Guanosine supplementation reduces apoptosis and protects renal function in the setting of ischemic injury (original) (raw)
Effect of intraperitoneal guanosine on plasma guanosine levels. We first examined the time course of changes in plasma guanosine levels following an intraperitoneal guanosine administration in mice subjected to bilateral ischemia/reperfusion. As shown in Figure 1, there was a significant increase in plasma guanosine that peaked 2 hours after guanosine administration. Three to 7 hours after the intraperitoneal injection, levels were significantly lower than those at 2 hours but still slightly higher than baseline. Thus, guanosine was administered 2 hours before surgery in all subsequent experiments except when stated otherwise.
Effects of intraperitoneal guanosine on plasma guanosine levels. Animals received 30 mg/kg guanosine intraperitoneally 2 hours before bilateral renal ischemia/reperfusion. Blood was collected at various time points and plasma guanosine levels determined by HPLC as detailed in Methods. Values are means ± SE (n = 3).
Effects of intraperitoneal guanosine on renal nucleotide levels during reperfusion. We have shown previously that renal GTP and ATP levels fall to less than 10% of control during renal artery clamp exceeding 20 minutes (12). After 6 to 8 hours of reperfusion, GTP and ATP normalized to preischemia levels. Furthermore, the decline in GTP levels during ischemia could not be prevented by guanosine supplementation (data not shown). Therefore, we investigated the effects of guanosine on GTP at an early time point during reperfusion. In the control kidneys, ATP was 18 ± 3 pmol/μg protein and GTP was 6 ± 2 pmol/μg protein. As shown in Figure 2, renal GTP levels were above 92% of control at 1 hour of reperfusion in the guanosine/ischemia mice (G/I) compared with 56% in normal saline/ischemia mice (NS/I). Furthermore, guanosine supplementation did not affect ATP levels significantly (44% in G/I versus 36% in NS/I). Thus, guanosine selectively increased GTP levels to near baseline values after only 1 hour of reperfusion. Guanosine had no significant effects on GTP or ATP in the sham kidneys. Similar results were obtained in rats: in control sham kidneys, ATP was 28 ± 4 pmol/μg protein and GTP 8 ± 2 pmol/μg protein. Guanosine increased GTP after 1 hour reperfusion to 107% of control compared with 65% in the NS/I (P < 0.01). ATP levels were 49% in the G/I group versus 58% in the NS/I (P = NS).
Effects of guanosine on renal nucleotide levels during reperfusion. Values are means ± SE. Mice received normal saline (NS, n = 5) or guanosine (G, n = 5) 30 mg/kg intraperitoneally, 2 hours before 32-minute bilateral renal ischemia (I) or sham surgery. Renal tissues were obtained from each kidney 1 hour after surgery and processed for nucleotide determination (*P < 0.02, when GTP in NS/I was compared with GTP in G/I).
Effects of guanosine and enhanced GTP recovery on renal histology at 24 hours. Histological sections of kidneys from NS-treated mice (NS/I group) removed 24 hours after ischemia showed extensive tubular necrosis, tubular dilation, and cast formation in the outer medulla (Figure 3). Sections from the guanosine-treated mice (G/I group) had similar histological features. Quantitation of these changes showed no significant differences between these groups (Table 1). In addition, staining with phalloidin demonstrated comparable disruption and thinning of apical actin rim between the two groups. The sham surgery group showed a well-formed, thick, and continuous actin rim (Figure 3). Thus, guanosine and enhanced GTP recovery after ischemia did not have a significant effect on the morphologic consequences of renal ischemia in this mouse model. The histology and actin architecture of the sham group that received guanosine were normal and indistinguishable from the NS-treated sham group.
Effects of guanosine and enhanced GTP recovery on renal histology at 24 hours. Mice received normal saline (NS) or guanosine (G) 30 mg/kg intraperitoneally 2 hours before renal ischemia (I) or sham surgery. Representative H&E-stained sections from sham, NS/I, and G/I 24 hours after surgery are shown. Compared with the sham group, sections from both NS/I and G/I show patchy tubular dilatation, necrosis, and extensive cast formation. DAPI (blue) and FITC-phalloidin (green) staining show significant loss and/or disruption of the apical actin rim in both NS/I and G/I as compared with sham.
