Immunosuppressant imprecision: multidirectional effects on metabolism and microbiome

SUMMARY

Transplant recipients require lifelong, multimodal immunosuppression to prevent rejection by dampening alloreactive immunity. These treatments have long been known to lack antigen specificity. Despite empirically selected long-term immunosuppression regimens, most allografts succumb to alloimmune responses that result in chronic inflammation and scarring. Additionally, immunosuppressive medications themselves contribute to unintended intestinal dysbiosis and metabolic disorders. This review focuses on the effect of immunosuppressant treatments on alloimmunity, gut microbiome, and metabolism, with a particular emphasis on the effects on metabolic disorders. We also outline the shared and unique microbial and metabolic signatures produced by each immunosuppressant class, underlining their distinct impacts on immunity and metabolic homeostasis. These observations underscore the need for a holistic understanding of these drugs’ on- and off-target effects to refine therapeutic strategies, enhance immunosuppression efficacy, and ultimately enhance graft and patient survival. By characterizing these complex interactions, strategies informed by the gut microbiome and host metabolism may offer a promising adjunctive approach to optimizing immunosuppressive regimens and promoting sustained graft acceptance.

INTRODUCTION

Organ transplantation has revolutionized the treatment of end-stage organ failure, offering hope to patients who would otherwise face limited survival prospects. Due to improvements in donor organ preservation and recipient selection, immunosuppressive regimens and monitoring, and infection prophylaxis, 1-year allograft survival rates have improved over the years, reaching over 90% for most organs (1). Despite these advances, long-term allograft survival rates beyond the first year have not experienced similar improvements. This discrepancy largely stems from the limitations of current immunosuppressive therapies. While lifelong immunosuppression is crucial for preventing organ rejection in transplant recipients, these treatments have long been known to lack antigen specificity. The lack of precision of immunosuppressants can lead to a range of adverse events. These complications arise from three main sources: antigen nonspecific effects that broadly suppress the immune system, off-target consequences affecting nonimmune cells, and on-target effects that impact nonimmune reactions of the targeted pathways (2–7).

The dilemma between the necessity of immunosuppression for graft survival and the imprecision of current therapies represents a critical challenge in improving long-term outcomes for transplant recipients. The imprecision of these immunosuppressants manifests in several ways. Their broad effects on the immune system increase the risk of infections (2) and certain cancers (3). Their off-target effects lead to organ-specific toxicities, including nephrotoxicity (4), gastrointestinal toxicity (5), and neurotoxicity (6). Perhaps most importantly for long-term patient health, these drugs can cause significant metabolic complications, such as diabetes and dyslipidemias (7).

Most mechanistic research on immunosuppressants has focused on their effects on T cells, B cells, and dendritic cells. However, their direct or indirect effects on other innate immune populations, as well as the structure and function of secondary lymphoid organs (SLOs) and lymph nodes (LNs), have received limited investigation. Recent research has begun to uncover another layer of complexity in immunosuppressant imprecision for its unintended effects on the gut microbiome and host metabolism. Emerging evidence suggests that immunosuppressants can significantly alter the composition and function of the gut microbiome, which in turn can influence drug metabolism, efficacy, and toxicity.

These multidirectional interactions between immunosuppressants, the microbiome, and host metabolism form a complex network that is only beginning to be understood. In this review, we explore the concept of “immunosuppressant imprecision,” which focuses on the metabolic and microbial effects of current immunosuppressive therapies. While many of the studies cited here are primarily animal models, understanding these intricate relationships is fundamental to paving the way for more precise, personalized immunosuppression strategies that can improve long-term outcomes for transplant recipients.

KNOWN IMMUNOSUPPRESSANT MECHANISMS

Current transplantation immunosuppressant therapies fall into four major categories: calcineurin inhibitors (CNIs), glucocorticoids, mechanistic target of rapamycin (mTOR) inhibitors, and antimetabolites (8). Each class has its specific mechanism of action and associated side effects. Calcineurin inhibitors, such as cyclosporine (CsA) and tacrolimus, suppress T cell activation by inhibiting the enzyme calcineurin, crucial for T cell activation (9, 10), but their use is associated with potential nephrotoxicity and other side effects (4). Glucocorticoids, including prednisone, offer broad anti-inflammatory and immunosuppressive effects by suppressing inflammatory gene expression and inducing T cell apoptosis (11), although long-term use can lead to significant complications like diabetes and osteoporosis (7, 12). mTOR inhibitors, like rapamycin (also called sirolimus) and everolimus, target the mTOR pathway to prevent T cell activation and B cell expansion and are valued for their antiproliferative benefits and reduced incidence of certain cancers (13), but can lead to dyslipidemia and renal injury (4, 7). Lastly, antimetabolites, such as mycophenolate mofetil and azathioprine, inhibit nucleic acid metabolism, which in turn inhibit cell proliferation (14, 15), but can induce gastrointestinal toxicity (16). Together, these immunosuppressants form the cornerstone of posttransplant care, aiming to prolong graft survival while managing the balance between efficacy and toxicity.

Calcineurin inhibitors

Calcineurin inhibitors (CNIs), including tacrolimus, inhibit calcineurin, which results in the inhibition of cytoplasmic nuclear factor of activated T cells (NFAT), thereby blocking T cell activation, Treg proliferation, and function (9, 10), and naïve B cell proliferation and plasmablast differentiation (8). Tacrolimus inhibits calcineurin through binding FK506-binding protein-12. Tacrolimus inhibits IgA class switching and secretion in naïve B cells (8). In vitro, CNIs suppress IL-21-mediated differentiation of T follicular helper cells (8). In the innate compartment, tacrolimus inhibits mast cell histamine release (17), impairs antigen-presenting cell (APC)-mediated phagocytosis (18), and inhibits dendritic cell (DC) MHC class I and II antigen presentation (19). Tacrolimus inhibits the synthesis of eosinophil cytokines (20). One study showed that tacrolimus is capable of suppressing M2 macrophage polarization through targeting JAK2 and inhibiting JAK2/STAT3 signaling (21), while another showed that tacrolimus shifted CD14⁺ monocyte polarization to an M2-like phenotype (22). Tacrolimus inhibits calcium-dependent events, such as IL-2 gene transcription, nitric oxide synthase activation, cell degranulation, and apoptosis (10). Furthermore, tacrolimus enhances the effect of glucocorticoids and progesterone through binding FK506-binding proteins in the hormone receptor complex and preventing degradation (10). Additionally, tacrolimus treatment reduces the ability of Kupffer cells to phagocytose bacteria, thereby increasing the likelihood of infections (18). Unlike tacrolimus, cyclosporine inhibits calcineurin through binding cyclophilin. Cyclosporine suppresses both IgA and IgE class switching and secretion (8) and has been suggested to downregulate the expression of TRAIL and FasL on natural killer (NK) cells, which helps to inhibit transplant rejection (23).

Glucocorticoids

Glucocorticoids, such as prednisone, bind to steroid receptors, which are then translocated to the nucleus where they regulate the transcriptional activity of numerous genes. In lymphocytes, glucocorticoids inhibit activation of cytidine deaminase in B cells, leading to dysregulation of immunoglobulin gene somatic hypermutation and class-switch recombination (24). Additionally, they induce apoptosis, skew Th1 to Th2 responses in CD4 T cells, and suppress activation by increasing the T cell receptor signaling threshold (11). In contrast, glucocorticoids modulate the suppressive function of Foxp3⁺ T cell (Treg) in vivo (25). Glucocorticoids also increase PD-1 expression on NK cells, which has an inhibitory effect on IFN-γ production (26). They exert a wide range of effects on the innate immune system, including inhibiting cyclooxygenase-2 (COX-2) synthesis via NF-kB transcription suppression, downregulating pro-inflammatory cytokine expression (IL-1, IL-6, and TNFα) in macrophages (MΦ), and reducing the number of circulating DC and their ability to produce IFNα (27, 28). Glucocorticoids induce apoptosis in eosinophils and inhibit apoptosis in neutrophils (29). A subset of glucocorticoid-induced M2 MΦ is important for the resolution of inflammation due to its ability to phagocytose early apoptotic neutrophils and release anti-inflammatory cytokines (30). Glucocorticoids also dampen immune cell adherence to vascular endothelium, impairing trafficking and homing of pro-inflammatory cells via downregulation of adhesion molecules ICAM-1, VCAM-1, and E-selectin as well as upregulation of endothelial junction proteins occludin, claudin-5, and VE-cadherin (31).

