New Insights into the Microbiome in Kidney Transplantation

Abstract

Purpose of Review.

Research in the past decade has revealed important implications for the microbiome in human health. Studies have defined a distinct gut microbiota in kidney transplant recipients and have recently linked the microbiota to infectious complications, similar to the allogeneic stem cell transplant population.

Recent findings.

In this review, we focus on the metabolism of immunosuppressive medications by the gut microbiota and on the urinary microbiome in the setting of infectious and immunological complications. We highlight seminal studies showing the role of specific gut microbiota in the direct metabolism of tacrolimus into a lesser effective immunosuppressant as well as the role of the gut microbiota in the reactivation of mycophenolic acid glucuronide. We describe distinct urinary microbiota patterns in kidney transplant recipients with interstitial fibrosis tubular atrophy, chronic allograft nephropathy, tolerance, and bacterial and viral complications.

Summary.

The microbiota has important implications for immunosuppressive medications and immunological outcomes in kidney transplant recipients. Further research is needed to better delineate the impact of the metabolism of tacrolimus and mycophenolic acid by gut bacteria and the role of the urine microbiota in the development of immunological and infectious complications.

Introduction

Recent advances in sequencing technologies have led to the burgeoning field of microbiome research. Technologies such as 16S rRNA sequencing identify microbial DNA and can classify cultivable and non-cultivable microbes. Indeed, several groups have profiled the gut microbiota in kidney transplant recipients. They have found a higher gut abundance of the common phylum, Firmicutes, and a lower abundance of Bacteroidetes in kidney transplant recipients (1, 2), which is in contrast to the human microbiome project (3). Importantly, several studies have linked specific gut microbiota to the development of infectious complications such as urinary tract infections and respiratory viral infections in kidney transplant recipients (4, 5). In this review, we focus on two newer developments in the field of microbiome research in kidney transplantation: the role of the gut microbiota in the metabolism of immunosuppressive medications, and the characterization of the urine microbiome and virome during infectious and immunological complications.

Metabolism of Tacrolimus and Mycophenolic Acid by the Gut Microbiota

The impact of the gut microbiota on altering medications has recently become a topic of interest. Sulfasalazine, an anti-inflammatory medication for inflammatory bowel disease, is a widely used pro-drug that requires colonic bacteria to cleave the azo-bond to produce the active drug (6). The gut microbiota, specifically Eggerthella lenta, has also been discovered to directly inactivate digoxin, a cardiac glycoside used to treat congestive heart failure (7). A recent study by Zimmerman et al. has taken the work a step further and evaluated the gut microbiota’s ability to metabolize a panel of drugs on a large scale (8). They tested 76 gut bacterial strains on 271 drugs and found that almost two-thirds of the drugs underwent chemical alteration. This study further implies that the variation in microbiome between individuals may account for the variability in drug metabolism seen clinically (8). The interaction between the gut microbiota and medications has thus important clinical implications both in terms of clinical efficacy as well as toxicity and this may be particularly true for tacrolimus and mycophenolate mofetil: two commonly used immunosuppressive medications in kidney transplantation.

