Figure 6. Transcripts of NAD⁺ biosynthesis are downregulated, while those of NAD⁺ consumption increase in susceptible strains, and this is reflected by changes in NAD⁺ levels and disease severity.

(A–C) Box plot of RNA-Seq–based gene expression of NAD⁺ de novo synthesis (A), NAD⁺ salvage (B), and NAD⁺ consumption (C) genes in the 3 most stereotypical strains after 1 week of folic acid administration. Other strains are summarized on Supplemental Figure 6. While the CAST/EiJ strain undergoes little variation upon FA treatment, in the PWK/PhJ and to a lesser extent in the C57BL/6J strain, there was a marked reduction in transcript expression of NAD⁺ biosynthesis accompanied by an increase in transcripts of the NAD⁺ consumers, Cd38 and Parp1, while Sirt1 underwent little variation. The box plot lower and upper hinges correspond to the first and third quartiles, and the center line is the median. The whiskers extend from the hinge to the largest value no further than 1.5× the interquartile range. FDR-adjusted P values are shown (***P < 0.001, **P < 0.01, *P < 0.05), using moderated 2-tailed t test. (D) Kidney NAD⁺ levels measured by HPLC-MS. This confirms that the reduction of NAD⁺ content is strongest in the PWK/PhJ strain and absent in the CAST/EiJ and A/J strains. FDR-adjusted P values are shown (***P < 0.001, **P < 0.01, *P < 0.05), using moderated 2-tailed t test. (E–J) Correlation between NAD⁺ levels and blood parameters, blood urea nitrogen (E), TIMP-1 (F), creatinine (G), and the kidney histology scores for fibrosis (H), tubule degeneration (I), and tubule dilation (F). The blue line is a linear fit of the data. Statistics significance is shown with Pearson r and FDR-corrected P values (***P < 0.001, **P < 0.01, *P < 0.05), using Pearson correlation test, corrected for multiple testing over all possible phenotype comparisons. These correlations and those pictured Supplemental Figure 7 indicate that NAD⁺ levels may be highly predictive of AKI susceptibility in mice.