Tacrolimus exposure windows responsible for Listeria monocytogenes infection susceptibility

Abstract

Tacrolimus is widely used to prevent graft rejection after allogeneic transplantation by suppressing T cells in a non-antigen-specific fashion. Global T cell suppression makes transplant recipients more susceptible to infection, especially infection by opportunistic intracellular pathogens. Infection followed by secondary challenge with the opportunistic intracellular bacterial pathogen, Listeria monocytogenes, was used to probe when tacrolimus most significantly impacts antimicrobial host defense. Tacrolimus treated mice showed no difference in innate susceptibility following primary infection, whereas susceptibility to secondary challenge was significantly increased. Modifying the timing of tacrolimus initiation with respect to primary infection compared with secondary challenge showed significantly reduced susceptibility in tacrolimus treated mice where tacrolimus was discontinued prior to secondary challenge. Thus, tacrolimus overrides protection against secondary infection primed by primary infection (and presumably live attenuated vaccines), with the most critical window for tacrolimus-induced infection susceptibility being exposure immediately prior to secondary challenge. These results have important implications for strategies designed to boost antimicrobial T cell mediated immunity in transplant recipients.

Graphical Abstract

Social Media:

Tacrolimus overrides anti-microbial T cell immunity primed by prior infection. We show the most critical window for tacrolimus-induced susceptibility is immediately prior to secondary challenge.

INTRODUCTION.

Transplantation is increasingly used in individuals with irreversible organ damage. An important consideration is preventing rejection of genetically foreign donor allograft tissues which requires life-long immune suppressive therapies. Allograft rejection is primarily mediated by T cells, and post-transplant therapies primarily target these immune cell subsets¹. The widely used calcineurin inhibitor, tacrolimus, is currently prescribed to >90% of solid organ transplant recipients^2,3. Tacrolimus primarily inhibits T cell activation, but does not discriminate between graft-specific cells that mediate donor tissue rejection from cells with other specificities that sustain host defense against microbial infection. Accordingly, transplant recipients are at increased infection risk, especially infection by intracellular pathogens such as Listeria monocytogenes (Lm) normally controlled by T cell mediated immunity^4–8.

Vaccination is an effective approach for boosting antimicrobial immunity against specific pathogens in defined developmental contexts. However, transplant recipients, especially in the first months following transplant, consistently show diminished responsiveness to both inactivated and live attenuated vaccines^9–11. Another consideration is the risk of infection from live attenuated vaccines, compared with inactivated vaccines, in transplant recipients on immune suppressive therapy. Administration of vaccines prior to transplantation and initiating immune suppressive therapy, is a potential work-around^12–14. However, interpretation of these results in clinical studies is limited by both interference of vaccine responsiveness by end-stage organ disease or use of only surrogate markers of protective immunity. Similar limitations apply to immunity primed by natural infection prior to transplantation. For example, solid organ transplant recipients remain at risk for disseminated and severe varicella zoster infection despite serological evidence of prior infection^15,16. To overcome these limitations, susceptibility was evaluated by adjusting when tacrolimus is administered to mice relative to Lm primary infection and secondary challenge.

RESULTS AND DISCUSSION.

The optimal dose of tacrolimus was first evaluated in dose-escalation studies. Elevated tacrolimus levels were found in the blood of mice after daily intraperitoneal dosing at 2mg/kg and 10mg/kg for three consecutive days (Figure 1a). The range (1–5 ng/mL) in mice treated with 10mg/kg was most consistent with the target therapeutic level in humans, and at the higher dosing range of prior studies in mice^17,18. Accordingly, the 10mg/kg daily dose was used in all further studies.

Figure 1. Tacrolimus selectively impairs immunity against secondary Lm challenge.

(a) Tacrolimus levels in whole blood from mice treated with each dosage for three consecutive days, and 24 hours after the last treatment. (b) Recoverable CFUs in the liver day 3 after primary infection or days 3 after secondary challenge (day 30+3) for tacrolimus treated (10mg/kg per day) compared with no treatment control mice. Data is from at least two independent experiments each with similar results, with each point representing data from an individual mouse. Bar, mean ± SEM.