Antimetabolites

Antimetabolites inhibit nucleotide metabolism and thus reduce cell proliferation. In particular, mycophenolate mofetil (MMF), an inosine-5′-monophosphate dehydrogenase inhibitor (IMPDHI), blocks the de novo purine nucleotide synthesis pathway by preferentially inhibiting guanosine synthesis and preventing T and B cell expansion as these cells lack the purine salvage pathway (14). Consequently, MMF suppresses the production of immunoglobulins (32). MMF inhibits the proliferation and cytotoxicity of NK cells (33). MMF also inhibits the glycosylation and expression of adhesion molecules, thereby blocking the recruitment of lymphocytes and monocytes to sites of inflammation (34). Upon administration, MMF is converted into its prodrug mycophenolic acid (MPA). MPA suppresses the maturation of dendritic cells, which in turn reduces antigen presentation to T lymphocytes, thereby preventing allogeneic T cell responses (32). The presence of MPA increases the expression of M2 surface markers on M1 macrophages (22). MMF also prevents infiltration and accumulation of macrophages via IL-17 (35, 36).

mTOR inhibitors

mTOR inhibitors, such as rapamycin, bind to and inhibit the mechanistic target of rapamycin (mTOR). mTOR is a kinase that phosphorylates threonine and serine residues in its substrates, such as S6K and 4E-BP, which are at the center of many crucial metabolic processes including cell growth and proliferation (13). Through the disruption of cytokines that signal lymphocyte growth and differentiation, rapamycin induces mid to late G1 arrest (37). Additionally, it induces T cell anergy in the face of TCR engagement and costimulation via inhibition of downstream signals from IL-2 receptor engagement (38). Furthermore, the drug promotes the expansion of Tregs (39). Rapamycin also inhibits B cell function through inhibiting immunoglobulin class switching and secretion (8). In innate cells, rapamycin enhances DC anti-inflammatory phenotype via increased mitochondrial metabolism and IL-10 production (40). Rapamycin also reduces metabolic activity and proliferation of NK cells (41). Rapamycin has also been shown to reduce IL-5-mediated survival of eosinophils in culture (20).

Other immunosuppressant therapies

Multiple monoclonal antibodies, such as alemtuzumab and antilymphocyte globulins (e.g., thymoglobulin), are also used for immediate and targeted immunomodulation. Alemtuzumab binds to CD52, resulting in the depletion of T and B lymphocytes (42). Monoclonal and polyclonal antibodies, though, are generally used for short time periods and thus will not be the focus of this review.

Sphingosine-1-phosphate (S1P) receptor agonists have immunosuppressive properties in the treatment of auto- and alloimmunity (43). FTY720 is a noncompetitive agonist/antagonist of S1P receptors that alters leukocyte migration. In vivo, FTY720 is phosphorylated into FTY720-P, which is proven to recruit β-arrestin to S1P1Rs resulting in prolonged S1P1R internalization and degradation (44). Lack of S1P1R on the surface of the cells prevents T and B cell egress from SLOs and thymus, reducing circulating lymphocytes and thereby preventing their migration towards S1P (45, 46). Recent studies on the use of FTY720 for multiple sclerosis have shown that the FTY720 treatment increases T cell factor 1 expression, which was integral in skewing T cell activation towards a less inflammatory phenotype, marked by the suppression of IFN-γ and granzyme B (47). FTY720 has also been shown to improve rheumatoid arthritis symptoms through the inhibition of DC migration to LNs (48).

JAK inhibitors specifically target members of the Janus tyrosine kinase (JAK) enzymes that are crucial to the signal transduction of several pro-inflammatory cytokines. The JAK family is comprised of JAK1, JAK2, JAK3, and TYK2, which are integral to the transduction of cytokine signaling (49). Many JAK inhibitors exist with differing selectivity against the JAK enzymes. Most are used in the treatment of rheumatoid arthritis and other autoimmune deficiencies, but one JAK inhibitor, ruxolitinib, is approved for use in treating graft versus host disease (GVHD). NK cell and T cell levels are reduced by JAK inhibition, while B cell counts increased in phase 1 kidney transplant clinical trials (50). JAK inhibitors in kidney transplant have also been shown to reduce MΦ infiltration and reduce CsA toxicity (50).

Costimulation blockers, such as belatacept, lack nephrotoxicity but have an increased risk of acute rejection and posttransplant lymphoproliferative disease (51). Belatacept binds to CD80 and CD86 on antigen-presenting cells, therefore inhibiting CD28-mediated costimulation of T cells. Additionally, the drug impairs crosstalk between B cells and T follicular helper cells resulting in reduced development of B cell germinal centers (52, 53). In vitro, belatacept also reduces plasmablast differentiation, Ig production, and Blimp-1 levels that are important to plasma cell function (52). Furthermore, treatment with belatacept also results in the downregulation of M1 MΦ markers in conjunction with the upregulation of M2 MΦ markers (54).

INTERACTIONS BETWEEN IMMUNOSUPPRESSANTS AND GUT MICROBIOME

Beyond their primary action on specific immune cell populations, immunosuppressants also significantly impact the gut microbiome, the largest immune organ in the human body, critically impacting alloimmunity and consequently promoting or inhibiting graft survival (55–59). These indirect effects on the microbiome and subsequent immunological consequences are crucial yet often under-investigated, representing a significant knowledge gap in organ transplant science.

Immunosuppressants are known to induce changes in the gut microbiota (Table 1). Most known interactions are from animal studies where individual drug effects can be elucidated, whereas human studies are limited to primarily determining the effect of clinical combinations of immunosuppressants. Methotrexate, an inhibitor of mammalian dihydrofolate reductase, also affects this conserved pathway in gut bacteria, leading to broad alterations in the gut microbiome in mice (60). Transferring human fecal microbiomes from pre- and post-methotrexate-treated patients into germ-free mice yields significant changes in B cells, Tregs, and CD4 T cells. Tacrolimus alters the composition of the gut microbiome in mice, and FMT from tacrolimus-treated animals extends immunosuppressive properties (42) and enhances skin allograft survival (61). A systematic review of articles related to immunosuppressant-induced changes in gut microbiota in either humans or animals shows 70% of articles mentioning changes in anaerobic bacteria levels including Ruminococcaceae, Lachnospiraceae, Firmicutes, Bacteroides, and Clostridiales (62). In mice, treatment with prednisolone decreases Bacteroidetes and increases Firmicutes in fecal samples and increases ileal Clostridium sensu stricto (63). In mice, combinations of immunosuppressants increase commensal Escherichia coli (62, 63). Administration of sirolimus or tacrolimus to rats results in reduced presence of Roseburia, Oscillospira, Mollicutes, Rothia, Micrococcaceae, Actinomycetales, and Staphylococcus, along with decreased bacterial diversity in the sirolimus group (64). MMF treatment ameliorates gut dysbiosis in spontaneously hypertensive rats marked by increased presence of Firmicutes and Bacteroidetes along with acetate- and lactate-producing bacteria (65). Rapamycin ameliorated dextran sodium sulfate (DSS)-induced colitis through changes to the gut microbiota in mice (66). In brief, immunosuppressants influence the gut microbiome, affecting Tregs and other immune cell populations, with evidence showing that changes in microbial composition and diversity can impact the efficacy of immunosuppressive therapies and alter gut immune cell dynamics.

TABLE 1.

Immunosuppressant effect on gut microbiota

Immunosuppressant	Host species	Effect on gut microbiota	Reference
Methotrexate	Mice	Broad alterations to the gut microbiome	(60)
Tacrolimus	Mice Rat Human	Alters gut microbiome composition ↑ Allobaculum, Bacteroides, Lactobacillus, and A. muciniphila ↓ presence of Roseburia, Oscillospira, Mollicutes, Rothia, Micrococcaceae, Actinomycetales, Staphylococcus, microbial diversity, and acetate- and butyrate-producing bacteria ↑Firmicutes/Bacteroidetes ratio ↑ Faecalibacterium prausnitzii	(42, 61, 64, 67, 68)
Cyclosporine	Rat	↑ F. prausnitzii ↓ Enterobacteriaceae, Clostridium clusters I and XIV	(69)
Glucocorticoids	Mice Human	↓ Bacteroidetes ↑ Firmicutes and ileal Clostridium sensu stricto ↑ Erysipelotrichales ↓ Clostridiales	(63, 70)
Dexamethasone	Rat Mice	↑ Proteobacteria ↓ Clostridiales and Lactobacillus ↑ Clostridiales ↓ Lactobacillaceae,	(71, 72)
Bromofuranone	Mice	↑ Anaerostipes ↓ Mucispirillum, Oscillospira, Bilophila, and Rikenella	(73)
Sirolimus	Rat	↓ presence of Roseburia, Oscillospira, Mollicutes, Rothia, Micrococcaceae, Actinomycetales, and Staphylococcus ↓ bacterial diversity	(64)
MMF	Rat Mice	ameliorated gut dysbiosis ↑ presence of Firmicutes and Bacteroidetes along with acetate- and lactate-producing bacteria ↓ overall gut diversity ↑ Proteobacteria	(65, 74)
Rapamycin	Mice	↑ Lactobacillus reuteri, Prevotellaceae, Paraprevotella, Christensenella, and Streptococcus ↓Romboutsia Bacteroides sartorii, and Peptostreptococcaceae, Marinilabiliaceae and Turicibacter	(66, 75)
Combinations of immunosuppressants	Mice Human	↑ commensal Escherichia coli ↑ Erysipelotrichales and Enterobacteriales, Enterococcus sp., Ruminococcaceae, Lachnospiraceae, Firmicutes and Bacteroides ↑ or ↓ Enterobacteriaceae depending on the study ↓ Eubacteria, Bifidobacterium sp., F. prausnitzii, Clostridium sensu stricto and Lactobacillus sp	(63, 70, 76–79)

