Sir

We welcome your call to debate whether journals should impose more stringent publication criteria for biological materials, notably cell lines (“Standards for papers on cloning” Nature 439, 243; 2006). Our mission at the DSMZ cell repository, the German national resource centre for biological material (http://www.dsmz.de) is to provide investigators with authentic, well-characterized biological material. But it is all too often compromised by users' reluctance to make their own materials freely available. About 90% of researchers contacted since 1994 after claiming to have established newly published cell lines have refused or simply ignored our requests for material. Only one in 580 contacts indicated a preference to deposit cells with another facility, and a search of a large sample of cell lines currently held by other major repositories revealed none previously denied to the DSMZ.

Even among the minority depositing human tumour-cell lines at the DSMZ 29% provided identifiably false, cross-contaminated cell lines (CCCL). Although reliable detection of CCCL is limited to institutions such as ours that hold cell lines en masse and DNA-profile all accessions, this is almost certainly an underestimate of the true level. The systematic use of CCCL is likely to generate distorted data: accordingly, the Sanger Institute has highlighted instances of CCCL among the NCI-60 reference cell line panel used to identify tumour-specific gene expression (http://www.sanger.ac.uk/genetics/CGP/NCI60). But if new cell lines are not deposited, they cannot be authenticated.

Initial descriptive publication is the only effective stage at which sanctions might be applied. Journals such as Nature and Science should adopt rigorous criteria requiring deposition of new cell lines, and encourage specialist journals to follow suit. Publications using existing cell lines should identify where these came from. Funding bodies could also insist that cell lines established with their support be submitted to public repositories. And those in charge of large collections could do more to clarify the availability and authenticity of their cell lines.