Chronic ethanol consumption reduces existing CD8 T cell memory and is associated with lesions in protection against secondary influenza A virus infections

Abstract

Chronic alcohol consumption is associated with increased incidence of disease severity during pulmonary infections. Our previous work in a mouse model of chronic alcohol consumption has detailed that the primary influenza A virus (IAV)-specific CD8 T cell response in mice that consumed ethanol had reduced proliferative capacity as well as the ability to kill IAV-target cells. Interestingly, recent studies have highlighted that human alcoholics have increased susceptibility to IAV infections, even though they likely possess pre-existing immunity to IAV. However, the effects of chronic alcohol consumption on pre-existing immune responses (i.e. memory) to IAV have not been explored. Our results presented herein show that IAV-immune mice that then chronically consumed alcohol (X31➞EtOH) exhibited increased morbidity and mortality following IAV re-exposure compared to IAV-immune mice that had consumed water (X31➞H₂O). This increased susceptibility in X31➞EtOH mice was associated with reduced IAV-specific killing of target cells and a reduction in the number of IAV-specific CD8 T cells within the lungs. Furthermore, upon IAV challenge recruitment of the remaining memory IAV-specific CD8 T cells into the lungs is reduced in X31➞EtOH mice. This altered recruitment is associated with a reduced pulmonary expression of CXCL10 and CXCL11, which are chemokines that are important for T cell recruitment to the lungs. Overall these results demonstrate chronic alcohol consumption negatively affects the resting memory CD8 T cell response and reduces the ability of memory T cells to be recruited to the site of infection upon subsequent exposures, therein contributing to an enhanced susceptibility to IAV infections.

Introduction

The abuse of alcohol by humans represents a significant health concern causing 1 in 20 deaths, worldwide (1). These deaths are due, in part, to the increased risk of severe pulmonary bacterial and viral infections that are associated with chronic alcohol consumption (2-6). A primary cause for the increased risk of severe pulmonary disease is the detrimental impacts that alcohol abuse has on the immune system. In humans that chronically abuse alcohol as well as following short periods of binge drinking, significant reductions in the numbers of T cells and B cells have been documented (2, 7, 8). This alcohol-induced lymphopenic state has been recapitulated in both rodent and non-human primate models of chronic ethanol (EtOH) consumption (2, 9-13). Although the overall reduced number in immune cell precursors likely contributes to disease severity following an infection, studies utilizing animal models of chronic EtOH consumption have revealed impairments in the development and function of pathogen-specific adaptive immune responses. During influenza A virus (IAV) infection in mice that chronically consume EtOH, the few primary effector IAV-specific CD8 T cells that do develop have a significantly reduced proliferative capacity, decreased cytotoxicity, and decreased production of interferon γ (14). This likely contributes to the increased magnitude and duration of a primary IAV infection observed in chronic EtOH consuming mice (10).

While studies have detailed the negative impacts of chronic alcohol consumption on the steady-state and developing immune responses during a primary infection, much remains unknown concerning the effects of chronic ethanol on pre-existing immune memory responses. Interestingly, heavy alcohol use in humans is associated with increases in the risk of severe IAV-associated disease (15). In part, this increase in disease may be related to the evidence discussed above of the lesions within the primary effector IAV-specific immune response; however, considering that the highest rates of alcohol dependence occur in adulthood (49), the issue becomes far more complex as these individuals likely had pre-existing T cell immunity. Studies show that there is a high prevalence of pre-existing T cell immune memory to IAV within the general population (16-19), including resident T cell memory within the lungs (20-23). Furthermore, studies examining immunity and protection during the 2009 IAV pandemic suggest that individuals that had pre-existing cross-reactive T cell immunity were more resistant to severe IAV-associated disease compared to those that did not (24-26). Thus, these studies indicate that chronic alcohol consumption may create additional lesions in pre-existing IAV-specific CD8 T cell immunity. Therefore, we determined the consequences of chronic alcohol consumption on existing memory CD8 T cell-mediated protection against IAV.

The composition of the CD8 T cell memory pool established following a primary infection consists of four known subsets. T central memory cells (T_CM) express both CD62L and CCR7, allowing them to traffic into the lymph nodes where they undergo rapid proliferation upon pathogen re-exposure. Conversely, T effector memory cells (T_EM) are absent from the lymph node due to their lack of expression of both CD62L and CCR7, and instead scan peripheral tissues, exhibiting potent cytotoxicity (27-29). In recent studies, a third subset termed T peripheral memory cells (T_PM) was identified that shares the proliferative capacity of T_CM as well as the cytotoxic capabilities of T_EM upon subsequent pathogen exposures. These three subsets are defined based on their expression of CX3CR1 and CD27, with T_CM (CX3CR1^loCD27^hi), T_EM (CX3CR1^hiCD27^lo), and T_PM (CX3CR1^intCD27^hi) differentially expressing these surface markers (30). While it is known that both T_CM and T_EM contribute to protection against IAV infection, the role that T_PM play has not been determined. However, it is likely that T_PM contribute to protection against IAV infection as they have been observed within the lung tissue (30). The fourth subset, known as tissue-resident memory cells (T_RM), are embedded within peripheral tissues, and act as a first-line of defense by rapidly responding upon pathogen re-exposure (31-34). This subset is distinguished from cells within the vasculature by intravascular (i.v.) antibody labeling and phenotypically characterized in the lungs by the expression of CD69 and CD103 (27, 35-37). Although originally characterized in animal models, highly proliferative and polyfunctional IAV-specific T_RM have also been identified within the human lung (21, 38, 39). Studies reveal that T_RM within the lung are important for mediating heterosubtypic protection, as mice lacking lung T_RM exhibit increased morbidity and mortality following heterologous IAV challenge (40, 41). Overall these studies indicate that multiple T cell memory subsets contribute to protection against secondary IAV infection and could be negatively impacted by chronic alcohol consumption.

