Transplantation

Kidney Function Restoration Program

Curing Kidney Problems Forever

Get Instant Access

The most commonly used technique to track BMDC and study their role of BMDC after renal injury is transplantation of bone marrow from rodents that are genetically distinguishable from the recipient, into bone marrow-ablated recipients. After bone marrow transplantation, the transplanted bone marrow cells stably reconstitute the recipient's bone marrow compartment. In practice, the efficiency of bone marrow ablation and reconstitution with donor bone marrow is never 100% and a certain degree of underestimation of, for example the actual number of renal infiltrated BMDC, is always present. Another issue to consider is that bone marrow transplantation experiments with newly introduced genetically modified BMDC can elicit immunological rejection. However, the possibility of an interfering immune response has not been studied in bone marrow transplantation experiments described in this article.

Genetic "Tracking Devices"

Y Chromosome Detection

Male to female bone marrow transplantation offers a reliable, but laborious approach for tracking BMDC in the kidney. Several studies describe detection of male BMDC using Y-chromosome in situ hybridization (Y-ISH) to determine the bone marrow origin of cells engrafting the female kidney [2;4;9-12]. Y-chromosome detection is also used to study extra-renal, host-derived cells in tissue biopsy or autopsy material after sex-mismatched kidney transplantation [13-15], or to study renal infiltrated BMDC in patients after sex-mismatched bone marrow transplantation [16]. Thus, beside the use in animals after sex-mismatched bone marrow transplantation, Y-chromosome detection is used for recipient-derived cell detection in humans, in which the introduction of transgene-expressing cells, by for example bone marrow transplantation, is not possible.

The ISH method is laborious and may yield variable results, explaining the contradicting results of detection [2;4;9;10;12]. In a study by Duffield and colleagues [9], stating that post-ischemic tubular epithelial restoration occurs independently of BMDC, confocal laser fluorescent microscopy was used to show that Y+ tubular cells often were artifacts. These artifacts were due to leukocytes overlying renal tubular structures, intratubular monocytes or nonspecific aggregates of fluorescent probe. This also suggests that the Y-chromosome positive tubular cells detected in previous studies in female mice following transplantation of bone marrow from male donor mice could be artifacts of imaging.

A drawback of this reporter system is that ISH is necessary to detect Y chromosome-positive cells in tissues. This technique time consuming and requires extensive pre-treatment of the sections, which affects tissue morphology and epitopes and therefore reduces the possibility to perform double-stainings for further characterization of the engrafted cells. Moreover, the method of Y-ISH for detection of BMDC can lead to an underestimation of Y-chromosome presence by distribution of the nucleus over multiple thin tissue sections, and loss of the Y chromosome for detection in some of the sections. Because of these bottlenecks, detection of the Y-chromosome in male tissue sections is never 100%, as it should be. Therefore, the exact number of Y chromosome-positive cells can only be estimated using a correction factor, as was done in a study by Direkze [17]. In this study radiation-induced injury was shown to elicit differentiation of BMDC to myofibroblasts in multiple organs, including the kidney. Using Y-ISH, the authors showed that only 50% of all male cells were

Y chromosome-positive, which should have been 100%. Therefore, the observed number of renal Y chromosome-positive BMDC in the female kidney was corrected by dividing by 0.5 to estimate the exact number of renal BMDC. However, this approach is not very accurate since this is only an estimation and not the exact number of BMDC. Together with the fact that bone marrow transplantation is never 100% efficient, this estimation can lead to incorrect interpretation of data. Another limitation of sex-mismatched BMDC tracking, especially in the clinical setting, is that female to male detection is more difficult and unreliable, thereby excluding female to male transplantations for detection of recipient-derived cells. Therefore, the use of sex-mismatched bone marrow transplantation for tracking of BMDC can be regarded as a limited and insensitive method.