Immunosuppressants can also affect the intestinal environment by directly eliciting changes in intestinal epithelial cells (IECs) and barrier functions (Table 2). Key functions of IECs include the physical segregation of commensal bacteria from host tissue, the integration of microbial signals, and mediators of intestinal homeostasis allowing the establishment of a permissive environment to the colonization of bacteria (80). Recent studies reveal links between increased bacterial translocation, due to a dysfunctional epithelial barrier, and a broad range of extraintestinal autoimmune and inflammatory conditions (81–83), systemic immune activation (84), progression of chronic infection (85), metabolic disorders (86, 87), and GVHD (88). In terms of immunosuppression interactions with IECs, MMF incubation with a monolayer of Caco-2 colorectal adenocarcinoma cells results in decreased tight junction protein claudin-1 and increased claudin-2, suggesting increased gut permeability (89). This aligns with observed MMF-related gastrointestinal toxicity in patients. Tacrolimus also results in increased gut permeability in rats through decreased occludin and Muc3 in the colon (42). Glucocorticoid treatment is linked to decreased expression of Muc2, a main component of colonic mucus (42). In terms of immunosuppression affecting innate immunity at the intestinal barrier, glucocorticoid and tacrolimus treatment in mice is associated with decreased intestinal bacterial proliferation through the reduction of C-type lectins RegIIIβ and RegIIIγ (42). Glucocorticoids also result in decreased biliary IgA in rats, leading to decreased bacterial IgA coating, which is important for neutralizing pathogenic bacteria (42). Tacrolimus also inhibits mucosal T cell and NK cell functions in mice (42). Recent single-cell transcriptomics unveils the remarkable heterogeneity within major IEC cell types and complex cellular interactions that underlie the autoimmune and inflammatory conditions (90–93). Understanding the drug effect on intestine epithelia, their interaction with gut microbiome, and the subsequent influence on immune activation is vital to holistically understanding the alloimmune responses.

TABLE 2.

Immunosuppressant effect on the intestinal barrier

Immunosuppressant	Effect on the intestinal barrier	Reference
MMF	↓ claudin-1 ↑ claudin-2 ↑ gut permeability	(89)
Tacrolimus	↑ gut permeability ↓ occludin and Muc3 ↓ bacterial proliferation ↓ RegIIIß and RegIIIγ Inhibits mucosal T cell and NK cell function	(42)
Glucocorticoids	↓ Muc2 ↓ bacterial proliferation ↓ RegIIIß and RegIIIγ ↓ biliary IgA ↓ bacterial IgA coating	(42)

Interactions between immunosuppressive treatments and the gut microbiome are bidirectional. The gut microbiota and its products contribute to complex metabolic interactions by impacting drug metabolism, affecting the efficacy, toxicity, and bioavailability (94) and contributing to high interindividual variability in drug metabolism and responses (95–98). Certain microbial enzymes have been found to be important in the metabolism of immunosuppressant drugs. β-glucuronidases have an important role in the metabolism of MMF: it is well established that hematopoietic cell transplant patients require higher doses of MMF than kidney transplant patients to reach therapeutic levels of its prodrug MPA; this difference is accompanied by two-fold lower β-glucuronidase activity in the stool of hematopoietic cell transplant patients compared to kidney transplant recipients (99). Cytochrome P450 enzymes have been implicated in tacrolimus and cyclosporine metabolism: CYP3A enzymes are responsible for the O-demethylation of tacrolimus in the small intestine, an important step in the metabolism of the drug (95). Esterases and amidases are also involved in prodrug activation: carboxylesterase 1 and 2 are involved in the hydrolysis of MMF to MPA (100). Reductases are involved in the conversion of immunosuppressants: 11β-hydroxysteroid dehydrogenase type 1 is a known cortisone reductase that has been shown to recycle glucocorticoids and convert their inactive forms back into active metabolites (101).

These microbiome-driven interactions with immunosuppressant drugs have far-reaching implications on the drug dosing schema for patients. No two gut microbiomes are identical, and these interindividual differences lead, in turn, to varying drug responses and the need for different dose requirements to reach the same therapeutic outcome in different patients. For example, Faecalibacterium prausnitzii directly metabolizes tacrolimus into less potent metabolites in vitro (67, 102). One study comparing kidney transplant patients who required dose escalation of tacrolimus to those who did not showed that Faecalibacterium prausnitzii abundance was higher in those requiring dose escalation, whereas it was only 0.8% in those whose dose remained stable (67). The gut microbiota can also reactivate the inactive form of mycophenolate mofetil and influence its pharmacokinetics (103, 104). These complex interactions lead to challenges in dosing and drug monitoring and strongly indicate the need for personalized immunosuppressive regimens. To that end, data-driven approaches that predict the interactions of drugs with the human microbiome are of growing interest. A recently developed machine learning framework has been developed that integrates genomic content and chemical properties to successfully predict drug-related impacts on the microbiome in animal models and clinical trials (105). While this framework has its drawbacks and simplifications, it provides proof of concept that such data-driven approaches can help aid in the development of personalized immunosuppressant regimens and help minimize adverse events associated with immunosuppressant drugs.

In our recent mouse study, we observed tacrolimus-induced shifts in the composition and structure of the gut microbiome. The most affected taxonomic groups are Clostriales and Verrucomicrobiae (i.e., Akkermansia muciniphila) with significantly reduced community diversity (106). In a continuation of our tacrolimus research, we expanded our investigation to compare gut microbiome alterations following treatment with various immunosuppressants, including mycophenolate mofetil (MMF), rapamycin, prednisolone, and tacrolimus. As shown in Fig. 1, by day 30 of treatment, the drugs demonstrated the most distinct changes in the induced gut microbiome composition and structure, compared to 3 or 7 days of administration, indicating a continuous and incremental drug effect. No taxonomic group in any of the drug treatment groups was persistently altered at all three time points. Rather, we observed distinct early, intermediate, and late drug effects on the gut microbiome, and this effect was specific to individual drugs. In particular, rapamycin resulted in an increased abundance in L. johnsonii on both days 3 and 7, while this disappeared by day 30. Interestingly, despite distinctive clustering of each drug’s gut microbiota after 30 days of treatment, all treatment groups similarly converged on key bacteria that were not present at earlier time points, including Duncaniella, Paramuribaculum intestinale, and CAG-873. All these taxa are from the family of Muribaculaceae, previously known as S24-7, a predominant commensal member of the mouse gut microbiome. Recently, pathobionts of Duncaniella and Paramuribaculum were recognized as both necessary and sufficient to induce diabetes (107). Overall, these results indicate that each immunosuppressant differentially affected the gut microbiome at different time points, a therapeutic fingerprint that may reflect alterations simultaneously in mucosal immunologic homeostasis, gut metabolism, and immunosuppressive biochemical mechanisms, an underappreciated consequence of these drugs. Furthermore, despite the different drug effects on gut microbiome, their elicited dysbiotic state after prolonged use converges. This convergence of the microbiome to a dysbiotic state, despite different drug treatments, could indicate a common mechanism or set of mechanisms related to immunosuppression that impacts the gut microbiome.

Fig 1.