Our results demonstrate that chronic EtOH exposure of IAV-immune mice increased morbidity and mortality upon a secondary IAV infection compared to IAV-immune H₂O control mice. Our findings further demonstrate alterations in the IAV-specific CD8 T cell memory response in IAV-immune EtOH mice compared to IAV-immune H₂O control mice. Upon re-exposure with IAV, we observed that IAV-immune EtOH mice had reduced accumulation of IAV-specific CD8 T cells within the lungs compared to IAV-immune H₂O controls. This reduced accumulation was associated with lower levels of two important CD8 T cell chemoattractants (CXCL10 and CXCL11) within the lungs. Overall, our findings highlight that chronic EtOH consumption negatively impacts existing T cell immunity to IAV infection and is associated with increased morbidity and mortality during a secondary exposure to IAV.

Materials and Methods

Mice

Six- to eight-week-old wild-type (WT) C57Bl/6 mice were originally obtained from the National Cancer Institute (Frederick, MD). B6.Cg-Tcra^tm1MomTg(TcrLCMV) 372Sdz CD90.1 (P14) mice whose T cells express a T cell receptor specific for the GP₃₃ epitope of LCMV were a kind gift from Dr. John T. Harty from the University of Iowa (Iowa City, IA). Mice were bred, housed, and maintained in the animal care facilities at the University of Iowa. All procedures performed were approved by and adhered to the regulatory standards and guidelines of the Institutional Animal Care and Use Committee of the University of Iowa.

IAV infection

Mouse adapted laboratory strains of X31, A/Puerto Rico/8/34 (PR8), and PR8-GP₃₃ (generated and kindly gifted by Dr. Ryan Langlois at the University of Minnesota, Minneapolis, MN) were prepared from stocks as previously described (42). X31-GP₃₃ was a kind gift from Dr. Steven Varga at the University of Iowa. Prior to intranasal (i.n.) infections, mice were anesthetized with isoflurane. The following doses of IAV strains were administered i.n. in 50μL of Iscove’s medium: 0.1LD₅₀ (1.11×10² TCIU₅₀) or 5LD₅₀ (5.54×10³ TCIU₅₀) A/Puerto Rico/8/34; 1.20×10³ TCIU₅₀ X31; 1.37×10⁴ TCIU₅₀ X31-GP₃₃; 2×10⁴ TCIU₅₀ PR8-GP₃₃.

Ethanol administration

Mice were administered ethanol using the Meadows-Cook model (10, 13, 43-45). Briefly, mice were acclimated to pharmaceutical grade ethanol supplemented in distilled water at the following concentrations: 10% for two days, 15% for five days, and 20% for seven weeks. These ethanol solutions were the only source of drinking fluid available to the ethanol group. Control animals received the same distilled water utilized for the ethanol solutions. Mice had free access to the appropriate drinking solution as well as laboratory chow during the treatment period. Once administered, mice were maintained on their respective drinking fluid through the remainder of the study (i.e. following IAV infection).

In vivo cytotoxicity assay

Procedures for the in vivo cytotoxicity assay were modified from the procedures previously described (46). Splenocytes from C57Bl/6 mice were resuspended in 14% Nycodenz (Axis-Shield, Oslo, Norway) that was overlaid with an equal volume of Iscove’s medium. Splenocytes were centrifuged at 2200 × g for 15 minutes at 4°C without brake or start assist. Mononuclear cells were isolated from the interface and labeled with 2μM PKH-26 (Sigma-Aldrich) at room temperature for five minutes. PKH-26 labeled mononuclear cells were split into two groups: one group was labeled with 1μM CFSE (CFSE^lo) and the other was labeled with 3μM CFSE (CFSE^hi) at room temperature for ten minutes. Labeling reactions were neutralized by adding equal volumes of fetal bovine serum (FBS). CFSE^hi and CFSE^lo cells were pulsed with 10μM of OVA₂₅₇ or 10μM IAV NP₃₆₆ peptides for 30 minutes at 37°C, respectively. CFSE^hi and CFSE^lo were mixed 1:1 and 50,000 cells were adoptively transferred intranasally into test mice. The ratio of CFSE^hi:CFSE^lo cells was determined eight hours after transfer. Percent specific killing was determined by the following equation ([100 - (%CFSE^lo cells/%CFSE^hi cells)]) and normalized to naïve controls.

Intravascular cell labeling

Cells within the vasculature and tissue were distinguished using an intravascular stain as previously described (35). Briefly, mice received 1μg of fluorophore-conjugated rat anti-mouse CD45.2 (clone 104; BioLegend, San Diego, CA) in 200μL of PBS by retroorbital intravenous injection three minutes prior to euthanasia.

Major histocompatibility complex class I tetramers

Major histocompatibility complex (MHC) class I tetramers specific for the H-2D(b) restricted influenza A virus nucleocapsid protein epitope NP_366-374/ASNENMETM were obtained from the National Institute of Allergy and Infectious Disease MHC Tetramer Core Facility (Atlanta, GA).

Flow cytometry

Lungs and lung draining lymph nodes (dLN) were harvested and digested for 30 minutes at 37°C in medium containing 1mg/mL Collagenase (Type 3; MP Biomedicals, Solon, OH) and 0.02 mg/mL DNase-I (MP Biomedicals). Single cell suspensions were made and 1×10⁶ cells/well were plated in 96-well round bottom plates. Cells were blocked with 2% rat and hamster serum for 30 minutes at 4°C. After blocking, antigen experienced CD8 T cells were identified as previously described (47). The following antibodies were used to identify antigen experienced CD8 T cell memory subsets: rat anti-mouse CD8α (53-6.7; BioLegend), rat anti-mouse CD11a (M17/4; BD Biosciences, San Jose, CA), rat anti-mouse CD103 (M290; BD Biosciences), rat anti-mouse CD69 (H1.2F3; eBioscience), rat anti-mouse CD127 (A7R34; BD Biosciences), rat anti-mouse CD62L (MEL-14; BioLegend), hamster anti-mouse KLRG1 (2F1/KLRG1; BioLegend), mouse anti-mouse CX3CR1 (SA011F11, BioLegend), and hamster anti-mouse CD27 (LG.3A10, BioLegend). Cells were fixed with BD FACS™ Lysing Solution per the manufacturer’s instructions and resuspended in PBS. Data were acquired on a LSRII (BD Biosciences) and analyzed using FlowJo software (Tree Star, Ashland, OR).