The introduction of tacrolimus, a calcineurin inhibitor, has significantly reduced acute rejection rates in kidney transplant recipients (9). Tacrolimus, however, has a narrow therapeutic range that requires frequent monitoring of trough levels in order to avoid acute rejection or nephrotoxicity (10). Intra-patient variability of tacrolimus trough levels is not completely understood, and recent work had suggested a role for the gut microbiota. A pilot study by Lee et al. investigated the relationship between the gut microbiota and tacrolimus. Utilizing 16S rRNA gene sequencing of the V4-V5 hypervariable region in fecal specimens from 19 kidney transplant recipients, the group discovered a relationship between a common gut bacterial species, Faecalibacterium prausnitzii, and tacrolimus dosing (11). The group theorized that high gut abundance of F. prausnitzii was associated with healthy colonic cells, which, in turn, led to intact P-glycoprotein functioning, explaining the positive correlation between F. prausnitzii and tacrolimus dosing (11). Based upon this work, Guo et al. further evaluated whether F. prausnitzii was involved in the direct metabolism of tacrolimus in vitro and demonstrated that intact, but not boiled, F. prausnitzii metabolized tacrolimus into a bacterial tacrolimus metabolite (M1) (12). They further found that many bacterial species, particularly in the Clostridiales order, directly metabolized tacrolimus like F. prausnitzii (12). Further studies on the dominant bacterial tacrolimus metabolite M1 revealed that M1 is less potent in inhibiting peripheral blood mononuclear cells by 15-fold when compared to parent tacrolimus (12). While this study shows that gut bacteria can metabolize tacrolimus to M1, it is not completely known whether gut microbiota metabolism of tacrolimus happens to a significant extent in vivo. Recently, the group had a follow up study and profiled blood M1 levels in 10 kidney transplant recipients after oral administration of tacrolimus and they detected the presence of blood M1 levels in all patients (13). While this pilot study was small, it nevertheless reveals that gut microbiota metabolism of tacrolimus occurs in vivo. The group, however, did not assess the impact of gut microbiota metabolism of tacrolimus on immunological outcomes such as acute rejection nor did the group assess the impact of antibiotic therapy on the gut microbiota metabolism of tacrolimus and tacrolimus trough levels. Further research will be needed to better understand the extent of gut microbiota metabolism of tacrolimus and its implications on immunological outcomes.

Mycophenolate mofetil (MMF) is another commonly used immunosuppressant in kidney transplant patients. It has been shown to decrease acute rejection rates in kidney transplant recipients (14). Early pharmacokinetic studies have shown a secondary peak of mycophenolic acid (MPA) after MMF administration, which has been thought to be secondary to enterohepatic recirculation due to gut bacteria (15). Notably, MMF is associated with gastrointestinal side effects such as diarrhea and more recent work has better defined the relationship between the gut microbiota and MMF. A study by Steven Greenway’s group demonstrated that an intact gut microbiota was necessary for MMF-induced toxicity. In this study, MMF-treated germ-free mice as well as MMF-treated mice that received broad-spectrum antibiotics did not display MMF-associated adverse effects when compared to control MMF-treated mice (16). MMF-treated mice were further found to have an increased abundance of members of the phylum, Proteobacteria, especially Escherichia and Shigella, indicating that MMF may have a role in the expansion of pathobionts (16). A follow-up study by this group demonstrated that administration of vancomycin reversed MMF-induced weight loss in a mouse model by reducing the activity of beta-glucuronidase (GUS) (17). Based on their findings, the group hypothesized that gut bacteria with GUS activity changes inactive mycophenolic acid glucuronide (MPAG) back to active MPA which could then contribute to MMF-induced gastrointestinal adverse effects. A follow up study by Zhang et al. profiled the gut microbiota and fecal GUS activity in 42 kidney transplant recipients with and without post-transplant diarrhea (18). They reported a positive correlation between fecal GUS activity and specific taxa, Coprococcus and Subdoligranulum (18). While they did not find fecal GUS activity significantly different between kidney transplant recipients with post-transplant diarrhea and without post-transplant diarrhea, they did report that higher fecal GUS activity was associated with prolonged duration of post-transplant diarrhea (18). Further studies are necessary to better understand the changes in fecal GUS activity over time in kidney transplant recipients and the relationship between elevated fecal GUS activity and adverse effects of MMF therapy.

Urinary Microbiome and Immunological Complications

While much focus has been placed on studying the microbial communities of the gastrointestinal tract, there is increasing interest in characterizing the urinary microbiome in kidney transplant recipients. In a longitudinal study by Fricke et al., the investigators described, for the first time, changes in the urinary microbiome over the first 6 months after transplantation (1). This initial study found detectable 16S rRNA in 33% of collected urine specimens. The 16S rRNA profiles showed that Lactobacillus was a dominant species with a greater than 80% relative abundance followed by Enterococcus, Bifidobacteriaceae, and Pseudomonas (1). The authors did not find a correlation between the identified species and urinary tract infections (UTIs); however, the cohort had only 4 cases (1).

