This poster (NOT released under an open license; merely published here with the permission of co-author Edwin Pena of Seton Hall University -- thank you!!!) was at this year's SETAC conference in Nashville, and describes some really interesting challenges with identifying PAHs (or as they call them, PAH-like substances) in fish oil. This includes a methodology to make fish oil from fish, which I reproduce here:

Fish oil was obtained from an “over the counter brand”: Nature’s Bounty (NB) (lot number: 371123-01), Lovaza (a prescription brand provided by Dr. John Sowa, Seton Hall University) DayBrook (DB) (commercial product provided from DayBrook industries), GILA (wild menhaden collected from the Barataria Bay, LA in 2010) and MVNJ (wild menhaden collected from the Delaware Bay, NJ in 2010). Wild menhaden were collected by NJ Fish and Wildlife and LA Wildlife and Fisheries.

Fish oil preparation: The head and tail of the fish were cut off, and the fish filleted and de-boned. The filets were cut into smaller pieces and pounded into meal using a glass tube inside a round bottom centrifuge tube. The meal was centrifuged in a round bottom tube for six hours at 10,000 rpm. Following centrifugation, two top layers could be seen, one oil and one aqueous. The bottom of the tube was punctured to separate the two layers.

The neat part was that the fish they used were menhaden from Barataria Bay, where Public Lab, GRN, and LUMCON folks have done a lot of work: http://publiclab.org/tag/barataria

Vitamin Standards/PAHs Standard Analysis To extract PAHs, the oil was thawed and mixed by vortexing. In a 1.5 ml microcentrifuge tube, 50 μl of fish oil and 1.15 ml of 75% ethanol (EtOH) were combined. The mixture was vortexed continuously for 1 minute and then the oil was separated from the EtOH by centrifuging for 20 minutes at 13,000 rpm. The oil went to the bottom of the tube. One milliliter of the EtOH was removed and placed in a quartz cuvette. Samples were analyzed for fluorescent compounds using two settings on a SpectraMax M5/M5 scanning fluorometer. The first setting involved holding the emission wavelength (Em) at 350 nm and scanning for excitation wavelengths (Ex) from 250 to 340 nm. This setting was best for aromatic hydrocarbons with one or two aromatic rings such as vitamin E (1 ring) and naphthol (2 rings). The second setting involved holding the Em at 450 nm and scanning for Ex from 250 to 430 nm. This setting was best for polycyclic aromatic hydrocarbons with 3, 4 and 5 rings such as hydroxypyrene and for vitamin A even though it has 1 aromatic ring. Figure 7.

So, the big takeaway is that vitamins E and A can give you false-positives for PAH contamination when using UV fluorescence to assess presence of PAHs, and those vitamins can be present in fishes because they occur in phytoplankton. Edwin and Carolyn make a recommendation for excitation/reading wavelengths of 350/450nm, which does not have the same issues with false positives due to vitamins.

We're not really quite at their level yet, and they are using a scanning fluorescence spectrometer while we're just using blue lasers, so I'm really most interested in the methodology for extracting fish oil from fish which you suspect are contaminated with PAHs. Edwin and Carolyn cite a parts-per-billion (ppb) range sensitivity in their tests!

Their full poster in PDF is here: edwin-pah-fish.pdf