Chemokine assay

Prior to euthanasia, blood was collected in non-heparinized capillary tubes (Fisher Scientific) and serum was collected as previously described (48). Lungs were homogenized using a Dounce tissue homogenizer and homogenates were purified by centrifugation. Serum and lung lysates were stored at −80°C until analysis. Serum and lung lysate samples were analyzed using the Bio-Plex Pro™ Mouse Chemokine Panel 33-Plex (Bio-Rad, Hercules, CA) per the manufacturer’s instructions. Data were acquired on a Bio-Plex 200 (Bio-Rad).

Statistical analysis

Experiments were repeated at least twice unless noted otherwise. All statistical analysis listed below were performed using GraphPad Prism software (La Jolla, CA). Comparisons between two groups was performed with a two-tailed student’s t-test. Comparisons between more than two groups at different time points were analyzed using two-way ANOVA with Holm-Sidak’s multiple comparison post-hoc test. Comparisons between more than two groups at a single time point were analyzed using a one-way ANOVA with a Tukey’s multiple comparison post-hoc test. Survival comparisons were analyzed using a Mantel-Cox Log Rank test. A P-value ≤ 0.05 was considered significant.

Results

Chronic ethanol consumption increases disease severity following secondary IAV infection

To determine the effects of chronic alcohol consumption on memory CD8 T cell-mediated immune protection against a secondary IAV infection, mice that were previously infected with a sublethal dose of IAV-X31 (H3N2) eight weeks prior were placed on 20% ethanol in distilled water (X31-immune→EtOH) using the Meadows-Cook protocol or distilled water alone (X31-immune→H₂O) for eight weeks (Figure 1A). IAV-X31 is a reassortant virus that contains the HA and NA gene segments from a H3N2 virus, while the remaining six gene segments are on an IAV-PR8 (H1N1) backbone. This allowed us to subsequently challenge X31-immune mice with IAV-PR8, to avoid HA3 and NA2- mediated antibody neutralization of the challenge infection, and ascertain the contribution of CD8 T cells to immunity and protection. Following the PR8 challenge, we observed substantial weight loss in naïve mice regardless of water or EtOH exposure (Figure 1B). However, similar to previous studies (10, 14), we observed that naïve EtOH consuming mice (Naïve →EtOH) exhibited enhanced mortality compared to naïve water consuming mice (Naïve→H₂O) (Figure 1C). As expected, disease associated weight loss was significantly ameliorated in X31-immune→H₂O compared to Naïve→H₂O and Naïve→EtOH mice; however, X31-immune→EtOH mice showed increased weight loss compared to X31-immune→H₂O mice (Figure 1B). Although no differences in mortality were observed for X31-immune→H₂O and X31-immune→EtOH mice challenged with a 0.1LD₅₀ exposure (Figure 1C), the increased weight loss in X31-immune→EtOH mice suggests that EtOH-induced deficits have occurred that decrease the ability to control IAV reinfection. To probe this lesion further, we determined the capabilities of mice to control an infection with a high-dose challenge (5LD₅₀) of PR8, which mimics the severity of many pandemic IAV infections. The high pathogenicity of this dose of PR8 was apparent as all naïve mice succumbed to the infection by day 10 (Figure 1E). While there were no differences in weight loss, 80% of X31-immune→EtOH mice succumbed to the high-dose PR8 challenge compared to 0% mortality in the X31-immune→H₂O group (Figure 1D & 1E). Given the known role in CD8 T cell mediated protection against subsequent heterologous IAV exposures(40, 41), these data suggest that chronic EtOH consumption may negatively impact pre-existing IAV-specific immunity and increase susceptibility to disease associated morbidity and mortality during subsequent IAV exposure.

Figure 1.

Chronic ethanol consumption detrimentally impacts susceptibility to influenza A virus infection in IAV-immune mice. (A) C57BL/6 mice were infected with 1.20×10³ TCIU₅₀ of X31 (H3N2) and allowed to establish immune memory for 8 weeks. At 8 weeks post X31 infection, mice were placed on the Meadows-Cook ethanol protocol or distilled water for 8 weeks. At 16 weeks post X31 infection, mice were challenged with either a (B,C) 1.11×10² TCIU₅₀ or a (D, E) 5.54×10³ TCIU₅₀ dose of heterologous A/Puerto Rico/8/34 (PR8; H1N1) and (B, D) weight loss and (C, E) survival were monitored. Error bars represent mean ± s.e.m. Data are from one independent experiment with n= 5 mice/group. X31→H₂O vs. Naïve→H₂O: ‡ P < 0.05; X31→H₂O vs. X31→EtOH: † P < 0.05; X31→EtOH vs. Naïve→H₂O: *P < 0.05 (Two-way ANOVA with Holm-Sidak multiple comparisons test).