MHC and ABO Blood Group Antigen Detection

Although all reporter markers that are incorporated in the genome can be detected using in situ hybridization, other, less complicated methods for detection of recipient-derived cells in human tissue biopsies or post-mortem tissue are possible, e.g. MHC or ABO blood group antigen detection by immunohistochemistry [14]. Besides in patient studies, this strategy was also applied in rat studies, e.g. by Rookmaaker et al. [6], who used a rat allogenic BM transplant model in BN rats with WR rats as bone marrow donors to generate rat BM chimeras. In the model of anti-Thy-1.1-glomerulonephritis, BMDC were traced using a donor-specific major histocompatibility complex class-I monoclonal antibody and were found to differentiate towards glomerular endothelial and mesangial cells [6]. Although it was not reported in this study, possible immune rejection of MHC mismatched bone marrow and the necessity for immunosuppressive therapy can be a limiting factor of this detection method.

Transgenic Reporters

Transgenic Reporters under the Control of Ubiquitous Promoters

Tracing of BMDC in the kidney using reporter genes relies on the expression of the reporter gene (i.e. the transgene) by the BMDC. Expression of the reporter gene at all times can only be accomplished if the gene is driven by a promoter of a ubiquitously expressed gene, such as mouse metallothionein [18;19], P-actin from human [20], rat [21] and chicken [22], ubiquitin [23], simian virus 40 [24], cytomegalovirus immediate-early [25] or ROSA 26 [26].

ff-Galactosidase

Bacterial f-galactosidase-transgenic mice expressing the E.coli-derived LacZ gene [27], are frequently used as a reporter for tracing BMDC. The presence of this reporter gene can be visualized by enzymatic detection of the bacterial f-galactosidase by X-gal staining or by immunohistochemical detection with an antibody against bacterial P-galactosidase. In the enzymatic detection, X-gal is converted by P-galactosidase to a blue reaction product which precipitates in situ. P-Galactosidase is among the most sensitive of reporter enzymes because only a few molecules of this enzyme readily convert X-gal in amounts that are detectable by light microscopy. An advantage of P-galactosidase is, that also soluble color substrates such as ONPG (o-nitrophenyl-P-D-galactopyranoside) exist that are employed to determine relative or total amount of reporter protein in tissue extracts.

Using this technique, previous studies have reported large numbers of X-gal+ tubular cells after ischemic renal injury [1;2], thereby suggesting that BMDC play an important role in repair of renal injury. However, many mammalian tissues, including the kidney [28;29], contain endogenous P-galactosidase, an enzyme important for the enzymatic digestion of glycolipids [30;31]. This mammalian enzyme has an acidic pH optimum [32;33], whereas the bacterial P-galactosidase enzyme has a neutral pH optimum [34]. In most published protocols, a weak buffer such as phosphate-buffered saline (PBS) was used, which may have become acidic during the exposure of fixed tissues to X-gal. Moreover, after ischemia the kidney becomes acidic due to disturbed pH regulation. At low pH, endogenous P-galactosidase is detected and, thereby, false positive cells. Several studies indeed showed that, using the enzymatic detection method of P-galactosidase at low pH (6.5), endogenous P-galactosidase could not be distinguished from the bacterial P-galactosidase in injured kidneys [9][35]. Therefore, X-gal+ tubular cells reported in previous studies [1;2] may have been detected as a result of increased intrinsic P-galactosidase activity, rather than the presence of P-galactosidase-positive bone marrow cells. Another disturbance in the detection of P-galactosidase is the presence of senescence-associated P-galactosidase (SA-P-gal), which is defined as P-galactosidase activity detectable at pH 6.0, in senescent cells [36].

Despite these detection problems, P-galactosidase can still be used as a reliable reporter, as long as bacterial P-galactosidase is clearly distinguished from endogenous mammalian P-galactosidase. This can be achieved by a simple modification of the X-gal method, raising the pH of the X-gal solution to weakly alkaline pH [29], or by using a commercially produced P-gal staining set which is designed to minimize staining from mammalian P-galactosidase [9]. Another method is based on immunolabeling with anti-bacterial P-galactosidase antibody instead of enzymatic detection of the reporter by X-gal staining. The study of Duffield [9] showed that, using anti-P-galactosidase, it was possible to discriminate between endogenous mammalian P-galactosidase activity and that resulting from LacZ gene expression. Therefore, P-galactosidase can be reliably detected without interference of endogenous mammalian P-galactosidase.

