Abstract
Background: The term “transrenal DNA” was coined in 2000 to signify that DNA in urine may come from the passage of plasma DNA through the kidney barrier. Although DNA in the urine has the potential to provide a completely noninvasive source of nucleic acids for molecular diagnosis, its existence remains controversial.
Methods: We obtained blood and urine samples from 22 hematopoietic stem cell transplant (HSCT) recipients and used fluorescence in situ hybridization, PCR for short tandem repeats, mass spectrometry, quantitative PCR, and immunofluorescence detection to study donor-derived DNA in the urine.
Results: All HSCT recipients exhibited high amounts of donor-derived DNA in buffy coat and plasma samples. Male donor–derived DNA was detected in supernatants of urine samples from all 5 female sex-mismatched HSCT recipients. Surprisingly, the amount of DNA in urine supernatants was not correlated with the plasma value. Moreover, cell-free urine supernatants contained DNA fragments >350 bp that were absent in plasma. Donor-derived polymorphs were detected in urine by fluorescence in situ hybridization. Coincidentally, donor-derived cytokeratin-producing epithelial cells were discovered in urine samples from 3 of 10 sex-mismatched HSCT recipients as long as 14.2 years after transplantation.
Conclusions: This report is the first to demonstrate the presence of donor-derived DNA in the urine of HSCT recipients; however, we show that much of this DNA originates from donor-derived cells, rather than from the transrenal passage of cell-free plasma DNA. Our discovery of donor-derived cytokeratin-producing epithelial cells raises interesting biological and therapeutic implications, e.g., the capacity of marrow stem cells to serve as an extrarenal source for renal tubule regeneration.
In 2000, Botezatu et al. proposed the “transrenal hypothesis,” that plasma DNA passes through the kidney barrier and enters the urinary tract to produce so-called transrenal DNA (Tr-DNA)1 (1). These investigators targeted a Y-chromosomal locus with nested PCR and detected male donor–derived DNA in urine samples from 5 of 9 women who received blood transfusions from male donors. In addition, such tumor-associated DNA markers as KRAS2 [v-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog (KRAS1)] and RASSF1A [Ras association (RalGDS/AF-6) domain family member 1A (RASSF1)] have been detected in urine samples from patients with pancreatic and colon cancers(1) and breast cancer patients(2), respectively. Other researchers have detected fetal-derived DNA in the urine of pregnant women in the first(1)(3) and third(3)(4) trimesters. Umansky and Tomei(5) have reviewed the evidence for the transrenal hypothesis; however, other investigators have challenged this hypothesis. Su et al. reported finding the KRAS mutation in the urine of colorectal cancer patients but not in plasma(6). In addition, 3 independent groups were unable to detect fetal DNA in maternal urine, even in pregnant women with a compromised kidney barrier function due to HELLP (hemolysis, increased liver enzymes, and low platelet count) syndrome(7)(8)(9).
The presence of target DNA at low concentrations may be one reason for the inability to detect fetal DNA in maternal urine. Fetal DNA is known to constitute only 2.3%–11.4% (mean, 6.2%) of the total circulating DNA in maternal plasma, even in late pregnancy (37–43 weeks of gestation)(10). The amount of fetal DNA in urine, if any, is likely to be low and may be difficult to detect. In contrast, we have shown that a median of 59.5% of the DNA in the plasma of hematopoietic stem cell transplant (HSCT) recipients is of donor origin(11); thus, the amount of donor-derived DNA in HSCT recipients is about 5- to 25-fold higher than that of fetal DNA in maternal plasma. In view of the distinct advantages of urine as a source for nucleic acid testing and the controversies surrounding the transrenal hypothesis, we followed up our previous observation and investigated the occurrence of nonhost DNA in the urine of HSCT patients. In phase 1 of the study, we used a mass spectrometry–based platform to detect, quantify, and compare the amounts of donor-derived DNA in urine and blood samples. In phase 2 of the study, we characterized the nature of urinary DNA with respect to size distribution and cellular origin.
Materials and Methods
phase 1
Patient recruitment.