Pre-existing IAV-specific CD8 T cell responses are reduced following chronic ethanol exposure

The increased susceptibility of X31-immune→EtOH mice to PR8 infection indicates deficiencies within the IAV-specific memory T cell response. Previous studies from our laboratory have associated defects in cytotoxic IAV-specific CD8 T cells with increased susceptibility and duration of a primary IAV infection in EtOH consuming mice (10, 14). As the challenge PR8 (H1N1) virus should evade the neutralizing X31-specific antibody response (i.e. H3N2) generated by the prior X31 infection, B cell responses are likely not contributing to the protection observed following PR8 infection of X31-immune mice. Thus, the deficiency in the IAV-specific memory response instead likely resides within the memory CD8 T cell compartment. To determine if limitations in cytotoxicity are present in the memory IAV-specific CD8 T immune response, we utilized an in vivo cytotoxicity assay. In both Naïve→H₂O and Naïve→EtOH mice, we observed no killing of target cells as equal frequencies of IAV peptide-pulsed (CFSE^lo) and control OVA peptide-pulsed (CFSE^hi) target cells were observed (Figure 2A). This highlights, as expected in non-IAV immune mice, the absence of a cytotoxic IAV-specific CD8 T cell response and that the lung environment of EtOH consuming mice did not result in selective lysis of either target cell population. However, lysis of IAV-specific target cells within the lungs of X31-immune→H₂O mice indicates the presence of a localized cytotoxic IAV-specific CD8 T cell response (Figures 2A and 2B). Lysis of IAV-specific target cells was significantly lower in X31-immune→EtOH mice compared to X31-immune→H₂O mice (Figures 2A and 2B), which suggests a defective cytotoxic IAV-specific CD8 T cell response and is consistent with the increased morbidity and mortality observed in X31-immune→EtOH mice upon secondary IAV infection (Figure 1). Since both groups of X31-immune mice should possess similar memory responses prior to EtOH or water exposure, these results suggest that chronic EtOH consumption negatively impacts the overall cytotoxic ability of the pre-existing CD8 T cell memory response.

Figure 2.

Chronic ethanol consumption negatively impacts existing IAV-specific CD8 T cell cytotoxic responses. Mice were infected then exposed to EtOH or water as in Figure 1. At 16 weeks post X31 infection, mice received 50,000 target cells i.n. from a 1:1 mixture of IAV NP₃₆₆ (CFSE^lo) and OVA₂₅₇ (CFSE^hi) peptide-pulsed splenocytes. Eight hours after i.n. transfer, (A) the ratio of CFSE^hi to CFSE^lo target cells within the lung were determined and (B) percent specific killing was calculated with the following equation: ([100 – (%CFSE^lo cells/%CFSE^hi cells)]). Data were normalized to the ratio/killing observed in Naïve→H₂O and Naïve→ EtOH controls. Error bars ± s.e.m. Data are representative of two combined experiments with n= 6-7 mice/group. P values were determined by two-tailed t-test.

The inability of X31-immune→EtOH mice to clear IAV-specific target cells from the lungs suggests a lesion within the IAV-specific CD8 T cell memory compartment. Since total CD8 T cells are numerically lower within human alcoholics and EtOH consuming animal models, we next determined if this numerical reduction extended into X31-immune mice. Consistent with prior findings (10), we observed a significant reduction in the total number of CD8 T cells in the lungs of Naïve→EtOH mice compared to Naïve→H₂O mice (Figure 3B). Consistent with the establishment of T cell memory prior to chronic EtOH consumption, total CD8 T cells were increased in the lungs of X31-immune→H₂O mice compared to naïve controls. However, subsequent reduction from that memory CD8 T cell set-point occurred in X31-immune→EtOH mice, as these mice had reduced numbers of total CD8 T cells compared to X31-immune→H₂O mice (Figure 3B). This reduction led to the number of total pulmonary CD8 T cell present in X31-immune→EtOH to be similar to Naïve→H₂O controls (Figure 3B).

Figure 3.

Numerical loss of existing CD8 T cell memory due to chronic ethanol consumption. Mice were infected then exposed to EtOH or water as in Figure 1. At 16 weeks post X31 infection, lungs were isolated and analyzed. (A) Representative gating strategies to enumerate (B) total CD8 T cells and (C) AgExp CD8 T cells (CD11a^hiCD8α^lo) within the lungs. (D) Lung-resident AgExp CD8 T cells (CD11a^hiCD8α^loCD45i.v. Ab^neg) were determined by i.v. antibody labeling. Error bars represent mean ± s.e.m. Data are representative of three combined experiments with n= 8-15 mice/group. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (One-way ANOVA with Tukey’s multiple comparisons test).

As EtOH consumption alters naïve CD8 T cells within the lungs (10), it is possible that this reduction in total CD8 T cells in X31-immune→EtOH mice could be due to alterations in the naïve CD8 T cell pool and/or the existing IAV-specific CD8 T cell memory pool within the lungs. Studies examining the IAV-specific CD8 T cell memory response have revealed that the epitope dominance is highly dynamic and changes over time (51). Therefore, in order to not bias our analysis and to encompass the entire IAV-specific CD8 T cell memory response, we used surrogate markers of activation to quantify the antigen-experienced CD8 (AgExp CD8) T cell memory pool (Figure 3A) (47). As expected, AgExp CD8 T cells were significantly higher in X31-immune→H₂O mice compared to naïve controls, indicating that prior X31 infection induced a pulmonary IAV-specific CD8 T cell response (Figure 3C). Interestingly, AgExp CD8 T cells were significantly reduced in X31-immune→EtOH mice compared to X31-immune→H₂O mice, suggesting a numerical loss of the IAV-specific CD8 T cell memory response due to EtOH exposure (Figure 3C). Furthermore, when examining the localization of these T cells within the lungs, our results demonstrate that although the frequency of lung-resident AgExp CD8 T cells was similar between X31-immune→H₂O and X31-immune→EtOH mice (Figure 3D), there was a significant reduction in lung-resident AgExp numbers in X31-immune→EtOH mice (Figure 3E). There was also a significant reduction in AgExp CD8 T cells in Naïve→EtOH mice compared to Naïve→H₂O mice indicating that loss of AgExp CD8 T cells from the lungs after EtOH was not unique to the IAV-specific CD8 T cell response (Figure 3C). However, the reduced frequency of lung-resident AgExp CD8 T cells in naïve mice compared to X31-immune→H₂O and X31-immune→EtOH mice suggests that IAV-specific CD8 T cell memory responses were negatively impacted by chronic EtOH exposure (Figure 3D). The loss of IAV-specific CD8 T cell memory responses was confirmed by quantifying NP₃₆₆-specific CD8 T cells in X31-immune→EtOH mice, which were reduced in numbers compared to X31-immune→H₂O mice (Supplemental Figure 1A). Further, when the IAV-specific CD8 T cell response was monitored using adoptively transferred donor P14 CD8 T cells (Supplemental Figure 1B), we observed a significant reduction in IAV-specific CD8 T cells in X31_gp33-immune→EtOH mice compared to X31_gp33-immune→ H₂O mice (Supplemental Figure 1C). Altogether these data verify that chronic EtOH consumption causes a numerical loss in the existing IAV-specific CD8 T cell memory pool within the lungs.