Differentially abundant taxonomic groups in the gut microbiome after 3-, 7-, and 30-day treatment. The gut microbiome taxonomic composition and structure were established using whole community metagenomic sequencing of intraluminal stool collected from colon mapped to the comprehensive mouse gut metagenome catalog (CMGM) (108). The relative abundance of microbial taxa in the gut lumen following treatment with each immunosuppressant (mycophenolate mofetil, rapamycin, prednisolone, and tacrolimus) was compared to no-treatment control at the time points of 3, 7, and 30 days. Taxonomic groups significantly more abundant in the treatment group compared to the control (CT), with a LogFC >1 and a P-value < 0.05, were marked in red. Conversely, taxa more abundant in the control than in the treatment group, with a LogFC <−1 and a P-value <0.05, were colored in blue. The parametric Wald Chi-squared test was used to calculate P values. Adapted from reference 109 (preprint, under a Creative Commons license) based on data from references 109 and 110 (preprint).

INFLUENCE OF GUT MICROBIOTA ON ALLOIMMUNITY

The gut microbiome modulates both innate and adaptive alloimmune functions (111–117), influencing pro- or anti-inflammatory circuits that direct alloimmune responses (118–120) (Table 3). Antibiotic-treated mice have impaired immunity to lymphocytic choriomeningitis virus (LCM) as evidenced by reduced CD8⁺ T cell response and IgG (116). Additionally, a study comparing the colonization of germ-free mice with either mouse or human microbiota revealed that a host-specific microbiota is important in conferring protection against infection; human microbiome colonized mice show low levels of CD4⁺ and CD8⁺ T cells (117). In the innate compartment, the beneficial microbe, Akkermansia muciniphila, reduces macrophage infiltration in atherosclerotic lesions, the expression of pro-inflammatory cytokines and chemokines, and circulating endotoxin levels in Apoe -/- mice (111). IL-22-producing innate lymphoid cells confine Alcaligenes in mouse lymphoid tissues to prevent systemic inflammation, which is marked by neutrophil infiltration, increased spleen size, and elevated serum IL-6 and TNFα levels (115). A study of LCM in antibiotic-treated mice also showed decreased macrophage-associated antiviral response genes (116). Together, these studies illustrate the complex interactions between the microbiome and host immune response.

TABLE 3.

Presence or absence of microbiome or bacterial components and their effect on immunity in mice

Treatment	Presence (+) or absence (−) of bacteria or bacterial component	Effect on immunity	Reference
Antibiotics	(−)	Impaired immunity to lymphocytic choriomeningitis ↓ CD8 T cell response ↓IgG ↓macrophage-associated antiviral response genes	(116)
Colonization of mice with human microbiota	(+)	Conferred protection against infection ↓CD4+ and CD8+ T cells	(117)
Akkermansia muciniphila	(+)	↓macrophage infiltration in atherosclerotic lesions ↓pro-inflammatory cytokines and chemokines ↓endotoxin levels	(111)
Depletion of innate lymphoid cells – leading to dissemination of Alcaligenes to host lymphoid tissues	(+)	↑inflammation ↑spleen size ↑neutrophil infiltration ↑serum IL-6 and TNFα	(115)
Germ-free vs. antibiotic-treated vs untreated convention mice	(+)	↑APC priming of alloreactive T cells and accelerated graft rejection	(55)
Flagellin – extracted from bacterial flagella	(+)	↑ Treg levels ↓GVHD	(56)
Ampicillin	(−)	↑GVHD (reversible by Lactobacillus johnsonii reintroduction)	(57)
Members of the Clostridium genus	(+)	↑intestinal Treg ↑ TGFβ ↓DSS-colitis ↓IgE response	(121)
Polysaccharides from the cell wall of Bifidobacterium bifidum	(+)	↑intestinal Treg via DCs expression TLR2	(122)
Commensal bacteria-induced Treg accumulation	(+)	↓Th17 and Th1	(123)
Antibiotics	(−)	↓Cx3CR1 + MΦ in perivascular space of small intestine lamina propria	(124)
Antibiotics	(−)	↓CX3CR1 + mononuclear phagocyte production of TH-1 regulating IL-10	(125)
Antibiotics	(−)	↓ TGFβ-responsive CD11c+CD121b+ MΦ and IL-1-responsive CD11c−CD206hi MΦ	(126)
Bacterial DNA	(+)	Stimulate CX3CR1 + mononuclear phagocytes to expand bone marrow hematopoietic progenitors	(127)

In transplantation, the microbiome critically impacts alloimmunity and, consequently, promotes or inhibits graft survival (55–59) (Table 3). A skin graft model of germ-free versus antibiotic-treated or untreated conventional mice showed that gut microbiota enhances APC priming of alloreactive T cells and accelerates graft rejection (55). Administration of flagellin, a TLR5 agonist extracted from bacterial flagella, to mice receiving hematopoietic stem cell transplantations elevates regulatory T cell levels and reduces GVHD (56). In murine models, ampicillin treatment before bone marrow transplant results in worsened GVHD, whereas reintroduction of Lactobacillus johnsonii prevents this pathology (57). In human small bowel transplant patients, distinct microbiome profiles exist between nonrejection and acute rejection samples, suggesting that gut microbiome dysbiosis could be involved in the rejection process (58). A comparison of pre- and post-lung transplant microbial communities in humans receiving lung transplants showed that the reestablishment of pretransplant microbes is protective, while de novo colonization of microbes is detrimental to the development of bronchiolitis obliterans syndrome (59). This research exemplifies the far-reaching importance of understanding the interactions the microbiome can have on immunity.

The microbiome also plays a crucial role in regulating innate immunity during the peri-transplant period, significantly influencing transplant outcomes. In clinical allogeneic hematopoietic stem cell transplantation, reduced microbiota diversity at the time of transplantation strongly correlated with an increased risk of transplant-related mortality (128, 129). Similarly, a study analyzing pre-, peri-, and postoperative stool samples from solid organ transplant patients observed low microbial diversity and increased abundance of pathobiont species both before and after transplantation, with the severity of dysbiosis associated with higher mortality rates (130). Specific bacterial species also differentially impact transplant outcomes in humans. For example, the depletion of anti-inflammatory Clostridia, shortly after transplant, increased the risk of GVHD in pediatric stem cell transplant patients (131). Clostridia-derived short-chain fatty acids (SCFAs), such as butyrate, are potent immunomodulators that regulate innate and alloimmune responses, promoting immune tolerance and reducing inflammation (132). Similarly, the presence of peri-transplant Blautia was associated with a lowered risk of GVHD-related death (133). Conversely, certain Actinobacteria and Firmicutes taxa present at neutrophil recovery correlated with increased GVHD incidence, while the presence of Lachnospiraceae was protective (134). Though the precise mechanisms underlying these associations are not fully understood, microbial metabolites like SCFAs have been shown to modulate innate immune cell function and trafficking (132, 135, 136). Additionally, microbiota influence intestinal barrier integrity and mucosal immunity, further contributing to their role in transplant outcomes (57, 88). Collectively, these studies underscore the critical influence of the microbiome during the peri-transplant period and its potential as a therapeutic target to improve transplant success.

Microbial-dependent induction of regulatory Foxp3⁺ T cells (Tregs) is key for immune homeostasis, and Tregs are critical in maintaining immunological tolerance (121–123, 137, 138) (Table 3). In mice, extrathymically induced Tregs are crucial to the maintenance of Th2 inflammation at mucosal sites (137). Moreover, members of the Clostridium genus have been proven efficient at inducing intestinal Tregs in mice via the accumulation of TGFβ, and early inoculation with these bacteria results in reduced occurrences of dextran sodium sulfate-induced colitis and IgE response in mice, suggesting the importance of the genus in maintaining intestinal immune homeostasis (121). In mice, polysaccharides from the cell wall of Bifidobacterium bifidum are also integral in promoting the expansion of intestinal Tregs via DCs expressing toll-like receptor (TLR) 2 (122). Furthermore, commensal bacteria-induced colonic Treg accumulation is essential in preventing aberrant expansion of Th17 and Th1 cells in mice (123). Additionally, in a murine lung transplant model, researchers showed that the recipient microbiome is essential for the induction of the Tregs, which are necessary for maintaining inflammation and preventing acute and chronic allograft rejection (138). These studies indicate that microbial interactions are crucial for the induction of regulatory Foxp3⁺ Tregs and essential for maintaining immune homeostasis and tolerance.

The microbiome affects more than just CD4⁺ T cell homeostasis (Table 3). In mice, gut microbiota is also required for the localization of CX3CR1⁺ MΦ to the perivascular space of small intestine lamina propria (124), the production of Th1-regulating IL-10 by CX3CR1⁺ mononuclear phagocyte (125), and the development of two distinct colonic MΦ populations (TGFβ-responsive CD11c⁺CD121b⁺ and IL-1-responsive CD11c⁻CD206^hi) (126). Other byproducts of microbiota, such as bacterial DNA, stimulate CX3CR1⁺ mononuclear phagocytes via endolysosomal TLRs to expand bone marrow hematopoietic progenitors in mice (127).