We recruited 22 HSCT patients from the Bone Marrow Transplant Clinic of the Department of Paediatrics, Prince of Wales Hospital, The Chinese University of Hong Kong, Shatin, Hong Kong SAR. All patients were in remission with respect to their primary conditions. Informed consent was obtained from patients or their parents. The study was approved by the Clinical Research Ethics Committee of The Chinese University of Hong Kong and was performed in accordance with the Helsinki Declaration.
Sample collection and preparation.
Peripheral blood was collected into EDTA-containing blood tubes. We harvested plasma by centrifuging blood samples twice to minimize contamination by blood cells, as previously described(12).
Fresh urine samples were collected into sterile plain bottles. A urine aliquot was tested with a urinalysis reagent strip (Multistix; Bayer) and assayed for creatinine on a Roche MODULAR analyzer. To inhibit possible nuclease activities, we mixed the sample with 0.5 mol/L EDTA, pH 8.0 (Invitrogen), to a final concentration of 10 mmol/L(1)(13). To separate the cell-free and cellular urine components, we centrifuged urine samples at 3000g at 4 °C for 10 min and filtered the supernatant through a 0.45-μm filter (Milex-GV; Millipore) to remove any remaining cells or cell debris. The cell-free urine supernatant was stored at −80 °C until DNA extraction. Pellets of urinary cells obtained after centrifugation were washed twice with 1× PBS (144 mg/L KH2PO4, 9 g/L NaCl, and 795 mg/L Na2HPO4 · 7H2O, pH 7.4; Invitrogen) and stored at −20 °C until DNA extraction.
Fluorescence in situ hybridization and DNA short tandem repeat analyses for peripheral blood chimerism.
Peripheral blood chimerism status was confirmed by fluorescence in situ hybridization (FISH) analysis for sex-mismatched HSCT recipients and by DNA short tandem repeat (STR) analysis for sex-matched HSCT recipients, as recommended by the American Society of Blood and Marrow Transplantation(14). Further details are in the in the Data Supplement that accompanies the online version of this article at http://www.clinchem.org/content/vol55/issue4 .
DNA isolation.
We used the QIAamp Blood Kit (Qiagen) according to the manufacturer’s spin protocol for blood and bodily fluids to extract DNA from 800 μL of plasma, 300 μL of buffy coat, and the urinary-cell pellet [resuspended in 200 μL of 1× PBS (Invitrogen)](10). For each 400 μL of fluid sample, we added 40 μL of Qiagen Protease and 400 μL of Qiagen Buffer AL, incubated at 56 °C for 10 min, and then added 400 μL cold absolute ethanol. We then transferred the mixture to a QIAamp Spin Column, centrifuged the column at 16 000g for 1 min, and washed the extraction column twice by centrifugation at 16 000g, first with Qiagen Buffer AW1 for 1 min and then with Buffer AW2 for 3 min. The DNA was eluted in 50 μL of deionized water and stored at −20 °C until analysis.
To extract DNA from the filtered cell-free urine supernatant, we mixed 15 mL of 6 mol/L guanidine thiocyanate (Sigma–Aldrich) and 1 mL of resin (Wizard Plus Minipreps DNA Purification System; Promega) with 10 mL of the processed urine(15) and incubated the mixture with gentle mixing at ambient temperature for 2 h. The resin-DNA complex was then isolated and washed on minicolumns with the wash buffer provided in the Wizard Plus Minipreps DNA Purification System. The urine DNA was then eluted in 100 μL deionized free water and stored at −20 °C until analysis.
Homogeneous MassEXTEND assay for zinc finger protein homologs.
The assay for zinc finger protein genes was designed to quantify the fractional concentration of male DNA in a mixture of male and female DNA. The assay used identical PCR primers to coamplify the genes for zinc finger protein homologs located on the X chromosome (ZFX, zinc finger protein, X-linked) and the Y chromosome (ZFY, zinc finger protein, Y-linked) in a single PCR (amplicon size, 120 bp). The amplicons were then differentiated with the homogeneous MassEXTEND assay (Sequenom) with an extension primer (5′-TCATCTGGGACTGTGCA-3′) designed to anneal at the position adjacent to a single-nucleotide site that differentiates the 2 genes. The extension primer was extended by 1 and 2 bases for ZFX and ZFY, respectively, with a selective combination of terminator nucleotides (i.e., dideoxynucleotides). The extended products (5′-TCATCTGGGACTGTGCAA-3′ for ZFX and 5′-TCATCTGGGACTGTGCAGT-3′ for ZFY) had distinct masses of 5498.6 and 5818.8 Da, respectively. The extended products were readily resolved by MALDI-TOF mass spectrometry. The ZFX signal represented total DNA, and the ZFY signal represented the male donor–derived DNA in female recipients. The fractional concentration of male DNA in the total-DNA preparation was calculated by assessing the ZFY peak height relative to that of ZFX in the mass spectrum. For further details of the PCR reaction, see the online Data Supplement.
