Skip to content. Skip to main navigation.


  • January 12 2011

    Why would anybody want to know the genomics of wet vs. dry earwax type?

    • By: ,

    Nobody. Well, almost nobody. The real answer is that the gene (ATP-binding cassette protein, sub-family 'C' member '11'; ABCC11) that controls earwax phenotype also controls other attributes of glands in the body, and there are genetic variations in this ABCC11 gene that affect the risk of conditions and the effectiveness of treatments in those other glandular tissues.

    In recent years, there is a tendency in various media toward ad hominem attacks disparaging direct-to-consumer genomic testing as "frivolous" or "useless". In a manner much like the email question that is the title of this post, such attacks are strong on attitude and posturing and often mention earwax and eye-color and other traits that are inconsequential with regard to health or other value-points. The aim of such attacks is to cast doubt upon the legitimacy of the firms offering the tests and upon the sensibility of those who spend quite a lot of money to purchase the tests. The rhetorical fallacy the attacks utilize is the Fallacy of Composition: the attacks imply that if one item or a few items [of the test offering] are of doubtful value then the entire package must be of dubious value.

    Not surprisingly, there is a tendency on the part of direct-to-consumer genomics firms themselves to down-play or under-communicate the valuable health uses to which the genomics microarray test results can be put. In part, the companies' reticence has to do with (1) the fact that claims pertaining to health risks or personalized medicine uses of the information are subject to regulatory agency requirements in the U.S. and other countries and (2) the fact that the companies are concerned about exposure to strict product-liability claims in the courts, if anyone perceives that they were harmed in the course of using the information for medical purposes. The companies to-date find it advantageous to under-communicate about the uses of the information, in which case their financial and legal exposures are very much reduced and their regulatory obligations may also be relatively small as well.

    In the instance raised in the email, we have this SNP:

    • rs17822931 (538G>A, Gly180Arg) in ABCC11 on Chr. 16 (DNA strand-orientation is inverted in the chip that 23andme uses, so the normal allele is 'C' instead of 'G')

    There are wide differences in the epidemiology of this ABCC11 variation among different racial/ethnic groups. Fewer than 2% of Caucasians have two copies of the variant; about 20% to 30% have one copy. But people of Hispanic descent have two copies about 9% of the time. Pacific Islander and other Asian populations have two copies of the variant 'A' allele more than 80% of the time.

    The normal wild-type (Gly180) of ABCC11 is associated with wet-type earwax, axillary "osmidrosis" (excessive underarm sweat secretion), and with the quantity of milk secretion from the mammary gland after pregnancy.

    The rs17822931 (538G>A) SNP variant causes substitution of an arginine in place of a glycine at position 180 in the ABCC11 protein, and this greatly affects the function and stability of the ABCC11 protein molecule.

    Microscopy studies of earwax cerumen gland-containing biopsies show that the normal 'wild-type' ABCC11 protein (people with two copies of 'GG'; 'CC' in 23andme) is localized in intracellular granules in the cells, as well as at the cell membrane of secretory cells in the cerumen glands.

    By contrast, localization at these sites in glands' cells does not occur for the SNP (Arg180) variant. This SNP variant does not get N-linked glucose molecules attached to it like the normal wild-type version of the protein does, and the cells recognize it as a "mis-folded" protein in the cells' endoplasmic reticulum. There it rapidly undergoes degradation, and this in turn determines the "dry" type of earwax as a Mendelian recessive trait.

    People with wild-type ABCC11 ('GG'; or 'CC' in 23andme) have higher risk of developing cancer and have higher risk of multi-drug resistance and less-than-optimal response to chemotherapy as well. This is because normal wild-type ABCC11 protein functions as an ATP-dependent pump that pumps organic chemotherapy drug anions out of the cancer cells.

    And the cells of people with the variant rs17822931 with the arginine substitution don't spit the chemotherapy drugs out so quickly or at all. This endows them with a better chance of success, at lower doses than would be needed in people who are wild-type 'GG'.

    So a major obstacle to the successful chemotherapy of human cancer is development of resistance to the drugs, and the effectiveness of chemotherapy drugs is marked by wide variability between individuals. Practical pharmacogenomics of rs17822931 is one way that the DTC test, ostensibly for "earwax type", can be very valuable.

    Toyoda's and Ishikawa's just-published review (link below) provides an overview of the impact of ABCC11 538G>A on patients' response to nucleoside-based chemotherapy and breast cancer risk. It is available on a pay-per-view basis (which is why I cannot attach a pdf copy of it with this blog post), but you may like to spend a little money and order yourself a copy of it online if this topic interests you.

    In summary, people need complete, above-board communication of the value and real uses of the genomics information they have purchased. Disavowing or obfuscating the real uses [for prognosis and personalized therapy optimization and prognosis prediction], purely to minimize the company's regulatory obligations or tort risks, disserves consumers and clinicians, and jeopardizes the future of genomics-based health industry as a whole. 

    Additional resources:  

    rs17822931 entry at dbSNP (NIH)

    Boumendijel A, et al., eds. ABC Transporters and Multidrug Resistance. Wiley, 2009.

    Gula R. Nonsense: A Handbook of Logical Fallacies. Axios, 2002.

    O'Connor R. The pharmacology of cancer resistance. Anticanc Res 2007;27:1267-72.

    Ota I, et al. Association between breast cancer risk and the wild-type allele of human ABC transporter ABCC11. Anticanc Res 2010;30:5189-94.

    Toyoda Y, Ishikawa T. Pharmacogenomics of human ABC transporter ABCC11 (MRP8): Potential risk of breast cancer and chemotherapy failure.

    Toyoda Y, et al. Earwax, osmidrosis, and breast cancer: why does one SNP (538G>A) in the human ABC transporter ABCC11 gene determine earwax type? FASEB J 2009;23:2001-13.

    Wang Z, et al. Mining potential functionally significant polymorphisms at the ATP-binding-cassette transporter genes. Curr Pharmacogenom Pers Med 2009;7:40-58.

    FDA letters in 2010 to DTC genomics firms:


    • Comments (0) | Rating: 0/5
  • Comment
  • Bookmark
  • Print