Histological scoring of mouse kidney sections after ischemia and administration of guanosine or saline
Similar changes were seen in rat kidneys. The majority of sections of outer medulla from both the NS/I and G/I rats showed evidence of tubular necrosis in >75% of tubules examined. The apical actin rim also showed comparable thinning and disruption in both groups (data not shown).
Effects of guanosine and enhanced GTP recovery on renal cell apoptosis at 24 hours. At 24 hours after ischemia, guanosine had a remarkable effect on the occurrence of apoptosis in renal tissues. As shown in Figure 4, kidneys from mice in the NS/I group showed extensive TUNEL-positive staining (Figure 4c), predominantly in the outer medulla. In contrast, kidneys from mice in the G/I group had very little TUNEL-positive staining (Figure 4d). Furthermore, because the TUNEL stain can occasionally be nonspecific, we examined nuclear morphology with DAPI staining at higher magnification (×60). In Figure 4e, nuclei from the NS/I group showed heavy staining and condensation of chromatin. The inset in Figure 4e shows ×120 magnification of a single nucleus. Fragmentation of chromatin into several smaller bodies is seen, a hallmark of apoptosis. Figure 4f is a ×60 magnification of nuclei from G/I mice. They do not show condensation and stain very faintly with DAPI.
Effects of guanosine and enhanced GTP recovery on renal cell apoptosis at 24 hours. Mice were treated as detailed in legend of Figure 2. Kidney sections were obtained at 24 hours and costained with TUNEL and DAPI. (a) A negative control (NC; TdT omitted) and (b) a DNase-treated positive control (PC) for TUNEL. All TUNEL-positive nuclei had bright green fluorescence. (c) A representative field showing the medulla from NS/I group. Extensive TUNEL-positive staining is seen and is localized primarily to the tubular epithelium (dark green autofluorescence). (d) A representative field from the medulla of the G/I group showing lack of TUNEL-positive staining. (e) A magnification of TUNEL-positive nuclei from NS/I mice, with only the DAPI channel turned on. ×60. Nuclei are condensed, dysmorphic, and show heavy staining. Inset shows a single nucleus. ×120. Typical apoptotic fragmentation of chromatin into four small bodies is seen. (f) A magnification of TUNEL-negative nuclei from G/I mice. ×60. They show lack of condensation and faint DAPI staining.
In addition, we confirmed the presence of apoptotic nuclei in the NS/I group by high-magnification (×100) light microscopy of H&E-stained tissues. As shown in Figure 5, a and b, very few if any apoptotic nuclei were observed in the G/I group. In contrast, multiple apoptotic nuclei with classic condensation and fragmentation can be seen in the NS/I group (Figure 5, c–f). Electron microscopic examination of tissues from the NS/I group clearly identified two types of nuclei in addition to normal ones: apoptotic nuclei (Figure 5g) and necrotic ones (Figure 5h). Thus, the apoptosis found by TUNEL was confirmed by fluorescence, light, and electron microscopic criteria.
Morphologic characterization of apoptosis by light and electron microscopy. Mice received normal saline (NS) or guanosine (G) 30 mg/kg intraperitoneally 2 hours before renal ischemia or sham surgery. Representative H&E-stained sections 24 hours after ischemia from the guanosine-treated group are shown (a and b). Nuclei with predominantly necrotic features are observed. They have degraded chromatin without formation of discrete, membrane-bound fragments. (c, d, e, and f) From the NS-treated group. Classic apoptotic bodies with dense condensation and fragmentation of chromatin into discrete fragments are seen (arrows). (g and h) Electron microscopic photomicrographs of sections from the normal saline group show clear apoptotic (g) and necrotic (h) morphologies.