To calculate the assay CV, we obtained genomic DNA from a healthy male volunteer and prepared artificial mixtures of 0%–100% male DNA by mixing a known concentration of male DNA with female genomic DNA. Measured and expected percentages of male DNA were compared; the mean and SD were calculated for 20 replicates.
phase 2
Real-time quantitative PCR for size analysis with LEP and SRY sequences.
Real-time quantitative PCR assays were used to quantify the LEP (leptin) and SRY (sex-determining region Y) genes in a 50-μL reaction mixture, as previously described(13). In female patients who received male stem cells, the SRY signal reflected donor-derived DNA, and the LEP signal reflected the total DNA from both the donor and recipient. For each locus, a common forward primer and different reverse primers (Integrated DNA Technologies) were used to generate 3 amplicons of different sizes. To control for urine concentration, we expressed the absolute DNA concentration in urine in genome equivalents per millimole of creatinine. For further details, see the online Data Supplement.
Combined FISH and immunofluorescence detection for the X and Y chromosomes and cytokeratin in urine samples.
We analyzed fresh unfractionated urine samples by microscopy according to the Rant–Shepherd method, as previously described(16). Urine samples with 0–15 white blood cells or red blood cells in a microscopy field produced with a ×20 objective were reported as negative for such cells. Only urine samples with negative microscopy results were used for further analysis. See the online Data Supplement for details of the combined FISH and immunofluorescence-detection protocol. In brief, Cytospin slides were prepared from urinary-cell pellets, fixed, and counterstained. A CEP X/Y DNA probe (Vysis/Abbott Molecular) was applied to each Cytospin slide according to manufacturer’s instructions. We mounted the slides with 4′,6-diamidino-2-phenylindole supplied in the CEP X/Y DNA Probe Kit and used a fluorescence microscope (Nikon Eclipse E600) to examine the slides for positive signals (orange for the centromere of chromosome X, green for the Yq12 region of chromosome Y and for CK). Two independent observers evaluated the stained slides (all areas under ×400 magnification); at least 500 cells were counted for each sample. The numbers of polymorphs and epithelial cells from the recipient and the donor were recorded. As external controls in each run, Cytospin slides were prepared from male and female urine samples with moderate numbers of polymorphs and epithelial cells (cases with inflammation).
statistical analyses
Statistical tests were performed with SigmaStat 3.0.1A software (SPSS).
Results
high percentage of donor dna in buffy coat samples from hsct recipients
The characteristics of the HSCT recipients are shown in Table S1 in the online Data Supplement. The 10 sex-matched HSCT recipients and the 12 sex-mismatched HSCT recipients were in complete remission with respect to their hematologic conditions. The analysis of chimerism status (i.e., the presence of lymphohematopoietic cells of nonhost origin(14) measured with FISH(17)(18) and DNA STR analyses as described above) revealed that 21 of the 22 patients had >99% donor lymphohematopoietic cells in the peripheral blood, fulfilling the criterion for full chimerism with complete lymphohematopoietic replacement(14). The percentages of donor DNA in buffy coat samples obtained with the mass spectrometry–based assay for the zinc finger protein genes were completely concordant with the FISH and STR results (see Table S1 in the online Data Supplement). In this mass spectrometry–based assay, the mean ratio of the ZFY signal to the ZFX signal for 20 replicates of male DNA at 50 ng per reaction was 0.46 (SD, 0.02; CV, 4.3%). Evaluation of the proportion of recovered DNA with a series of artificial mixtures containing 0%–100% male DNA indicated high correlation of the measured fractional concentration of male DNA with the expected value (r = 0.975; see Fig. S1 in the online Data Supplement).