INFLUENCE OF IMMUNOSUPPRESSANTS ON GUT METABOLISM

The direct effect of immunosuppressants on gut metabolism is an underdeveloped area of study. In our recent mouse study, we observed tacrolimus-induced rapid and profound metabolic phenotypes in both circulation and gut lumen within 2 days of treatment, prior to previously mentioned alterations in gut microbiota composition and structure (106). The results support the intricate relationships among gut microbiome, metabolic activities, and immune cells in an immunosuppressed environment. The phenotypes were marked by changes to interconnected amino acid metabolisms, bile acid conjugation, glucose homeostasis, and energy production after only 2 days of tacrolimus treatment. A similarly diverse metabolic profile, including multiple changes to amino acid and lipid metabolism, is seen in the metabolome of patients with GVHD receiving cyclosporine (139). The metabolic profile and gut microbiome after 7 days of treatment were distinct from those after 2 days of treatment, indicating continuous drug effects on both gut microbial ecosystem and host metabolism. Further changes to amino acid metabolism occurred after 7 days of treatment, accompanied by pronounced and correlated shifts in the gut microbiome, as mentioned earlier. Crucial mechanisms demonstrated in this study involve increased local catabolism and production of amino acid-derived metabolites via synergistic reactions among the microbiome, intestinal epithelia, and host organs connected through circulation. These results show that the effects of tacrolimus on the gut microbiota and metabolome are not immediate but accrue over time and thus are progressive. While it is crucial to grasp the long-term effects of such immunosuppressants on metabolic disorders, it is also important to identify early indicators of the onset and progression of these changes. Understanding these early alterations is crucial for developing interventions before irreversible changes in metabolic responses occur and affect graft function.

In addition to our tacrolimus research, we observed distinct gut metabolic profiles as a result of other immunosuppressant treatments (Fig. 2A). We noted strikingly similar patterns of alterations in both essential and nonessential amino acids (Fig. 2B). Tacrolimus affected around twice as many metabolites as the other drugs, suggesting it had the most impact on gut metabolism. More than half of the metabolites affected by rapamycin (54%) are also affected by tacrolimus, most of which were amino acids, which suggest shared features in altering amino acid metabolism by these two drugs, as indicated in the biplot in Fig. 2C. Asparagine, phenylalanine, tyrosine, leucine, histidine, tryptophan, isoleucine, methionine, valine, arginine, glutamic acid, serine, and aspartic acid were the most influenced by tacrolimus and rapamycin (Fig. 2D). These results indicate that these drugs affect essential amino acids, leading to their accumulation in the gut. Previous studies have shown that mTOR and calcineurin are essential in regulating the biogenesis of several amino acids, supporting the idea that immunosuppressants inhibiting these molecules can, in turn, alter amino acid levels (140, 141). The mTOR system is also integral in protein synthesis through the initiation of translation and ribosomal biogenesis, further affecting the availability of amino acids (140). These amino acids, together with a few other dipeptides, form a positively correlated network, implicating concerted network interactions among functionally related compounds due to immunosuppressant treatments.

Fig 2.

The luminal metabolome after 7-day treatment. (A) Sparse partial least squares discriminant analysis (sPLS-DA) (142) to demonstrate the clusters by treatment groups. (B) The loadings plot shows the variables selected by the sPLS-DA model for component 1. The variables are ranked by the absolute values of their loadings. (C) Biplot with loading vectors and principal components labeled. (D) Metabolite concentration of amino acids in each treatment group. Data plotted from references 109 and 110 (preprints).

Though there are shared metabolites influenced by different immunosuppressant treatments, distinct metabolic activities unique to each drug were also identified, emphasizing the specific metabolic impact of individual immunosuppressants. There were 20 compounds significantly suppressed by at least three of the immunosuppressant drugs, including amino acids and their derived metabolites: neurotransmitter related compounds, phospholipids, organic acids related to lipid and protein metabolism, and pyridoxine (aka. vitamin B6). These compounds are related to various pathways, indicating that the impacts of immunosuppressant drugs on these metabolic pathways are multifaceted. Gut trimethylamine (TMA) levels were significantly reduced in three treatment groups. Since TMA is produced predominantly by the gut microbiome, reduced TMA levels demonstrate the significant impact of these drugs directly on the microbiome. Among all four immunosuppressant treatments, MMF appeared to be the most distinct in terms of the metabolic effects elicited. MMF suppressed a wide range of pathways in lipid, amino acid, carbohydrate, and nucleotide metabolism, including increased starch and sucrose metabolism. Similar results were seen when intestinal cells were exposed to MPA: MPA exposure altered 35 proteins, most of which are related to nucleotide and lipid metabolisms (143). Prednisolone, different from other drugs, showed increased tryptophan, arginine, proline, glycerophospholipid, vitamin B6, and histidine metabolism. In a study comparing patients with GVHD receiving either steroids, cyclosporine, or both, steroid treatment alone was associated with increased lipid metabolites (139). Asparagine was extremely elevated in the rapamycin and tacrolimus groups and moderately elevated by prednisolone. These observations suggest that each drug not only contributes to common metabolic pathways but also exerts unique effects on the gut metabolome, highlighting the complexity of their interactions with host metabolism.

Together with the gut microbiome results that demonstrated the most pronounced changes at day 30 but mild changes at days 3 and 7, these results support the notion that metabolic changes due to immunosuppressant treatments occur prior to changes in gut microbiome composition and structure. This comprehensive approach that combines microbiome and metabolome allows us to delineate the specific and potentially unique metabolic consequences each drug has on the gut microbiome, thereby enhancing our understanding of their roles in posttransplantation outcomes and providing insights into the broader implications of immunosuppressive therapy on gut health and systemic immunity.

INFLUENCE OF METABOLIC CHANGES ON ALLOIMMUNITY

The products of gut microbiota metabolism, or gut metabolome, also influence the global immune state of the host (Fig. 3). Short-chain fatty acids (SCFA) (i.e., acetate, propionate, and butyrate), long-chain fatty acids (i.e., palmitic acid and stearic acid), secondary bile acids (BA) (i.e., isoallolithocholic acid, lithocholic acid, deoxycholic acid, and isodeoxycholic acid) (144), and amino acid and vitamin derivatives (i.e., indoles and polyamines) (145) are known to be involved in these processes. As an example of this crosstalk between metabolites and the immune system, colonic group 3 innate lymphoid cells (ILC3) express an SCFA-sensing receptor, Ffar2, which specifically responds to bacterial metabolites, driving IL-22 production and gut barrier function in mice (146). In a murine model, the SCFAs butyrate and propionate induce the generation of Tregs (147). Acetate has also been shown to induce Treg development in mice (148). In mice, butyrate promotes Th1 cell development through increased IFNγ and T-bet expression and suppresses the Th17 lineage through suppression of IL-17, RORα, and RORγT (149). Two derivatives of lithocholic acid, 3-oxoLCA and isoalloLCA, modulate Th17 and Treg differentiation in mice through the binding of RORγT or the production of mitochondrial reactive oxygen species, respectively (150). Another secondary bile acid, isoDCA, increases Foxp3 expression in mice through suppressing the immunostimulatory capacity of DCs (151). In a murine model, inosine upregulates il12rb2, Tbx21, and ifng via the GPCR adenosine A_2A receptor on T cells, in turn promoting their activation to a Th1 state (152). BAs were previously shown in mice to signal Tregs, Th17, or DC in immune-mediated disorders (144, 153, 154). Polyamine (i.e., putrescine, spermidine, and spermine) metabolism intersects with glutamine, arginine, and tryptophan metabolism pathways, critically regulating immunity and inflammation (155–160). Polyamines modulate systemic and mucosal adaptive immunity by controlling T cell differentiation (160–163), altering the cytokine milieu and governing inflammation (164–167). In mice, arginine in combination with Bifidobacterium increases polyamine biosynthesis, resulting in decreased TNF, IL-6, and inflammation (168). A recent cohort study reported polyamine metabolites, N-acetyl putrescine and spermidine, are increased in allo-hematopoietic stem cell transplantation (HSCT) recipients without GVHD compared to those with GVHD (136). These studies support the mechanisms of the immunosuppressants that impose off-target immunomodulatory effects through altering gut metabolic activities.