The distribution of TUNEL-positive staining between outer medulla and cortex in the four groups is shown in Figure 6. The highest TUNEL-positive staining was in the medulla of the NS/I group and represented 28% of all nuclei stained in this region. In the G/I group, the number of TUNEL-positive nuclei was reduced significantly to 3%. There was less TUNEL positivity in the cortex of the NS/I group (8%), but this was still reduced in the G/I group to less than 2%. Similar results were observed in rats. The NS/I group showed 33% TUNEL-positive nuclei in the medulla. This was reduced in the G/I group to less than 5%. There was minimal cortical TUNEL positivity (<2%) in all groups.
Quantitative evaluation of TUNEL-positive nuclei in cortex and medulla. Values are means ± SE and represent number of TUNEL-positive nuclei expressed as percentage of total number of nuclei. Mice were treated as detailed in the legend of Figure 2. Kidney sections were obtained at 24 hours and processed for TUNEL and DAPI staining as detailed in Methods. Number of fields counted: 20 for G/I, 35 for NS/I, 8 for G/sham, and 8 for NS/sham. (*P < 0.01 when medullary TUNEL-positive nuclei in NS/I were compared to all other groups).
Distribution of TUNEL-positive nuclei among various medullary structures. We next examined the distribution of TUNEL-positive nuclei in the NS/I group among various medullary segments of the nephron. As detailed in Methods, this was done by overlay of TUNEL-stained sections with directly contiguous sections stained with phalloidin and anti-THP Ab (17). As shown in Figure 7, TUNEL-positive nuclei (Figure 7a) and proximal tubules, distal tubules, as well as loop segments (Figure 7b), could easily be overlaid to yield a composite picture (Figure 7c) showing the TUNEL distribution among various segments.
Distribution of TUNEL-positive nuclei among various nephronal segments. (a) A 5-μm thick section from NS/I mice stained with TUNEL and DAPI. The TUNEL positive nuclei were pseudocolored white instead of green to allow clear distinction from the green FITC-phalloidin. (b) A 5-μm section immediately contiguous to the one shown in a is stained with the actin marker FITC-phalloidin (bright green) to identify proximal tubules (P). Distal tubules (D) show minimal or no staining. The section was costained with anti-THP Ab and a Cy5-conjugated secondary Ab (red). This is predominantly a marker for Henle’s loop segments (LH). (c) An overlay of a and b allows the localization of TUNEL-positive nuclei to various tubular segments. (d) A quantitative estimate of the distribution of TUNEL-positive nuclei normalized to total number of DAPI-positive nuclei for each particular tubular segment or lumen. Other refers to interstitial or vascular TUNEL-positive nuclei.
Figure 7d shows a quantitative estimate of the TUNEL distribution obtained by counting ten fields similar to that shown in Figure 7c. On average, there was no difference between proximal, distal, and loop (THP+) segments, with each having 26–29% of their nuclei TUNEL positive. Of note is the presence of about 44% TUNEL-positive nuclei in tubular lumens (normalized to all luminal nuclei). Few TUNEL-positive nuclei were noted in interstitial and vascular structures.
Effects of guanosine and enhanced renal GTP levels on renal function at 24 hours. The results presented above show that guanosine and increased GTP levels did not alter the histological findings at 24 hours but did reduce significantly medullary apoptosis. To test if this reduction in apoptosis had any impact on function, we measured serum creatinine in all groups at 24 hours. As shown in Figure 8, the NS/I group had a mean creatinine of 2.1 ± 0.3 mg/dl. This was significantly reduced to 0.76 ± 0.12 mg/dl in the G/I group. In fact, the creatinine in the G/I group was comparable to that of the two sham groups. Guanosine and enhanced GTP levels had a similar protective effect in the rat: at 24 hours, creatinine was 3.2 ± 0.3 mg/dl in the NS/I group compared with 1.4 ± 0.2 mg/dl in the G/I group (n = 3, P < 0.01).