sex-mismatched hsct recipients had very high percentages of donor-derived dna in plasma
We then used the mass spectrometry–based zinc finger protein assay to examine the percentages of male DNA in plasma samples (see Table S2 in the online Data Supplement). In female sex-mismatched HSCT recipients (n = 5), the percentage of male DNA represented the amount of donor-derived DNA, and the mean fractional concentration of male donor–derived DNA in these patients was 79.3%. In male sex-mismatched HSCT recipients, the proportion of male DNA (i.e., recipient-derived) was only 27.2%; therefore, the fractional concentration of donor-derived DNA was 72.8%. All 12 sex-mismatched HSCT recipients had very high contributions of donor-derived DNA (mean fractional concentration, 76.1%).
high percentages of donor-derived dna in urine supernatants
We then examined urine supernatants for the presence of donor-derived DNA. Sex-matched HSCT recipients were recruited as controls (see Table S2 in the online Data Supplement). Typical mass spectrometry tracings are shown in Fig. S2 in the online Data Supplement. Male sex-mismatched HSCT recipients had a mean fractional male-DNA concentration of 92.3% in urine supernatants (Fig. 1B ). Such a high percentage is expected from the lysis of the host’s (male) urinary epithelial cells shed into the urine. Interestingly, all 5 female sex-mismatched HSCT recipients had male donor–derived DNA in urine supernatants. The mean fractional concentration reached 38.3% (range, 26%–88.1%; Fig. 1A ).
Fractional concentrations of male DNA in samples of plasma and cell-free urine supernatant from 5 female (A) and 7 male (B) sex-mismatched HSCT recipients.
correlations between the amounts of male dna in urine supernatants, urinary-cell pellets, and plasma
We next examined if the amount of donor-derived male DNA in urine supernatants was correlated with the plasma value. Surprisingly, we found no significant correlation for the 5 female sex-mismatched HSCT recipients (Spearman rank order correlation, r = −0.3; P = 0.683) or for the entire group of sex-mismatched HSCT recipients (r = −0.427; P = 0.178) (Fig. 2A ). The fractional concentration of donor-derived male DNA in urine supernatants, however, was significantly correlated with that in urinary-cell pellets (P = 0.0186; Fig. 2B ).
Relationships of fractional concentrations of male DNA in urine and plasma.
(A), Cell-free urine supernatant and plasma. (B), Cell-free urine supernatant and urinary-cell pellets. No correlation was found in (A), but a positive correlation (r = 0.682, Spearman rank order correlation; P = 0.0186) was demonstrated in (B).
urine supernatants of sex-mismatched hsct recipients contained donor-derived dna fragments >350 bp that were absent in plasma
In view of the unexpected lack of correlation between plasma and urine supernatant DNA values, we proceeded to phase 2 of the study. Two gene loci were chosen to study the size of DNA fragments in plasma and urine supernatants. SRY signals represent the amount of male donor–derived DNA in female sex-mismatched HSCT recipients, and LEP signals represent both donor- and recipient-derived DNA. Because urine DNA had previously been shown to contain fragments of 150–200 bp(5), we designed primers to produce amplicons shorter than 150 bp to maximize the yield of Tr-DNA(19). We also designed primers that produced amplicons of >200 bp to study the contribution of non–Tr-DNA. We used amplicons of comparable fragment lengths (63, 107, and 377 bp for SRY and 63, 105, and 356 bp for LEP). Consistent with the results for the assay for the zinc finger protein genes (see Table S2 in the online Data Supplement), all 5 female sex-mismatched HSCT recipients had male donor–derived DNA (SRY) in both the plasma and the urine supernatants (see Table S3 in the online Data Supplement). Interestingly, 4 of these 5 patients had SRY and/or LEP fragments in urine supernatants that were larger than 350 bp. These DNA fragments were absent in the corresponding plasma samples (see Tables S3 and S4 in the online Data Supplement).
fish evidence for donor-derived cells in urine
To investigate the origin of these longer DNA fragments, we conducted cellular analyses of fresh urine samples obtained from 10 of the 12 sex-mismatched HSCT recipients. We prepared urinary-cell pellets and counted a minimum of 500 cells for each patient (mean, 929 cells; range, 510–1981 cells). In a microscopical analysis, we identified donor-derived epithelial cells and polymorphs from both donors and recipients in all 10 patients. Intriguingly, we noted that a small proportion of the donor cells in some samples had a rounded nucleus and had assumed the morphology of urinary epithelial cells.