Chronic ethanol consumption causes a numerical reduction in CD8 T cell memory subsets

The memory CD8 T cell pool consists of a diverse population of cells that is established following a primary infection. CD8 T cell memory has traditionally been categorized into T central memory (T_CM) and T effector memory (T_EM). However, recent studies have expanded this dogma with the identification of T peripheral memory (T_PM) and tissue-resident memory (T_RM) subsets (23, 30-34, 36, 37, 52). The representation of each of these subsets within the CD8 T cell memory pool changes over time following a primary infection. For the populations within the circulation, T_EM predominate early in the CD8 T cell memory response, but the CD8 T cell pool eventually shifts to being dominated by T_CM and T_PM (30). Similarly, while T_RM can be found within the lung tissue early during the CD8 T cell memory response(52), their numbers wane over time (53). In order to better understand the EtOH-induced loss of CD8 T cell memory, we next determined the effects of chronic EtOH consumption on the composition of the CD8 T cell memory pool by examining the representation of T_RM (CD69^posCD103^pos), T_CM (CD103^negCD27^hiCX3CR1^lo), T_EM (CD103^negCD27^loCX3CR1^hi), and T_PM (CD103^negCD27^hiCX3CR1^int) within the AgExp CD8 T cell pool in the lungs prior to subsequent IAV exposure (Figure 4A). Since prior studies have shown that T_EM are largely excluded from the tissue and are primarily within the circulation (30), we began by analyzing the entire lung (i.e. circulation and tissue). As expected, X31-immune→H₂O mice had significantly increased numbers of AgExp T_CM, T_EM, T_RM, and T_PM memory subsets in the lung compared to both naïve controls (Figure 4B). Interestingly, the numerical loss of AgExp CD8 T cells in X31-immune→EtOH mice was widespread, as there were reductions among all CD8 T cell memory subsets examined compared to X31-immune→H₂O mice (Figure 4B). When we analyzed each memory subset for their tissue-residence based on i.v. antibody labeling, we observed that T_CM, T_RM, and T_PM, but not T_EM, of X31-immune→H₂O and X31-immune→EtOH mice had an increased frequency of lung-resident cells (i.v. Ab^neg) compared to naïve controls (Figure 4C). This observation agrees with previous studies that determined that T_EM are devoid from most tissues and instead are within the circulation (30). Interestingly, we observed decreased frequencies of lung-resident T_PM in X31-immune→EtOH mice compared to X31-immune→H₂O mice, but frequencies of lung-resident T_RM were similar (Figure 4C). The current understanding is that while T_PM survey peripheral sites by migrating in and out of peripheral tissues from the circulation (30), T_RM remain embedded within tissue and do not egress into the circulation (31, 33, 36). Therefore, the decreased frequencies of lung-resident T_PM may indicate that the ability of circulating memory to gain access into the lung tissue is altered, which would be consistent with the altered chemokine environment observed in chronic EtOH animal models (11, 12, 45, 54). Nevertheless, even though the frequency of lung-resident T_RM were similar between X31-immune→H₂O and X31-immune→EtOH mice, there was a significant reduction in the number of lung-resident T_RM, as well as T_CM and T_PM, in X31-immune→EtOH mice (Figure 4D). Overall, these data suggest that chronic EtOH consumption detrimentally impacts multiple IAV-specific CD8 T cell memory subsets within the lungs.

Figure 4.

Numerical loss of existing IAV-specific CD8 T cell memory due to chronic ethanol consumption occurs across multiple T cell memory subsets in the lungs. Mice were infected then exposed to EtOH or water as in Figure 1. At 16 weeks post X31 infection, lungs were isolated and analyzed. (A) Representative gating strategies to (B) enumerate AgExp CD8 T_RM(CD11a^hiCD8α^loCD69^posCD103^pos), T_CM(CD11a^hiCD8α^loCD103^negCX3CR1^loCD27^hi), T_EM(CD11a^hiCD8α^loCD103^negCX3CR1^hiCD27^lo), and T_PM(CD11a^hiCD8α^loCD103^negCX3CR1^intCD27^hi) within the lungs. (C) Frequency of lung-resident (CD45i.v. Ab^neg) memory was determined by i.v. antibody labeling. Error bars represent mean ± s.e.m. Data are representative of two combined experiments with n= 10-15 mice/group. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (One-way ANOVA with Tukey’s multiple comparisons test).

Previous studies have determined that the lung dLNs are an important site not only for the maintenance of the IAV-specific CD8 T cell memory response following the resolution of primary IAV infection (55), but also for their reactivation and expansion upon secondary IAV challenge (51). Therefore, existing AgExp CD8 T cell memory responses within the lung dLNs were quantified (Figure 5). Similar to the trends observed in the lungs, total CD8 T cells and AgExp CD8 T cells in the lung dLNs of X31-immune→EtOH mice were significantly reduced compared to X31-immune→H₂O mice (Figures 5A & 5B). There were also reductions in the number of total IAV-specific CD8 T cells when responses were examined by an individual epitope specificity in the lung dLNs of X31-immune→EtOH mice compared to X31-immune→H₂O mice (Supplemental Figure 1A & 1C). When we analyzed the individual subsets that make up the complete AgExp CD8 T cell memory response, we observed reductions in T_PM in X31-immune→EtOH mice compared to X31-immune→H₂O mice, but T_CM responses appeared unaltered (Figures 5C & 5F). T_EM and nonlymphoid organ (NLO) T_RM responses were not detected in large numbers within the lung dLNs, as these subsets are generally not found in secondary lymphoid organs (27, 30) (Figures 5D & 5E). A recent study has reported on a secondary lymphoid organ (SLO) resident CD8 T cell subset (56); however, we observed no difference in the number of SLO resident CD8 T cells in X31-immune→EtOH mice compared to X31-immune→H₂O mice (Figure 5G). Altogether these results suggest existing IAV-specific CD8 T cell memory responses in the lungs and lung dLNs are negatively impacted by chronic EtOH consumption.