Effects of guanosine and enhanced renal GTP levels on renal function at 24 hours. Values are means ± SE. Creatinine levels at 24 hours are shown after administration of normal saline (NS) or guanosine (G) 2 hours before sham surgery or bilateral renal ischemia. Mean creatinine was significantly higher in the NS/I mice than in the other groups. (*P < 0.01, n = 6 for NS/I and G/I groups).
We further examined the window of efficacy of guanosine. We found that guanosine was most protective when given 2 hours before ischemia. Mean creatinine was no different than in the control postischemic group when guanosine was given 4 or 8 hours before ischemia or 4 or 8 hours after ischemia. When guanosine was given at the time of ischemia or 2 hours after ischemia, the improvement in mean creatinine did not reach statistical significance. Peak guanosine levels occur 2 to 3 hours after an intraperitoneal injection (Figure 1), coinciding well with the enhanced GTP levels observed 1 hour after ischemia (3 hours after guanosine administration).
Effects of guanosine on nucleotide content, cell viability, and apoptosis after chemical anoxia in LLC-PK1 cells. Because of the cellular complexity of kidney tissues, we investigated if the reported changes in nucleotides and cell death can be reproduced in tubular cells in culture. For this purpose, we used the well-characterized LLC-PK1 proximal tubular cell line to model the effects of guanosine. Resting ATP and GTP levels were 36 ± 4 pmol/μg and 8 ± 2 pmol/μg protein, respectively (n = 20). The addition of up to 200 μM guanosine for 2 hours increased ATP to 40 ± 3 pmol/μg and GTP to 9 ± 2 pmol/μg protein (n = 5). Both changes were not statistically significant.
Next we investigated the effects of guanosine on nucleotide recovery after antimycin A–induced chemical anoxia. As shown in Figure 9a, 45 minutes’ exposure to 0.1 μM antimycin A in depleted media reduced ATP to 5% ± 2% and GTP to 8% ± 2% of control levels. After 2 hours’ recovery in regular media, ATP was 33% ± 3% and GTP was 49% ± 4% of control values. When 200 μM guanosine was included in the recovery medium for 2 hours, ATP was 35% ± 2% and GTP was 88% ± 5% of control (Figure 9b). Thus, guanosine enhanced early selective GTP recovery after chemical anoxia.
Effects of guanosine on nucleotide levels and apoptosis after chemical anoxia recovery in LLC-PK1 cells. Values are means ± SE. (a) LLC-PK1 cells were treated for 45 minutes with 0.1 μM antimycin A in depleted media followed by recovery for 2 hours. Nucleotides were measured at 45 minutes of depletion (n = 7) and at 2 hours of recovery (n = 6). (b) Cells were treated identically except for the addition of 200 μM guanosine to the recovery medium (n = 6). (c and d) Representative fields of confocal microscopic images of LLC-PK1 cells at 24 hours after recovery from chemical anoxia. Cells were costained with Hoechst 33342 and propidium iodide as detailed in Methods. In c, cells recovered in regular media. Apoptotic features such as condensation and fragmentation of chromatin is seen in most cells in the field. In d, cells recovered in the presence of 200 μM guanosine. Most nuclei showed normal morphology, and only few had apoptotic features.
We then quantitated cell viability after chemical anoxia and recovery for 24 hours. The presence of guanosine significantly increased cell viability at 24 hours: viability was 70% ± 8% in the guanosine-treated group as compared with 44% ± 6% in the control group (P < 0.02). As shown in Figure 9, c and d, the increase in viability in the guanosine group primarily was due to a reduction in apoptotic cell death. Indeed, the guanosine-treated group showed less than 10% apoptosis at 8, 12, and 24 hours as compared with 40% in the control group (the representative fields shown in Figure 9 are from a 24-hour time point). Neither group showed any significant necrosis at these three time points as evidenced by exclusion of propidium iodide. Thus, guanosine supplementation reduces apoptosis and increases viability after in vitro chemical anoxia in LLC-PK1 proximal tubular cells.