donor-derived cells in urine carried epithelial signatures
To characterize the phenotype of these epithelial-like cells, we carried out combined FISH and immunofluorescence detection with CK as an epithelial marker. Male and female urine samples were included as controls to establish probe efficiency. Seventy-five percent of the male cells positive for 4′,6-diamidino-2-phenylindole in these control samples had positive signals for chromosomes X and Y, and 80% of the female cells had 2 chromosome X signals. No chromosome Y signal was detected in the female urine samples. For immunofluorescence detection of CK in control samples, we correlated the number of CK-positive cells with that obtained by Papanicolaou staining, a technique commonly used in conventional cytology. Seventy-two percent of the Papanicolaou-stained epithelial cells were CK positive with this protocol. Our probe efficiencies (75%, 80%, and 72% for X, Y, and CK signals, respectively) compared favorably with those reported for studies that used similar techniques(20)(21).
All 10 sex-mismatched HSCT recipients had recipient-derived CK-positive epithelial cells and donor-derived CK-negative polymorphs in their urine (Table 1 ). Remarkably, 3 patients had donor-derived CK-positive epithelial cells in their urine (Fig. 3 ). These donor-derived cells constituted 1.3%, 0.4%, and 0.4% of all the epithelial cells in these 3 patients (Table 1 ). Patients 5 and 16 had received bone marrow transplants, and patient 9 had received a peripheral blood stem cell transplant. At the time of urine collection, these 3 patients had received their transplants 3.7, 14.2, and 2.3 years before, respectively.
Combined FISH and immunofluorescence detection of CK and chromosomes X and Y in a fresh urine sample from patient 5, a male patient who had received an HSCT from a female donor.
A donor-derived CK-positive cell with an epithelial morphology (ED) was seen. Indicated are staining of the nucleus by 4′,6-diamidino-2-phenylindole (blue), the X chromosome (orange), and the Y chromosome and CK (green). Original magnification, ×400. PR, recipient-derived polymorph; PD, donor-derived polymorph.
Combined FISH and immunofluorescence detection of fresh urine samples from sex-mismatched HSCT recipients (n = 10).
| Patient No. | Polymorphs from donor, % | Epithelial cells from donor, % | Time from HSCT to urine analysis, years |
|---|---|---|---|
| 5 | 79.1 | 1.3 | 3.7 |
| 6 | 91.0 | 0.0 | 14.7 |
| 7 | 9.1 | 0.0 | 16.2 |
| 8 | 77.8 | 0.0 | 5.0 |
| 9 | 9.8 | 0.4 | 2.3 |
| 10 | 34.6 | 0.0 | 5.1 |
| 11 | 53.2 | 0.0 | 2.1 |
| 12 | 60.0 | 0.0 | 7.3 |
| 15 | 36.2 | 0.0 | 4.1 |
| 16 | 60.2 | 0.4 | 14.2 |
| Patient No. | Polymorphs from donor, % | Epithelial cells from donor, % | Time from HSCT to urine analysis, years |
|---|---|---|---|
| 5 | 79.1 | 1.3 | 3.7 |
| 6 | 91.0 | 0.0 | 14.7 |
| 7 | 9.1 | 0.0 | 16.2 |
| 8 | 77.8 | 0.0 | 5.0 |
| 9 | 9.8 | 0.4 | 2.3 |
| 10 | 34.6 | 0.0 | 5.1 |
| 11 | 53.2 | 0.0 | 2.1 |
| 12 | 60.0 | 0.0 | 7.3 |
| 15 | 36.2 | 0.0 | 4.1 |
| 16 | 60.2 | 0.4 | 14.2 |
Combined FISH and immunofluorescence detection of fresh urine samples from sex-mismatched HSCT recipients (n = 10).
