Immunomagnetic brads-PCR (IM-PCR), positively-charged virosorb filters (F),
or a combination of both methods (F-IM-PCR) were used to capture, concentr
ate and rapidly detect hepatitis A virus (HAV) in samples of lettuce and st
rawberries experimentally contaminated. Direct reverse transcriptase-polyme
rase chain reaction IRT-PCR) amplification of the collected HAV-beads compl
ex showed a detection limit of 0.5 plaque forming units (PFU) of the virus
present in 1-ml of wash solution from the produce, which was several hundre
d-fold more sensitive than that demonstrated by RT-PCR. In separate trials,
virus-containing wash solutions from the produce were passed through the f
ilters and the captured virus was eluted with 10 ml volumes of 1% beef extr
act. Of the 62% filter-captured HAV, an average of 34.8% was eluted by the
1% beef extract. PCR amplification of 2 mu l from this eluate failed to pro
duce a clear positive band signal. As little as 10 PFU, present on each pie
ce of the lettuce or strawberry, was detectable by the F-IM-PCR, which was
almost 20 times less sensitive than the detection limit of 0.5 PFU by the I
M-PCR. However, considering the large volumes ( less than or equal to 50 ml
) used in the F-IM-PCR, the sensitivity of detection could be much greater
than that of the IM-PCR, which was restricted to less than or equal to 20 m
l volumes. These data indicate that the F-IM-PCR method provides the potent
ial for a greater sensitivity of detection than the IM-PCR, since low level
s of virus could be detected from large volumes of sample than possible by
the IM-PCR method. Although positively-charged filters captured a greater a
mount of virus than both the IM-PCR and F-IM-PCR methods, direct PCR amplif
ication from beef extract eluates was not successful in detecting HAV from
produce. (C) 2000 Elsevier Science B.V. All rights reserved.