Gender-related disparities in the frequencies of PD-1 and PD-L1 positive peripheral blood T and B lymphocytes in patients with alcohol-related liver disease: a single center pilot study

Introduction

Alcohol abuse remains a huge problem in the Western world. The available body of evidence indicates that it is associated with the increased prevalence of different chronic disorders i.e., cancers, lung and cardiovascular diseases as well as liver cirrhosis (Yoon & Chen, 2018, https://pubs.niaaa.nih.gov/publications/surveillance111/Cirr15.pdf). According to data from the National Institute on Alcohol Abuse and Alcoholism, over 80,000 deaths per year in the United States are attributed to alcohol misuse (National Instituteon Alcohol Abuse and Alcoholism, 2017, https://www.niaaa.nih.gov/alcohol-health/overview-alcohol-consumption/alcohol-facts-and-statistics). Also, the high mortality rate due to liver cirrhosis in Poland and other Central and Eastern European countries, as well as in the United Kingdom, Ireland, and Finland (World Health Organization (2014), http://www.who.int/healthinfo/statistics/mortality_rawdata/en/; Bosetti et al., 2007) is related mainly to the high per capita alcohol consumption.

There is a sufficient body of evidence, that alcohol-attributable end-stage liver disease and liver cancer are entirely preventable (Sheron, 2016; Rehm, Samokhvalov & Shield, 2013). Therefore, it is vitally important for policy planning to chart the methods leading to the reduction of harmful alcohol consumption. Furthermore, also systematic research in order to implement new diagnostic and therapeutic tools and reach new conclusions should be foster. In comparison to huge advances made in the management of viral hepatitis (vaccines and oral therapies for HBV, oral regimes for HCV), alcohol-related liver disease (ALD) management has lagged. Moreover, patients with ALD are generally identified at the late stages of the disease, and programs for early detection are scarce (Ndugga et al., 2017).

Exposure to chronic and excessive alcohol consumption, its breakdown metabolites and gut-derived endotoxins dysregulate immune signaling and give rise to activation of the local and systemic pathways of inflammation (Gao & Tsukamoto, 2016; Szabo & Saha, 2015). Toxic ethanol metabolites including reactive oxygen species (ROS) activate T and B cell clones against self- and modified proteins. Activated B cells can produce immunoglobulins directed both to haptens and native antigens (Szabo & Saha, 2015; Molina et al., 2010). As a result, the non-resolving inflammatory response may occur leading to ALD development and progression. Furthermore, ethanol misuse contributes to the release of circulating modulators of immunity and inflammation, that affect multiple organs and tissues, potentially causing their failure (Louvet & Mathurin, 2015; Hernaez et al., 2017). Therefore, the aforementioned responses need to be properly controlled to maintain the mechanisms of tolerance and immune homeostasis.

Recent studies have elucidated the relevance of the programmed cell death 1 (PD-1) receptor and its ligand PD-L1 in inhibition of self-reactive and effector cells and the protection against immune-mediated tissue damage. As a negative regulator, PD-1 exerts a suppressive effect on previously activated T and B cells by binding to its ligands PD-L1/PD-L2. It results in inhibition of antigen-specific immune cell proliferation, cytokine release, and cytolytic function. PD-L1 is also constitutively expressed on T and B cells, dendritic cells (DCs), macrophages, mesenchymal stem cells and bone marrow-derived mast cells (Yamazaki et al., 2002; Boussiotis, Chatterjee & Li, 2014). PD-1 conducts signals only when it is cross-linked with B- or T-cell antigen receptors. These coinhibitory signals modulate the intensity and duration time of immune reactions and may restrict immune-induced tissue damage, control the resolution of inflammation, and maintain peripheral immune tolerance (Bardhan, Anagnostou & Boussiotis, 2016; Saresella et al., 2012).

The PD-1/PD-L1 pathway has been shown to play an important role in a variety of diseases, like cancer, autoimmune conditions, and chronic infections. It has already been investigated using experimental animal models in systemic lupus erythematosus, encephalomyelitis, myasthenia gravis, diabetes mellitus, myocarditis, inflammatory bowel diseases, systemic sclerosis, and rheumatoid arthritis. The critical role of PD-1/PD-L1 signaling in the prevention of disorders caused by impaired immune activation has become evident (Dinesh, Hahn & Singh, 2010; Zamani et al., 2016). However, its impact on the development and progression of ALD has not been entirely explained yet. It is likely that alterations of the checkpoint inhibitor surveillance may be responsible for long-lasting activation of the inflammatory response in the liver, as well as the systemic character of the disease. Recently Markwick et al. (2015) reported that PD-1, and T-cell immunoglobulin and mucin domain-containing protein 3 (TIM3), as well as their ligands i.e., PD-L1 and galectin-9, represented relevant elements of innate and adaptive immunity in the course of alcoholic hepatitis (AH). T lymphocytes from AH patients showed higher expression of PD-1 and PD-L1 in comparison with control T cells. They indicated that PD-1 and TIM3 blockade might create the potential new approach for the treatment of the disease.

