View metadata, citation and similar papers at core.ac.uk brought to you by CORE provided by Elsevier - Publisher Connector Kidney International, Vol. 68 (2005), pp. 2694–2703 Calcium-binding proteins annexin A2 and S100A6 are sensors of tubular injury and recovery in acute renal failure CHAO-WEN CHENG,ABDALLA RIFAI,SHUK-MAN KA,HAO-AI SHUI,YUH-FENG LIN,WEI-HWA LEE, and ANN CHEN Graduate Institute of Life Sciences, Graduate Institute of Medical Sciences, Department of Internal Medicine, Department of Pathology, Tri-Service General Hospital, National Defense Medical Center, Taiwan, Republic of China; and Department of Pathology, Rhode Island Hospital, Rhode Island Calcium-binding proteins annexin A2 and S100A6 are sensors Acute tubular necrosis is the most common pathologic of tubular injury and recovery in acute renal failure. entity responsible for the clinical state of acute renal fail- Background. Rise in cellular calcium is associated with acute ure (ARF) [1, 2]. The two main causes of acute tubu- tubular necrosis, the most common cause of acute renal failure (ARF). The mechanisms that calcium signaling induce in the lar necrosis are ischemic and toxic injuries [3]. In the quiescent tubular cells to proliferate and differentiate during latter type, a variety of renal environmental substances acute tubular necrosis have not been elucidated. that include heavy metals such as mercury, lead, and ura- Methods. Acute tubular necrosis induced in mice by sin- nium are known to cause ARF.Nephrotoxic acute tubular gle intravenous injection of uranyl nitrate and examined af- necrosis is histologically evident as epithelial cell necrosis, ter 1, 3, 7, and 14 days. Renal function was monitored and kidneys were evaluated by histology, immunohistochemistry, mainly in the proximal convoluted tubules, with preser- Western blotting, in situ hybridization, and real-time reverse vation of the tubular basement membrane, and usually transcription-polymerase chain reaction (RT-PCR). Models of intact distal tubular segments [4]. folic acid induced-ARF and ischemic/reperfusion (I/R) injury Although severely damaged by toxin, the kidney were similarly investigated. has the ability to completely recover structurally and Results. Analysis of mRNA expression of intracellular cal- cium and phospholipid-binding proteins demonstrated selective functionally. Normally, quiescent cells undergo dedif- expression of S100A6 and Annexin A2 (Anxa2) in the renal cor- ferentiation and regain their potential to divide after tex with marked elevation on day 3, and gradually decline on enhancement of DNA synthesis in acute tubular necrosis. day 7 and further attenuation on day 14. Similarly, the expres- Consequent to cell proliferation, the new cells differenti- sion of both proteins, as demonstrated by immunohistochem- ate to restore the functional integrity of the nephron [5]. istry and Western blot analysis, was increased and reached the peak level on day 7 and then gradually declined by day 14. Vi- Little is known of the mechanism(s) by which regenera- mentin, a marker of dedifferentiated cells, was highly expressed tion of renal tubules is mediated. The observations that during the recovery phase. Combined in situ hybridization im- hepatocyte growth factor (HGF), epidermal growth fac- munohistochemistry revealed colocalization of both S100A6 tor (EGF), and bone morphogenetic protein-7 (BMP-7) and Anxa2 with proliferating cell nuclear antigen (PCNA). The are among the potent regulators of kidney organogenesis universality of this phenomenon was confirmed in two other mouse acute tubular necrosis models, the ischemic-reperfusion and that these agents can also promote tubular regener- injury and folic acid-induced ARF. ation after a variety of insults [6] are consistent with the Conclusion. Collectively, these findings demonstrate that idea that the regeneration process may be partially con- S100A6 and Anxa2 expression, initiated in response to tubu- trolled by a mechanism similar to that operating during lar injury, persist in parallel throughout the recovery process of development. tubular cells in acute renal failure. Tubular cell calcium concentration and content are in- creased following acute renal injury induced by ischemic + and toxic insults. The divalent calcium cation (Ca2 ) sig- naling system operate by binding to effector molecules, + Ca2 binding proteins, that mediate stimulation of numer- Keywords: acute tubular necrosis, S100A6, annexin A2, calcium- 2+ binding proteins. ous Ca dependent processes such as transcription and + cell proliferation. Two large families of the Ca2 binding Received for publication May 4, 2005 and in revised form June 22, 2005 proteins are the annexins and the EF-hand motif S100 Accepted for publication July 20, 2005 proteins [7]. The annexins are a family of phospholipid- binding proteins that share a common property of in- C 2005 by the International Society of Nephrology teracting with membranes and target molecules in a 2694 Cheng et al: Calcium-binding proteins sensing acute renal failure 2695 + Ca2 -dependent manner [8, 9]. S100 proteins represent analyzed before folic acid or vehicle administration. The + the largest subgroup in the EF-hand Ca2 binding protein mice were sacrificed, and their kidneys were collected at family. A unique feature of S100 proteins is that individ- 1, 3, 7, and 14 days at each time point. ual members are localized in specific cellular compart- + ments from which some are able to relocate upon Ca2 + activation, thus transducing the Ca2 signal in a temporal and spatial manner by interacting with different targets Renal function specific for each S100 protein [10]. Interactions between Blood samples collected through the retro-orbital ve- annexins and S100 proteins are now known to include nous plexus were centrifuged (3000 × g,10minutes), and several members of these protein families. the supernatant containing the serum was withdrawn and ◦ During the course of our studies on global gene expres- stored at −70 C until assayed. Renal function was as- sion profiling in a mouse model of uranyl nitrate–induced sessed by measuring the elevation in plasma levels of cre- ARF, we observed dramatic up-regulation in the expres- atinine and blood urea nitrogen (BUN). These analyses, sion of S100A6 (calcyclin) and annexin A2. Accordingly, which utilized Fuji DRI-CHEM 3030 (Fuji Photo Film we postulated these interactive effector molecules might Co. Ltd., Tokyo, Japan), were obtained within 15 minutes play a role in the pathophysiology of ARF and their once the serum samples thawed. expression might be useful biomarkers of the temporal events and processes of the tubular epithelial cell in acute tubular necrosis. Renal histopathology The formalin-fixed renal tissues were dehydrated in a graded series of ethanol solutions and embedded in paraf- METHODS fin as described elsewhere [13]. Three micron sections Animal models were obtained and stained with hematoxylin and eosin. Uranyl nitrate–induced acute tubular necrosis. Female For typical lesions of acute tubular necrosis, tubular cell 8-week-old C57BL/6 mice were purchased from the Na- necrosis, of which some of them were sloughed into the tional Laboratory Animal Breeding and Research Center tubular lumina accompanied by casts [14]. Quantitative (Taipei, Taiwan). All mice received a single tail vein in- analysis of renal tubular necrosis was performed by opti- jection of uranyl nitrate [(UO2 (NO3)2·6H2O] (100 lg cal microscopy. Briefly, 100 intersections were examined in 100 lLof5%NaHCO3). The mice were sacrificed at for each kidney and a score from 0 to 3 was given for 0, 3, 7, and 14 days after the injection. Blood and urine each tubular profile involving an intersection: 0, normal were collected for clinical evaluation, and kidneys were histology; 1, tubular cell swelling, brush border loss, nu- removed for molecular and histopathology studies. Sam- clear condensation, with up to one third of tubular profile ples intended for histopathology were fixed in formalin showing nuclear loss; 2, as for score 1, but greater than according to a standard protocol. one third and less than two thirds of tubular profile shows Ischemic/reperfusion (I/R) injury. This model was in- nuclear loss; and 3, greater than two thirds of tubular pro- duced as previously described [11]. Briefly, bilateral I/R file shows nuclear loss. The total score for each kidney injury was generated in female Balb/c mice (25 to 30 g) was calculated by addition of all 100 scores with a max- by occluding the renal pedicles with microvascular clamps imum score of 300. There was also evidence of tubular for 30 minutes under ketamine-xylazine anesthesia. Com- cell regeneration, it was defined as flattened epithelial pleteness of ischemia was verified by blanching of the cells with hyperchromatic nuclei and mitotic figures [14]. kidneys, signifying the stoppage of blood flow. The blood Quantitative analysis of renal tubular regeneration was flow to the kidneys was reestablished by removal of the performed by optical microscopy. Briefly, 100 intersec- clamps (reperfusion) with visual verification of blood re- tions were examined for each kidney and a score from 0 turn. Mice subjected to sham operation (identical treat- to 3 was given for each tubular profile involving an inter- ment except that the renal pedicles were not clamped) section: 0, normal histology; 1, flattened epithelial cells were used as controls. During the procedure, animals with hyperchromatic nuclei and mitotic figures with up were well hydrated and their