Enhanced Green Fluorescent Protein (EGFP)

Green fluorescent protein (GFP) is responsible for the green bioluminescence of the jellyfish Aequorea Victoria. Using this protein, transgenic mouse lines were generated, in which all tissues emitted green light under excitation. However, the first generation GFP transgenic mice proved to be unsuitable for use in experimental renal disease models, since GFP was, in the healthy animal, not expressed in all renal components [7; 37]. If, in the setting of GFP+ bone marrow transplantation, BMDC regulate GFP in a similar way as kidney cells and BMDC differentiate to kidney cells, this would result in an underestimation of GFP expressing, bone marrow-derived, kidney cells. Therefore, mutants of the first generation of GFP transgenic mice were constructed that had about 35-fold brighter fluorescence, termed 'enhanced' GFP (EGFP) [37]. These mutants were the result of double amino acid substitutions in the wild-type GFP cDNA construct, placed under the control of the same chicken P-actin promoter and a cytomegalovirus enhancer as in the first generation of GFP transgenes. EGFP transgenic mice do express EGFP in all renal components, as well as in other tissues, with the exception of erythrocytes, hair [38] and some leukocytes (personal observation).

Chimeric mice reconstituted with EGFP bone marrow are commonly and successfully used to trace bone marrow-derived cells in renal disease models [39-43]. The advantage of EGFP as a reporter is that introduction of a substrate is not required, unlike other commonly used reporters such as P-galactosidase and alkaline phosphatase, allowing to monitor the presence of EGFP by sole illumination of tissue sections or living cells [38]. Moreover, EGFP is not disturbed by expression of endogenous EGFP and because the excitation optimum for EGFP is close to 488 nm, the transgenic cells can also be analyzed by flow cytometry. A drawback of EGFP is that in tissue sections EGFP is often too weak for direct fluorescence microscopy, making antibody labeling necessary for detection (personal observation, confirmed by other researchers). Moreover, the utility of EGFP as a reporter in renal disease models is limited by the fact that the kidney possesses intensive auto-fluorescence, which complicates detection of EGFP+ cells, unless confocal microscopy is used. This problem can also be overcome by using one of the spectral variants of GFP, emitting blue, cyan or yellow light [37;44;45]. Moreover, these spectral variants of GFP can be very useful for achieving in vivo double-labeling.

Besides EGFP transgenic mice, rats expressing EGFP ubiquitously were generated. EGFP transgenic rats were originally established using the same construct and technique described for the production of EGFP transgenic mice [38]. Rats are, in comparison to mice, preferable in certain renal disease models, such as experimental kidney transplantation (easier microsurgical procedure) and anti-Thy1 antibody-mediated glomerulonephritis model (Thy1 is a well-established rat-specific mesangial marker, EGFP rat BM chimeras with Thy1 nephritis are described in references [5;46;47]). Therefore, the EGFP transgenic rat is an important tool for studying BMDC in renal disease.

Several characteristics of EGFP, its reliable detection, possibility of detection without introduction of a substrate in living cells and the availability of spectral variants of EGFP to avoid green autofluorescence or to simultaneously label multiple cell types, make this reporter an attractive option for use in BMDC tracking in renal disease models.

Human Placental Alkaline Phosphatase (hPAP)

Alkaline phosphatase (AP) dephosphorylates many types of phosphorylated molecules, for example nucleotides and proteins. Besides this functional property, which is extensively used in molecular biology, AP is used as a label in enzyme immunoassays. In humans, AP is present in all tissues throughout the body, but is particularly concentrated in liver, bile duct, kidney, bone and placenta. The advantage of human placental alkaline phosphatase (hPAP) is its heat stability, which allows to distinguish the placental form from other forms of AP [48;49].