On this background, we aimed to explore possible alterations in the expression of PD-1/ PD-L1 proteins on peripheral T and B lymphocytes, their eventual correlations with conventional markers of inflammation and the severity of alcoholic hepatitis (AH), as well as with the severity of liver dysfunction classified according to the criteria (Child-Turcotte-Pugh, Model of End-Stage Liver Disease, modified Maddrey’s discriminant function) widely used for this purpose in clinical settings for ALD patients.

Since a great body of evidence indicates that compared with their male counterparts, women are more susceptible to the toxic effects of ethanol in the liver for any given dose of alcohol (Chou, 1994; Greenfield, 2002; Grant et al., 2017; Lowe et al., 2019), and also symptoms and signs in ALD demonstrate sexual dimorphisms, possible gender-related differences in the PD-1/PD-L protein expression were also assessed. Sex hormones have been suggested to influence immune response. These hormones exert their biological effects by binding to the inherent receptors on immune cells, what results in further modification of gene expression, lymphocyte proliferation, antigen presentation, and cytokine secretion (Whitacre, 2001). Females are known to produce higher titers of circulating immunoglobulins, as well as a major variety of autoreactive antibodies. Women also demonstrate a more pronounced humoral immune response against infections in comparison with men (Arroyo & Montor, 2011). Understanding the background of gender-related differences in the ALD course may be relevant for individually tailored therapies.

Materials & Methods

Results

Analysis of the PD-1 and PD-L1 positive T and B lymphocyte frequencies in the peripheral blood of patients with ALD in comparison with the control group and gender-related differences

The general frequencies of PD-1 and PD-L1 positive T and B subsets did not differ between the ALD and the control group (see Table S1). Although, when the ALD group were analyzed based on their gender, significantly higher frequencies of PD1 and PD-L1 positive CD19+ B cells in ALD females compared to female controls were observed (Fig. 1; Table 2).

Table 2:

Comparison of the frequencies (%) of PD-1/PD-L1 positive T and B lymphocytes inALD and controls based on patients’ gender (a Mann-Whitney test).

Variable	ALD group				p1*	Controls				p2*	p3*	p4*
	Males ( n = 38)		Females ( n = 18)			Males ( n = 16)		Females ( n = 9)
	Median	5–95 percentile	Median	5–95 Percentile		Median	5–95 Percentile	Median	5–95 Percentile
PD-1 positive cell subsets
CD4+	18.44	1.83–3.28	18.07	1.43–4.03	0.73	21.07	8.40–9.64	14.20	6.55–30.67	0.17	0.18	0.60
CD8+	19.40	2.68–6.57	23.06	0.67–6.70	0.99	25.57	7.55–7.51	21.11	11.28–43.08	0.68	0.22	0.63
CD19+	3.27	0.36–4.20	7.08	0.69–0.76	0.02	2.54	0.78–9.39	1.66	0.35–14.70	0.47	0.59	0.04
PD-L1 positive cell subsets
CD4+	0.65	0.04–2.76	0.90	0.09–8.04	0.04	1.06	0.26–5.10	1.83	0.10–4.03	0.80	0.09	0.44
CD8+	0.44	0.03–1.66	1.14	0.02–7.21	0.02	0.97	0.13–2.43	0.89	0.18–2.88	1.00	0.07	0.63
CD19+	0.71	0.01–6.40	1.99	0.01–8.88	0.04	0.69	0.25–8.77	0.45	0.07–1.52	0.16	0.62	0.03

DOI: 10.7717/peerj.10518/table-2

Notes:

p1

ALD males vs ALD females

p2

female controls vs male controls

p3

ALD males vs male controls

p4

ALD females vs female controls

There were no statistically significant differences in the frequencies of PD-1/PD-L1 positive T and B cells between the ALD and control groups of the studied men (Table 2).

Furthermore, females with ALD presented with significantly higher frequencies of PD-1 and PD-L1 positive B cells, as well as PD-L1 positive T cells (both subsets) in comparison with ALD males (Table 2).

