Seeing is believing

Abstract

Ischemic injury to the kidneys is a prevalent clinical problem, contributing importantly to chronic kidney disease. Yet, underlying molecular mechanisms are elusive. To address the possible role of autophagy, we engineered a novel strain of mice harboring a ubiquitously expressed CAG-RFP-EGFP-LC3 transgene. Using this tool, we examined the post-ischemic kidney and detailed the dynamics of renal tubular epithelial autophagy. In addition, we defined the role of MTOR in the resolution of autophagy during epithelial survival and kidney repair.

The human kidneys receive about 20% of cardiac output and consume 7% of daily energy expenditure to support their diverse functions, including active ion transport in renal tubules. This high metabolic demand renders the kidney, especially the mitochondria-rich proximal tubules, susceptible to ischemic injury.

Mouse models of renal ischemia-reperfusion injury induced surgically by clamping renal blood vessels to arrest blood supply, followed by release of the vascular clamp to allow reperfusion, have provided critical insight into the pathogenesis of kidney injury and repair. In response to injury, some tubular epithelial cells undergo cell death mediated by necrosis or apoptosis. Careful examination, however, reveals that a majority of the surviving proximal tubular cells sustain sublethal injury. Mitochondrial fragmentation and formation of autophagic vacuoles can be visualized easily by electron microscope. Also, tubular injury triggers cellular repair processes, with epithelial proliferation as a major regenerative mechanism. Although a growing body of evidence suggests that autophagy is beneficial to the kidney following injury, the dynamics of autophagy involved in the initial cell survival phase, and subsequent regeneration phase, are not fully understood.

To probe the role of autophagy in ischemia-induced injury of renal tubular cells, we engineered a line of mice harboring a tandem fluorescence LC3 transgene. Based on the differential pH sensitivity of RFP (pK_a 4.5), which is stable in acidic pH, and EGFP (pK_a 5.9), which is quenched in the acidic lysosomal environment, autophagosomes can be distinguished from autolysosomes. Expression of the transgene is driven by the ubiquitously expressed CAG promoter, so that multiple cell types could be evaluated. Combined GFP and RFP fluorescence yields a yellow signal within phagophores and autophagosomes. By contrast, GFP is quenched in autolysosomes, and they emit a red (RFP) signal.

Few RFP and EGFP puncta are detected in the proximal and distal nephron under basal conditions. However, as expected, renal epithelial cells display a time-dependent appearance of EGFP and RFP puncta after deprivation of glucose and amino acids in cell culture. In this setting, more than 85% of the EGFP fluorescence signal is lost, whereas the RFP signal persists in autolysosomes. Furthermore, we discovered for the first time that acid-secreting intercalated cells of the collecting ducts harbor abundant fluorescent puncta, which is likely due to a high turnover rate of membrane vesicles. In agreement with findings in GFP-LC3 mice, podocytes that withstand high levels of stress as the glomerular filtration barrier also exhibit a high level of autophagy, suggesting that the basal autophagic activity varies within the heterogeneous cell population of the kidney to maintain specific cellular functions.

With starvation, there is a significant increase in the number of EGFP and RFP puncta in epithelial cells within all nephron segments, pointing to a robust in vivo response to a physiological stimulus. Similar to epithelial cells in culture, EGFP and RFP puncta in renal tubules display differential pH sensitivity, allowing us to study the initiation, progression, and resolution of autophagy in the kidneys following ischemic injury.

Interestingly, arresting renal blood flow for 45 min does not result in the appearance of EGFP and RFP puncta, either at the end of the ischemic insult or 4 h post reperfusion. At 8 h post reperfusion, EGFP and RFP fluorescence are visibly increased in surviving tubular cells. While EGFP puncta reach a peak at 24 h and return to baseline levels at 3 d, RFP puncta also peak at 24 h but persist at high levels for 3 d before returning to baseline levels by 7 d.

There are several possibilities that could account for these dynamics. As epithelial ion transport is a major determinant for renal oxygen consumption, stopping renal blood flow and ceasing glomerular filtration alleviates the need for tubular transport. This reduction in metabolic demand may lead to better tolerance of hypoxia. It is well established that renal oxygen delivery is reduced during reperfusion due to vasoconstriction and capillary congestion. Renal oxygen consumption may increase after reperfusion due to inefficient epithelial ion transport and ATP generation. These events could trigger further hypoxia and increased AMP to ATP ratios, which together signal to induce autophagy. Another possibility is that autophagy may function to remove damaged organelles as tubules launch a repair phase. The fact that EGFP puncta, which represent early autophagic vacuoles, peak at 24 h and RFP puncta persist at high levels for 3 d, suggests that stimuli for autophagy induction are reduced after 24 h, at which time autophagosomes proceed to fuse with lysosomes for clearance.

Proximal tubular cells are poised in the G₁ phase of the cell cycle, ready to reenter the cell cycle for tubular repair. However, we found that Ki67 expression is significantly lower in cells with activated autophagy, whereas no difference is observed in the expression of phospho-histone H3 between cells with and without autophagy. As Ki67 identifies proliferating cells in the late G₁ to M phases and phospho-histone H3 marks cells in the late G₂ and M phase, these results suggest that activation of autophagy is associated with suppressed reentry and/or early progression through the cell cycle. Therefore, autophagic cells may be less likely to function as precursors for cell division, which is essential for tubular repair.

To begin to understand the mechanism governing the resolution of autophagy during tubular repair, we examined the activity of a known autophagy inhibitor, MTOR. The majority of cells manifesting MTOR activation contain no fluorescent dots and harbor signs of active cell proliferation. Within the population of cells manifesting activated autophagy, higher levels of Ki67 expression are detected when MTOR is activated. Inhibition of MTORC1 with rapamycin results in increased numbers of early autophagic vacuoles and reduced cell proliferation.

In aggregate, our dissection of the dynamics of autophagic flux and the associated molecular events support a model of MTOR-dependent feedback to release cells from autophagy, which is important for cell survival following injury and cell cycle re-entry for tubular repair (Fig. 1). Together, our findings raise the enticing prospect that manipulation of autophagy could reduce renal epithelial injury and accelerate renal repair.

Figure 1. Renal epithelial autophagy following ischemia-reperfusion injury. Autophagy is induced to promote cell survival in response to injury. Using a novel tandem fluorescence probe, autophagosomes can be detected as yellow fluorescent puncta (RFP + GFP), whereas autolysosomes appear red (RFP only) due to quenching of EGFP fluorescence in the acidic environment. Activation of MTOR plays a role in autophagy resolution and cell proliferation for tubular repair.

Li L, Wang ZV, Hill JA, Lin F. New Autophagy Reporter Mice Reveal Dynamics of Proximal Tubular Autophagy. J Am Soc Nephrol. 2014;25:305–15. doi: 10.1681/ASN.2013040374.