Retraction Watch

Tracking retractions as a window into the scientific process

Contaminated cells force retraction of Blood paper

with 6 comments

blood414Blood has an interesting retraction of a 2011 paper on what a group of authors claimed was a new cell line — but which proved, apparently, to be a chimera.

The article, “Oxygen-regulated expression of the erythropoietin gene in the human renal cell line REPC,” came from a team at Universität Duisburg-Essen, in Germany, and has been cited 21 times, according to Thomson Scientific’s Web of Knowledge. Here’s the abstract:

Erythropoietin (EPO), the key hormone in red blood cell renewal, is mainly produced in the adult kidney. Anemia and hypoxia substantially enhance EPO expression to increase erythropoiesis. Investigations of the cellular physiology of renal EPO production have been hampered by the lack of an adequate human cell line. In the present study, we present the human kidney cell line REPC (for renal Epo-producing cells), established from an explanted human kidney exhibiting EPO gene expression and release of the EPO protein in an oxygen-dependent manner. Hypoxic induction of EPO mRNA showed the typical transient increase and peak in expression after 36 hours under continuous conditions of hypoxia. Bioactive EPO protein accumulated in the culture supernatant. The induction of EPO gene expression in REPCs critically depended on the activation of hypoxia-inducible transcription factors (HIFs). SiRNA treatment revealed that the expression of EPO was largely dependent on the activation of the transcription factor complex HIF-2. In addition, hepatic nuclear factor 4α was shown to be critically involved in hypoxia-induced renal EPO expression. Using the human kidney cell line REPC, we provide for the first time a powerful tool with which to study the cellular and molecular regulation of renal EPO production.

Turns out those cells weren’t what the researchers thought they were. Per the retraction notice:

The Editors of Blood wish to retract the above-mentioned publication. The paper reported on a permanent human renal cell line, which was established from the tumor-free renal tissue of an anonymized male patient suffering from a kidney tumor. The cells were capable of hypoxia-inducible erythropoietin production – therefore termed Renal Erythropoietin Producing Cells (REPC) – showed features of neuronal cells, and were negative hepatitis virus B.

After publication the authors have freely provided the cells to a number of researchers who requested them. One of the collaborators has brought to the authors’ attention that REPC were contaminated with human Hep3B cells. Upon this, the authors have externally tested the cells including frozen aliquots of the youngest passage that was left after treating all REPC for mycoplasma infection. DNA profiling using 8 different and highly polymorphic short tandem repeat (STR) loci revealed that the REPC were cross-contaminated with Hep3B. STR profiles of the used cells were matching the STR reference profile of the cell line Hep3B as indicated by a search of the STR database from cell banks ATCC (USA), HPACC (UK), JCRB (Japan), RIKEN (Japan), KCLB (Korea) and DSMZ (Germany). Upon this, all groups who received cells were immediately informed about the contamination and have been asked to send back aliquots of the cells. These were again externally tested, but the contamination was confirmed. One collaborating group had identified different cell phenotypes in the REPC cultures by immunohistochemistry and tried to isolate single clones from the phenotypically different appearing cells. The latest STR profiling of the two clones, however, also revealed persisting cross-contamination with Hep3B.

Thus, the authors notified Blood that they can currently neither provide cells that were described in the original publication nor reproduce the original results due to the lack of the cells. Because the cells should have served as an important new in vitro model or renal cells to study erythropoietin production, the authors and the Journal feel obliged to retract the above publication to avoid any further confusion about the existence of a permanent human renal erythropoietin-producing cell line. The authors apologize for any inconvenience that the research community may have faced with REPC obtained directly from them or from other sources.

All authors agree to the retraction.

This is far from the first retraction we’ve seen for incorrect cell lines. In fact, it’s not even the first one we’ve seen in the past week for mixed-up kidney cells.

 

Comments
  • Deidentified April 29, 2014 at 10:11 am

    One of the side stories of HeLa is its ability to cross-contaminate other cell lines. Cancer lines already have a number of transformations that make them different, and this makes their in vitro behavior unpredictable if they are already immortalized.

    From the paper, 1g of kidney tissue was cut up, digested by collagenase, put through a mesh and cultured in RPMI, FBS and pen-strep. I’ve always tried to get by without using pen-strep and being as aseptic as possible…since bacterial contamination will warn me if I was sloppy or if cross-contamination has entered my system. Pen-strep will cause me to lose less flasks to bacterial contamination (e.g the yellow flask of doom), but will mask things like mycoplasma and cell line contamination.

    Looks like this was a primary culture that they had immortalized, so the presence of Hep3B would suggest contamination through shared materials from another cell line being grown up in the same lab. It’s often a prudent practice to make up complete media for just a particular project (versus having a master bottle for feeding your Hep3B and your primary cell line). And if you are sufficiently paranoid, a separate incubator

    I notice: “Upon this, the authors have externally tested the cells including frozen aliquots of the youngest passage that was left after treating all REPC for mycoplasma infection.”

    Did this suggest mycoplasma was also in the cell lines, or some kind of prophylactic procedure where they assumed their strange results came from mycoplasma contamination? If mycoplasma did appear it may be a proxy for contaminated equipment, incubator, etc.

    • Deidentified April 29, 2014 at 2:50 pm

      Disregard first paragraph, though HeLa was a contaminant. Going to ATCC’s website they have a helpful website about misidentified cell lines: http://www.atcc.org/en/Products/Cells_and_Microorganisms/Cell_Lines/Misidentified_Cell_Lines.aspx

      It sounds like they had serious problems simply trying to separate the REPC from the HepB. If only they had saved some of that kidney from the beginning. My guess is that they spun down and froze aliquots from primary, but eventually discovered mycoplasma and treated the infection, then discarded previous aliquots. I suspect that even if they had saved them the Hep3B contamination probably entered the same time as mycoplasma. I would also suspect they checked the previous aliquots and detected mycoplasma all the way back (then threw those aliquots out), but have no proof.

    • Stewart April 29, 2014 at 6:45 pm

      Junk Science

      Yes, the images from the Biochemical Journal (2006) and the Journal of Immunology (2009) have identical specific markers and are identical.

      The 2006 paper authors appear to believe “Total RNA was prepared, transcribed into cDNA and expression of the cDNAs for CD14 (‘CD 14’) and TLR4 (‘TLR 4’) were quantified by PCR”. We all know, of course, that PCR cannot quantitate cDNAs, but rather qualitate whether a gene is expressed or not. There is no quantitation, despite the rather odd claim that there is.

      The authors, later in the 2009 paper state “Total RNA was extracted and expression of TLR4 and CD14 was analyzed by qualitative PCR”, which is correct.

      • Mel May 4, 2014 at 8:08 am

        They might have meant they quantitative PCR (aka qPCR or RT-PCR). Using this method, assuming the proper controls, it is possible to determine the amount of starting RNA from a sample in a quantitative, not just qualitative, way. That’s not necessarily what they did, but it is possible that what they said is correct.

  • Dave April 29, 2014 at 7:05 pm

    Kids: this is why primary cells should always be prepared and cultured in a separate hood using separate media and reagents. And mycoplasma would not be a problem if they weaned themselves off of heavy and routine antibiotic use. Poor lab practices have led to this retraction, nothing more, nothing less.

  • Post a comment

    Threaded commenting powered by interconnect/it code.