There were no gender-related differences in the frequencies of PD-1/PD-L1 positive T and B lymphocytes in the control group (Table 2).

Discussion

Innate and adaptive immune responses represent a driving force in the initiation and development of ALD leading to liver fibrosis and subsequent conversion to cirrhosis with eventual further progression to liver cancer in about 2% of cirrhotics (Orman, Odena & Bataller, 2013). Regarding advanced liver disease, ALD is the main cause of cirrhosis worldwide, accounting for 50% of the cases (Poznyak & Rekve, 2014). Liver dysfunction with loss of the ability to clear immunogenic cellular residues and ethanol metabolites from the systemic circulation results in persistent stimulation of the immune system. Patients who suffer from ALD present not only with the increased titer of circulating antibodies but also with other lymphocyte-mediated responses triggered by antigens originating from oxidative stress (Sutti, Bruzzì & Albano, 2016). Paradoxically, these patients present with enhanced immune activity and exacerbated inflammatory responses, but are unable to cope with bacterial infections. Their immune effector cells are primed, but antibacterial functions are switched-off (Riva & Chokshi, 2018). The PD-1/PD-L axis represents an important co-inhibitory pathway, which modulates immune system activation and tolerance. However, overexpression of these molecules may contribute to excessive inhibitory signals and lead to an impaired immune response. The PD-1/PD-L1 signaling has been shown to play a pivotal role in different disorders including chronic infections, autoimmune diseases, and cancers (Riva & Chokshi, 2018; Pardoll, 2012). To date, the effects of checkpoint blockade have not been fully explained in patients with ALD, the disease with no currently approved pharmacological treatment. Although many studies on immune alterations in ALD have been conducted during the last decade, the available body of evidence for the role of the PD-1/PD-L1 signaling is fairly scarce.

In general, our results revealed no differences in the frequencies of PD-1/PD-L1 positive T and B lymphocytes between ALD patients and healthy controls (Table S1). Interestingly, the analysis based on patients’ gender revealed significant disparities. The studied women with ALD presented with significantly higher frequencies of PD-1 and PD-L1 positive B cells (CD19+) in comparison with female controls (Fig. 1, Table 2). Recently, the effect of PD-1 engagement in inducing B cell dysfunction has been also described in HIV infection (Moir & Fauci, 2014; Boliar et al., 2012). Furthermore, Huang et al. (2016), reported that the density of CD20 positive B cells was significantly increased in the liver tissues of patients with chronic liver disease (CLD) of different etiologies compared to normal liver tissues. CLD patients with a higher hepatitis grade presented with significantly more intense B cell infiltration compared to those with a lower grade of inflammation. Enrichment of B cells in hepatic diseases may implicate this subpopulation in their pathogenesis. Since the development of collateral circulation, as a consequence of portal hypertension, facilitate the interaction of gut-derived antigens and endotoxins (that bypass the liver) with antibody-producing cells, hyperimmunoglobulinemia is frequently seen in patients with ALD (Husby et al., 1977; Tomasi & Tisdale, 1964). Alterations in the B cells number in heavy alcoholics and patients with ALD have been reported in previous studies (Matos et al., 2013; Zhang & Meadows, 2005). B cells through the immunoglobulin secretion may mediate both T cells-dependent and—independent immune responses (Li et al., 2019; Parra, Takizawa & Sunyer, 2013). Short-term ethanol treatment at high dose down-regulated splenic macrophages and DCs activity via enhancing B cells function as the antigen-presenting cell, and eventually facilitating a microenvironment that leads to increased activation of CD4⁺ T cells (Andrade et al., 2009). The relevance of B cells in patients with ALD deserves to be further clarified.

Furthermore, significant differences were found between females and males in the ALD group. They differed in terms of the frequencies of PD-1 and PD-L1 positive B cells, as well as the PD-L1 positive all T subsets. The aforementioned gender-related immune disparities occurred as a result of the disease because they were not observed in the control group (Table 2).

Our findings are consistent with recently published reports that address the role of sex hormones in modulating PD-1/PD-L1 signaling and gender-related efficacy of their blockage (Bardhan, Anagnostou & Boussiotis, 2016; Wu et al., 2018; Capone et al., 2018). Sex hormones influence both innate and adaptive immune responses by their impact on T and B cells, antigen presentation and cytokine secretion (Klein & Flanagan, 2016). It shed light on the gender-related discrepancy observed in immune disorders. Also, our results confirmed the presence of the sexual dimorphism in immune reactivity during the course of ALD due to a divergence between males and females in the frequencies of PD-1/PD-L1 positive T and B lymphocytes. Previous reports have already provided some evidence of the gender-related differences in morbidity among alcohol abusers, but underlying mechanisms are not well understood (Chou, 1994; Greenfield, 2002).