Transgenic rats have been generated in which hPAP is placed under control of the ubiquitously active ROSA26 gene promoter [26]. The hPAP transgenic rat shows ubiquitous expression of hPAP in the kidney (Figure 1A) and is therefore suitable for BM transplantation experiments and subsequent tracking of BMDC in renal disease models.

We have used hPAP as a marker molecule to track BMDC in the post-ischemic rat kidney and showed BMDC differentiation towards tubular epithelial cells and myofibroblasts [50;51]. To allow for renal BMDC tracking, we reconstituted lethally irradiated F344 rats with ROSA26-hPAP transgenic bone marrow cells and subsequently studied infiltrating BMD (hPAP+) cells in the kidney after unilateral ischemia/reperfusion injury. The heat-stability of the hPAP enzyme allowed reliable detection of hPAP+ cells, without interference of endogenous alkaline phosphatase, which is abundantly present in the kidney. In our model, we showed that heat-inactivation of endogenous alkaline phosphatase resulted in complete absence of substrate conversion by this enzyme, without destroying the reactivity of hPAP (Figure 1B) [50;51].

Detection of hPAP+ BMDC in renal tissue sections by (immuno)histochemical staining was confirmed by BMDC labeling and fluorescence-activated cell sorting (Figure 2). It has also been described that the expression of hPAP is stable in isolated cells in culture [52].

Bmdc Culture

Figure 1. Renal expression of ROSA26-hPAP in transgenic rats and bone marrow chimeras. Expression of the ROSA26-hPAP gene was assessed by enzymatic hPAP staining on kidney sections. hPAP is ubiquitously expressed in the kidney of a healthy ROSA26-hPAP transgenic rat (A). ROSA26-hPAP expression can be easily detected on renal infiltrating hPAP+ BMDC (black) in ROSA26-hPAP bone marrow chimeric rats 7 days after induction of ischemia (B). No interference of endogenous AP was observed. Lens magnification 200x (B) and 400x (A).

Figure 1. Renal expression of ROSA26-hPAP in transgenic rats and bone marrow chimeras. Expression of the ROSA26-hPAP gene was assessed by enzymatic hPAP staining on kidney sections. hPAP is ubiquitously expressed in the kidney of a healthy ROSA26-hPAP transgenic rat (A). ROSA26-hPAP expression can be easily detected on renal infiltrating hPAP+ BMDC (black) in ROSA26-hPAP bone marrow chimeric rats 7 days after induction of ischemia (B). No interference of endogenous AP was observed. Lens magnification 200x (B) and 400x (A).

10° 101 102 103 10" FITC

Figure 2. hPAP expression determined by FACS. ROSA26-hPAP transgenic bone marrow cells were detected by flow cytometry after isolation and labeling with anti-hPAP and a FITC labeled conjugate. The graph shows FITC expression in bone marrow cells of a non-transgenic F344 rat (left, transparent), and of a ROSA26-hPAP rat (right, red).

10° 101 102 103 10" FITC

Figure 2. hPAP expression determined by FACS. ROSA26-hPAP transgenic bone marrow cells were detected by flow cytometry after isolation and labeling with anti-hPAP and a FITC labeled conjugate. The graph shows FITC expression in bone marrow cells of a non-transgenic F344 rat (left, transparent), and of a ROSA26-hPAP rat (right, red).

The stable and ubiquitous expression of hPAP, its applicability in rat models and simple detection methods, make the ROSA26-hPAP transgene a reliable reporter for studies on the fate of BMDC after renal injury.