Fig 3.

Interactions of gut metabolome with host immunity.

ADVERSE EFFECTS OF IMMUNOSUPPRESSANT DRUGS

Immunosuppressants have complex and multidirectional interactions with the microbiome, metabolome, and immune system. These complex interactions lead to a variety of adverse events (Fig. 4).

Fig 4.

Overview of interactions between immunosuppressants, gut microbiome, metabolism, intestinal epithelial cells, and host responses.

Infection

Due to their suppressed immune systems posttransplant, transplant patients are at higher risk of infection and more severe infections. Because immunosuppression is antigen-nonspecific, multiple immune responses are suppressed to many pathogens, enabling them to infect patients following transplantation. One of the most common opportunistic pathogens is cytomegalovirus (CMV), which infects most humans and then persists as a latent virus, which can reactivate and infect susceptible patients such as the immunocompromised (169). Patients treated with CNIs are more predisposed to gingival infections, intracellular pathogens, and enhanced herpesvirus replication; corticosteroids can cause susceptibility to bacteria, fungi, and hepatitis B; mTOR inhibitors put patients at risk of poor wound healing and idiosyncratic interstitial pneumonitis (2). Patients with depletion of T lymphocytes are predisposed to viral infection, while those with B-lymphocytic depletion are susceptible to encapsulated bacteria (2). Furthermore, patients undergoing plasmapheresis are at risk of line infections and infections by encapsulated bacteria (2).

Cancer

Cancer is another widely recognized adverse effect of immunosuppression, with solid-organ transplant recipients experiencing a 2- to 4-fold increase in incidence over the general population (170). Chronic immunosuppression impairs antitumor immune surveillance as well as increases the susceptibility of patients to oncogenic viral infections, thereby increasing their risk of developing cancer posttransplantation (171). The most common cancers in transplant recipients are nonmelanoma skin cancers (172). Some other cancers seen in transplant recipients are non-Hodgkin lymphoma, lung cancer, liver cancer, kidney cancer, and Kaposi sarcoma (173). Risk is also high for anogenital cancers caused by human papillomavirus and liver cancer caused by hepatitis C and B viruses (174). Overactivation of the mTOR pathway is common in cancer, allowing for increased growth, proliferation, survival, and motility of cancer cells (170).

Gastrointestinal complications

Certain gastrointestinal complications of transplantation have been associated with different immunosuppressive regimens. MMF has been proposed to inhibit colonic crypt cell division and disrupt the villous structure of the duodenum, causing diarrhea, as well as slowing intestinal turnover, leading to ulceration (16). Additionally, MMF has been suggested to reduce the proliferation of enterocytes and has been shown to impair ZO-1 and occludin expression in mice, which contributes to MMF-related intestinal damage (175). Steroids have been implicated as a cause of spontaneous perforation of the intestine (16). Cyclosporine has been linked to an increased risk of gallstones due to reduced bile flow and is potentially associated with acute pancreatitis (16). In rabbits, CNIs have also been shown to reduce the size of gut-associated lymphoid follicles as well as reduce the uptake and transport of fluorescent particles and immunoreactive CD43⁺ and MHC-II⁺ cells into Peyer’s patches (176). In mice, mTOR activation is critical for intestinal stem cell-driven renewal and repair of the intestinal epithelium; therefore, disruption of mTOR through inhibitors, like rapamycin, can negatively impact intestinal health (177, 178).

Metabolic disorders

Due to the off-target drug effects, immunosuppressants are well-accepted to cause a variety of metabolic toxicities and metabolic syndrome, with the risk increased in those who are older, have presurgical diabetes, or have a high BMI (7). The onset of metabolic syndrome can be accelerated by the transition from a catabolic state marked by organ failure, anorexia, and weight loss to an anabolic state of recovery (7). These disorders, while common, lack clinical management.

Obesity

The transition from a catabolic state to an anabolic state, along with the increased appetite, limited physical activity, and the immunosuppressants themselves, can lead to a high risk of weight gain and obesity posttransplant. One study on posttransplant weight gain showed that the weight gain is positively correlated with cumulative corticosteroid dose (179). Corticosteroid-related weight gain occurs through increased appetite and fluid retention. In another study comparing everolimus to tacrolimus, it was noted that everolimus administration results in less weight gain than tacrolimus (180). Another study in rats showed that rapamycin results in decreased body weight, while MMF has no effect on weight gain (181). Gut microbiota composition is important to the occurrence of weight gain due to its importance in enabling the human body to acquire nutrients and regulate energy usage (182). The microbiota of obese individuals are also altered from those of healthy individuals, suggesting that gut dysbiosis may contribute to the occurrence of weight gain in posttransplant patients (182).

Dyslipidemia

Another frequent immunosuppressant-induced metabolic side effect is the imbalance of various lipids, typically resulting in increased total cholesterol. Corticosteroids are commonly associated with increased appetite, weight gain, and insulin resistance, all of which play a part in inducing hyperlipidemia, which in turn can lead to increased lipolysis and cause the liver to uptake increased amounts of free fatty acids (7). Similarly to corticosteroids, CNIs, like cyclosporine and, to a lesser extent, tacrolimus (12), impair insulin secretion and lead to increased free fatty acids, but they also inhibit the binding of low-density lipoprotein (LDL) to its receptor, impact levels of lipases, and decrease bile acid synthesis, allowing for the accumulation of various lipids (7). mTOR inhibitors contribute to dyslipidemia through the lowered levels of hepatic lipase and the reduced degradation of apolipoprotein B-100 (7). Early studies of MMF in kidney transplant recipients show a beneficial effect on serum lipid levels (12). Hepatic steatosis can accompany dyslipidemia, leading to nonalcoholic fatty liver disease. This effect is transient if corticosteroids are discontinued within 3–6 months post-liver transplant, but has also been seen with tacrolimus-based immunosuppression (7). Microbiota metabolites, such as SCFAs, bile acids, and trimethylamine N-oxide (TMAO), influence cholesterol balance, highlighting the role that altered gut microbiomes, like those in immunosuppressed patients, can have on the development of dyslipidemia (183). A recent study in mice showed that sirolimus-induced dyslipidemia is associated with reduced thickness of intestinal mucosa, increased intestinal permeability, and functional changes to the microbiome relating to carbohydrate and lipid metabolism (184). Under dysbiotic conditions, the growth of normal gut flora is altered, which can lead to dyslipidemia. In turn, dyslipidemia can inhibit the ability of beneficial bacteria to flourish, leading to further dysbiosis (183).

Metabolic syndrome

Metabolic syndrome refers to a set of metabolic dysregulations that lead to an increased risk of developing cardiovascular diseases and diabetes. The criteria that make up the syndrome are insulin resistance, hypertension, central obesity, and dyslipidemia—specifically high triglycerides and low high-density lipoprotein cholesterol. Renal transplant patients, in particular, already have a high risk for metabolic syndrome from the dialysis treatment prior to transplant, and their risk only increases with weight gain and other adverse effects of immunosuppressive drugs (185). Corticosteroids cause insulin resistance, weight gain, hypertension, and dyslipidemia (7, 185). CNI treatment can inhibit insulin secretion, cause hypertension, and increase cholesterol levels (7, 12, 185). mTOR inhibitors are implicated in hyperlipidemia and insulin resistance (7, 185). One study showed that microbiota-induced Th17 cells can protect against the development of obesity and metabolic syndrome through regulating lipid absorption across the intestinal epithelium (186).

Diabetes mellitus

Posttransplant diabetes mellitus (PTDM) is a common metabolic consequence of organ transplantation. Corticosteroids can induce insulin resistance, upregulate gluconeogenic enzymes, and cause pancreatic β cell toxicity which in turn can lead to hyperglycemia and PTDM (7). CNIs can lower blood magnesium levels, which results in increased cellular calcium and impairs insulin secretion, in turn reducing glucose uptake through insulin-regulated glucose transporters (7). CNIs can also reversibly impact the secretion of proliferation genes in β cells through inhibiting the activation of NFATc (7, 12). While there is limited clinical evidence that mTOR inhibitors cause PTDM, in vitro studies have shown that sirolimus causes β-cell apoptosis, impairs the proliferation of ductal cells, and may contribute to hyperglycemia and insulin resistance with chronic use (7, 12). MMF also shows no clinical diabetogenic effects and may be harmful to β cells, but it has been shown to be protective in diabetes-prone rats that develop diabetes as soon as the MMF treatment ends (12). Comprehensive analysis of gut microbiota from PTDM patients shows significant alterations from healthy controls and those with preexisting diabetes mellitus: at the phylum level, Proteobacteria abundance decreases while Bacteroidetes increases (187). A second study similarly resulted in increases in pathogenic bacteria and decreases in probiotic-producing SCFAs, revealing a potential microbial phenotype causing PTDM (188). In addition to altering the microbiome, immunosuppressants are known to affect the integrity of the intestinal barrier as well as associated organs, such as the pancreas, which, when combined with changes to microbial patterns, could contribute to the development of PTDM (189).

