Kidney International paper retracted after lab records “were improperly filed”

A group of University of California, Davis kidney researchers have retracted a paper after being unable to verify key parts of it.

Here’s the detailed retraction notice for “Proteinuria decreases tissue lipoprotein receptor levels resulting in altered lipoprotein structure and increasing lipid levels,” published by Limin Wang, George Kaysen, and colleagues in Kidney International last July:

A post-publication reader inquiry prompted a review of the western blots contained in Figure 4 and showed that an error in image selection indeed occurred. Laboratory records for these data were improperly filed and we now have insufficient evidence to verify these data. The data had reported differences in the levels of receptors for lipoproteins in nephrotic rats (nephrotic syndrome (NS)) compared with control animals and rats with hereditary analbuminemia (Nagase analbuminemic rat (NAR)). Owing to the uncertainty of the Figure 4 data, we cannot verify the conclusions made in the paper with regard to the mechanism responsible for the structure of lipoproteins present in NS compared with NAR and control animals.

We determined that lipid levels are increased in NS significantly more so than in NAR. We determined that NS results in lipoproteins that are significantly enriched in triglycerides and triglyceride to phospholipid (PL) content, as a measure of lipoprotein surface area, and have depleted apo E content on the surface of both very-low-density lipoprotein (VLDL) and low-density lipoprotein (LDL) compared both with control and NAR. We determined that apo E levels are lower on HDL of NAR than in control animals, but significantly lower in NS than in either NAR and control animals.

We determined that the reduction in apo E content in these lipoproteins occurs despite an overall increase in total serum apo E both in NS and NAR (Table 3). We determined that the mechanism of increased apo E levels in both NS and NAR was reduced clearance (Table 3), with no significant contribution by increased synthetic rate. This occurred in the context of increased VLDL size compared with control animals, resulting in large, triglyceride rich relatively apo E poor VLDL in NS (Figure 1) despite the expanded apo E pool.

As we had shown previously that apo E had an important role in the interaction between triglyceride rich lipoproteins and lipoprotein lipase, we had hypothesized that the affinity of apo E for lipoproteins was reduced in NS but not in NAR. We measured the affinity of apo E for HDL, LDL and VLDL but found no difference in affinity of apo E for any of the lipoproteins when we compared control animals with NS or NAR (Figure 3, Table 4). We also found no difference in the affinity of apo E by lipoprotein type.

We have previously shown that clearance of triglyceride (TG) rich lipoproteins is significantly impaired in NS compared with NAR and controls, and postulated that the mechanism responsible for this difference was decreased uptake of triglycerides in tissues in NS compared with NAR providing a potential mechanism for the relative enrichment of TG in NS lipoproteins.

Our inability to verify the conclusions communicated by the western blots is important in as much as we have not clarified the mechanism for the increase in triglyceride content in NS compared with NAR. We had postulated that the reason that triglyceride content was increased was a result of decreased uptake of triglycerides in peripheral tissues in NS compared with NAR resulting in the accumulation of large apo E poor triglyceride rich lipoproteins in the NS. We cannot substantiate that mechanism given that we are unable to verify the western blot data. We deeply regret being unable to verify these data.

We have reviewed the rest of the data contained in the paper. The data contained in Table 1 is verifiable. The data contained in Table 2 with regard to all but the ratio of ApoB/PL and ApoC/PL is verifiable. The ratio of apoB/PL and apoC/PL is not, however, but this has no impact on the principal hypothesis proposed in this paper.

In light of this, the authors would like to retract the article.

The study has yet to be cited, according to Thomson Scientific’s Web of Knowledge.