Firefly Luciferase

Transgenic mice ubiquitously expressing luciferase from the North-American firefly (Photinus pyralis) have been generated, in which the firefly luciferase gene is controlled by promoter and enhancer elements of the human cytomegalovirus major immediate early gene [53]. Firefly luciferase, an enzyme that causes light emission from yellow to green wavelengths in the presence of a substrate (luciferin), oxygen, ATP and magnesium [54], is the most commonly used bioluminescent reporter in biomedical research. The fast rate of firefly luciferase enzyme turnover in the presence of the substrate luciferin and its short halflife allows for real-time measurements, because firefly luciferase does not accumulate intracellularly to the extent of other reporters [55]. Moreover, luciferase expression has been shown to be stable in isolated cells in culture [54].

The greatest advantage of the use of the bioluminescent firefly luciferase gene as a reporter protein is that the internal biological light source provided by luciferase can penetrate relatively easy through tissues, allowing in vivo detection using in vivo imaging techniques. Therefore, luciferase-produced bioluminescence can be non-invasively and repetitively measured in real-time in the same animal, thereby reducing the interference of animal-to-animal variation and requiring fewer animals per study [56].

A short-coming of the currently available techniques is that it is not possible to accomplish the imaging of real-time luciferase expression within individual cells within living organisms [57]. In vivo imaging can provide information on the renal localization of luciferase, but cellular localization must be determined, similar to other reporters, in (postmortem) tissue sections or cell lysates. Nevertheless, sensitive noninvasive imaging of firefly luciferase gene expression makes this reporter suitable for studying BMDC recruitment to the circulation and homing to the injured kidney in a living organism.

Despite these advantages, bone marrow transplantation of ubiquitous luciferase-expressing bone marrow to study renal infiltrating BMDC has not been described. However, the use of a murine mesenchymal stem cell line transfected with a retroviral construct encoding firefly luciferase to study homing of these cells to the ischemic kidney was recently described (abstract by Kielstein J.T. et al. J. Am. Soc. Nephrol. 2007; 17: 527A). Detection of selectively luciferase-expressing BMDC was reported in the kidney [35] and will be discussed in the section on transgenic reporters under control of tissue or cell type-specific promoters. Moreover, the recently generated firefly luciferase/EGFP double transgenic mouse [58] will be very useful in studies on BMDC fate in the renal disease.

Transgenic Reporters under the Control of Tissue or Cell Type-Specific Promoters

Phenotypical changes or functional properties of renal infiltrating BMDC are mostly studied by combined immunohistochemistry detecting the reporter marker in conjunction with a cell type-specific marker. Another option is to perform bone marrow transplantation with transgenic bone marrow in which the reporter gene is driven by a cell type-specific promoter. Since differentiation of BMDC to that specific cell type will elicit activation of the promoter and expression of the reporter marker, this allows, in addition to mere localization of the BMDC, to evaluate potential differentiation of the BMDC to a certain cell type in a more reliable way. A problem that can be encountered when using these tissue or cell type-specific promoters is that activation of the promoter and thus expression of the reporter is not as specifically as it should be. However, when expression is cell-type-specific within the organ of interest, the reporter can be utilized. For example, in the first version of the GFP expressing 'green' mice, GFP was expressed in the renal podocytes, skeletal muscle, pancreas and heart [37]. Although GFP expression was not exclusively observed in renal podocytes, these mice could be used to study differentiation of BMDC to podocytes [7;59].

Epithelial-Specific Expression of EGFP

Lin et al. [42] crossed two mouse strains to accomplish conditional tubular epithelial specific expression of EGFP to study the contribution of intra-renal and BMDC to post-ischemic tubular regeneration. The first mouse strain, Z/EG, is a double reporter mouse. The first reporter, lacZ, is linked to a ubiquitous promoter and is flanked by two loxP sites. The second reporter, EGFP, resides further downstream. The EGFP gene will only be activated in the presence of cre recombinase, resulting in recombination between the loxP sites and subsequent deletion of the lacZ sequence. Once activated, EGFP expression is irreversible and inheritable, irrespective of the continued presence of cre recombinase.