| Patient No. | Polymorphs from donor, % | Epithelial cells from donor, % | Time from HSCT to urine analysis, years |
|---|---|---|---|
| 5 | 79.1 | 1.3 | 3.7 |
| 6 | 91.0 | 0.0 | 14.7 |
| 7 | 9.1 | 0.0 | 16.2 |
| 8 | 77.8 | 0.0 | 5.0 |
| 9 | 9.8 | 0.4 | 2.3 |
| 10 | 34.6 | 0.0 | 5.1 |
| 11 | 53.2 | 0.0 | 2.1 |
| 12 | 60.0 | 0.0 | 7.3 |
| 15 | 36.2 | 0.0 | 4.1 |
| 16 | 60.2 | 0.4 | 14.2 |
| Patient No. | Polymorphs from donor, % | Epithelial cells from donor, % | Time from HSCT to urine analysis, years |
|---|---|---|---|
| 5 | 79.1 | 1.3 | 3.7 |
| 6 | 91.0 | 0.0 | 14.7 |
| 7 | 9.1 | 0.0 | 16.2 |
| 8 | 77.8 | 0.0 | 5.0 |
| 9 | 9.8 | 0.4 | 2.3 |
| 10 | 34.6 | 0.0 | 5.1 |
| 11 | 53.2 | 0.0 | 2.1 |
| 12 | 60.0 | 0.0 | 7.3 |
| 15 | 36.2 | 0.0 | 4.1 |
| 16 | 60.2 | 0.4 | 14.2 |
Discussion
In this study, we have demonstrated the presence of donor-derived DNA in urine samples from sex-mismatched HSCT recipients. Our data highlight the important contribution of DNA in the supernatants from donor-derived cells. Further characterization of these cells led to the discovery of an unexpected population of donor-derived cells carrying epithelial signatures.
Recent reports that favor the transrenal hypothesis have generally based their conclusion on the urinary detection of a target previously known to be present in plasma. The detection of donor-derived DNA in the urinary tracts of the HSCT recipients we studied would have lent support to the transrenal hypothesis. In the present investigation, we had applied more stringent criteria and hypothesized that if the transrenal hypothesis were correct, then (a) the amount of donor-derived DNA in the urine supernatants would correlate with that in the plasma and (b) the sizes of donor-derived DNA fragments in the urine would be limited by the physical properties of the kidney barrier and would not be larger than their hypothetical sources in the plasma.
Our data show that both of these inferences do not hold. First, our MALDI-TOF mass spectrometry data show that the amounts of donor-derived DNA in urine supernatants do not correlate with those in plasma. Second, our size analysis shows that urine supernatants contain long DNA fragments that are absent in the plasma. Furthermore, we detect donor-derived cells by in situ hybridization.
In this HSCT model, the amounts of donor-derived DNA in urine supernatants could be affected by several factors: (a) the number of donor-derived cells present in the urine; (b) the amount of cell-free recipient DNA from urinary cells dying in situ or in the urine; and (c) the amount of donor-derived DNA in plasma, the hypothetical source of Tr-DNA. The presence of variable percentages of donor-derived white blood cells in the urine of these HSCT recipients could explain not only the lack of correlation between the donor-derived cell-free DNA in urine and plasma but also the existence of long donor-derived DNA fragments in urine supernatants. Although the results of the present study do not support or contradict the Tr-DNA concept, they do suggest that the positive evidence for Tr-DNA in the current literature may be confounded by the presence of nonhost cells in the urine. Without a detailed analysis of the cellular contribution to urine DNA, the previous reports on Tr-DNA should be interpreted with caution. It is hoped that future model systems can be developed that allow one to separately explore the phenomena of cellular and cell-free DNA transfer into the urine.
During the search for donor-derived urinary DNA, we discovered donor-derived cytokeratin-positive epithelial cells in the urine of HSCT recipients. This in vivo finding is novel, given the current literature. The observation of these cells in 3 of 10 sex-mismatched HSCT recipients suggests that this line of differentiation was not a sporadic occurrence but possibly a common developmental path taken by bone marrow–derived stem cells (BMSCs).
The results of studies with animal models and in vitro experiments have suggested the potential of BMSCs in kidney regeneration(22)(23)(24)(25)(26)(27). BMSCs tagged with green fluorescent protein were recruited into the glomerular mesangium in both physiological(28) and pathologic(23) states. In addition to mesangial cells, Kale et al. showed that murine BMSCs could differentiate into proximal tubular cells in ischemic kidneys(25). Human mesenchymal stem cells have been shown to generate metanephroi and to express podocyte- and tubular epithelial cell–specific genes in rodent whole-embryo culture(29). Human in vivo data regarding the role of extrarenal stem cells in kidney regeneration are relatively limited to the detection of recipient-derived, Y chromosome–positive renal tubular cells in female allografts transplanted into male recipients(22)(30)(31). Cytokeratin was used as a tubular cell marker by Gupta et al.(30). These reports provide indirect evidence that BMSCs may be a source of the extrarenal stem cells in humans.