Figure 5.

Numerical loss of existing IAV-specific CD8 T cell memory due to chronic ethanol consumption occurs across multiple T cell memory subsets in the lung dLNs. Mice were infected then exposed to EtOH or water as in Figure 1. At 16 weeks post X31 infection, lung dLNs were isolated and analyzed. (A) Total CD8 T cells, (B) AgExp CD8 T cells (CD11a^hiCD8α^lo), (C) T_CM(CD11a^hiCD8α^loCD103^negCX3CR1^loCD27^hi), (D) T_EM(CD11a^hiCD8α^loCD103^negCX3CR1^hiCD27^lo), (E) NLO T_RM(CD11a^hiCD8α^loCD69^posCD103^pos), (F) T_PM(CD11a^hiCD8α^loCD103^negCX3CR1^intCD27^hi), and (G) SLO T_RM (CD11a^hiCD8α^loCD69^posCD62L^neg) were enumerated within the lung dLNs. Error bars represent mean ± s.e.m. Data are representative of two combined experiments with n= 8-10 mice/group. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 (One-way ANOVA with Tukey’s multiple comparisons test).

IAV-specific CD8 T cell recall responses are altered in chronic ethanol consuming mice

Following subsequent exposures of IAV, a critical component of protection is the early proliferation and accumulation of IAV-specific memory CD8 T cells within the lungs and lung dLNs (51). Interestingly, in mouse models of chronic EtOH consumption, the developing IAV-specific CD8 T cell effector response has a blunted proliferative capacity that ultimately contributes to increased disease severity (14). Considering we observed increases in disease severity in X31-immune→EtOH mice compared to X31-immune→H₂O mice (Figure 1), we next determined if EtOH consumption altered the expansion of existing IAV-specific CD8 T cell memory in X31-immune→EtOH mice following IAV re-exposure. To this end, X31-immune→H₂O and X31-immune→EtOH mice were challenged with a 0.1LD₅₀ dose of PR8 and AgExp CD8 T cells were quantified in the entire lung, the lung parenchyma (CD45 i.v.Ab^neg), and the lung vasculature (CD45 i.v.Ab^pos) at day 3 p.i., i.e. a time point preceding the arrival of de novo effector T cells in the lungs (Figure 6) (57-59). Within the total lungs (i.e. interstitium + vasculature) of X31-immune→EtOH mice, the total AgExp CD8 T cell response was significantly lower in number three days after PR8 challenge compared to X31-immune→H₂O mice (Figure 6A); however, the fold expansion (i.e. accumulation) of total AgExp CD8 T cells was similar in X31-immune→EtOH mice and X31-immune→H₂O mice based on the fold-increase versus the numbers present in pre-challenged mice (7.6x vs. 7.7x, Figure 6A). This translated into little to no defects in the fold-increase of T_CM, T_EM, T_RM, and T_PM within the lungs of X31-immune→EtOH mice following PR8 challenge (Figures 6B – 6E). Interestingly, when we examined the AgExp CD8 T cell responses within the lung vasculature (i.e. CD45 i.v.Ab^pos) of X31-immune→EtOH mice, the fold-increase of total AgExp CD8s, as well as in T_CM, T_EM, and T_PM, were higher compared to X31-immune→H₂O mice (Figures 6F, 6G, 6H, & 6J). However, little to no difference was detected in the fold-increase of lung-resident T_RM in X31-immune→EtOH mice (Figure 6I). Conversely, when we assessed cells within the lung tissue (i.e. CD45 i.v.Ab^neg) we observed substantial reductions in the fold-increases of total AgExp CD8s, T_CM, T_EM, and T_PM of X31-immune→EtOH mice compared to X31-immune→H₂O mice (Figures 6K, 6L, 6M, and 6O) and little to no change in T_RM (Figure 6N). This reduced fold-increase of T_CM, T_EM, and T_PM within the lung tissue, when coupled to the increases observed within the lung vasculature of X31-immune→EtOH mice, suggests that EtOH consumption causes defects in migration of these T cell subsets into the tissue from the vasculature and/or their proliferation once in the lung interstitium (Figure 6).

Figure 6.

Recall response of existing IAV-specific CD8 T cell memory are altered in the lungs of EtOH consuming mice. Mice were infected then exposed to EtOH or water as in Figure 1. At 16 weeks post X31 infection, mice were challenged with heterologous PR8 (H1N1) and lungs were isolated and analyzed at three days post PR8 challenge. (A, F, K) AgExp CD8 T cells, (B, G, L) T_CM, (C, H, M) T_EM, (D, I, N) T_RM, and (E, J, O) T_PM were enumerated within the (A-E) total lung, (F-J) within the lung vasculature (i.e. CD45 i.v. Ab^pos), or (K-O) within the lung tissue (i.e. CD45 i.v. Ab^neg). Data were compared to pre-challenge numbers from Figure 5 (Day 0) to calculate fold-increase. Error bars represent mean ± s.e.m. Data are representative of two combined experiments with n= 4-5 mice/group. *P < 0.05, **P < 0.01, ***P < 0.001, NS= not significant (student’s t test).