The second transgenic mouse strain, creksp, expresses cre recombinase under the control of the renal tubular epithelial-specific Ksp-cadherin promoter. Crosses between these two mouse strains result in the creksp;Z/EG transgenic mice which specifically and permanently express EGFP in mature tubular epithelial cells [42]. In creksp;Z/EG bone marrow chimeric mice, differentiation of BMDC to tubular epithelial cells would be visible by EGFP expression. However, the creksp;Z/EG transgenic mice that were used in this study showed a mosaic expression pattern of EGFP, due to inefficient cre/loxP recombination and non-ubiquitous expression of the Z/EG promoter. Therefore BMDC fate could not be determined reliably using this reporter marker [42].

Fibroblast-Specific Expression of EGFP

To determine the source of renal interstitial fibroblasts in renal fibrosis, Iwano et al. [60] used FSP1.GFP bone marrow chimeras. In these chimeras, renal infiltrating BMDC express GFP under control of the FSP1 (fibroblast specific protein 1) promotor, and therefore, express GFP upon conversion to a fibroblast phenotype. Although Iwano demonstrated differentiation of BMDC to FSP-1+ cells after unilateral ureter obstruction (UUO), the question remains if FSP-1 is specifically expressed by fibroblasts.

The use of transgenic mice constitutively expressing a reporter molecule under the control of an endothelial-specific promoter for bone marrow transplantation studies has, to the best of our knowledge, not been described in renal disease models. In cardiovascular research however, the use of bone marrow chimeras in which P-galactosidase expression is transcriptionally regulated by endothelial-specific promotors Flk-1 or Tie-2 has been proven suitable to study BMDC differentiation to an endothelial phenotype [61]. Moreover, the use of transgenic mice in which a reporter gene is placed under the control of a podocyte-specific promoter, such as nephrin or podocin, will also be a useful tool to extend our knowledge on BMDC differentiation in renal injury models.

Pro-Collagen 1A2-Specific Expression of Luciferase/B-galactosidase

The use of reporters driven by cell-type specific promoters for BMDC detection can give information about differentiation fates of BMDC. However, another important question is whether BMDC are functional in their differentiated state. This question was addressed in a study by Roufosse [35], using the UUO model of renal fibrosis to study the functional contribution of BMDC to fibrosis by determination of their capacity to produce collagen. To this end, a transgenic mouse was generated that expressed both luciferase and P-galactosidase reporter genes under the control of a promoter and enhancer element of the gene encoding pro-COL1A2 (coding for the a2 chain of the pro-collagen type 1). Roufosse demonstrated the unreliability of P-galactosidase in this model (see also P-galactosidase section). However, detection of pro-collagen 1 could still be accomplished by luciferase. The presence of luciferase was determined by measurement of luminescence or luciferase protein, or by in situ hybridization for luciferase mRNA, the latter allowing the authors to determine the exact location of luciferase activity and co-expression with other markers [35].

In the study of Roufosse, in vivo imaging techniques were not used to determine luciferase activity. When performed on living animals in vivo bioluminescence imaging would have provided information on renal luciferase activity and would have allowed for non-invasive tracing of luciferase activity in time in the same animal.

Stability of Transgene Expression

It may occur that a transgene is not expressed, that expression disappears in subsequent generations of the transgenic rodent, despite the presence of the transgene in the genome, or that expression is or becomes variable in different tissues. This phenomenon can be caused by gene silencing, which causes the loss of transcription of the particular gene [62]. Gene silencing can be the result of DNA methylation and/or histon deacetylation, causing alterations in chromatin structure by as yet unknown but cell-type restricted mechanisms. Both cause the shutting-off of gene transcription [63].

Gene silencing can also take place on a post-transcriptional level [62]. This occurs when mRNA of the transgene is degraded or blocked prior to translation e.g. by microRNA. Alternatively, the stability of transgenic protein may be affected, causing a high degradation rate, e.g. by increased ubiquitination.

The possibility of gene silencing in transgenic bone marrow transplantation models is often not considered and is likely to be an underestimated problem.

Was this article helpful?

0 0

Post a comment