Our study provides the first direct evidence that cytokeratin-producing epithelial cells can be derived extrarenally from marrow or peripheral blood stem cells. Although donor-derived cytokeratin-producing epithelial cells were detected at a modest contribution of 0.4%–1.3% in 3 of 10 patients, our data show unequivocally that this line of differentiation is possible. Moreover, the detection of these cells in HSCT recipients as long as 14.2 years after transplantation suggests that the contribution of kidney epithelial cells from hematopoietic stem cells is stable, durable, and continual.
This phenomenon has interesting biological and therapeutic implications, particularly for patients with chronic kidney diseases, for whom renal-replacement therapy is far from a cure. A larger cohort of sex-mismatched HSCT patients will be needed to confirm this observation. Examination of renal biopsies from sex-mismatched HSCT patients will help elucidate the actual contribution of extrarenal stem cells in the maintenance of renal tubular architecture. Further understanding of the mechanisms involved in the generation of these epithelial cells from extrarenal sources may advance our knowledge in regenerative medicine.
Author Contributions:All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revising the article for intellectual content; and (c) final approval of the published article.
Authors’ Disclosures of Potential Conflicts of Interest:Upon manuscript submission, all authors completed the Disclosures of Potential Conflict of Interest form. Potential conflicts of interest:
Employment or Leadership: None declared.
Consultant or Advisory Role: Y.M.D. Lo, Sequenom.
Stock Ownership: Y.M.D. Lo, Sequenom.
Honoraria: None declared.
Research Funding: Earmarked Research Grant (CUHK4436/06M) from the Hong Kong Research Grants Council. Y.M.D. Lo, Sequenom.
Expert Testimony: None declared.
Other: Y.M.D. Lo, S.S.C. Chim, N.B.Y. Tsui, and R.W.K. Chiu hold patents or have filed patent applications on aspects of circulating nucleic acid–based diagnostics.
Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.
Acknowledgments: The authors thank E.S. Lo for her kind assistance in the slide examination.
Some of these data were presented in the 3rd Trainee Presentation Session of the Hong Kong College of Pathologists, Hong Kong SAR, China, November 24, 2007.
Nonstandard abbreviations: Tr-DNA, transrenal DNA; HELLP, hemolysis, increased liver enzymes, and low platelet count; HSCT, hematopoietic stem cell transplant; FISH, fluorescence in situ hybridization; STR, short tandem repeat; CK, cytokeratin; BMSC, bone marrow–derived stem cell.
Human genes: KRAS, v-Ki-ras2 Kirsten rat sarcoma viral oncogene homolog; RASSF1A, Ras association (RalGDS/AF-6) domain family member 1A; ZFX, zinc finger protein, X-linked; ZFY, zinc finger protein, Y-linked; LEP, leptin; SRY, sex-determining region Y.
References
Zhong XY, Hahn D, Troeger C, Klemm A, Stein G, Thomson P, et al. Cell-free DNA in urine: a marker for kidney graft rejection, but not for prenatal diagnosis?.
Antin JH, Childs R, Filipovich AH, Giralt S, Mackinnon S, Spitzer T, Weisdorf D. Establishment of complete and mixed donor chimerism after allogeneic lymphohematopoietic transplantation: recommendations from a workshop at the 2001 Tandem Meetings of the International Bone Marrow Transplant Registry and the American Society of Blood and Marrow Transplantation.
Shepherd M. A revision of the microtitre tray method for urine microscopy.
Ito T, Suzuki A, Imai E, Okabe M, Hori M. Bone marrow is a reservoir of repopulating mesangial cells during glomerular remodeling.
Imasawa T, Utsunomiya Y, Kawamura T, Zhong Y, Nagasawa R, Okabe M, et al. The potential of bone marrow-derived cells to differentiate to glomerular mesangial cells.