Previous studies show that effector IAV-specific CD8 T cells in chronic EtOH mice have a reduced proliferative capacity (14). Interestingly, based on Ki67 staining, we observed a significant reduction in the frequency of proliferating AgExp CD8 T cells within the lung tissue of X31-immune→EtOH mice prior to secondary challenge. However, no differences in the proliferation of AgExp memory CD8 T cells was observed within the lungs of X31-immune→EtOH mice and X31-immune→H₂O mice at 3 days post challenge (Supplemental Figure 2). Furthermore, the fold-increase in total AgExp CD8 T cells within the lung dLNs in X31-immune→EtOH mice was similar to X31-immune→H₂O mice (Figure 7A). T_CM and T_PM appear to equally contribute to this expansion, as we observed similar fold-increases in both subsets in X31-immune→EtOH mice and X31-immune→H₂O mice (Figures 7B & 7C). Overall these data strongly suggest that proliferation of memory CD8 T cells is not affected by EtOH consumption following secondary IAV challenge, but that the defect in T_CM, T_PM, and T_EM lies in their reduced initial numbers as well as their ability to migrate out of the blood and into the inflamed lung tissue in X31-immune→EtOH mice.

Figure 7.

Recall response of IAV-specific CD8 T cell memory cells in the lung dLNs of EtOH consuming mice. Mice were infected then exposed to EtOH or water as in Figure 1. At 16 weeks post X31 infection, mice were challenged with PR8 (H1N1) and lung dLNs were isolated and analyzed at three days post PR8 challenge. (A) AgExp CD8 T cells, (B) AgExp T_CM and (C) AgExp T_PM were enumerated within the lung dLNs. Data were compared to pre-challenge numbers from Figure 6 (Day 0) to calculate fold-increase. Error bars represent mean ± s.e.m. Data are representative of two combined experiments with n= 4-5 mice/group.

Chronic EtOH consumption significantly alters the chemokine milieu during secondary IAV infection

It is well established that chemokine and chemokine receptors have an essential role in controlling the migration patterns of immune cells, including T cells during the steady-state and during inflammation/infection(60-66). Studies show that following secondary respiratory virus exposure, CCR5 and CXCR3 play crucial roles in the trafficking of T cell memory to the lungs and lung airways (60, 62-66). Furthermore, it is known that the chemokine environment is significantly altered following chronic EtOH exposure in non-human primates (11, 12, 54) and that the ligands for CCR5 (CCL5) and CXCR3 (CXCL10, CXCL11) are significantly altered in the lungs of human alcoholics (67). Thus, we determined the expression of CCR5 and CXCR3 on AgExp CD8 T cells in the lungs of X31-immune→H₂O and X31-immune→EtOH mice following a secondary IAV challenge. Unexpectedly, we observed little to no differences in the expression of CXCR3 or CCR5 on AgExp CD8 T cells from X31-immune→EtOH and X31-immune→H₂O mice (Figure 8A & 8C). We next determined the expression of the chemokine ligands for CXCR3 (CXCL10 and CXCL11) and CCR5 (CCL3, CCL4, and CCL5) within the lungs and serum of X31-immune→EtOH and X31-immune→H₂O mice following secondary IAV challenge. Although there were no differences in chemokine levels prior to secondary challenge (Figure 8B & 8D), we observed significant reductions in CXCL10 and CXCL11 within the lungs of X31-immune→EtOH mice at day 3 post challenge (Figure 8B). Conversely, we observed no difference in the level of CCR5 ligands within the lungs of X31-immune→EtOH mice compared to X31-immune→H₂O mice (Figure 8D). Interestingly, we observed similar levels of CXCL10, CXCL11, CCL4, and CCL5 within the serum of X31-immune→EtOH and X31-immune→H₂O mice (Figure 8B & 8D). Although we did observe a significant reduction in CCL3 within the serum of X31-immune→EtOH mice (Figure 8D), the difference was marginal (i.e. only ~4pg/mL difference). These data indicate that chronic EtOH consumption causes reductions in CXCL10 and CXCL11 within the lungs following IAV re-exposure, which likely impacts the migration of AgExp CD8 T cell memory into the lungs from the vasculature due to reduced signaling through CXCR3.

Figure 8.

Chronic EtOH consumption causes alterations in the chemokine milieu within the lungs of IAV-immune mice following secondary IAV challenge. Mice were infected then exposed to EtOH or water as in Figure 1. At 16 weeks post X31 infection, mice were left unchallenged or challenged with heterologous PR8 (H1N1) and lungs were isolated and analyzed at three days post PR8 challenge. Chemokine receptor expression of (A) CXCR3 and (C) CCR5 was determined for CD45 i.v. Ab^pos and CD45 i.v. Ab^neg AgExp CD8 T cells within the lungs. Serum and lung levels of (B) CXCR3 ligands (CXCL10 and CXCL11) and (D) CCR5 ligands (CCL3, CCL4, and CCL5) were measured by Bio-Plex. Error bars represent mean ± s.e.m. Data are from one independent experiment with n= 3-4 mice/group. *P < 0.05, NS=not significant (student’s t test).

Discussion

Chronic abuse of alcohol has significant impacts on the innate and adaptive branches of the immune system. Not only does chronic alcohol consumption result in an overall numerical reduction in several immune cell subsets, it has also been reported to alter immune cell function. During a primary IAV infection, IAV-specific effector CD8 T cell numbers are reduced in EtOH consuming mice. Further, these cells have a reduced proliferative capacity and decreased cytotoxic ability (10, 14). This dysfunctional IAV-specific effector CD8 T cell response likely contributes to the increased burden of disease and mortality observed during primary IAV infection in chronic EtOH consuming mice and humans (1-7, 10, 13, 68). Yet, unlike immunologically naïve mouse models, humans likely have pre-existing IAV-specific memory responses prior to their chronic alcohol consumption. The data presented herein demonstrate that mice with established IAV-specific CD8 T cell memory within the lungs that then chronically consumed EtOH (i.e. X31-immune→EtOH) exhibited increased morbidity and mortality following a subsequent IAV challenge (Figure 1). The increase in IAV-disease severity in X31-immune→EtOH mice was associated with reduced T cell cytotoxic activity within the lungs, as X31-immune→EtOH mice had a reduced capacity to lyse IAV-specific target cells within the lungs compared to X31-immune→H₂O (Figure 2). This finding during memory is consistent with our previous studies examining the IAV-specific CD8 T cell response during a primary IAV infection, which determined that effector IAV-specific CD8 T cells from chronic EtOH mice had a reduced cytotoxic activity (14).