Drinking alcohol presents a health challenge for women. Even small amounts of alcohol affect women in a more severe way than men and are much more risky for them. Also, our study introduced the evidence, that females with the severe course of ALD defined by MELD>20, CTP class C or mDF>32 presented with the higher frequencies of PD-1 and PD-L1 positive B cells (CD19+) and the PD-L1 positive T cell subsets (CD4+, CD8+) in comparison not only with female controls, but also with ALD females whose MELD was <=20 and mDF<=32, as well as with their male counterparts with MELD>20, CTP class C and mDF>32 (Tables 4, 5 and 6; Figs. 2, 3 and 4). Moreover, women with ALD classified as the CTP class C had significantly higher frequencies of PD-L1 positive CD8+ T lymphocytes in comparison to ALD females with CTP class B (Fig. S3). Also, frequencies of CD4+PD-1+ were significantly lower in females with CTP class B compared with femalres classified as CTP class A and C (Fig. S2). Our findings may suggest the impairment in the regulation of immune suppressive mechanisms in women with the severe ALD course. Further functional tests should be performed to confirm our assumption. The results are also consistent with previous reports indicating that PD-L1 is ubiquitously expressed at low levels and strongly induced by proinflammatory signals (Liang et al., 2003). Similar data were obtained by Markwick et al. (2015), who performed the ex-vivo evaluation of immunological features of patients with acute alcoholic hepatitis (AAH) and reported hyperexpression of the PD-1 and PD-L1 proteins on T cells. Moreover, the researchers demonstrated an impaired function of T lymphocytes obtained from AAH patients what was manifested by decreased production of gamma interferon. Increased numbers of IL-10 synthesizing T cells in response to chronic endotoxin exposure were also observed. Blocking PD-1 and TIM3 reversed the aforementioned effects and restored the antimicrobial activity of both T lymphocytes and neutrophils. The findings link the PD-1 pathway to the regulation of antibacterial immunity.

The PD-1/PD-L1 expression on peripheral T and B lymphocytes was also found to be increased in other inflammatory disorders. Jia et al. (2016), reported an upregulation of PD-1 expression on CD4+ and CD8+ T cells in severe septic patients and in patients with type 2 diabetes mellitus (T2DM). However, patients with severe sepsis showed higher PD-1 expression in comparison with T2DM patients.

The increased expression of PD-1 on T cells and PD-L1 on monocytes was observed in patients with septic shock in comparison with healthy controls (Zhang et al., 2011). However, in contrast to our study, the PD-1 expression on B cells (CD19+) and the PD-L1 expression on T (CD4+, CD8+) and B cells (CD19+) were not changed.

There are many similarities between severe ALD and sepsis in the context of the immune derangement. Patients with alcohol-associated liver failure, alike septic patients, present with an increased susceptibility to infections as a result of concurrent immune activation and immune exhaustion that seems to be related to the overexpression of negative immune checkpoints causing immunosuppression (Riva & Chokshi, 2018). It becomes more clear now, why several attempts to attenuate inflammatory response by the administration of antibodies directed against tumor necrosis factor-alpha have brought disappointing results in severe AH (Naveau et al., 2004).

The PD-1 overexpression on neutrophils in patients with systemic lupus erythematosus (SLE) correlated with the disease activity and severity. Similarly in our study, the frequencies of PD-L1 positive CD8+ correlated with mDF in ALD females (Table 7). The results of Liu, Weng & Weng (2009), showed that patients with SLE had significantly increased percentages of PD-1-expressing CD3+T cells and CD19+B cells, PD-L1-expressing CD19+ B cells, and PD-L2-expressing CD14+ monocytes. ALD females in our study also presented with significantly higher frequencies of the PD-1/PD-L1 positive B cells (CD19+) compared to female controls (Fig. 1, Table 2).

During the time of regular immune response, the upregulation of PD-1 expression occurs in response to T cell activation. Then interaction with PD-L1 and PD-L2 appears to deliver negative signals and eventually lead to the apoptosis of activated lymphocytes (Sharpe et al., 2007). Several reports reveal that persistent hyperexpression of the co-inhibitory molecules promotes T cell exhaustion with a progressive loss of their effector function. Exhausted T cells express multiple inhibitory receptors, that take part in the negative immune regulation. Therefore, the checkpoint inhibitors affect the host response to infection by modulating the balance between effective immune defense and immune-mediated tissue injury (Fallon et al., 2018).

