The carcinogenic effects of aristolochic acid (AA) and ochratoxin A (OTA) are widely described. Despite many trials aiming to discover the mechanism of their involvement to nephropathy progression, the sequence of events is still not clear.
The two main components of AA, AAI and AAII are responsible for nephropathy progression, however AAI is more potent cytotoxic agent towards kidney epithelium (Arlt et al., 2002; Liu et al., 2009
). Nephrotoxic activity of OTA is well-documented, however, species-dependent discrepancies between man, pig and rodents are underlined. Such variations may be caused by the differences in the binding of OTA to serum proteins, oral bioavailability, the half-life of OTA in serum as well as in the different plasma clearance between species (reviewed in Petzinger and Ziegler, 2000
). In the present study, porcine renal proximal tubule epithelial cells (LLC-PK1), a well characterized cell line often used in toxicological studies (Dietrich et al., 2001
) was chosen as a model for investigation. Importantly, the high susceptibility of pigs towards OTA and their importance for livestock production is well-known and pork as well as food products from pigs fed with contaminated grain may also be a source of OTA (International Programme on Chemical Safety, 1990
). Moreover, it is well established that proximal tubular cells are the main site of toxicity of both OTA and AAI. Proximal tubule injury is observed in aristolochic acid nephropathy in rats (Mengs, 1987; Lebeau et al., 2005
) and analysis of both kidney functions and renal biopsies from AAN patients showed increased tubular proteinuria, impairment of proximal tubule functions and tubular necrosis (Depierreux et al., 1994
). OTA was shown to be removed by tubular, but not glomerular filtration to the urine and in vivo
studies underlined that OTA affects the proximal part of the nephron (Groves et al., 1998
In AAN (Depierreux et al., 1994; Yang et al., 2005
) and other kidney diseases (Neusser et al., 2010
) tubulointerstitial damage observed during kidney fibrosis may be the effect of blood vessel injury. In the proper vessel functioning an important role plays vascular endothelial growth factor (VEGF), which in kidneys is expressed both in podocytes and additionally in proximal tubular cells (Baderca et al., 2006
), which are the main site of AAI as well as OTA injury. Moreover, both tubular and glomerular VEGF may play an important role in the maintenance of peritubular or glomerular capillaries. Diminished VEGF production may lead to decreased endothelial survival and angiogenesis as well as tubular damage by ischemia (reviewed in: Schrijvers et al., 2004
). The importance of the alterations in VEGF expression in epithelial cells of proximal and distal tubules was shown in human diabetic nephropathy patients (Lindenmeyer et al., 2007
) as well as in patients with progressive proteinuric renal failure (Rudnicki et al., 2009
We investigated the effect of AAI and OTA on VEGF, the potent pro-angiogenic factor, which is claimed to affect the nephropathy progression. The data concerning the role of VEGF in development of AAN are still not clear, although it seems that regulation of VEGF expression plays an important role in this disease. VEGF expression was reported to be down-regulated in rats with chronic AAN (Sun et al., 2006b
) as well as in acute AAN rat model (Wen et al., 2008
). In contrast, it was shown that in AA-induced acute tubular necrosis (AA-ATN) VEGF expression is elevated in renal tubules compared to control group, nevertheless, the expression was lower than in antibiotic-induced ATN (Yang et al., 2007
). In our study we observed the elevation of VEGF transcription as well as protein expression after AAI treatment in LLC-PK1 cells. Interestingly, we showed that OTA has different effect on VEGF production compared to AAI in short-term treatment as potent inhibition of VEGF expression in LLC-PK1 cells was observed after OTA stimulation. In male F344/N rats treated with OTA no alterations in urinary level of VEGF was found (Hoffmann et al., 2010
), however, the level of VEGF in urine may differ from ones present in organs or in serum. As VEGF may be produced by podocytes as well as epithelial tubular cells in order to better understand the effects evoked by AAI/OTA it is important to determine the production of VEGF by both types of cells in in vivo
In order to investigate the mechanism of differential effect of AAI and OTA on VEGF production we verified the effect of both toxins on the activity of transcription factors, which binding sites are present in VEGF promoter, such as HIFs, AP-1, NFκB or SP-1 (Pages and Pouyssegur, 2005
). Our data indicate that both toxins exert complex effect on various transcription factors, and as the result they may differentially regulate VEGF expression. AAI treatment caused SP-1 and HIFs up-regulation, whereas AP-1 was inhibited after 24 h of toxin delivery. Similarly, OTA treatment diminished AP-1 activity, but it also potently down-regulated SP-1 and HIFs activity. Moreover, the activity of NFκB was influenced neither by AAI nor by OTA. Such complicated data show that, although both toxins induced kidney damage, the mechanisms are different and should be carefully investigated in details. Additionally, the effect may be cell-type dependent as in human HKC-8 cells only the effect of OTA on HRE and AP-1 activity was the same as in LLC-PK1 cells, whereas NFKB was induced and SP-1 activity was not affected by this toxin (Fig. S3
). In pheochromocytoma PC-12 cells the inhibition of AP-1 (Oh et al., 2004
), whereas in 12-day rat embryo midbrain cells the activation of this factor by OTA was observed (Hong et al., 2002
). In contrast to our data, where we did not observe the alterations in NFκB activity after OTA delivery, such activation was shown in proximal OK cells (Sauvant et al., 2005
), in immortalized human kidney epithelial cells (IHKE) (Rached et al., 2006
) as well as in 12-day rat embryo midbrain cells (Hong et al., 2002
). On the other hand, in LPS-activated RAW264.7 macrophages DNA binding activity of NFκB was considerably lower after AAI treatment in comparison to control cells (Liu et al., 2011
). However, such differences may be caused by the dose and time of stimulation in individual experiment. In case of SP-1, there are no data concerning the effect of OTA on activity of this transcription factor, so we have shown for the first time the diminishment of SP-1 activity after OTA. Moreover, our results indicating inhibitory effect of OTA on HRE activity and HIFs transcription factors are also unique. To our best knowledge, only one paper showing increased mRNA level for HIF-1α in rat proximal tubule cells after OTA treatment was published already but the protein level was not investigated (Luhe et al., 2003
). However, in case of HIF proteins, the protein stability is much more crucial, therefore these data and our data do not exclude each other.