A consistent finding in our results is that the reduction in the number of existing IAV-specific CD8 T cells was observed across all T cell memory subsets within both the lungs (Figures 3, 4, and Supplemental Figure 1) and the lung dLNs (Figure 5 and Supplemental Figure 1) of X31-immune→EtOH mice compared to X31-immune→H₂O mice. Interestingly, based on Ki67 staining, we did observe a reduction in the proliferation of IAV-specific CD8 T cells within the lungs of IAV-immune EtOH mice prior to PR8 challenge, which likely contributes to the reduced numbers we observed (Supplemental Figure 2). Previous studies have determined that the CD8 T cell memory compartment is maintained by both antigen-dependent and antigen-independent mechanisms (23, 51, 55, 62, 69-77). In particular, maintenance of the existing IAV-specific CD8 T cell memory compartment relies on both IL-15 (73) and antigen-bearing rDCs continually migrating out of the lungs and into the lung dLNs (55, 78). Interestingly, the number of IL-15 producing cells within the spleen is significantly reduced in chronic EtOH mice (79). Furthermore, recent studies from our laboratory have determined that the migration of CD11c^pos DCs out of the lung and into the lung dLNs is reduced in EtOH consuming mice following IAV infection (80). Therefore, the reduction in the numbers of existing IAV-specific CD8 T cell memory in X31-immune→EtOH we observed could be dictated by either reduced expression of homeostatic cytokines, such as IL-7 and/or IL-15, or reduced access to antigen.

Our data further reveal that the negative impacts of chronic EtOH consumption not only affect resting CD8 T cell memory subsets, but also alter the recall response following subsequent IAV challenge. While the numbers of IAV-specific CD8 T cells in the lung interstitium of IAV-immune consuming mice were reduced at day three post challenge compared to IAV-immune water consuming mice, the fold-increase in the total lung (interstitium + vasculature; Figure 6) was similar between these two groups. Unlike what was observed for naïve/effector CD8 T cells during a primary infection (14), this and our analysis of Ki67 (Supplemental Figure 2) suggests that proliferation of IAV-specific memory CD8 T cells in IAV-immune EtOH consuming mice is unaffected following secondary IAV challenge. However, we did observe a reduction in the fold-increase of IAV-specific CD8 T cell subsets resident in the lung interstitium (Figure 6K – 6O) and a higher fold-increase of IAV-specific CD8 T cells within the lung vasculature (Figure 6F – 6J) of IAV-immune EtOH consuming mice compared to IAV-immune water controls. Altogether, these data suggest that trafficking of IAV-specific CD8 T cell memory from the blood into the lungs is reduced by chronic EtOH consumption.

Consistent with this concept of reduced migration, our results further demonstrate that the chemokine milieu within the lungs is significantly altered in IAV-immune mice following EtOH exposure and virus challenge. Although CCR5 and CXCR3 expression on IAV-specific memory CD8 T cells was not affected, we observed significant reductions in the levels of CXCL10 and CXCL11 within the lungs of IAV-immune EtOH consuming mice on day three following secondary IAV challenge (Figure 8). Interestingly, in addition to airway epithelial cells and alveolar macrophages, CD8 T cells have been observed to secrete CXCL10 upon TCR engagement and contribute to the CXCL10 pool within the lungs during IAV infection (81, 82). Furthermore, it is well known that CD8 T_RM possess a “sensing and alarm function” that rapidly induces local chemokine expression, including CXCL10, by cells of the innate immune system (34, 83). Therefore, the reduced numbers of total and IAV-specific CD8 T cells within the lungs of X31-immune→EtOH mice (Figure 3-6) likely contributes to the reduced levels of CXCL10 that we observed following secondary IAV exposure (Figure 8). Although the specific role of CXCL10, and its receptor CXCR3, in the trafficking of CD8 T cell memory within the lungs following a secondary infection is not well defined, CXCR3 expression is important for the effective localization of CD8 T cell memory to the lung airways (63, 65). Interestingly, a recent study demonstrated CXCR3-deficient CD8 T cells have a reduced capacity to enter into areas of infection and have reduced interactions with virally infected cells within the skin (61). Thus, the reduced expression of CXCL10 and CXCL11 within the lungs of IAV-immune EtOH consuming mice may deter the egress of IAV-specific CD8 T cells out of the blood and into the infected lung tissue, ultimately contributing to reduced virus control and increased disease severity.

Altogether the data presented herein describe for the first time the negative impacts that EtOH consumption has on existing IAV-specific CD8 T cell memory responses. The consumption of EtOH led to an overall numerical reduction within the four known CD8 T cell memory subsets (i.e. T_CM, T_EM, T_PM, and T_RM) and was associated with increased disease severity upon secondary IAV challenge. Furthermore, the recall response following secondary IAV challenge was altered in X31-immune→EtOH mice, which had a reduced fold-increase in IAV-specific CD8 T cells within the lung tissue that was associated with a significantly altered chemokine milieu within the lungs. These results support continued investigation into the effects of EtOH consumption on existing IAV-specific CD8 T cell memory pool and the impact it has on the protection against severe IAV-associated disease.

Key Points:

1. Chronic EtOH consumption alters the existing memory T cell compartment.

2. EtOH induced changes in existing memory T cells increase susceptibility to IAV.

3. Chronic EtOH alters memory T cell mediated killing and recruitment into the lungs.

Abbreviations used in this paper:

i.n.

intranasal

EtOH

ethanol

dLNs

draining lymph nodes

T_CM

central memory

T_EM

effector memory

T_RM

T resident memory

T_PM

T peripheral memory

IAV

influenza A virus

AgExp

antigen experienced