FUTURE DIRECTIONS

Immunosuppressant drugs form a complex network of interactions with the host immune system, gut microbiome, and metabolism. These drugs act on various cells throughout the body, affecting not only the immune system but also a multitude of microbes and metabolic processes. Our research on tacrolimus, for example, revealed rapid and profound effects on gut microbes and their metabolic functions, which in turn impacted local, regional, and systemic levels of immune cells (106). This intricate interplay highlights the need for integrated approaches in transplant medicine to optimize treatment and manage side effects. However, delineating the specific pathways leading to adverse effects from immunosuppressants remains a significant challenge. To address this, there is a critical need for sophisticated technologies that can elucidate underlying mechanisms of drug action and side effects, provide continuous monitoring of patient responses, and identify biomarkers to predict individual outcomes, ultimately enabling more personalized and effective management of immunosuppression in transplant recipients.

Live biotherapeutics (aka probiotics), prebiotics, postbiotics, and synbiotics

Gastrointestinal symptoms and postoperative infections are significant causes of morbidity and mortality in solid organ transplant recipients, often necessitating prolonged antibiotic use in addition to immunosuppressants (190, 191). The growing concern over multidrug-resistant pathogens, such as Clostridium difficile, coupled with the lack of standardized guidelines for infection prevention and control, has spurred interest in alternative approaches to managing these complications. Live biotherapeutics (LBP), also known as probiotics, have emerged as promising interventions to improve gut health and enhance graft survival in transplant recipients. When used in combination with immunosuppressants and prophylaxis antibiotics, they may offset side effects and enhance the healing process posttransplantation. LBPs, particularly those containing Bifidobacterium and Lactobacillus strains, have demonstrated anti-inflammatory and pro-tolerogenic immune effects in this context. Several studies in humans have shown encouraging results. A few, small clinical studies have shown that LBPs result in improved graft function and injury as measured by serum bilirubin concentration and transaminase activity (192), prevent Clostridium difficile infections (193), and lower creatinine levels along with improved estimated glomerular filtration rate (194). These studies provide proof of concept that LBPs should be further investigated for the beneficial effects they can provide on restoring gut health and enhancing graft survival. Combined with a data-driven approach, future LBP interventions could be targeted or engineered to enhance drug metabolism and reduce the side effects, such as gastrointestinal toxicity, associated with immunosuppressive drugs.

Prebiotics, defined as nondigestible dietary substances selectively utilized by host microorganisms (195), and synbiotics, which combine pre- and probiotics, have also shown promise to improve the gut dysbiosis seen in transplant patients. The International Scientific Association for Probiotics and Prebiotics Synbiotics combines prebiotics and probiotics to synergistically improve gut health. Clinical studies have shown that synbiotics are capable of improving transplant outcomes by inducing regulatory T cells and reducing the risk of severe acute GVHD, following allogeneic hematopoietic stem cell transplantation (196), lowering plasma p-cresol concentrations in kidney transplant recipients (197), and lowering infection rates in liver transplant recipients (198). As these studies show, prebiotic interventions are an exciting avenue in which beneficial bacteria can be promoted in order to optimize drug efficacy and reduce side effects.

Postbiotics were defined in 2021 by the International Scientific Association of Probiotics and Prebiotics (ISAPP) as “a preparation of inanimate microorganisms and/or their components that confers a health benefit on the host” (199). Postbiotics are nonliving microbial cells that have been deliberately inactivated and can include metabolites or cell components. They can be derived from a variety of sources, including heat-killed bacteria or fermented foods that have been heat-treated to increase their shelf life (200). Thus, postbiotics possess the beneficial biological activities of probiotics, while no longer having the concern of probiotic viability, addressing safety issues such as bacteremia, transmitting virulence genes, or spreading antibiotic resistance genes to the gut microbiota via horizontal transfer (201, 202). Additionally, the postbiotic preparations can be easily and stably stored at room temperature over the years without the need to consider the progressive reduction in biological activity due to loss of bacterial viability. Small molecule inhibitors of microbial enzymes, such as β-glucuronidase inhibitors, have the potential to modulate immunosuppressant metabolism and, in turn, reduce drug-induced side effects like diarrhea in transplant patients (99). In a new, ongoing area of research, these functional and physical attributes of postbiotics have spurred considerable interest among investigators.

However, several limitations and challenges must be addressed before the clinical implementation of LBPs, prebiotics, and synbiotics in combination with immunosuppressant drugs. Many existing studies have limited sample sizes and insufficient statistical power. There is also a lack of standardization in dosing, measurement methods, and outcome reporting across studies. Further, safety measures and potential adverse effects require more thorough investigation, especially in immunocompromised patients. Moreover, long-term effects of probiotic use remain unclear, necessitating extended follow-up studies. Further research should focus on addressing these limitations and conducting large-scale, long-term randomized controlled trials to evaluate both efficacy and safety profiles in transplant recipients. It is of particular importance to establish standardizing protocols for probiotic administration, measurement, and outcome reporting to enable more robust meta-analyses and establish evidence-based guidelines for this immunocompromised population. Meanwhile, investigating the mechanisms of action of specific probiotic strains and prebiotic compounds in the context of transplantation is also crucial. The potential to reduce antibiotic dependence and improve graft outcomes through microbiome modulation represents an exciting frontier in transplant medicine, warranting continued investigation and development.

Monitoring risk in transplant medicine

A critical gap in transplant medicine is the lack of continuous, real-time monitoring to assess the risk of graft health, predict potential complications, and guide timely adjustments of immunosuppressant regimens. Traditional methods, such as biopsies and periodic blood tests, are invasive and provide only intermittent snapshots of graft status, limiting the ability to fine-tune immunosuppression on an as-needed basis. This gap in continuous monitoring can lead to either over immunosuppression, increasing the risk of infections and other side effects, or under immunosuppression, potentially resulting in graft rejection. Recent advances in sequencing technologies, such as Oxford nanopore sequencing, offer promising avenues to address this challenge through noninvasive measurement of the gut microbiome and selected immune markers (203). Nanopore technology has several advantages that make it particularly suited for transplant monitoring: real-time, long-read sequencing capabilities; portability and flexibility for in-field use; and cost-effectiveness compared to traditional sequencing, such as the Illumina platform. Recent applications of nanopore sequencing demonstrate its potential in transplant medicine. These include malaria surveillance from dried blood spots and pathogen surveillance from ticks (204–206). Additional research has shown the feasibility of nanopore sequencing in pathogen identification from stool samples (207, 208). Studies have also used nanopore sequencing for rapid donor-specific single nucleotide variation detection of ex vivo lung perfusate cell-free DNA (cfDNA) (209). Others validated nanopore implementation in peripheral blood human leukocyte antigen (HLA) class I typing and illustrated that it can be used in routine diagnostics with high accuracy (210). A recent study used nanopore-targeted sequencing to rapidly and accurately identify infection present in the blood of deceased donors (211). While these applications suggest the great potential of continuous monitoring, major limitations and challenges must be addressed. These include the lack of predefined gut microbiome features indicative of homeostasis or dysbiosis in transplant patients; the need for consequence-informed microbial biomarkers specific to immune activation; and the complexity of data interpretation and integration into clinical decision-making. Future implementation of nanopore technology could enable early detection of microbiome changes signaling immune activation, allograft rejection risk, or infection. This continuous monitoring would facilitate timely interventions and personalized immunosuppressant regimens, potentially improving long-term transplant outcomes. Integrating these advanced technologies into routine care could usher in a new era of proactive, precision-based transplant medicine, enhancing both graft longevity and patient quality of life.