Since this initial study, several recent studies have investigated the urinary microbiome in kidney transplant recipients with immunological complications, such as chronic allograft dysfunction (CAD), and with interstitial fibrosis and tubular atrophy (IFTA). Wu et al. investigated a cohort of 67 kidney transplant recipients from the Deterioration of Kidney Allograft Function cohort (19). They performed 16S rRNA sequencing of the V4 region and compared 35 male and female kidney transplant recipients with CAD to 32 female kidney transplant recipients with non-CAD (19). They found that the abundance of Corynebacterium was significantly higher in the CAD group than the non-CAD group (19). A limitation of the study is the comparison to an only female non-CAD cohort. Similarly, another study led by Modena et al. evaluated the urinary microbiome associated with IFTA (20). The authors profiled the urinary microbiome in 25 kidney transplant recipients with IFTA, 23 with normal biopsies, and 20 healthy non-transplant controls and reported several findings (20). Healthy female controls had significantly higher abundance of Lactobacillus and lower abundance of Streptococcus than healthy male controls (20). The urinary abundance of Streptococcus was significantly lower in male kidney transplant recipients with IFTA or with normal biopsies than in male healthy controls (20). More recently, a urinary microbiome profile has emerged from tolerant kidney transplant recipients. Colas et al. profiled the urine in spontaneously tolerant kidney transplant recipients, kidney transplant recipients with stable allograft function, kidney transplant recipients on minimal immunosuppression, and healthy volunteers (21). Using a DESeq2 algorithm, a program that allows for microbial comparisons using count data, the authors reported 34 operational taxonomic units (OTUs) that were significantly different between the tolerant kidney transplant recipients and healthy controls (21). At the family level, many of the taxa that were higher in the tolerant group were composed of the phylum, Proteobacteria (21). Further analysis also revealed different taxa associated with immunosuppressive regimens. The family Streptococcaceae was negatively correlated with mTOR inhibitors and the family Staphylococcaceae was negatively associated with calcineurin inhibitors (21).

Altogether, a distinct urinary microbiome is associated with several allograft complications and even transplant tolerance. However, it is uncertain whether the urinary microbiome is associated with the development of these transplant outcomes or whether it just reflects the environment during these complications. Future mechanistic studies using specific strains of identified bacteria are needed to better understand the relationship between the urinary microbiome and these allograft complications. It is also important to understand that the urinary microbiome is sensitive to contamination and most of these specimens were collected by clean catch. Future studies are also needed with catheterized urine specimens.

Urinary Microbiome and Infectious Complications

Common infections of the urinary tract in kidney transplant recipients include urinary tract infections (UTI) and viral infections such BK infection. While 16S rRNA sequencing can provide comprehensive bacterial identification, a major limitation of this method is the inability to profile viral and fungal species. Metagenomic sequencing, on the other hand, can provide more comprehensive evaluation and can identify bacterial, viral, and fungal species and can thus be useful for monitoring infections in the urinary tract of kidney transplant recipients. Burnham et al. evaluated urinary cell-free DNA (cfDNA) using metagenomic sequencing of 141 urine specimens from 82 kidney transplant recipients (22). They found excellent congruence of urinary cell-free DNA profiles and urine cultures that were positive by conventional culturing (22). Using the cfDNA metagenomic sequencing data, they were able to detect antimicrobial resistant genes associated with antimicrobial resistance profiles and were able to estimate bacterial growth given the uneven coverage of metagenomic sequencing (22). In addition, urinary cfDNA profiling detected a plethora of viruses such as BK virus and importantly, rare viruses not routinely screened for, such as adenovirus and parvovirus, could be detected days to weeks prior to clinical presentation (22). The authors were also able to estimate kidney allograft injury by estimating the donor and recipient proportion, suggesting the utility of cell-free DNA metagenomic sequencing as a comprehensive test to monitor infections in the urinary tract (22).