To investigate how tacrolimus impacts antimicrobial host defense and T cell immunity, Lm infection susceptibility was compared between tacrolimus treated and no treatment control mice. We reasoned that Lm infection is ideally suited for evaluating functional shifts in antimicrobial T cells given the established role of this lymphocyte subset in protection against Lm secondary challenge after primed by primary infection, and the increased susceptibility of transplant recipients to infection by this opportunistic pathogen^4–8. Recombinant Lm engineered to stably express the MHC class I H-2K^b:OVA_257–264 peptide (Lm-OVA) was utilized to facilitate the identification of CD8+ T cells with surrogate H-2K^b:OVA_257–264 Lm specificity¹⁹. Using this approach, we first showed that tacrolimus does not significantly impact the susceptibility of mice to primary infection. Similar numbers of Lm were recovered from the liver of tacrolimus treated compared with no treatment control mice 3 days after primary Lm-OVA infection with an inoculum (2 × 10⁴ CFUs) that is 20% of the LD₅₀ (Figure 1b). These results are in agreement with the limited inhibition of innate immune components, including natural killer, monocyte and macrophage cells by tacrolimus, that protect against primary Lm infection^20,21.

This analysis was extended to investigate how tacrolimus impacts the susceptibility of mice to secondary Lm challenge where protection is known to be mediated by CD8+ T cells primed by primary infection^4,22. Thirty days after primary Lm-OVA infection, mice maintained on daily tacrolimus treatment were re-infected with the same Lm-OVA dosage. As expected, sharply reduced Lm CFUs were found in the liver of no treatment control mice after secondary challenge compared with primary infection (Figure 1b). Tacrolimus suppressed protection against Lm secondary challenge in this context since 80-fold increased Lm was recovered from tacrolimus treated compared with no treatment control mice (Figure 1b). Thus, tacrolimus selectively impairs host defense against secondary, but not primary, Lm infection. An important distinction between these results compared with prior studies which reported failure of tacrolimus to suppress the induction of T lymphocyte populations capable of mediating adoptive transfer of immunity and expression of pre-existing protection against Lm infection likely reflects our increased tacrolimus dosing required to achieving a therapeutic range²³.

Increased susceptibility of tacrolimus treated mice to secondary challenge paralleled significantly reduced in vivo cytotoxic CD8+ T lymphocyte activity. Lysis of congenially discordant OVA_257–264-peptide loaded donor cells was significantly reduced after adoptive transfer into tacrolimus treated compared with control mice 30 days after primary Lm-OVA infection (Fig 2A). Interestingly, despite this reduction in cytolysis of OVA_257–264-specific target cells, the percent and absolute number of OVA_257–264 specific CD8+ T cells identified by H-2K^b:OVA_257–264 tetramer staining amongst splenocytes were similar between tacrolimus treated compared with control mice (Figure 2b). This discordance likely reflects qualitative differences in the activation/protective function of Lm-OVA primed CD8+ T cells in tacrolimus treated mice. For example, PD-1 expression was sharply reduced amongst OVA-specific CD8+ T cells in tacrolimus treated compared with control mice in agreement with tacrolimus-induced inhibition of the transcriptional regulator, NFAT, required for PD-1 expression^24,25 (Figure 2c). Expression levels of others canonical T cell-intrinsic activation (CD27, CD25, CD69), co-inhibitory (KLRG1) and exhaustion (LAG-3) markers were similar between OVA-specific CD8+ T cells between tacrolimus treatment compared with control mice (Figure 2c).

Figure 2.

Memory CD8+ T cells primed by Lm-OVA infection in tacrolimus treated mice. (a) Percent lysis of OVA_257–264 pulsed compared with unpulsed control donor CD45.1 cells after intravenous transfer into tacrolimus treated compared with no treatment control mice each 30 days after primary Lm-OVA infection. (b) Representative FACS plots and composite data showing the percent OVA-specific CD8+ T cells identified by H-2K^b:OVA_257–264 tetramer staining and total number amongst splenocytes 30 days after primary Lm-OVA infection for tacrolimus treated compared with no treatment control mice. (c) Relative expression of each marker by H-2K^b:OVA_257–264 tetramer staining CD8+ T cells (blue or red) compared with tetramer negative CD8+ splenocytes 30 days after primary Lm-OVA infection for tacrolimus treated compared with no treatment control mice. (d) Relative intensity of H-2K^b:OVA_257–264 tetramer staining by OVA-specific CD8+ T cells and H-2K^b:OVA_257–264 avidity after normalization for levels of T cell receptor (CD3) expression amongst splenocytes 30 days after primary Lm-OVA infection for tacrolimus treated compared with no treatment control mice. (e) Cytokine production after OVA257–264 peptide stimulation compared with no stimulation controls by CD8+ splenocytes 30 days after primary Lm-OVA infection for tacrolimus treated compared with no treatment control mice. Data is from at least two independent experiments each with similar results, with each point representing data from an individual mouse. Bar, mean ± SEM.

