Kidney cell-specific knockdown: anything but simple.

USING GENE KNOCKOUT STRATEGIES to help define the physiological and/or pathophysiological role(s) of factors produced by specific renal cell types presents substantial challenges. Renal cell-specific gene targeting, involving a Cre recombinase (Cre)-expressing transgene (either constitutively expressed or inducible) and a loxP-flanked (floxed) target allele, has been achieved in vivo in mouse podocytes (7), proximal tubules (9), thick ascending limbs (10), principal cells (1), and intercalated cells (6). However, these approaches are time consuming, expensive and risky, particularly if the transgenic and/or floxed mice are either not yet made or are in the wrong mouse strain (various renal disease models can require specific mouse strains). Furthermore, renal cell-specific gene knockout has been largely confined to mice; although genetic engineering approaches in rats [transcription activator-like effector nucleases (TALEN), zinc finger nucleases, and clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated nucleases (Cas)] (2) are being more widely applied, they are not yet reliably used to achieve cell-specific gene targeting. A potentially quicker, speciesand strain-independent, and less expensive approach to obtain renal cell-specific targeting involves using vectors expressing short hairpin RNA (shRNA; administration of siRNA cannot yield renal cell-specific targeting unless the carrier were engineered to bind to only unique renal cell populations). Typically, shRNA, driven by a ubiquitously expressed promoter, have been administered via direct renal administration (3, 5); while a given renal cell type may have a high degree of transfection, other cell populations are also affected. To obtain renal-specific shRNA production, mice with transgenic Cre-inducible shRNA were subjected to ultrasound microbubble delivery of endothelial (tyrosine kinase receptor-2)or distal nephron (Ksp-cadherin)-specific promoter-driven Cre (4). However, this approach still involves a transgenic animal; in addition, RNA knockdown occurred in renal cells in which the Cre-driven promoter should not be active, suggesting an issue with the ultrasound treatment per se and/or possibly shRNA transfer between cells. While the above are only a few examples of the renal gene silencing studies being performed, they illustrate the difficulties in obtaining renal cell-specific targeting in a reliable, timely and costeffective manner. The study by Nam et al. (8) in a recent issue of the American Journal of Physiology-Renal Physiology describes a new and potentially easier alternative to achieve renal cellspecific targeting. The authors generated two lentiviral constructs: 1) U6 promoter-loxP-cytomegalovirus promoterenhanced green fluorescent protein (GFP)-loxP-aquaporin-3 (AQP3) shRNA; and 2) Homeobox 7 (HoxB7) promotermCherry-Cre. While the two constructs can enter cells throughout the body, the HoxB7 promoter and hence Cre and mCherry (confers red fluorescence) should be selectively active/expressed in the collecting duct. The shRNAcontaining construct expresses GFP constitutively, but the floxed CMV-GFP acts as a stop sequence and prevents shRNA transcription. In the presence of Cre, CMV-GFP is excised, GFP expression is lost, and AQP3 shRNA is expressed. Thus, at least in theory, cells will be green and lack AQP3 shRNA or will be red and contain AQP3 shRNA. The lentiviruses were administered by hydrodynamic tail vein injection (essentially rapid administration of a large volume) three times over the course of a week using a viral titer of 4 10 transfection units (optimal dosing determined empirically). GFP expression was highest in the liver, spleen, and kidney 6 wk after the first injection, while mCherry expression was only found in the kidney. No cells appeared to express both GFP and mCherry, suggesting efficient Cre-mediated recombination; however, while the images suggest collecting duct-specific staining, collecting duct mCherry expression was not specifically described (e.g., by showing samples costained for AQP2). Administration of both lentiviral constructs caused a marked decrease in renal AQP3 immunofluorescence and total renal AQP3 protein was reduced by 75% at 6 and 12 wk after the first injection. Importantly, knockdown of AQP3 was associated with doubling of urine volume, and urine osmolality fell by half. Taken together, and despite the issue mentioned above, the findings in this study are quite impressive. The potential benefits of the Nam et al. (8) method include the relative ease and low cost of achieving renal cell-specific knockdown, the ability to obtain such knockdown in adulthood, thereby avoiding developmental effects, the effectiveness of the treatment in achieving a phenotype, the persistence of the effect (at least 3 mo), and the potential usefulness of this approach in species other than mice. If proven widely applicable and reproducible in the hands of other laboratories, this approach has significant potential as an alternative to existing genetic engineering approaches. That said, there are some potential issues that may present challenges in moving this new system forward. First, several laboratories have found that hydrodynamic vein injection has not consistently resulted in a high degree of renal expression; whether the dosing prescription and lentiviral vectors used in the Nam et al. study will resolve this issue remains to be seen. Second, future successes will depend on sufficient target knockdown in specific cell types; this issue could become more important when less readily measurable or more ubiquitously expressed targets than AQP3 are involved. Third, relatively low animal-to-animal variability will need to be obtained with regard to the degree of knockdown. Despite these caveats, Dr. Nam and colleagues are to be congratulated on developing this system; while hardly simple, it does hold out hope for making renal cell-specific knockdown achievable by many more laboratories. Address for reprint requests and other correspondence: D. E. Kohan, Div. of Nephrology, Univ. of Utah Health Sciences Center, 30 N 1900 E, Salt Lake City, UT 84132 (e-mail: [email protected]). Am J Physiol Renal Physiol 309: F1007–F1008, 2015; doi:10.1152/ajprenal.00434.2015. Editorial Focus