Multi-omics technologies towards personalized immunosuppression

The complex nature of transplant rejection and immunosuppressant responses necessitates a comprehensive approach to understanding underlying mechanisms and identifying predictive biomarkers. Multi-omics technologies, which integrate data from genomics, epigenomics, transcriptomics, metabolomics, and proteomics, offer a powerful tool for characterizing the biological processes involved in graft survival and immunosuppressant drug responses (212, 213). Metabolomics, in particular, has emerged as a valuable noninvasive method to assess graft function. By analyzing small bioactive molecules in various body fluids (e.g., serum, urine, saliva, feces, sweat, breath), metabolomics provides insights into graft function without the need for invasive biopsies. Recent studies have demonstrated the potential of this approach. One study employed untargeted liquid chromatography-mass spectrometry (LC-MS) and revealed discriminative metabolites, such as creatinine, kynurenine, and uric acid, in acute graft rejection among renal transplantation patients (214). LC-MS also reveals that a high presence of asymmetric dimethylarginine in the preservation medium of donor livers is a good indicator of poor outcome post-orthotopic liver transplantation (215). Nuclear magnetic resonance spectroscopy (NMR)-based metabolomics has been used to show increased levels of lipids and lipoproteins in the plasma of heart transplant recipients when rejection occurs (216). “Pharmacometabolomics”, a term coined in 2006 as “the prediction of the outcome of a drug or xenobiotic intervention in an individual based on a mathematical model of preintervention metabolite signatures,” represents an exciting strategy to personalize drug treatments (217). In this strategy, postdrug analysis of the metabolome can be applied to assess the toxicity of drug compounds and classify individuals as responders or nonresponders to the drug. This can then guide predrug metabolome analysis to determine biomarkers and biochemical signatures that predict how well a transplant patient will respond to and metabolize the immunosuppressant drugs. A recent study on the pharmacometabolomics of everolimus in heart transplant recipients revealed that a higher dose requirement was associated with co-administration of MMF, the CYP3A5 expressor genotype, and the presence of lysophosphatidylcholine, whereas a lower dose requirement was associated with patients receiving vitamin K antagonists (218). Furthermore, as observed in our study comparing the effects of tacrolimus, antibiotics, and combined treatment on mice, distinct “metabotypes” or metabolic phenotypes can be associated with different treatments and time points (106). These metabotypes provide unique insight into the development of metabolic changes and could represent potential therapeutic and diagnostic targets in managing the myriad metabolic disorders that can arise from prolonged immunosuppressant usage.

Other emerging technologies of interest are single-cell RNA sequencing (scRNAseq) and spatial transcriptomics for biopsies from transplant patients. scRNAseq provides a previously unavailable precision to studying gene expression in cells within a tissue. One study on 16 kidney transplant biopsies revealed a relationship between FCGR3A⁺ monocytes and NK cells and the degree of inflammation (219). Another study comparing a kidney allograft with fibrosis and glomerulopathy to one successfully treated for active antibody-mediated rejection showed that the fibrotic kidney expressed high levels of infiltrative, recipient-derived fibroblasts, whereas the other graft showed mostly donor–organ-derived fibroblasts (220). These studies show promising impact, but this technique has very low throughput and does not provide spatial information on the cells. Spatial transcriptomics allows for visualization of gene activity to be mapped back to intact tissue sections, providing insight into the complex spatial organization of biological processes. One study using spatial transcriptomics on kidney allograft biopsies revealed that renal interstitial regions are enriched with rejection-associated genes, while the tubular areas were not, suggesting that rejection is primarily driven by biological processes in the renal interstitium (221). Another study, utilizing single-cell RNA sequencing followed by a multiplex immunofluorescence analysis, demonstrated the importance of FcγRIII⁺ NK and FcγRIII⁺ nonclassical monocytes in antibody-mediated rejection and further highlighted the specificity of this rejection to the glomerular areas (219). Genetic variants, gene expression, protein expression, metabolite levels, and epigenetic modifications can be used alone or in combination to characterize individual variability, help predict optimal drug dosage, and classify groups of individuals that may respond or not, may have sensitivities, or may experience toxic side effects to a drug (213).

While these advanced technologies and approaches offer unprecedented opportunities to personalize immunosuppression strategies for improving transplant outcomes, several limitations and challenges must be addressed. The integration of multi-omics data remains complex, with issues of data heterogeneity and interpretation posing significant hurdles. This is in addition to challenges in standardized protocols and the requirement for robust bioinformatics pipelines and infrastructure to handle large volumes of data in multiple data types. Additionally, the cost and technical expertise required for these advanced technologies may limit their widespread adoption, particularly in resource-constrained settings. Moreover, the need for clinical validation through large-scale, prospective studies is not dispensable. While significant challenges remain, the potential benefits of these approaches in tailoring treatments, minimizing adverse effects, and enhancing graft longevity warrant continued investigation and development. As these technologies mature and become more accessible, they have the potential to transform transplant medicine, moving towards precision-based patient care in transplant medicine.

Biographies

Allison Kensiski is a Laboratory Research Technician in the Bromberg Lab in the University of Maryland School of Medicine helping on projects that vary from the network of interactions between microbiota, immunosuppression, and host immune systems to how immunological scarring affects donor-specific alloimmunity and graft outcomes. She previously worked for a histology lab as a laboratory assistant. She received her Bachelor of Science in NanoEngineering from the University of California, San Diego in 2016, her Master of Science in Biotechnology from Johns Hopkins University in 2020 and plans to start work towards a PhD.

Dr. Samuel J. Gavzy is a resident physician in general surgery at the University of Maryland Medical Center. He graduated from the University of Chicago with a dual bachelors in Biological Science and History of Science, followed by a Master of Science in Biomedicine from the Icahn School of Medicine at Mount Sinai, and a MD from Rutgers New Jersey Medical School. He has worked in multiple immunology laboratories with translational projects spanning basic, mucosal, and transplant immunology. During surgical residency, Dr. Gavzy spent his two-year research post-doctoral fellowship in the laboratory of Dr. Jonathan Bromberg studying the network interaction of gut microbiota, immunosuppressant therapeutics, and the mucosal immune system as it relates to transplant allograft rejection.

Dr. Long Wu is a research associate in Dr. Bromberg’s lab at the University of Maryland School of Medicine. She earned her Ph.D. from Jinan University, China, focusing on cell-based immunotherapy. During her postdoctoral fellowship at the University of Maryland, she expanded her expertise to bone marrow transplantation and immunotherapy. Her current research investigates how immunologic scarring and non-allograft-related events influence long-term donor-specific alloimmunity and graft outcomes, aiming to advance understanding of immune regulation in transplantation. Long Wu has been working in the field of immunology for over ten years, driven by a passion to uncover mechanisms that improve graft survival and patient outcomes.

Dr. Valeria Mas is a Professor of Surgery, Microbiology and Immunology at University of Maryland School of Medicine. She received her Master’s from National University of Cordoba, Argentina, and PhD from National University of San Luis. She completed her postdoctoral fellowship at Virginia Commonwealth University then joined as an Assistant Professor. She is the former director and founder of the Transplant Research Institute, University of Tennessee Health Science Center, where she was appointed Endowed Professor of Transplant Research and Surgery. Previously, she was the Director of the Molecular Transplant Research Laboratory at University of Virginia, and the Director of the Transplant Genomics Laboratory at Virginia Commonwealth University. Her research focuses on improving organ donor utilization by better understanding donor organ biology and response to injury and increasing longevity of grafts by evaluating cells, cell-cell interactions and associated pathways that differentiate wound healing vs. impaired repair and loss of graft function.

Dr. Bing Ma is currently an Assistant Professor in the Institute for Genome Sciences and the Department of Microbiology and Immunology at the University of Maryland School of Medicine. She earned her PhD from the University of Wisconsin-Madison, where she specialized in ‘omics’ technologies. Following her graduate training, she has held research positions focusing on the development and application of advanced sequencing methodologies. Her work explores the intestinal ecosystem's complex dynamics and its essential role in health and disease. Dr. Ma has a strong interest in leveraging cutting-edge sequencing technologies to generate mechanistic insights and translate these findings into actionable strategies for disease prevention, diagnosis, and treatment. With over a decade of experience in this rapidly advancing field, she remains committed to bridging innovative research and practical applications to improve human health.

Dr. Jonathan S. Bromberg is a Professor of Surgery, Microbiology and Immunology and the Vice Chair for Research at the University of Maryland School of Medicine. He received his Master’s and PhD from Harvard before doing postgraduate studies at University College, London. He served as intern, resident, and chief resident in the Department of General Surgery at the University of Washington Affiliated Hospitals before doing his fellowship at the Hospital of the University of Pennsylvania. He has focused on T cell immunobiology for over 25 years. For over 15 years, he has also focused on migration, trafficking, secondary lymphoid organ structure and function, and lymphatic structure and function, and how these processes and structures influence T cell immunity and tolerance in models of cardiac and pancreatic islet transplantation. He has concurrently maintained an active clinical practice in solid organ transplantation exposing him to the problems of patients and their immune system.