Interestingly, OVA-specific CD8+ T cells in tacrolimus treated mice paradoxically expressed increased levels of CD44 (a mark of antigen-experienced cells) and CXCR3 (the chemokine receptor associated with effector CD8+ T cells tissue trafficking) (Figure 2c). Expression of these markers by microbe-specific CD8+ T cells in tacrolimus treated mice may indicate a terminally differentiated phenotype associated with pathological tissue inflammation previously described after lymphocytic choriomeningitis virus infection¹⁷. Further evaluation showed expanded CD8+ T cells with surrogate Lm-OVA specificity in tacrolimus treated compared with control mice had similar affinity, based the intensity of H-2K^b:OVA_257–264 tetramer staining, and avidity after normalization for T cell receptor expression based on anti-CD3 staining intensity (Figure 2d); and similar production of effector cytokines IFN-γ and TNF-α in response to cognate peptide ex vivo stimulation (Figure 2e). Thus, reduced CD8+ T lymphocyte cytolytic activity and unique qualitative differences in the expanded memory pool of CD8+ T cells with surrogate Lm-specificity are associated with increased susceptibility of tacrolimus treated mice to secondary Lm infection. Important next steps will be to investigate the generalizability of these findings to antigen-specific T cells primed by other pathogens during tacrolimus treatment given the more profound impacts of tacrolimus treatment in suppressing the expansion and cytokine production by virus-specific T cells after primary lymphocytic choriomeningitis virus infection¹⁷.

Having validated this model showing that tacrolimus functionally suppresses protective immunity against secondary Lm infection, we further investigated the most critical timing of tacrolimus initiation with respect to primary infection compared with secondary challenge. In particular, we asked whether susceptibility to Lm secondary challenge in tacrolimus treated mice reflects impaired priming and maintenance of protective memory T cells, or alternatively their impaired re-activation in response to secondary challenge. These parameters were experimentally reproduced in separate groups of mice by discontinuing tacrolimus treatment 3 days prior to secondary Lm challenge, or initiating tacrolimus treatment 3 days prior to secondary of Lm challenge, respectively (Figure 3a). Each group of tacrolimus treated mice showed increased susceptibility compared with no treatment control mice (Figure 3b). Interestingly however, mice with tacrolimus treatment restricted to only prior to and the first 27 days after Lm primary infection (discontinued treatment 3 days prior to secondary challenge) showed significantly reduced Lm susceptibility (>6-fold less recoverable Lm in the liver; P = 0.03) compared with mice maintained continuously on tacrolimus treatment throughout primary infection and secondary challenge (Figure 3b). Comparatively, Lm susceptibility was not significantly different between mice initiated on tacrolimus treatment only beginning 3 days prior to Lm-OVA secondary challenge compared with mice maintained continuously on tacrolimus treatment (Figure 3b). Thus, tacrolimus overrides protection against secondary infection primed by primary infection (and presumably live attenuated vaccines), and the most critical window for tacrolimus-induced infection susceptibility in this context is administration immediately prior to secondary challenge.

Figure 3.

Tacrolimus administration immediately prior to secondary Lm challenge is sufficient for impaired immunity. (a) Schematic outlining when each group of mice are initiated on daily tacrolimus relative to primary infection and secondary challenge. (b) Recoverable CFUs in the liver day 3 after secondary challenge for each group of mice described in panel A compared with no treatment control mice. Data is from at least two independent experiments each with similar results, with each point representing data from an individual mouse. Bar, mean ± SEM.

