.

The current study was limited by its small cohort size and short duration of follow-up. Further work will therefore be required to confirm the preliminary findings of this pilot study.

Highlights.

• Suppressor cells were assessed for the first time in dogs with myasthenia gravis

• Numerical alteration was not observed in myeloid-derived suppressor cells

• Numerical alteration was not observed in regulatory T cells

• No correlation was found between suppressor cell frequency and clinical course