The knowledge about AA influence on the activity on transcription factors is also very limited. We have presented for the first time that AAI exerts opposite effect than OTA on SP-1 and AP-1, enhancing and diminishing their activity, respectively. The already published data about the effect of AA on NFκB is contradictory, as inhibition in human HK-2 cells (Chen et al., 2010
) and induction in kidney of Hupki (human TP53 knock-in) mice (Arlt et al., 2011
) has been observed. Interestingly, in our hands activity of NFκB was not affected but we observed HIF induction after AAI delivery. These data are in accordance with results from animal studies. The presence of hypoxia was also observed in male Wistar rats treated with AAI for 4 days (Cao et al., 2010
). In rat model AA evoked elevated nuclear staining for HIF-1α with concomitant reduction in VEGF production in long (8–16 weeks) (Sun et al., 2006a,b
) and short (4–7 days) term (Wen et al., 2008
) experiments. Moreover, this increase of nuclear HIF-1α was present in the tubular cells in damaged area (Wen et al., 2008
). However, in our studies concomitantly with HIF stabilization we observed elevation of VEGF production. The discrepancies between our results and published data may come from different time of stimulation and species-dependent differences in response. Additionally, it is possible that in case of longer AA treatment other transcription factors known to regulate VEGF expression, like AP-1, may play a role. Therefore, it seems that regulation of VEGF expression after delivery of AAI is much more complex. Thus, the understanding of the sequence of events evoked by AA is important to identify the origin of AAN development and still needs to be clarified.
The most important part of our study is the discovering of the possible mechanism of AAI/OTA action on VEGF production. The augmentation of HIFs and SP-1 transcription factors activity by AAI was paralleled with the up-regulation of VEGF transcription and protein level. By the use of mithramycin A, an inhibitor of SP-1 activity, and chetomin, an inhibitor of HIFs, we showed that AAI-elevated VEGF production is reversed after inhibition of SP-1 and HIFs, what confirms the role of these transcription factors in the effect of AAI on VEGF expression.
The next salient finding of our study is that hypoxia attenuated the inhibitory effect of OTA on VEGF production. In the kidney the localization of HIF isoforms depends on cell type with HIF-1α presence in the tubular epithelia, whereas HIF-2α expression mostly in endothelial, glomerular and interstitial cells (Rosenberger et al., 2005
). Although different role of HIF isoforms in kidney development may be the result of divergent localization in cells, it is well documented that HIF-1 and HIF-2 also differs in regulation of gene expression (reviewed in Loboda et al., 2010
). HIF stabilization elevates angiogenesis and therefore it may attenuate adverse effects of toxins delivery. On the other hand, HIF triggers also the expression of connective tissue growth factor (CTGF), which exhibit profibrotic effects (Higgins et al., 2004
). Thus, long-term activation of HIF may lead to fibrosis development. Therefore the proper balance in HIF activation is crucial for therapeutic effect.
Our data indicate that in prevention of diminishment of VEGF production evoked by OTA in kidney proximal tubular epithelial cells HIF-2α but not HIF-1α plays a crucial role. Additionally, we found that HIF-1α overexpression diminished VEGF production, whereas only AdHIF-2α transduction resulted in elevation of VEGF expression. Therefore, it seems that two isoforms of HIF may play a distinct role in regulation of VEGF production in porcine proximal tubular epithelial cells, which are the major target of OTA action. Moreover, only HIF-2 exerts protective effect, especially against short-term acute kidney injuries. These results are in accordance with studies showing that HIF may be protective in acute renal injuries whether in case of chronic ones they exert opposite effect (Manotham et al., 2004
). Still, the role of each HIF isoform in different kidney cell types may be various. Additionally, also the other factors, such as AP-1 and SP-1, should be investigated in this context.
In conclusion, we have shown complicated pattern of VEGF regulation by different toxins affecting kidney biology. To our knowledge, the influence of AAI and OTA on some transcription factors have not been investigated before and further investigations are necessary to analyze this intriguing effects.