The translational implication is that tacrolimus negatively impacts both the recall protective function and the priming-maintenance of T cells stimulated by infection or live attenuated vaccines, with the most significant impacts for individuals maintained on tacrolimus therapy at the time of infection. A rodent model of tacrolimus treatment and infection allowed these parameters to be evaluated in isolation. An inherent limitation with any preclinical model is direct clinical applicability. Nonetheless, with regards to transplant recipients maintained on tacrolimus therapy, the more modest tacrolimus blood levels (1–5 ng/ml) we achieved in mice, despite relatively high dosages, compared to clinical targets suggest that sustained tacrolimus treatment, even at low levels, would still functionally suppress protective T cell-mediated immunity against secondary infection. In the larger translational context, these results highlight the clear need for immunological therapies that discriminate between harmful effector T cells with allograft specificity and protective antimicrobial T cells; or that consistently promote sustained long-term allograft tolerance without the need for ongoing administration of immune suppressive agents. Until these goals are achieved, reduced dosing of immune suppressive agents will likely be the primary means of broadly augmenting antimicrobial immunity in transplant recipients. With regards to the current most commonly used anti-rejection therapy, tacrolimus, our results suggest that discontinuation would selectively only augment host defense against infection by pathogens where there is pre-existing T cell mediated immunity. Importantly however, unleashing the protective benefits of vaccine-primed immunity in transplant recipients will require innovative approaches that allow discontinuation, even transiently, of tacrolimus while preserving graft tolerance^26,27.

METHODS

Mice.

C57BL/6 (H-2^b; CD45.2) and B6-Ly5.1 (H-2^b; CD45.1) mice were purchased from the NCI colony at Charles River Laboratories, housed and bred at Cincinnati Children’s Hospital, and used at 6–8 weeks of age. All experiments were performed in accordance with the Institutional Animal Care and Use Committee approved protocols.

Tacrolimus.

For administration to mice, tacrolimus (LC Laboratories; Woburn; Woburn, MA) was dissolved in DMSO, and resuspended into sterile saline for intraperitoneal injection into mice at either 2mg/kg or 10mg/kg. Tacrolimus levels in whole blood were evaluated with ARCHITECT chemiluminescent microparticle immunoassay (Abbot Diagnostics; Abbot Park, IL).

Listeria monocytogenes.

For infection, Lm-OVA was grown in brain heart infusion media at 37°C, back diluted to early log-phase (OD₆₀₀ 0.1), washed and re-suspended in sterile saline, and intravenously injected (2 × 10⁴ CFUs per mouse). Three days after infection, recoverable Lm CFUs were enumerated by spreading serial dilutions of the liver homogenate onto agar plates.

In vivo cytotoxicity assay.

Splenocytes from CD45.1+ congenic mice were mock stained or stained with CFSE (50 nM) for 10 minutes at room temperature, pulsed with OVA_257–264 peptide (CFSE^low) or no peptide control (CFSE^high), mixed at a 1:1 ratio, transferred i.v. into recipient mice, and harvested 20 hours thereafter. Percent killing was calculated as = 100 − [(percent OVA_257–264 peptide pulsed cells recovered/percent unpulsed cells recovered)/(percent OVA_257–264 peptide pulsed cells transferred/percent unpulsed cells transferred)] × 100 as described²⁸.

Cell staining and stimulation.

For tracking CD8+ T cells with H-2K^b:OVA_257–264 specificity, single cell suspensions were prepared from the spleen and peripheral (axillary, brachial, cervical, inguinal, mesenteric, pancreatic, para-aortic/uterine) lymph nodes. Cells were stained with PE-conjugated H-2K^b:OVA_257–264 MHC class I tetramer and enriched with anti-PE microbeads (Miltenyi Biotec). Bound fractions were eluted and stained with fluorochrome-labeled cell surface antibodies For cytokine production, splenocytes were stimulated ex vivo with OVA_257–264 peptide (1 μM) for 5 hours in media supplemented with Brefeldin A (GolgiPlug, BD Biosciences).

Data analysis.

Differences between two groups were analyzed using the Student’s t-test (normal distribution of data) or Mann Whitney test (data not normally distrusted), and for three of more groups the one-way ANOVA test (normal distribution of data) or the Kruskal-Wallis test (data not normally distrusted) (Prism, GraphPad). For each analysis, a P-value of ≤ 0.05 was taken as statistical significance.

Abbreviations:

Lm

Listeria monocytogenes