Microorganisms may take advantage of the inhibitory signals in order to escape immune surveillance and favor chronic infection (Keir et al., 2008). Therefore, overexpression of PD-1/PD-L1 is a major feature of chronic infections in humans (Golden-Mason et al., 2007; Day et al., 2006; Boni et al., 2007). Consequently, findings in our female subgroup may also suggest that their antibacterial defense is impaired. On this basis, we are tempted to speculate that women with ALD may be more than men prone to infections, that frequently complicate the course of the disease (Jalan et al., 2014). The issue needs to be explored in future trials.

Xu et al. (2014), demonstrated that PD-1 expression on T cells was changing along with the HBV infection progression and its variation was correlated with HBV virus load as well as liver function. We also observed differences in the frequencies of PD-1/PD-L1 positive T and B cells associated with the progression and severity of the ALD course in females (Tables 4, 5 and 6; Figs. 2, 3 and 4). Therefore, the expression patterns seem to change with the disease evolution, from the early stage of steatosis through alcoholic hepatitis to liver cirrhosis, and finally to liver cancer, and await further exploration.

Moreover, the available body of evidence indicates that the PD-1/PD-L1 pathway can impact the immune clearance of antigen-presenting cells and T lymphocytes, and promote cancer development. Therefore, it has emerged as a promising new target for cancer therapy due to a good treatment response seen in metastatic renal and lung carcinomas, as well as melanomas (Philips & Atkins, 2015). Whether the expansion of the PD-1/PD-L1 positive lymphocytes demonstrated in our studied women with ALD is relevant in terms of the risk of neoplasm development needs to be further clarified.

Our results are particularly noteworthy in the light of alarming findings of the recent publication authored by Grant et al. (2017). The researchers reported increases in the proportion of women who drink alcohol and in high-risk drinking patterns among women, as well as the higher 12-month prevalence of alcohol use disorders among women.

Also, a recent study by Lowe et al. (2019), points to gender differences in alcohol preference and indicates that inhibition of various steps in inflammasome signaling can reduce alcohol consumption in females. Therefore, targeting an inflammatory cascade in ALD seems to open new insights into the development of modern treatments focused on individually tailored therapy.

If the results of our study are confirmed, several potential prospects emerge from them.

1. Since gender-related disparities in regulatory immune mechanisms exist in ALD patients, there is a strong need to re-arrange the treatment regimens.

2. Host-directed immunomodulatory therapies aimed at re-establishing the impaired immunity create an encouraging approach to the management of patients with ALD.

3. Targeting the PD-1/PD-L1 pathway seems to be helpful in amelioration of immune-dysregulation and therefore requires further in-depth investigations in ALD.

Our findings should be interpreted with caution in the context of potential limitations. First, this was a single-center and a pilot study, therefore the overall sample size is relatively small. Second, we have not evaluated the changes of the PD-1/PD-L1 expression in the same individual over time, therefore, the impact of treatment effects cannot be assessed. Third, since the present study was not designed to predict the mortality of ALD in relation to the PD-1/PD-L1 expression, the relatively small number of non-survivors (only 5 persons) makes the evaluation likely biased. Fourth, alcohol consumption was self-reported in our study. As demonstrated by Stockwell & Stirling (1989), most people are not able to accurately assess the volume and power of a drink, so the real ethanol intake may lack precision. The impact of recall bias, as well as a deliberate misreporting of alcohol consumption also cannot be ruled out.

Conclusions

Our research adds new evidence to a limited number of studies conducted on the role of checkpoint inhibitors in alcohol abusers with ALD. The findings show that gender-related differences in the frequencies of PD-1/PD-L1 positive T and B cells in the peripheral blood exist and might potentially be related to different susceptibility to ethanol-induced liver damage in men and women. Since the upregulation of the frequencies of PD-1/PD- L1 positive lymphocytes paralleled both the severity of AH and liver dysfunction in females with ALD, the pathway seems to play a pivotal role in the disease progression. Results from our study may be further exploited to re-define therapeutic targets and create sex-tailored interventions for ALD treatment according to individual patient needs. Consequently, our findings represent an initial step in the exploration of ethanol- and sex-related alterations in immune cells reactivity modulated by the PD-1/PD-L1 pathway.

Supplemental Information