Studies by Rani et al. have also confirmed the use of metagenomic sequencing to better characterize the microbiome and virome. They profiled the urinary microbiome in 21 kidney transplant recipients and 8 healthy controls using metagenomic DNA shotgun sequencing and demonstrated a decrease in microbial diversity in the kidney transplant group (23). Transplant recipients had an increased abundance of E. faecalis and E. coli, which are common pathogens that cause UTIs and had increased abundance of enzymes in the folate pathway, suggesting resistance to trimethoprim-sulfamethoxazole, a common prophylactic antibiotics for Pneumocystis jirovecii (23). The same group conducted a follow up study which investigated the urinary virome using metagenomic sequencing in 22 kidney transplant recipients with or without BK viremia (24). In addition to detecting BK virus in the urine of kidney transplant recipients with BK viremia, they also detected JC virus and torque teno virus (24). A complementary study using liquid chromatography-mass spectrometry analysis was performed on the urine of 133 kidney transplant recipients and 8 healthy controls (25). Using this method, 37 unique viruses were identified and there were significant differences in the distribution of viral proteins among kidney transplant recipients with stable allograft function, acute rejection, CAN, and BK virus nephritis (25).

Further advances are being made in metagenomic sequencing which can further our understanding of the host’s response to infections. A follow up study by De Vlaminck’s group utilized whole genome bisulfite sequencing in 51 urine supernatants from kidney transplant recipients to examine host’s injury (26). Whole genome bisulfite sequencing takes advantage of genome-wide CpG methylation marks detected within the cfDNA to identify the origins of the tissues (26). Using this technique, the group was able to detect kidney tissue injury in the urine of patients with confirmed infection, in addition to identifying the pathogens involved (26). The study demonstrated a higher amount of kidney tissue in kidney transplant recipients with BK virus nephropathy (BKVN) compared to those with normal allograft biopsies (26). Furthermore, the amount of bladder tissue in the urine was also higher in kidney transplant recipients with UTIs compared to those without UTIs (26). The whole genome bisulfite sequencing technique thus allows quantification of tissue injury in the urine of kidney transplant recipients and could provide clinicians with an additional tool to differentiate true symptomatic UTI from colonization.

While these newer profiling techniques in the urine are providing new insight into the microbiome and virome of the urinary tract, several questions still exist. Detection of microbes is not synonymous with infection. Sigdel et al. revealed 37 unique viruses in the urine and Burnham et al. also detected several viruses not classically pathogenic in the urine (22, 25). Future studies are needed to better understand the host’s response to these detected microbes. In addition, studies have not systematically tested metagenomic sequencing for estimating antimicrobial susceptibility testing. In order to provide a comprehensive clinical test, further studies are needed to show the sensitivity and specificity of metagenomic sequencing to the gold standard testing in urine.

Conclusion

Our understanding of the role of the microbiota in complications in kidney transplant recipients continues to expand. Recent studies have elucidated a novel role for the gut microbiota in the metabolism of tacrolimus and mycophenolate mofetil, two key immunosuppressants commonly used in kidney transplant recipients. The extent to which the microbiota impact levels of these immunosuppressants are not known and further research is needed. However, if they are found to have a significant impact, profiling the gut microbiota may be helpful in providing personalized approaches for immunosuppressive medications. In addition, recent work in the urine microbiota has revealed a distinct gut microbiota in kidney transplant recipients with IFTA, CAN, and tolerance. Furthermore, newer techniques such as urinary cfDNA sequencing can provide a wealth of information beyond just identification of pathogens and commensals. Further research is needed to better understand the host-pathogen interaction in the urine of kidney transplant recipients.

Key points.

• The gut microbiota can metabolize tacrolimus and mycophenolate mofetil, two of the most common immunosuppressive medications used in kidney transplant recipients

• Kidney transplant recipients with interstitial fibrosis tubular atrophy, chronic allograft nephropathy, and tolerance have a distinct urine microbiota

• Urinary cell-free DNA sequencing can provide comprehensive microbial profiles as well as characterization of bacterial growth rates and antibiotic resistant genes